KR100556638B1 - Method of Forming a Contact Assembly - Google Patents

Abstract

Contact structures exhibiting elasticity or compliance with respect to various electronic components couple the free end of wire 502 to substrate 508, form wire 530 in the shape of a wire stem 502 with a spring shape and wire stem 530 is formed by overcoating this wire stem 530 with at least one layer of material 522. In one embodiment, the free end of the wire stem 530 is coupled to the connection region of the substrate 508, the wire stem 530 is formed to have a spring shape, the wire stem 530 is freestanding by electric discharge Cut into shapes, the freestanding wire stem is overcoated by plating. Various materials are known for the materials of the wire stem 530 (functioning as a false walk) and the overcoating (540, functioning as a superstructure above the false walk). The elastic contact structure of the present invention is ideal for making "temporary" probes to electronic components such as semiconductor dies for preliminary and functional tests.

Contact structures, semiconductor dies, wire stems, sacrificial substrates, probe connections

Description

Method of Forming a Contact Assembly

DETAILED DESCRIPTION Hereinafter, detailed description will be made of preferred embodiments of the present invention, examples of which will be described in the accompanying drawings. Although the present invention has been described as the content of these suitable embodiments, it should be understood that they are not intended to limit the technical spirit and scope of the present invention to these particular embodiments.

1 is a perspective view of a supply wire having a free (base) end coupled to a substrate by a coupling head (capillary) of a wire assembly according to the prior art.

1A is a side view of a wire having a free end coupled to a terminal on a substrate, in accordance with the present invention.

1B is a side view of a wire having a free end coupled to a terminal on a substrate through an opening in the photosensitive layer, in accordance with the present invention.

1C is a side view of a wire having a free end coupled to a metal layer applied to a substrate through an opening in the photosensitive layer, in accordance with the present invention.

FIG. 1D is a side view of the substrate of FIG. 1C with the wires protectively coated in accordance with the present invention. FIG.

1E is a side view of the substrate of FIG. 1D, with the photosensitive layer removed and the metal layer partially removed, in accordance with the present invention.

1F is a side view of a wire having a free end coupled to a substrate, which may be a sacrificial substrate, in accordance with the prior art.

Fig. 2 is a perspective view of a wire whose free end is joined to a substrate and formed into a certain shape in accordance with the present invention, and schematically illustrates certain components of a suitable wire assembly (elastic mechanism) according to the present invention.

2A is a side view of a wire formed in a shape supporting a resilient protective coating, with a free end coupled to a substrate, in accordance with the present invention.

Figure 2B is a side view of a wire formed in a shape supporting a resilient protective sheath with a free end coupled to a substrate, in accordance with the present invention.

2C is a perspective view of a wire having a free end coupled to a substrate and formed into a shape supporting the elastic protective sheath according to the present invention.

FIG. 2D is a side view of a wire having a free end coupled to a substrate and formed into a shape that supports an elastic protective coating in accordance with the present invention.

2E is a side view of a wire having a free end coupled to a substrate and formed into a shape that supports an elastic protective coating in accordance with the present invention.

Figure 2F shows a free (base) end coupled to a substrate in a "loop" embodiment of the present invention in which a wire loop can be protectively coated with solder in accordance with the present invention, the opposite end of which is also coupled to the substrate. Side view of the wire.

Fig. 2G is a side view of a wire with a wire loop having a free (base) end coupled to a substrate and shaped to support an elastic protective sheath in accordance with the present invention, the opposite end of which is also coupled to the substrate.

Figure 2H is a side view of a wire formed in a shape supporting a resilient protective coating, with a free end coupled to a substrate, in accordance with the present invention.

Figure 3A illustrates the "problem" associated with shaping the wire into a predetermined shape in accordance with the present invention such that the wire "guides" or "delays" the trajectory taken by the capillary.

Figure 3B illustrates the "answer" to the "problem" set forth in Figure 3A in accordance with the present invention, illustrating the trajectory of the capillary that can be offset (changed) with care to ensure that the wire is fabricated to a desired shape. Drawing.

4A is a side view of a wire bond head (capillary tube) raised on a substrate, with the electrode for performing the EFO cutting the wire, in accordance with an aspect of the present invention.

4B is a side view of a wire bond head (capillary tube) raised on a substrate after the wire has been cut by the technique of FIG. 4A, in accordance with an aspect of the present invention.

4C is a side view of a wire bond head (capillary tube) raised on a substrate after the wire is cut by another (FIG. 4B) technique, in accordance with one aspect of the present invention.

FIG. 4D is a side view of a wire bond head (capillary tube) raised on a substrate after the wire is cut by another technique (FIGS. 4B and 4C) in accordance with one aspect of the present invention.

Figure 5 is a side view of a wire that is formed to have any shape and that is protective coated with a multilayer (at least two) material to make an elastic contact structure, in accordance with the present invention.

FIG. 5A is a side view of a wire that is formed to have a “springable” shape and is fully protected with a single layer material to make an elastic contact structure, in accordance with the present invention. FIG.

5B is a side view of a wire that is formed to have a springable shape and is partially protective coated with a single layer material to make an elastic contact structure, in accordance with the present invention.

5C is a side view of a wire that is formed to have a springable shape and is fully protected with a single layer of material for making an elastic contact structure having microprotrusions on its surface in accordance with the present invention.

5D is shaped to have an elastically deformable form in accordance with the present invention and is fully overcoated (“surrounded”) with (at least two) multiple layers of material to form an elastic contact structure having fine protrusions on its surface. Side view of the wire.

Fig. 5E is a side view of a wire shaped to have a form that is elastically deformable in accordance with the present invention and fully overcoated (“enclosed”) as a layer of material and inserted into a material with good electrical conductivity such as an elastomeric material.

5F is a diagram showing the deformation performance of the contact section according to the applied strain force F, which is a schematic diagram of a typical elastic contact structure shaped and overcoated to be elastically deformed in accordance with the present invention (not shown in this figure). .

FIG. 5G is a side view of a wire shaped to have an elastically deformable form and fully overcoated (“enclosed”) with at least one coating while applying heat during an overcoating (eg, plating) process in accordance with the present invention. FIG.

FIG. 5H is a side view of a wire shaped to have a straight pin shape in accordance with the present invention and fully overcoated (“enclosed”) with at least one coating while applying heat during an overcoating (eg, plating) process. FIG.

Figure 5I is a side view of one embodiment where two wire stems are mounted and overcoated on a single terminal in accordance with the present invention.

6A is a side view of an application to the contact structure according to the present invention showing that the elastic contact structure may “start” from various levels on one or more components and “end” at one common plane level.

6B is a side view of an application to the contact structure according to the present invention showing that the elastic contact structure may “start” from one level on one electronic element and “end” at various levels on one or more electronic elements.

6C shows that the elastic contact structure may “start” from the surface of the first electronic element and “end” at the surface of the second electronic element, wherein the surface of the second electronic element is not parallel to the surface of the first electronic element. Side view of an application example for the contact structure according to the present invention shown.

FIG. 7A is a side view of the ceramic semiconductor package with an arrangement of elastic contact structures disposed in rows on an outer surface of the ceramic semiconductor package in accordance with the present invention. FIG.

FIG. 7B is a plastic semiconductor package having an array of elastic contact structures arranged in rows on an outer surface of the plastic semiconductor package and further including a plurality of elastic contact structures connecting the semiconductor mold within the package in accordance with the present invention. Side cross-sectional view in the form of a partial perspective view.

Figure 7C is a side view of the package with an arrangement of elastic contact structures disposed in rows on the outer surface of the PCB base semiconductor package in accordance with the present invention.

Figure 7D is a side view of an overmolded package overcoated with a joining wire to improve its strength and in particular stability during overmolding in accordance with the present invention.

FIG. 8A is a side view of a wire shaped to form a loop with the distal end of the wire joined to a removable element in accordance with the present invention; FIG.

Figure 8B is a side view of the looped wire of Figure 8A after overcoated in accordance with the present invention.

Figure 8C is a side view of the looped overcoated wire of Figure 8B after the removable element has been removed in accordance with the present invention.

8D is a side view of the looped wire of FIG. 8A in accordance with an alternative embodiment of the present invention after the removable element has been removed and before the wire is overcoated.

Figure 9A is a side view of an overcoated wire shaped into a form in accordance with the present invention in one embodiment of an elastic probe;

Fig. 9B is a side view of an overcoated wire shaped into a form in accordance with the present invention in an alternative embodiment of an elastic probe.

Figure 9C is a side view of a multilayer contact pad for an elastic contact structure in accordance with the present invention.

10A-10C are side views of a first step (step-1) of a process of forming a chip-probing card with elastic contacts in accordance with the present invention;

10D-10G are side views of a second step (step-2) of the process of forming a chip-probe card with elastic contacts in accordance with the present invention;

10H and 10I are side views of an alternative embodiment of a second step (step-2) of the process of forming a chip-probe card with elastic contacts in accordance with the present invention;

Figure 10J is a side view of one embodiment of an overcoated wire shaped to function as a probe in accordance with the present invention.

10K is a side view of another embodiment of an overcoated wire shaped to function as a probe in accordance with the present invention.

11A-11F are side views illustrating a process for mounting an elastic contact structure to a removable substrate in accordance with the present invention.

12A-12C are side views of a "gang transfer" technique for mounting an elastic contact structure to an outer surface of a semiconductor package in accordance with the present invention.

12D is a side view of a technique for mounting an elastic contact structure in a recess in one surface of a semiconductor package in accordance with the present invention.

12E is a side view of a technique for manufacturing a probe card in accordance with the present invention.

Figure 12F is a side view of a technique for placing an outer layer on the structure to make it more stable on a substrate on which the elastic contact structure is mounted in accordance with the present invention.

Figure 13A is a side view of a semiconductor package having a resilient contact structure that allows for a temporary (eg, elastic) connection to a test substrate for testing and preliminary testing in accordance with the present invention.

FIG. 13B is a side view of the semiconductor package of FIG. 13A enabling permanent (eg, soldered) connection to a printed circuit board in accordance with the present invention. FIG.

14A-14E are side views of a technique for mounting an elastic contact structure in a semiconductor mold in accordance with the present invention.

14F and 14G are side views of a technique similar to that described for FIGS. 14A-14E for mounting an elastic contact structure to a semiconductor mold prior to separation from the wafer in accordance with the present invention.

Figure 15 is a partial perspective view of multiple elastic contact structures mounted at multiple mold locations on a semiconductor wafer in accordance with the present invention.

Figure 15A is a partial perspective view of a number of elastic contact structures mounted to a semiconductor mold and increasing the effective pitch of "pin out" (bond pad spacing, as is commonly used herein).

16A-16C are perspective views of a process for forming an elastic contact structure on a mold (wafer or cube) in accordance with the present invention.

Figure 16D is a perspective view of an alternative process (relative to Figures 16A-16C) for forming an elastic contact structure on a mold (wafer or cube) in accordance with the present invention.

16E and 16F are side views of a technique for assembling an elastic contact structure in a manner suitable for stacking chips (semiconductor molds) from the top one by one in accordance with the present invention.

Figure 17A is a perspective view of one embodiment of an interposer in accordance with the present invention.

Figure 17B is a perspective view of one embodiment of an interposer in accordance with the present invention.

Figure 17C is a perspective view of one embodiment of an interposer in accordance with the present invention.

Figure 17D is a perspective view of one embodiment of an interposer in accordance with the present invention.

Figure 17E is a side view of one embodiment of an interposer in accordance with the present invention.

Figure 18A is a side view of one embodiment of an interposer in accordance with the present invention.

Figure 18B is a side view of the interposer of Figure 18A disposed between two electronic elements in accordance with the present invention.

Figure 19A is a side view of another embodiment of an interposer in accordance with the present invention.

Figure 19B is a side view of the interposer of Figure 19A disposed between two electronic elements in accordance with the present invention.

20A is a side view of another embodiment of an interposer in accordance with the present invention.

Figure 20B is a side view of the interposer of Figure 20A disposed between two electronic elements in accordance with the present invention.

Figure 21 is a perspective view of another embodiment of an interposer suitable for carrying out engineering changes in accordance with the present invention.

22A and 22B are side views of additional embodiments of interposers in accordance with the present invention.

Figure 22C is a side view of another embodiment of an interposer in accordance with the present invention.

22D-22F are side views of another embodiment of an interposer in accordance with the present invention.

Figure 23A is a side view of an embodiment of a semiconductor package using an elastic contact structure mounted in accordance with the present invention.

Figure 23B is a side view of another embodiment of a semiconductor package using an elastic contact structure in accordance with the present invention.

Figure 23C is a side view of a via-less PCB substrate with flip-chip having external contacts in accordance with the present invention.

Fig. 24A is a partial perspective view of an embodiment of a loop type contact formed on a terminal of a substrate according to the present invention.

Fig. 24B is a partial perspective view of an embodiment of a contact in the form of a solder ridge formed on a terminal of a substrate according to the present invention.

Figures 24C and 24D are partial perspective views of embodiments of terminal connections formed on terminals of a substrate according to the present invention.

Fig. 24E is a perspective view of another embodiment for providing an electronic element having a heat-dissipation structure according to the present invention.

Figure 25 is a side view of an embodiment of a dual printed circuit board assembly having electronic elements mounted on both sides of an elastic contact structure according to the present invention.

Figure 26 is a side view of an embodiment of a dual printed circuit board assembly having electronic elements mounted on both sides of an elastic contact structure in accordance with the present invention.

Figure 27 is a side view of an embodiment of a printed circuit board assembly having one electronic element mounted with an elastic contact structure in accordance with the present invention.

Figure 28 is a side view of an embodiment of a printed circuit board assembly having one electronic element mounted with an elastic contact structure in accordance with the present invention.

Figure 29 is a side view of an embodiment of a printed circuit board assembly having electronic elements mounted with an elastic contact structure in accordance with the present invention.

30 is a side view of an embodiment of a printed circuit board assembly having electronic elements mounted with an elastic contact structure in accordance with the present invention.

Figure 31 is a side view of an embodiment of a printed circuit board assembly having electronic elements with other electronic elements in accordance with the present invention and equipped with an elastic contact structure.

Figure 32 is a side view of an embodiment of a printed circuit board assembly having electronic elements with other electronic elements in accordance with the present invention and equipped with an elastic contact structure.

Figure 33 is a side view of an embodiment of an electronic element packaging method according to the present invention.

Figure 34 is a side view of an embodiment of an electronic element packaging method according to the present invention.

Figure 35 is a side view of an embodiment of an electronic element packaging method according to the present invention.

Figure 36A is a side view of an embodiment of a printed circuit board assembly having electronic elements mounted with an elastic contact structure in accordance with the present invention.

36B is a plan view of an electronic element assembly in accordance with the present invention.

36C is a plan view of an electronic element assembly in accordance with the present invention.

Figure 37 is a side view of an embodiment of a printed circuit board assembly having an electronic element with an elastic contact structure in accordance with the present invention and using another electronic device.

Figure 38 is a side view of an embodiment of a printed circuit board assembly having electronic elements mounted with an elastic contact structure in accordance with the present invention.

Figure 38A is a detailed side view of the embodiment of Figure 38 in accordance with the present invention.

Figure 39 is a partially expanded side view of a test interface arrangement in accordance with the present invention.

Figure 40A is a side view of an embodiment of the present invention using an insulating (dielectric) material to form a wire stem according to the present invention.

Figure 40B is a side view of an embodiment of the present invention using an insulating (dielectric) material to form a wire stem according to the present invention.

41A-41C are side views of alternative techniques for mounting wire stems in contact regions (eg, terminals) on electronic elements in accordance with the present invention.

Figure 42 is a side view of an alternative technique for shaping a wire stem with an external tool in accordance with the present invention.

42A and 42B are side views illustrating alternative techniques for shaping two or more wire stems with external tools in accordance with the present invention.

Figure 42C is a perspective view of an alternative technique for shaping a plurality of wire stems with an external tool in the form of a comb according to the present invention.

Figure 43A is a side view of the capillary of the wire coupler according to the present invention.

Figure 43B is a side view of an alternative embodiment of the capillary of the wire coupler according to the present invention.

Figure 44 is a plan view of an electronic element mounted with a plurality of elastic contact structures in a form arranged to accommodate thermal expansion of the electronic element according to the present invention.

Figure 45 is a perspective view of an elastic contact structure mounted to a substrate to receive a compressive force F into a torsional action force T in accordance with the present invention.

46A-46F are side views illustrating a technique for reducing the overall inductance of an elastic contact structure in accordance with the present invention.

Figure 46G is a cross sectional view of a prior art insulated wire that may be used in certain embodiments of the present invention.

Figure 47 is a side view of a wire stem having a modified thickness (eg, diameter of a circular wire) in accordance with the present invention.

48A-48E are side views, in partial cross-section, of a technique for removing the wire stem after overcoating in accordance with the present invention;

49A-49C are side views, in partial cross-section, of a technique for forming a eutectic lead prepared tip on a contact structure in accordance with the present invention;

50A and 50B are perspective views of a technique for providing a topographical (approximate) tip to a contact structure in accordance with the present invention.

51A-51C are partial cross-sectional side views of a technique for forming a partially overcoated contact structure in accordance with the present invention.

52A-52D are perspective views of a technique for fabricating an elastic contact structure suitable for making a connection in an exposed intermediate portion of the wire stem according to the present invention.

Figure 52E is a perspective view of a technique for fabricating multiple free standing contact structures without cutting wire stems in accordance with the present invention.

Figure 52F is a perspective view of an alternative technique for fabricating multiple free standing contact structures without cutting wire stems according to the present invention.

53A and 53B are perspective views of alternative techniques for multiple free standing contact structures that do not cut wire stems according to the present invention.

53C and 53D are side views illustrating a technique for manufacturing a free standing upright wire stem according to the present invention without electronic heating from a loop.

Figure 54 is a side view showing a part of a contact structure connected to an external element according to the present invention.

In the side views shown above, some of the side views are often shown in cross-section for clarity. For example, in many of the drawings, the wire stem is shown completely in bold solid lines, while overcoating is shown in substantially cross section (often without hatching).

In the figures shown above, the size of certain elements is often enlarged (not the actual size relative to other elements in the figure) for clarity.

In the figures shown above, components often have numbers in the figures as “prefix” and “suffix” refer to the same element (eg, element 108 in FIG. 1). Is the same substrate as element 208 of FIG. 2 and element 808 of FIG. 8A.

The present invention relates to contact structures for making electrical connections between electronic components, in particular microelectronic components, in particular contact structures having elasticity and / or compliance.

The parent application of this application is U.S. Patent Application No. 08 / 340,144, co-filed November 15, 1994, and PCT Patent Application No. PCT / US94 / 13373 (in progress; The application is hereby commonly cited as CASE-2), all of which are partly filed in US patent application Ser. No. 08 / 512,812, filed November 16, 1993.

Gold is the "best material" for electrical connections between electronic components because of its excellent conductivity and non-corrosiveness. For example, it is well known to make a plurality of wire bond connections between conductive pads of a semiconductor die and inner ends of lead frame branches. This is known as an example for making a permanent connection between a first "active" electronic component (die) and a second "inactive" electronic component (lead frame).

The present invention advantageously uses wire-bonding equipment as appropriate, in which generally wires (eg gold wires) are fed from a spool through a capillary (or “bonding head”) to bond to the substrate. do. In general, the properties of the coupling head will be determined by the properties of the coupling made thereby. When the coupling head engages in ball engagement, it is generally a "capillary tube". When the coupling head engages wedges, this is generally a “wedge”, which terms are recognized in the art. To simplify the subject, hereafter mainly the term “capillary tube” will be used to indicate a bonding head suitable for making ball or wedge bonds that apply thermal energy and / or pressure during bonding.

The following US patents, incorporated by reference herein (patented by patent number, inventor's name, published month / year, US classification / subsection), indicate the status of the wire bonding technology area.

(a) U.S. Patent Application No. 5,110,032 (Akiyama et al .; 5/92; U.S. Classification 228/102) entitled Wire Bonding Method and Apparatus, discloses a wire 13 supplied from a wire spool 12 through a capillary tube 10. ) Is disclosed. [In the present invention, the wire 13 is insulated] A control unit 20 is shown that includes a CPU and a memory unit (software instruction storage). The control unit controls the movement of the capillary and the discharge power circuit 18 used to cut the wire with the discharge voltage in relation to the discharge electrode 17.

(b) US Patent Application No. 3,460,238 (Christi et al .; 8/69; US Classification 227/111), entitled Wire Cutting of Wire Joining Machines, states that the wire cutting operation of the wire coupler is sufficient to frictionally bond the wires. And moving the bond needle (or “capillary tube” as used herein) to a state of insufficient holding pressure to deform the wire in the bond zone. The present invention is cited as an example of the fact that wire bonding has been known for decades and that it is generally not desirable to "strain" the wire while moving the capillary.

(c) U.S. Patent Application No. 5,095,187 (Glyga et al .; 3/92; U.S. Classification 219/68) entitled Wire Weaking, supplied through a wire coupler, provides one or a combination of heat, pressure, and vibration. To provide a wire bonding technique in which a wire is coupled to a contact portion of an electronic component. The present invention discloses the weakening or cutting of the wire by local heating, and how the cutting operation can be an enlarged part on the cut end of the wire. The cutting heat can be applied to the wire by the electrode and can be made to extend the electric field from the electrode to the wire, creating an arc between the electrode and the wire. The present invention describes a cutting technique in which the arc-shaped first portion becomes the first polarity for weakening the wire and the arc-shaped second portion becomes the reverse polarity for controlling the dispersion of charged particles emitted from the wire.

(d) U.S. Patent Application No. 4,860,433 (Miura, 8/89; U.S. Classification 29/605), entitled Inductance Element Fabrication, discloses a technique for winding a coil of fine copper wire to a spool member on a substrate. The copper wire is insulated. As is the result of pre-bonding, it is known that the insulation is removed from the ends of the wire when an electronic flame off (hereinafter EFO) spark cuts the wire. The end of the wire is coupled to the conductive pad of the substrate. Thus, either the capillary or the substrate is rotated and the table supporting the substrate can be moved in the vertical direction, thereby winding a coil of fine copper wire on the spool member. Eventually the opposite end of the wire is coupled to another conductive path of the substrate.

Packaging is an important field in realizing a plurality of electrical connections between the first electronic component and the second electronic component. For example, in a ceramic package, the semiconductor die is arranged in a cavity of the ceramic package and wire bonded to a (typically) extending conductive trace. A plurality of pins (or pin arrays) are arranged on the outer surface of the package such that the pins are connected to internal traces (patterned with a conductive layer) and biased against the jointly extending conductive traces. The package is then mounted to a printed circuit board (PCB) with a corresponding plurality of holes (eg, an arrangement of holes) in which each hole receives a corresponding respective package pin. These pins are typically soldered to the plated through holes to realize a permanent connection between the first electronic component (packaged semiconductor device) and the second electronic component (PCB). Alternatively, the package is housed by a socket having a corresponding plurality of holes (eg, an array of holes) in which each hole receives one of the package pins so that the first electronic component (packaged semiconductor device) and the second electronic Temporary connection is realized between parts (sockets).

In general, it is known to protect semiconductor devices against moisture. To this end, different packages exhibit varying degrees of sealability, for example, ceramic packages generally provide excellent sealing properties to the environment at relatively high cost, and include plastic (eg resin) and PCB (packed) The package is relatively inexpensive to the environment at a relatively low price. It is known to coat the joining wires and their peripheral connections (eg, for dies and lead frames) in order to have the best of both fields, ie good sealing at low cost. One example can be found in US Patent Application No. 4,821,148 (Kobayashi et al .; 4/89; US Classification 361/392), a resin packaging semiconductor device entitled Metal-Organic Mixture Protective Layer Mid. In the present invention, the silver electrode 4 on the lead frame 2 is coupled to the aluminum electrode 5 on the silicon chip 1. The final assembly is immersed in a benzotriazole (BTA) solution of ethyl alcohol. The Ag-BTA film is formed on the surface of the silver electrode, the Al-BTA film is formed on the surface of the aluminum electrode, and the Cu-BTA film is formed on the surface of the copper wire. These three metal-BTA films protect wires and electrodes from damage from ambient humidity. Importantly, in the present invention, the interconnection already achieved (prior to coating) by the metal bond between the metal-BTA film and the wire bond itself is not important. Suitably, such "wet" coatings are not conductive because they wear down the device (die).

Fins, ie electrically conductive elements with elongated stiffness, are known and are generally brazed to a pad of an electronic package (including a pin carrier).

U.S. Patent Application No. 3,373,481 (Rins et al .; 3/68; U.S. Classification 29 / 471.3), entitled `` Conductor Interconnect Method '', discloses an integrated circuit device 10 by means of a thermopressing gold sphere 13 (FIG. 2). The formation of the pin-shaped gold base structures 13 above the terminal portion 12 and the formation of the spherical body with the heating vacuum holder 14 is disclosed. The use of ultrasonic binders in vacuum holders is discussed here. As shown in FIG. 3 of this invention, a bond is formed between the foundation structure 13 and what is seen as the entire surface of the terminal portion 12.

U.S. Patent Application No. 4,418,857 (Einslier et al .; 12/83; U.S. Classification 228/124), titled High Melting Point Process for Au: Sn = 80: 20 Brazing Alloys for Chip Carriers, describes the use of brazing pins for chip carrier substrates. Exemplary techniques are disclosed.

In U.S. Patent Application No. 4,914,814 (Bunun et al. 4/90; U.S. Classification 29/843), titled Circuit Package Assembly Process, an array of pin holes is registered in a pin mold that is effectively registered as an array of conductive pads on one side of the chip carrier. Filling the array of pin holes with lead / tin solder, heating the lead-tin solder in the pin mold so that the solder melts and coalesces with the array of conductive pads on the chip carrier, thereby joining a small array of conductive pads on the chip carrier. Disclosing forming an array of pins. This allows the formation of elongated solder terminals having a controlled height and a relatively high aspect ratio.

US Patent Application No. 4,955,523 (Carlo Magno et al .; 9/90; US Classification 228/179) entitled Electronic Component Interconnect discloses a technique for interconnecting electronic components, wherein the interconnect wire is a contact material. And contacts to the first part (such as semiconductor die 1) without using materials other than wires and wires. These wires are then cut into predetermined lengths between two and twenty times (2d to 20d) of the wire diameter, and are joined to the contacts on the second component 21 by a conductive material such as soldering. These wires, once joined to the first part, are cut to a predetermined length (as long as the joining head 9 of the wire assembly) through the opening 13 in the side wall of the joining head. For this reason, the electrode 51 is inserted into the opening 13. As shown in the present invention, the free standing wires 7 allow their ends to be inserted into the pool 27 of conductive material, such as soldering in the recesses of the second part. See also US Patent Application No. 5,189,507 (Carlomagno et al .; 2/93; US Classification 257/777) entitled Electronic Component Interconnect.

Surface mounting technology solves some of the problems associated with forming interconnections between electronic components, but in itself has proved to fail to achieve the once-overcoming full overcome. Generally, in surface mount technology (SMT), including flip-chip technology, solder bumps are formed on the surface of the first electronic component, and pads are provided on the surface of the first electronic component, the component After they are faced to each other, heat is applied to flow the solder of the solder ridge. Attempts to realize reliable surface mounting are sometimes referred to as “solder columns” and include forming solder ridges having a controlled shape and a high aspect ratio (a ratio of height to width).

U.S. Patent Application No. 5,130,779 (Agawala et al .; 7/92; U.S. Classification 357/67), titled Soldering with Conductive Cover Arrangement, describes high aspect ratio "stretched" brazing rows. On the “electron carrier (substrate)”, a first solder is deposited and a pad covered by the metal layer is formed. Another (second) solder is formed on the first solder. Another (third) braze can be formed on the second braze.

Prior testing and testing is the field of other attempts to use a plurality of temporary connections formed between a first electronic component, such as a semiconductor die, and a second electronic component, such as a test card. As described in detail below, this typically involves packaging the semiconductor die first, and on a package with a special test fixture such as a "nail bed" arranged in a pattern corresponding to the pattern of pins on the semiconductor package. Contacting the pins. The fabrication of such test fixtures in particular requires different test fixtures for all alignment patterns, resulting in expensive and time consuming.

Forming elastic interconnects between electronic components is generally known to be desirable and has been the goal of an ongoing attempt. Mainly (eg, typically), external connection to a semiconductor device is achieved by rigid pins arranged on the outer surface of the package relatively. Many attempts have been made in the past to implement elastic interconnect structures. In some cases, elastic connections have been achieved as metal elements. In other cases it has been realized as a combination of metallic elements and elastomeric elements.

One recent attempt to achieve an elastic connection is described in a paper describing the "Elasticon Connector" described in the 1993 ICEMM Bulletin 341-346, Elastomer Connector for MCM and Test Equipment. . Ilesticon connectors use solid gold or gold alloys for conductive elements embedded in an elastomeric material (eg, a liquid elastomer resin injected into a mold cavity), typically a plurality of chip modules (MCMs) or a single Used for interconnect requests for land grid array (LGA) packages for needle modules (SCM). Together with the properties of the elastomeric material, the size, shape and wire spacing can be tailored for special applications including high density PCB and IC chip testing equipment as well as MCM and SCM packaging, board-to-board and cable-to-board interconnects. Can be. Solid gold wires and silicone elastomer materials are not damaged by corrosion. Figure 1 of the paper describes a basic embodiment of an elesticon connector, in which a plurality of wires are ball-bonded to a rigid substrate and extend linearly at an angle (eg 45-85 °) from the surface of the substrate. do. Attaching the base end of the wire to the substrate is an angled flying tip wire bonding process that uses the compressive force and ultrasonic energy applied through the capillary tip and the thermal energy delivered through the heated stage on the wire assembly. Is achieved by The capillary and substrate are arranged such that the shear blade mechanism can cut the wire at a predetermined height and angle at the substrate surface. The EFO is used to melt the wire extending from the capillary tip to begin the next ball bonding (of the base end of the next wire to be bonded to the substrate). After all the wires are mounted on the substrate, ball-shaped contacts are formed at the distal ends of each wire by a laser ball forming process, and the plurality of wires are embedded in the elastomeric material. The ball-shaped centrifugal ends prevent the wires from loosening due to vibration and shortening between the contacts. As suggested in this paper, since the elesticon connector is compressed between two parallel surfaces, each forming direction of the conductor is essential to minimize the plastic deformation of the wires. Each forming direction also provides a "wiping" contact surface that allows the wire to rotate and slide relative to the mating contact surface when the connector is compressed. This paper discusses the use of gold / palladium alloys and platinum for wires. Figure 3 of this paper describes wires clustered into groups of four wires at one contact pertaining to the formation of grooves in the elastomer between the wires of each group. Various embodiments of the elesticon connector discussed in this paper require a substrate of ceramic, metal, silicon or epoxy-glass layer material, while embodiments of the interposer have a thin layer of gold on the top surface. Etchable substrate materials such as copper provided are required. Figure 8 of this paper describes the integrated probe contacts and properly explains that the test capability for known good dies is one of the barriers to overcome in MCM packaging. As shown here, the probe matrix uses 0.005 cm (0.002 inch) diameter gold wires in the array. The probe may be permanently attached to the test module or may be fabricated in an interposer structure. U.S. Patent Application No. 5,386,344 (Bearman et al., 1/95; US Classification 361/785), entitled Flexible Circuit Card Elastomer Cable Connector Assembly, discloses a related "ilestpack" turnable polymer cable connector.

Other descriptive attempts at forming the resilient connections have been described in the United States, which discloses independently bendable springs having horizontal resilience, including sine, helical, cantilever and buckling beam shapes, in sheet and wire shapes, in the form of spring array connectors. Patent application 5,229,939 (Walker et al .; 4/94; US classification 439/74). The connectors thus formed have practical flexibility to provide reinforcement against changes including manufacturing tolerances, alignment tolerances, and thermomechanical expansion. The manufacture of the spring connector generally proceeds as follows. The three-dimensional mandrel defines the inner surface shape of the spring 12. An insulating layer 50 is applied on the spring 12. The conductive layer 14 is applied on the insulating layer at a predetermined position. The plurality of springs 12 are a single sheet of elastic material 12 that can be easily deposited (on the mandrel), such as nickel (96%)-phosphorus (4%) and nickel (97%)-cobalt (3%). It is formed as a layer. This mandrel is removed after depositing the spring layer 38. Alternatively, this mandrel may be a "sacrifice" as is known in the field of "lost wax" casting. The use of elastic wires as a spring layer (e.g., as an alternative to using thin plates) is discussed, covered by insulating layer 21, covered by segmented conductive layer 212, and an outer insulating layer ( Wire 204 covered by 214. The final spring connector is essentially intended to form a contact (to the attachment) (if not single). As appropriately set in the present invention,

“There must be a difference between the two types of electrical connections, namely attachments and contacts. Attachment is a relatively permanent connection and typically includes durable metal connections such as soldering, micro brazing or micro welding connections. Contacts are relatively temporary connections and mean separate connections between mating connectors and generally rely on compressive forces without durable metal connections. ”

Other descriptive attempts at forming elastic connections can be found in U.S. Patent Application No. 4,067,104 (Tracy et al .; 1/78; U.S. Classification 29/626), entitled "Making Flexible Metal Interconnect Arrays for Joining Small Electronic Components". Can be. The invention discloses fabricating an array of cylindrical columns, such as several indium layers, over an active device such as a semiconductor, which makes the columns flexible to accommodate different thermal expansions.

In US Patent Application No. 4,793,814 (Jipcock et al. 12/88; US Classification 439/66) entitled Electrical Circuit Board Interconnect, electrical interconnections between corresponding contact pads of opposing first and second circuit boards are described. A connector arrangement is disclosed to provide an electrically nonconductive support member comprising a resiliently elastomeric material. The electrically conductive interconnecting element extends through the thickness of the support member and has a pair of pad engagement surfaces arranged to engage the respective contact pads.

U.S. Patent Application No. 4,330,165 (Apostles et al .; 5/82; U.S. Classification 339/59) entitled Press Contact Interconnect discloses an elongated rod member (1) made of rubbery elastomer and the rod ( It consists of a plurality of linear conductive bodies 2 buried in the rod member 1 and two thin plate members 3 and 3 joined to the side of the rod member 1. The ends of the linear bodies 3 appear as opposing faces, i.e. as the top and bottom surfaces of the rod member 1, forming contact surfaces with the circuit board so that the interconnecting substrate is sandwiched therebetween, such as a sandwich. Is mounted. The linear conductive bodies 2 are each ribbons of metal, for example.

U.S. Patent Application No. 4,295,700 (Apostles et al .; 10/81; U.S. Classification 339/61), entitled Interconnect, discloses a press-type interconnect having a rectangular connection piece of sheet or film of elastic material, It is an assembly of an alternating electron conductive elastic material and an electrically insulating material, the rectangular connecting piece having a plurality of electrically conductive paths (ie, electronically conductive strips) as a whole. The rectangular connecting piece is sandwiched between two retaining members made of an insulating material.

U.S. Patent Application No. 3,795,037 (Rootmer; 5/74; U.S. Classification 29/628), entitled Electrical Connector Device, includes a plurality of curved or curved portions buried in and extending between surfaces of elastomeric insulating material. Ten stretched flexible conductors are disclosed. Exemplary flexible flexible conductors are formed of any suitable electrically conductive elastic material such as, for example, phosphor bronze, have an exemplary cross-sectional size of approximately 25 microns (μm), and may have a free length of 2 mm.

US Patent Application No. 5,067,007 (Kanji; 11/91; US Classification 357/54), which is a semiconductor device having a tip for mounting on a surface of a printed circuit board, discloses improving the reliability of a surface mount package. In the case where the package is mounted to the wire substrate, the tip pins axially exerted exhibit a smaller bending force than the coupling force at the joint. For this purpose, lead pins are made of a material having great elasticity, such as a fiber reinforced material, a strain pseudo elastic material, an ultra high tensile material, or a heat resistant ultra high tensile material. The invention also shows that all pins in a pin grid arrangement (PGA) package are not necessarily equal in length—finally some pins do not contact pads (eg, on a test substrate). 7A and 7B of the present invention are of particular interest, where the fins 20 have a special shape and are formed of a special material. These pins 20 are made of a material having a Young's Modulus of less than 15 × 10 10 Pa and have a curved portion in the center of the arc so that displacement from the axial direction is less than half the diameter of the pin. It becomes big. Examples of materials that satisfy the Young's Modulus formula include elongated fine elasticity, bound with a soft material such as copper as high purity copper (Cu), high purity iron (Fe), high purity nickel (Ni), a copper alloy, and a bonding material. It is a composite wire composed of wire. With this shape and formed from this material, the strain strength (yield strength) of the pin is greater than the bond strength of either the brazing metal 12 (keeping the pin in the package) or soldering (13, connecting the pin to the PCB). Becomes small. When thermal or mechanical stress is applied on the fins, they undergo elastic deformation to reduce the stress. In the example shown and described in FIG. 1D, tungsten (W) molybdenum, carbon amorphous metal and fine wires with large elasticity are described. Fine wire is a composite wire that is bundled with a soft material such as copper as a bonding material. The composite wire has a plating metal 11b composed of gold (Au) or gold / nickel. As explained above, gold is very soft and inelastic in nature. As particularly mentioned in the present invention, “The thickness of the plating should be small so that the effect of bending strength (of the composite wire 11A) can be neglected. Plating is effective in facilitating soldering, in particular having a thickness of 1 to 4 μm for nickel and 0.1 to 1 μm for gold ”(see paragraph connection lines 7-8).

US Patent Application No. 5,366,380 (Raymond; 11/94; US Classification 439/66), entitled Spring-Spring Contact Elements for Electrical Connectors and Integrated Circuit Packages, discloses contact elements for electrical connectors and integrated circuit packages (ICP). This is especially useful for surface mount applications. The contact element has a base, a spring portion having a spiral spring element at least partially, and a tapered contact portion that is joined in a convenient manner with the conductive rim of the hole, the flat sheet being a punching, rolling, and / or forming operation, with a thin wall It can be manufactured in the drawn portion, or module portion. The contact elements are kept compressed by a hold-down mechanism. The contact element is coupled with an insulating housing and / or spacers that function as alignment, compression limiting, contact support, and mounting fixation.

U.S. Patent Application No. 4,764,848 (Simpson; 8/88; U.S. Classification 361/408), a surface-mounted array strain relief device, discloses a pin hole in a substrate on which an integrated circuit chip or similar electronic component or device 18 is mounted. Discloses a mounting "route" of a relatively thin finned electrical conductor extending through. These pins 22 are made of an electrically conductive material, such as copper, each pin extending between this root and this tip to provide strain relief when the tip of the pin is connected to the surface.

U.S. Patent Application No. 5,317,479 (Pie et al .; 5/94; U.S. Classification 361/773), entitled Flexible Plated Lead, provides mechanical and electrical connections between a board contact on a circuit board and a chip contact associated with the circuit chip. A curved lead that provides is disclosed. The curved leads are virtually completely plated by soldering and formed into a single piece of conductive material. Generally, the curved leads have two parallel surfaces, one for connecting with the chip contacts and the other for connecting with the substrate contacts, with at least one curved portion therebetween. The leads are suitably, for example, 0.0472 cm (0.018 inch) wide and 0.1178 cm (0.070 inch) overall length (before the bend) and are formed of a thin strip of material, suitably for example from 0.00762 to It is formed of thick beryllium copper and cobalt alloys of 0.0127 cm (0.003 to 0.005 inch) thick. The leads are carefully controlled so that they are covered with the solder in such a way that excess solder does not interfere with the compliance of the leads.

U.S. Patent Application No. 4,989,069 (Hockins et al., 1/91; U.S. Classification 357/74), which is a semiconductor package having a lead broken from the support, is caused by a semiconductor package having a different coefficient of thermal expansion and a printed circuit on which the package is mounted. Disclosed is a stress buffer frame having a flexible metal lead that reduces stresses to be reduced. The leads are mainly flat ribbon-like leads, which are curved to extend from the buffer frame and bent again to have a portion 38 parallel to the buffer frame. Nickel / iron alloys are discussed as the material for the frame (and leads).

The reason for providing some degree of compliance with electrical connections is to absorb stresses caused by thermal cycling. In general, when a semiconductor device (die) is operated, it generates heat. This causes the die to expand at a different speed than the package on which the die is mounted, and similarly causes it to expand at a different speed than the electrical connection between the package and the die. (Assuming that the heat generated by the die is adequately uniform through the die, the die is likely to expand around its center, or center of symmetry.) This is the substrate on which the die is mounted (and the electricity between the die and the substrate) It is similar to the case of a surface mounted die in the state of expansion at a different speed than the connection). The difference in coefficient of thermal expansion between the die and the peripheral elements (eg, package, substrate, electrical connection) creates mechanical stress. In general, it is desired to resolve and reduce these thermally induced mechanical stresses. For example, U.S. Patent Application No. 4,777,564 (Duffney et al .; 10/88; U.S. Classification 361/405), in the form of a lead for use with surface-mounted components, extends from electronic components and is highly resistant to thermal cycling. An electrical connection connected to a printed circuit board having stress fracture resistance is disclosed.

In U.S. Patent Application No. 5,086,337 (Noro et al .; 2/92; U.S. Classification 357/79), titled Connection Structures of Electronic Components and Electrical Devices Using Such Structures, LSI chips (semiconductor dies with large degree of integration) and wires The degree of desirability for having strain (degree of freedom) and flexibility (spring) in the vertical direction (z) under bonding between the substrates is discussed (for example 50 to 55 lines on four sides). This invention discloses a connection structure for electrically connecting electrical parts such as LSI chips to a substrate such as a wire substrate having a function of absorbing different thermal expansion in the horizontal direction and deformation ability in the vertical direction. A representative embodiment of the connecting structure is a connecting structure having a flat shape, a spiral spring shape, one end of the flat spiral spring connected to one electronic component, and the other of the flat spiral spring connected to the other electronic component. a) and 2 (b). This allows the two parts to deform in the z direction, while maintaining the connection between them. Flat, spiral spring connectors are generally formed of Cr-Cu-Cr "sandwich", annealed and coated with Au to increase their solder-wetability. (Cr is relatively non-wetting compared to Au.) Also disclosed is U.S. Patent Application No. 4,893,172 (Matsumoto et al .; 1/90; U.S. Classification 357/79), titled Connection Structure of Electronic Components and Manufacturing Method. It is. (In this patent application, the term "flexibility" is generally a term related to plasticity and the term "elasticity" is related to a spring.)

As a general proposal, due to the brittle nature (weakness) of semiconductor dies, it is much easier to fabricate elastic contact structures on electronic components than semiconductor dies. This has led to the development of "passive" electronic components arranged (embedded) between two electronic components having elastic contact structures bonded therein and having corresponding contact pads in use.

In US Patent Application No. 4,642,889 (Grabe et al .; 2/87; US Classification 29/840), entitled Flexible Interconnect and Method thereof, an interposer having an interconnect area between the circuit board and the device is provided with a conductive strip on the circuit board. And positioning said surface area and said interconnect area arranged between pads on a surface mounting apparatus. The interconnect area of the interposer has arranged therein a plurality of fine wires with fluid and solder. It is disclosed to have a high soldering foundation so that practical compliance is useful for compensating for mismatches in thermal expansion coefficients, and it is also disclosed to install such a high soldering foundation in an interposer.

US Pat. App. No. 3,509,270 (Dub et al .; 4/70; US Classification 29/625), entitled Interconnects and Fabrication for Printed Circuits, discloses an insulator having an opening 4 extending with a compression spring 6 inserted therein. (2) is disclosed. The insulator is sandwiched between two electrically conductive circuit members 8, 10 coupled to the insulator by the resin layer 12 with the spring inserted into the opening and out of the opening. The springs are presented as being made of a gold alloy containing platinum, silver, copper and zinc, and a brazing sheath is provided.

U.S. Patent Application No. 3,616,532 (Beck et al .; 11/71; U.S. Classification 174 / 68.5), entitled Multilayer Printed Circuit Interconnect Device, provides a similar interconnect arrangement (Dube et al.), Wherein coil springs are used for melt soldering. It is compressed and inserted into the port and then withdrawn from the port to solidify the solder to keep the coil of the compression spring close to each other with respect to the return force of the spring. These compressed springs are then inserted into the opening of the interposer layer 24 arranged between the two printed circuit boards 10, 12. The assembly (substrate, pre-compressed spring, interposer with substrate) is heated such that the solder holding the coil of the spring weakens and releases the multiple spring tension. These springs thus expand between the printed circuit level (substrates) and form a contact.

U.S. Patent Application No. 4,667,219 (Lee et al .; 5/87; U.S. Classification 357/68) titled a semiconductor chip contact surface includes a plurality of openings arranged under a semiconductor chip 18 having contact arrays 44, 46, and 48. A connector plate 80 having 82 is disclosed. The openings of the plate are aligned with the contacts of the chip. A plurality of flexible conductors (described as S-shaped copper wire 84 in FIG. 8) extend through each opening of the connector plate and are arranged below the chip contacts and the connector plate (transfer elements 64, 67, 72). Thus, each contact on the chip is flexible and electrically coupled to one of the terminals, respectively.

U.S. Patent Application No. 4,705,205 (Allen et al .; 11/87; U.S. Classification 228/180), titled Chip Carrier Mounting Device, describes an "interconnect preform placement device" (known as an "interposer"). . In Figures 11A-11C and Figures 12 and 13 of the present invention, a variety of soldering "preliminaries having S-shaped (FIGS. 11-11C) or C (FIG. 12) or spring shapes of coils (FIG. 13) are shown. Molded body ”is disclosed. These preforms are arranged in openings (holes) through a support (“hold”, “hold”) layer (eg, 50, 52 in Fig. 11A). A preform of C type is arranged around this support layer. These preforms are "filled solder composition" materials or "supported solder" containing a filler of discrete particles (sold supported by a support strand or tape arranged around the outer surface of the solder preform shape). It retains its shape during solder melting or reflow. For example, the particles of filled solder composition have a melting point above the melting point of the solder. Filling materials such as high temperature polymers or glass membranes coated with copper, nickel, iron, and metal have been proposed, and as described above, separate particles (e.g. powders) and continuous as single strands or many strands in each preform. Length can be in the shape of.

As mentioned above, the electronic components are connected to printed circuit boards (PCBs), also known as printed wire boards (PWBs). Printed circuit boards (PCBs) are technologically advanced and generally involve the formation of conductive traces on insulating substrates to realize complex interconnections between electronic components mounted or plugged into the PCB. PCBs having conductive traces on both sides of the substrate as well as multilayer arrays of insulating and conductive layers are known. It is also generally known to realize a connection from layer to layer within a PCB. A more in-depth discussion of basic PCB technology can be found in the printed circuits of Linden's spatial technology printed in Prentis-Hole, 1962, which is presented here by reference. In any of the embodiments described below for installing PCB substrates, these substrates may be formed of materials other than "traditional" printed circuit board materials. For example, a “PCB” substrate may be optionally formed of one or more plastic materials, such as polyimide, with a conductive foil layer sandwiched therebetween as is known. The term “circuited” will also appear below, indicating the conduction pattern on the PCB substrate.

US Patent Application No. 4,532,152 (Elade et al .; 7/85; US Classification 427/96), a printed circuit board fabrication with a metallized channel titled, discloses at least on its side to define a predetermined set of conductive paths. An injection molded plastic substrate is provided to provide a pattern of channels. Substrate metallization, electrodeless plating, electrolysis, gas plating or vacuum deposition by flame spraying are presented. The substrate (eg, 20 in FIG. 12) is provided with a plated through hole 58 that allows electronic components mounted on one side of the printed circuit board to be electrically interconnected with conductive paths on the other side of the printed circuit board. This invention is only cited as an exemplary plated through hole.

In general, the above-described techniques for realizing electrical connections between electronic components require their own “methodology”, ie each requires a distinct form of the connection structure itself (eg, coupling wires, pins, etc.). .

In addition, each of the techniques described above is impaired due to inherent limitations. For example, replacing a first electronic component permanently connected to a second electronic component typically requires careful desoldering from the second electronic component, after which the first electronic component is replaced with a second electronic component. Carefully solder it. Sockets aim to solve this, but add cost (sometimes too expensive) to the entire system. Also, plugging into the socket shows poor connection reliability when compared to soldered connections. Surface mounting techniques as described above, which provide solder balls to the first electronic component and conductive paths to the second electronic component, require a carefully controlled process to reliably realize and prevent them from disassembling well ( Replacement of one of the electronic components).

In terms of the frequency of using gold for connection between electronic components, although gold exhibits excellent electrical conductivity, it is damaged by certain disadvantages associated with the present invention. For example, gold has very low yield strength and properties that make it extremely counterintuitive for operating gold wire in elastic contact structures. In simple terms, when physically stressed, gold (for example, gold wire) deforms and tends to maintain its deformation shape. This is the so-called plastic deformation.

Another drawback of using gold wire as an interconnect medium is the nature of the gold to interact with the tin content of the solder, ie common lead-tin solder. Notwithstanding this fact, some eutectic solders are known which exhibit desirable interconnectivity, such as gold-tin. Eutectic materials and properties will be described in more detail below.

It is a general object of the present invention to provide an improved technique for interconnecting electronic components, in particular small electronic components.

Another object of the present invention is to provide an improved technique for mounting contact structures on electronic components, in particular semiconductors.

It is another object of the present invention to provide an improved technique for manufacturing an elastic contact structure.

It is another object of the present invention to provide an improved technique for manufacturing electronic assemblies.

It is another object of the present invention to include improvements in cutting bond wires and to provide a technique for the formation of wire bond connections in electronic components.

Another object of the present invention is to provide an improved technique for plating purposes.

According to the present invention, a flexible elongate member (such as a gold wire) is mounted to a contact area on the surface of an electronic component (such as a semiconductor die, printed circuit board or semiconductor package, etc.) and has a “wire stem” having at least one bend. Springable shape).

The elongated member functions as a "falls walk" for the next "upper structure" built on the fall walk. The superstructure includes at least one layer applied, such as plating, on the poleswalk (or previously applied superstructure layer) to apply not only the poleswork but also the remaining areas of the contacts on the substrate. The upper structure (overcoating) is a continuous structure covering the contact area and the wire stem.

It can be firmly mounted to an electronic component and used to realize a temporary and / or permanent connection to another electronic component of the electronic component, or can be used to realize a permanent (or soldered) connection of the electronic component to another electronic component. The final elastic and / or flexible (elastic) contact structure is thereby formed.

In general, as used herein, compliance includes elasticity (elastic deformation) and flexibility (plastic deformation). In some cases, both elastic and plastic deformation are required. In other cases (such as contact structures used for probes), only elastic deformation is required, but some plasticity may be acceptable. In other cases, a complete spring, which exhibits purely elastic deformation, is not desired because too high a load can be applied on the substrate to be mounted. In other cases (such as contact structures on interposers), pure firing may be desired to accommodate non-flatness. The present invention allows the plastic and elastic deformation of the contact structure to be adapted for the application in which it is used.

According to this feature of the invention, the stem plays almost no role once the superstructure is made, but in terms of being driven at home, in some embodiments of the invention, the structural or electrical properties of the final elastic contact structure are greatly increased. Without reducing the stem can be partially or completely removed after fabricating the superstructure. Also, as described in detail above, the stem need not be electrically conductive. In general, once overcoated, the stem has little or no structural impact on the mechanical properties of the final contact structure (if not removed).

In most embodiments of the invention presented herein, the stem has a springable shape so that it can function as an "initial" elastic structure. However, in some embodiments presented herein, if the elasticity of the final contact structure is not a concern in any of these embodiments, the stem is only formed in a straight line.

In an exemplary embodiment of the invention, the poleswalk (wire stem) is a gold wire having a diameter in the range of 0.0018 to 0.0051 cm (0.0007 to 0.0020 inch), and the upper structure (overcoating) is 0.0025 to 0.0254 cm (0.0001 to 0.0100). nickel plating with a thickness in the range of inches).

According to one aspect of the invention, the plurality of elastic contact structures are mounted to an electronic component having a plurality of contact areas. For example, up to hundreds of elastic structures may be mounted to a semiconductor die or semiconductor package having up to several hundred bonding pads or package pads, respectively.

According to one aspect of the invention, a plurality of elastic contact structures can be established at different levels of contact areas on one or more electronic components, which are all intended to be terminated in a common plane for connection to other electronic components having a flat surface. Can be.

According to another aspect of the invention, two or more elastic contact structures can be fabricated on a single contact area of an electronic component.

According to one aspect of the invention, a first group of elastic contact structures can be fabricated on a first side of an electronic component (such as an interposer substrate), and an electrical connection between two electronic components on which the interposer substrate is arranged In order to realize this, a second group of elastic contact structures can be fabricated on the second, opposite side of the interposer substrate.

Alternatively, both groups of resilient contact structures may begin on the same side of the interposer substrate and one group extends through the opening of the interposer substrate to protrude from the opposite surface of the interposer substrate.

Alternatively, the resilient contact structure may be formed through a hole extending through the support member, which extends beyond the opposing face of the support member to contact the electronic component arranged with respect to the opposing face of the support member.

According to one aspect of the present invention, certain improvements in wire bonding are presented, which provide ultrasonic energy while using wire from the capillary of the wire coupler, and cut the wire by EFO (Electronic Flame Off) technology. While there is something like providing ultraviolet light.

In accordance with one aspect of the invention, the stiffness (stability) of a "conventional" wire bond (such as a bond wire extending between the bond pad and lead frame fingers on the semiconductor die) is improved by overcoating the bond wire with the superstructure. Can be. This is important, for example, to prevent shortening of adjacent wires during transfer molding of the package body.

In spite of many other embodiments described herein, the wire is generally coupled to one end of the terminal, the pad or the like is springable and cut to be a free-standing wire stem. The wire itself, which may be a conventional gold bonding wire, does not function as a spring. It is readily formed (for example in S-shape) as opposed to its ability to function as a spring in general.

Freestanding wire stems are protective coated (eg, plated) with a conductive metallic material that serves two principal purposes.

(a) Since this is an elastic material such as nickel, the final protective coated wire stem (contact structure) can act as a spring.

(b) It covers the terminals (or pads, etc.) to securely fasten the wire stem (and elastic contact structure) to the terminals.

In most of the embodiments presented herein, it is the protective sheathing material to realize the desired electrical connection. However, in some embodiments, the wire stem is exposed to the tip (central end) of the contact structure and realizes (increases) the electrical connection.

Various applications for contact structures made in accordance with the techniques of the present invention are described, including mounting the contact structures to semiconductor dies, packages, PCBs, and the like. The use of elastic contact structures as probes is also described. Coupling the wire stem to the sacrificial element (or member, or structure) is also described. Multilayer coatings on wire stems are also described.

Various additional features and embodiments of the present invention are shown below, and it should be understood that certain features described for certain embodiments are not necessarily limited to those embodiments.

Advantages of the present invention include, but are not limited to, the following features.

(a) Elastic and / or flexible contact structures may be mounted directly to the semiconductor die prior to separating the die from the semiconductor wafer or after separating the die from the semiconductor wafer.

(b) Elastic contact structures can be used to temporarily connect to an electronic component for processes such as pretesting and testing of the electronic component, and the same elastic contact structure can be used to permanently connect to the electronic component.

Other objects, features and advantages of the present invention will become apparent in light of the following description.

The first case described above (CASE-1), which is used herein as a reference, discloses a method of manufacturing a raised (protruding) elastic electrical contact having a rigid form on a variety of electrical elements, the contact having various aspects. It has physical, metallurgical and mechanical properties, sizes and surfaces that are adjusted to meet the electrical, thermal and geometrical conditions required for the electronic component connection applications of the device.

Generally, (a) a wire stem is formed on a terminal of an electronic element, (b) the wire stem is shaped in three dimensions to define a skeleton of the final contact structure, (c) the wire A multistage method is described wherein the stem is coated in at least one lamination step with a coating that completely or partially surrounds the backbone and the terminals. The common coating not only helps to “anchor” the protruding contact structure to the terminal, but also (1) characterizes the protruding contact for stability over long periods of mating contact with the connecting electronic element. And (2) help to determine the solder assembly properties as well as the long term effects from the contacts by soldering, and (3) the mechanical properties of the final projecting contact structure.

In general, the wire serves as the "skeleton" of the final protruding contact (i.e., the "proto" elastic contact) and the coating functions as the "muscle". In other words, the (shaped) wire defines the shape of the final contact, and the coating over the wire provides the contact with some elastic (and conductive) properties. In this respect, the coating is regarded as a major component of the contact, and for this purpose some combination must be made for thickness, yield strength and modulus of elasticity to ensure the desired force-strain characteristics of the final spring contact. . An important property of the coating is that it is continuous. It is also disclosed to use multiple (eg, two) wire stems to form a framework for an elastic contact.

The first case includes (a) ceramic and plastic semiconductor packages, (b) thin printed circuit boards (PCBs), Teflon (tm) base circuit boards, multilayer ceramic substrates, silicon base substrates, various composite substrates, and those known to those skilled in the art. Elastic to various electronic elements including other substrates for integration of electronic systems, (c) semiconductor devices such as silicon and gallium arsenide devices, and (d) passive devices such as resistors and capacitors. A method of mounting a contact structure is disclosed.

In the first case, the following wire materials and dimensions are disclosed.

(a) gold, aluminum or copper alloyed (or doped) with beryllium, cadmium, silicon or magnesium, (b) soldering, in particular lead-tin soldering wires, (c) alloys of silver and platinum, and (d) 0.00127 to 0.0127 cm (0.0005 to 0.0050 inch) in diameter.

The first case describes the following wire coupler and technology.

(a) a ball-wedge conjugate, (b) a wedge-wedge conjugate, and (c) applying pressure, temperature or ultrasonic energy, or a combination thereof to form the conjugate, and (d) Capillary movement control technology to set the shape, (e) programming the control system software of the wire coupler to rule out a common coupling step for cutting the wire, and instead, ball formation prior to ball engagement. Is used to cut the wire to a predetermined height, and (f) the electron flame off (EFO) electrode has a high potential difference after the capillary is moved upward to a predetermined position. Generated to result in the generation of sparks to melt cut the wire stem and to facilitate the feeding purpose (from the supply spool) of the wire for subsequent stem joining steps.

According to the invention there is provided (a) an electron flame off (EFO), (b) a mechanical means such as a knife, (c) a laser and (d) an oxygen torch for cutting wire stems. Any suitable means can be used.

The first case discloses a method of laminating tin on a gold wire stem as a first layer using subsequent reaction of gold and tin at a temperature below the temperature of the gold-tin eutectic. Gold-tin alloys are generated, which are principally stronger than (pure) gold.

The first case describes the following coating materials and dimensions.

(a) nickel, copper, cobalt, iron, and alloys thereof; (b) corrosion or semi-corrosive metals or alloys selected from the group consisting of gold, silver, rhodium, copper, and the like; (c) a thickness between rhodium, ruthenium or other elements of the platinum group, and their alloys having gold, silver and copper, (d) tungsten, and (e) 0.0000762 to 0.01778 cm (0.00003 to 0.007 inch).

According to the invention, palladium is a suitable coating material.

In the first case, the following coating process is started.

(a) wet electromechanical plating of metals on the skeleton and terminals, such as through electrolytic or non-charged liquid solutions, and (b) nickel (with a tensile strength exceeding 5,624 kg / cm 2 (80,000 psi)) and its Conductive coatings by electroplating nickel from the alloy, (c) wave soldering or electrolytic lamination soldering, (d) physical and chemical vapor deposition, and various lamination processes of gas, liquid or solid precursors. Lamination method.

The first case discloses a method of applying a multilayer coating over a wire stem. For example, (a) overcoating nickel (with strong oxidizing properties) with a non-corrosive or semicorrosive coating such as gold, silver, platinum elements and alloys thereof, and (b) protruding contacts in certain applications. It is a process of selecting multiple layers of the coating to follow the negative properties.

The first case discloses that the wire stem (skeleton) can be overcoated with an elastic coating having local protrusions. Such coatings may be produced through dendritic growth of the electroplated laminate or through immersion of foreign particulates into the conductive laminate. Optionally, a regular uniform first layer of lamination to provide elastic properties can be applied, and the layered top layer then uses local protrusions or particulates to complete the conductive overcoating. The local protrusions significantly increase the local pressure acting on the connecting terminal by the elastic protruding contact during interconnection coupling, and increase the contact resistance when contacting a material on which the surface is easily stabilized and oxide formed. Decrease.

The first case discloses that an improvement in the mechanical strength of the final contact is achieved when a coating having an internal compressive stress is laminated which effectively increases the level of stress required to deform or break the final projecting electrical contact. Internal compressive stress in the coating not only improves the elastic properties of the final contact, but can also control its strength and ductility levels.

In a first case, (a) the top point (end) of the contact can be made to extend a vertical coordinate (over the surface of the electronic element) that is about the same (for all contacts on the electronic element), and (b) The contact may start from two or more electronic elements and the distal end may be planar with respect to a subsequent planar other electronic element, and (c) the contact may start from various levels of the electronic element and the distal end may be made flat. And (d) the vertical coordinates of the leading end (end) of the contact can be easily controlled by a software algorithm used in the wire coupler system.

In the first case a wire stem can extend between the terminal and the removable substrate, which can be temporarily used in processes such as electroplating of the wire stem.

In the first case it is disclosed that a soldering column having a controlled (eg, large) aspect ratio and shape can be formed. In this respect, a barrier layer of 100 to 1000 microinches thick nickel alloy between the wire stem (especially gold) and the solder overcoating (especially a lead-tin alloy such as lead-tin, which is nearly eutectic) It discloses that a method of applying is described. These solder columns protrude but are not elastic. The use of multiple (eg, two) wire stems per soldering column is as effective as the opposite wetting method (with one wire stem) in connecting the solder.

The first case discloses that a "loop" can be formed by joining both ends of the wire stem to a common terminal and the loop can be used as the skeleton of the solder column.

In the first case a controlled control capable of elastic coupling with a connection terminal on an element or substrate for testing a pre-tested or releasable electrical connection in use by suspending the contact (such as with a removable arrangement). It is possible to form an elastic contact.

In the first case, the contact may be used as an "alternative technique" for standard techniques for attaching the pins to plastic and ceramic packages. Its usefulness is due to the fact that the pin shaped contacts produced by the method according to the invention do not require tools or molds of a particular shape.

The first case discloses that the skeleton can be formed as a loop-like process, such as a fence, which is filled in the defined area by soldering by defining a terminal. Although not elastic, it provides a large solder pad on the electronic element for thermal connection to the heat absorber or substrate.

The invention described in the first case differs significantly from the gist of the prior art for overcoated bond wires in the following points. That is, (a) typically overcoating according to the prior art is for the purpose of non-conductive or corrosion protection, and (b) the elastic contacts according to the invention form a skeleton (which is essentially not elastic and in fact non-conductive). Effectively combining the steps and overcoating the skeleton with an elastic coating (the outer layer must be conductive) to obtain an elastic contact structure.

The effect of the elastic contact structure disclosed in the first case is that (a) the same contact structure can be used for removable or permanent attachment of an electronic element, and (b) the substrate and element in which the elastic contact has a connection form of a terminal. It can be used as a standard means of connection between and on an interposer structure for interconnecting two (or more) electronic elements.

The aforementioned second case (CASE-2), which is used herein as a reference, discloses contact structures for interconnects, interposers and semiconductor assemblies and methods for manufacturing them. In general, the disclosure of the present case is advantageous over the first case as far as the manufacturing performance of the elastic contact structure is concerned.

The second case discloses a number of specific uses of the elastic contact structure, namely various embodiments relating to the interposer, wherein (a) the elastic contact structure is mounted on one side of the interposer substrate in a plated through hole, Mounted on the opposite side of the interposer substrate in the plated through hole, (b) the elastic contact structure is mounted on one side of the interposer substrate in the plated through hole, and the looped contact structure is plated through hole Is mounted on the other opposite side of the interposer substrate, and (c) the elastic contact structure has a portion extending from the plated through-hole that raises one side of the interposer substrate and also downwards the interposer substrate. And (d) a molded plastic interposer substrate having a set of holes extending from one side into the substrate. Another set of holes extending from the opposite side into the substrate. Each pair of holes is somewhat offset (eccentric) from the others. The hole is plated with a conductive metal. Due to the offset above, each hole has a shoulder (at the bottom thereof) on which a supplemental contact structure can be formed. In this way, the first set of contact structures extends out one side of the interposer substrate, the other set of contact structures extends out the other side of the interposer substrate, and the contact structures are interconnected by respective through hole plating methods. Electrically connected.

The second case discloses a method of forming an overcoated contact structure by soldering and comprising a plurality of (three) framework structure spaces on a single contact pad to form a solder post.

The second case discloses a method of forming a composite contact structure in which a plurality of (2 or 3) elastic contact structures are mounted on a single pad, and also discloses a method of fixing the ends of the plurality of contact structures with soldering. Doing.

The second case discloses a method of forming a contact structure that can act as a probe, and also discloses a probe contact having a dielectric metal layer disposed on the metal layer to provide a controlled impedance to the shield contact.

The second case discloses a method of forming a contact pad at a free end of a probe contact structure having a plurality of sharp points and the free end of the probe contact structure having a probe-type tip.

The second case discloses a method of arranging an elastic contact structure on a semiconductor device (electronic element), wherein the tips of the contact structures are aligned in two rows, and the top tip of each row of alternate contact structures fan outs. Offset from the top tip of the other contact structure to provide a staggered arrangement.

The second case discloses a contact structure having various shapes of large bent portions or small bent portions on an electronic element. The second case also discloses a method of staggering the respective "shaped" contact structures in opposite directions. The use of such a shape for carrying out preliminary testing and testing of a semiconductor device is disclosed by deformably coupling a connection contact pad on a pretest substrate. The use of registration or alignment pins is also disclosed. The possibility of having a fine pitch on a contact pad of a semiconductor device on which a contact structure is mounted and a coarse pitch for the tip of the contact structure is disclosed.

In a second case, a plurality of semiconductor package assemblies are disclosed in which the elastic contact structure is provided on one or two semiconductor devices provided on both sides of a printed circuit board in the case of two semiconductor devices. The use of an elastic contact structure as a spring clip for securing semiconductor device (s) on a printed circuit board is disclosed. A method of forming an elastic contact structure that is properly (removably) inserted into a plated through hole is disclosed. The use of alignment pins for semiconductor package assemblies is also disclosed.

The second case discloses the use of additional elements, such as capacitors, in the large plated through holes of the printed circuit board in the semiconductor package assembly.

The second case discloses mounting additional elements, such as capacitors, on a printed circuit board in a semiconductor package assembly.

The second case discloses a number of further assemblies using elastic contact structures.

In the following description, for the sake of complete description, some overlapping portions are (must be) in the disclosure of the first and second cases. In a continuous portion, bonding the wire to the substrate, shaping the wire into a form of “elastic deformable” (suitable to act as a spring when overcoated with an elastic material), the wire to be a wire stem There will be a description of the step of cutting and overcoating the wire to form an elastic contact structure that is well fixed to the substrate.

How to Bond Wires to a Region on a Substrate

Figure 1, similar to Figure 1 in the first case, shows a wire 102 fed through a capillary tube 104 (shown in cross section) of a wire coupler (not shown in this figure and fully shown in Figure 2) and have. Wire 102 is fed from the spool 106 to the capillary. The capillary 104 is directed towards the surface 108a of the substrate 108, such that the free end 102a of the wire 102 comes into contact with the surface 108a of the substrate 108 and is bonded in any suitable way. do. Methods of joining the free ends of the wires to the surface of the substrate are known and do not require further explanation.

As shown in FIG. 1, the free end 102a of the wire 102 is coupled to the surface 108a of the substrate 108 in an arbitrarily formed “contact area 110” (shown in dashed lines). The contact region 110 may have any shape (which may be rectangular, circular or any other shape), with the combined free end 102a of the wire 102 being the surface 108a of the substrate 108. ) Is apparently larger than and enclosed in the (relatively small) position.

As will be described in detail below, in many subsequent embodiments the joined end 102a of the wire 102 becomes the “middle end” of the final “wire stem”.

1A shows that the free end 102a of the wire 102 can be coupled to the conductive terminal 112 on the surface 108a of the substrate 108. Methods of forming conductive terminals (or "pads" or "bonding pads") on the surface of a substrate and methods of coupling wires thereto are known. In this case, the terminal 112 forms a contact region 110 (as shown in FIG. 1). In Figure 1A, capillary 104 is shown and shaped in dashed lines.

FIG. 1B shows a substrate 108 (typically non-conductive or semi-conductive) on an opening 116 where the free end 102a of the wire 102 is etched into an overlapping layer of photoresist 118. It may be coupled to a metal (conductive) layer 114. In this case, the opening 116 in the photoresist forms a contact region 110 (as shown in FIG. 1). In FIG. 1B, the capillary 104 is shaped with dashed lines.

1C, 1D and 1E illustrate a technique for coupling the free end 102a of wire 102 to the surface 108a of the substrate 108, as shown in FIG. 102 is a preferred method of bonding to a semiconductor substrate. In Figure 1C, the capillary 104 is shown and shaped in dashed lines. In FIG. 1C, it can be seen that the conductive layer 120 is disposed on the top surface of the substrate 108 (as shown in FIG. 1B). Layer 120 may also be a top metal layer, which is typically intended to be bonded out of the mold as formed by openings in passivation layer 124 (typically nitrogenous). In this way, a bond pad can be formed that can be an area corresponding to the area of the opening 122 in the passivation layer 124. Typically, (according to the prior art) a wire may be coupled to the bond pad.

According to the present invention, a blank layer 126 of metal material (eg, aluminum) is stacked over the passivation layer 124 so that the conductive layer 126 follows the phase of the layer 124 and “dipped into the opening 122. dipping ”and electrically contacting said layer 120. The photoresist patterned layer 128 is applied over the layer 126 with the opening 132 aligned over the opening 122 in the passivation layer 124. The main feature of the present technology is that the opening 132 is larger than the opening 122. As is apparent, the bonding area (formed by the opening 132) is larger than that otherwise formed by the opening 122 on the semiconductor mold 108. The free end 102a of the wire 102 is coupled to the top surface (in the direction shown) of the conductive layer 126 in the opening 132. After the wire is shaped to have a shape (as shown in FIGS. 2 and 2A-2H) and cut to form a “wire stem” (as shown in FIGS. 4A-4D), the wire stem is removed. Overcoated (as shown in FIGS. 5 and 5A-5F). [For this purpose, the overcoated wire stem is named “elastic contact structure” 130.) Typically, in a typical manner, the wire stem is shown in bold solid lines as shown in FIGS. 1D and 1E, for example. Material 134 overcoating the wire 102 in its preformed form completely surrounds the wire stem and covers the conductive layer 126 in the area formed by the opening 132 in the photoresist 128. do. The photoresist 128 is removed (by chemical etching or cleaning) and the substrate is a conductive layer (except for a portion of the layer 126 covered by the material 134 overcoating the wire stem). 126 is subjected to selective etching (eg, chemical etching) to remove all of the above materials. Thus, the structure shown in FIG. 1E can be formed, the main advantage of which is that the contact structure 130 becomes the contact area of the bonding pad (ie, the opening 122 in the passivation layer 124) (in the prior art). Is fixedly bonded (by the coating material 134) to a region (formed by the opening 132 in the photoresist) that can be more easily manufactured. In this case, the area formed by the opening 132 in the photoresist 128 is the contact area 110.

1F is a conductive (e.g., hatching of solid line) where the free end 102a of wire 102 may be a removable substrate (e.g., capable of dissolving after joining the wires) as described in more detail below. Metal as shown) to the substrate 108. In this case, the contact region 110 is not shown, but may be formed arbitrarily as shown in FIG. In Figure 1F, the capillary 104 is shown and shaped in dashed lines.

In all cases described above (not shown in FIGS. 1, 1A, 1B, 1C-1E, and 1F) that are typical and not limited, the free end 102a of the wire 102 is the formation region 110 on the substrate. Is coupled within. However, as can be readily seen from the figure, the coupling itself (of the middle end of the wire) occupies a relatively small area within the forming region 110. By way of example, the small area of the engagement itself may be only 5% to 50% of the total area of the contact area 110, but preferably may be a larger area centered in the contact area.

As described below, the wire stem can be overcoated with a material that can provide elasticity and secure itself to the entire contact area.

Properties of wire

The wire is an elongate element having a material and size that are easily assembled into a (eg, “flexible”) form as described below. As is apparent, it is not particularly important in the present invention that the wire can be electrically conductive between two electronic elements as it is overcoated entirely with a conductive material (in most embodiments described below). However, it is within the scope of the present invention that the wire is made of a conductive material.

In general, according to the present invention, the current properties of the wire (eg, the ability to be shaped and overcoated) tend to be superior to its structural and electrical properties. Moreover, since the wire stem is overcoated with an elastic conductive material, the wire stem becomes very large.

A typical material for the wire is gold, which is circular (cross section) with a diameter (thickness) of approximately 0.001 inch (1 mil). The diameter includes, but is not limited to, in the range of 0.7 to 2.0 mils. The wire is preferably in the range of 0.0005 to 0.0030 inch (0.5 to 3.0 mil). Such a wire is very easy to shape (in a predetermined shape), has excellent conductivity, is easy to use, and exhibits corrosion resistance for a long time.

Wires made of gold are available from a variety of sources and come in a variety of sizes (eg, a diameter of 0.001 inch), for example having the following composition.

99.99% gold and the rest beryllium;

99.99% gold and the rest copper;

1% silicon aluminum alloy;

-1% magnesium aluminum alloy.

Preferably, pure copper wire is preferably available and readily available for use as the wire stem of the present invention.

Other suitable materials for wires (other than gold) in which a similar range of diameters may be used include aluminum; Copper alloyed by containing small amounts of other metals to obtain certain physical properties such as beryllium, cadmium, silicon and magnesium; Gold, copper or aluminum alloyed with materials such as lead, tin, indium, cadmium, antinomy, bismuth, beryllium and magnesium; Silver, palladium, platinum; Metals or alloys such as platinum-based metals; And lead, tin, indium and their alloys.

Typically, a wire made of any material that is flexible to the bond (using temperature, pressure, and / or ultrasonic energy to perform the bond) may be suitable for practicing the present invention.

Preferably, the material of the wire is gold, aluminum, copper or alloys thereof. For example: ① gold or cadmium doped (alloyed) with beryllium (eg 12 ppm, preferably 5-7 ppm), ② aluminum doped with silicon or magnesium, optionally silver or copper, and ③ copper / Platinum / palladium mixed with silver.

As described in more detail below, in many embodiments, the wire is cut to have a distal end and a certain length. The wire has any desired length, but has the same size length to be used with the semiconductor device and package having a small shape such that the length is from 0.010 to 0.500 inch. The wire need not have a circular cross section, but this is better. The wire may have a rectangular cross section and may have a different cross section.

As described in detail below (see FIGS. 51A and 51B), the wire stem may be formed of a nonmetallic material such as plastic and may be overcoated to form an elastic contact structure.

Conventional wire bonds require rigid bonds fabricated at both ends of the wire bond loop to avoid cohesion. Bond strength is very important and process constraints are relatively stringent.

In accordance with the present invention, the constraints of bonding the wire stem on the substrate are very relaxed (eg, as in conventional wire bonding). Generally, the bond strength is “sufficient” as long as the wire stem is in place during shaping and overcoating. In this way, it is possible to ensure the various wire materials described above, which materials are usually selected from properties different from the stiffness of the joints formed therewith.

Forming and Geometry of Wire Stem

Once the free (proximal) end 102a of the wire 102 is coupled to the substrate 108, the capillary 104 is moved generally upward (z-axis) from the surface of the substrate and is typically (not shown). The base mounted on the xy table is moved in the x and y directions. This applies a relative movement between the capillary and the substrate described below as a capillary moved in three axes (x-axis, y-axis, z-axis). As the capillary moves, the wire “plays out” at the end of the capillary.

According to the present invention the relative movement between the capillary and the substrate is controlled and the desired shape is applied to the wire.

In a typical wire bonding operation, the free end of the wire is bonded (eg to a bond pad on a semiconductor die), the capillary moves upward (to a predetermined height above the surface of the substrate), the substrate moves upwards (typically It is the substrate that is moved to apply relative movement in the xy plane between the substrate and the capillary.), The capillary moves downward (e.g., to a bonding site on a conductive trace such as a leadframe, semiconductor package, etc.). The wire is played out of the capillary during this relative movement of the capillary. This (for example, up / over / down) movement of the capillary (prior art) is generally arc-shaped, rather than wire (even straight wire can be thought to have “shape”). Shape ”, but the“ shaping ”of the present invention is quite different.

Apparently the wire stem is shaped such that the contact structure including the wire stem overcoated by plating or the like has a function as an elastic contact structure. With this in mind, the concept of "shaping" of the wires according to the invention is completely different from the "shaping" of the prior art (ie not intended to establish the resulting elastic contact structure). The concept of "shaping" of a wire as used herein differs dramatically from prior art shaping in the manner in which the wire is shaped in both the inherent purpose of shaping the wire and the geometry of the resulting wire shape.

As will be explained in greater detail below, once the proximal end of the wire is coupled to the substrate, it is carefully “formed”, “fashioned” or corresponds to the shape of the wire. And a shape having a defined and preferred geometry that serves as the "skeleton" as referenced in the parent application to establish the physical form of the subsequent coating on the wire that applies elasticity to the subsequent contact structure of the coated wire. Other advantages of the coating besides applying elasticity to the shaped wire stem are described below). Therefore, the terms “configuring”, “fashioning”, “forming”, “shaping”, etc., as used herein, are considered to be elastic (spring properties) in the contact structure (coated wire) made. It is particularly intended to be interpreted as having a meaning describing the wire's ability to establish the resulting form of coating to which it is applied.

Reading this patent specification, a person of ordinary skill in the art to which the present invention is closest to knows that once configured and overcoated, the wire stem itself is essentially unnecessary—electrical conductivity requiring overcoating material, and made elastic contacts. It will be appreciated that it provides the desired mechanical properties of the structure. Obviously, however, in certain embodiments of the present invention it is necessary that the wire stem is electrically conductive, since an electrical contact is made to it.

In some sense, the wire stem of the present invention has a function similar to the temporary pole or support used in the structure to establish the resulting shape of the "falls walk"-stone or brick arch. As an approximation of this, overcoating can be thought of as functioning as a "superposition".

In another sense, the wire stem has a function similar to a “mandrel” —a core from which other materials can be formed.

In another sense, wire stems have the same function as “templates” —patterns or models for parts that are made or synthesized.

It is understood that this approximation is not perfect in that the poleswalk, mandrel and template are typically removed after their original purpose. In the case of the wire stem of the present invention, embodiments are disclosed in which the wire stem is removed, but it is not necessary to remove the overcoated wire stem.

Perhaps a more suitable approximation is that the wire stem has the function of "outline" in a similar way that outlines can be made before writing a book. The outline explains what the book will be and can be included in the book or discarded after it is written. In either case, the outline establishes the form in which the book is made.

Shaped wire stems in the parent application were described as “skeleton” —supporting structures or frameworks. This is also a selection of useful terms in that the skeleton is typically left in place. Like the human skeleton (skeleton) that determines the shape of the tissue thereon, the wire stem of the present invention establishes the shape of the contact structure to be made. However, unlike the human skeleton, which must remain in place to perform its intended function of the tissue thereon and contributes significantly to the "mechanical" properties of the human body, the wire stem of the present invention is a material on which The coating need not remain in place to perform its intended function, nor does it significantly contribute to the mechanical properties of the contact structure to be made.

One of the significant advantages of using easily deformable, malleable, and compliant materials for wire stems is that they do not significantly alter the physical properties (eg tensile strength, elasticity, etc.) of the resulting elastic contact structures. Rather, it can be easily shaped to establish the shape of the overcoating applied thereto. As long as the wire stem has the function of an important first step in the overall (initiated, but not completed) process of manufacture of the final contact structure, the wire stem can be characterized as an “unfinished” contact structure.

Since the present invention is primarily intended to fabricate elastic contact structures for interconnecting electronic components such as semiconductor dies, the elastic contact structures are almost invisible to the naked eye in many cases, and the shape and contribution of the wire stems (copper wire of gauge 14 or Take a predetermined length of the malleable wire (such as 15.2 cm (6 inch)) of the corresponding gauge lead-tin solder wire, mount one end of the wire into the hole of the wooden block, 2A, FIG. 2B, FIG. 2E) It may be easier to visualize by manually forming a free upright wire stem having any of the forms disclosed herein. If a compressive force (for example, a few ounces applied by the user's fingertip) is applied toward the block at the tip of the wire stem, it can be seen that this results in the wire stem deforming. This is because the wire stem itself has very few "spring characteristics". The overcoated wire stems formed by the present invention are more similar to steel wires (such as, for example, 15.24 cm (6 inch) long coat hanger wires) formed in the same way as solder, in which case the steel wires have a compressive force therewith. It will exhibit notable spring characteristics (or spring back) when applied. (It is understood that steel wires are not “overcoating” on deformable wire stems.) Overcoating (eg nickel) on easily shaped wire stems (eg gold) exhibits similar elasticity to coat hanger wires. .

In other words, prior art wire bonding, including joining the ends of the wire at one location, moving up (up, up, and down) to another location, and subsequent joining and cutting, is a wa It can be thought of as applying a “shape” to a word, and the shape that is (usually arc-shaped) is relatively accidental. In contrast, according to the present invention, the wire is basically "fashioned" or "shaped" along its entire length to intentionally (not accidentally) have a specific functional (proto spring) shape. Another useful term describing the application of a shape that must be elastic (when overcoated) to a wire is convoluting (or shaping) the wire into a “convoluted” shape (or a shape). It is.

Although the prior art wire bonding loops are overcoated (for example with nickel), they generally do not function as elastic contact structures, inserting both ends of the coat hanger wires described above into the two holes of the wood block, the same compressive force By applying it to the top of the curve (not the tip). Even in this “crude” model, loss of spring characteristics can be easily observed.

In many of the embodiments described below, the wire is made (shaped) with at least two bends, which also makes the “making” (shaping, shaping, shaping) of the shape of the present invention an accidental and typically one of the prior art. Differentiate from bending shape. In another sense, the invention is intended to "developing" a shape that has the function as a spring in the wire (once overcoated).

In many embodiments described below, the wire starts in a particular direction (typically away from the surface of the substrate) and then bent in one direction, then bent in the other direction, and then terminated in the same specific direction in which it started (ie far from the surface of the substrate). It is shaped to have a shape to be.

2 illustrates a surface 208a of the substrate 208 (compare 108) in accordance with one of the techniques described with respect to FIGS. 1, 1A, 1B, 1C-1E, and 1F (or other techniques described below). Wire 202 (compare 102) with its free ends 202a (compare 102a) coupled within the determined contact area 210 (compare 110) on (as compared to 108a). The initial position of the capillary 204 (compare 104) is shown in dashed lines. The final position of the capillary 204 is shown in solid lines. The surface 208a of the substrate 208 is located in the x-y plane (although the entire surface of the substrate need not be planar). The final position of the capillary 204 is shown in FIG. 2 in solid line, displaced from the surface of the substrate in the positive z direction. The wire 202 is fed from the supply spool 206 (compared to 106) through the capillary 204 and shaped (to have a shape) in the following manner.

The free (near) end 202a of the wire 202 is joined at the point marked “a” to the surface 208a of the substrate 208 and the capillary 204 is at its initial (dashed) position. The capillary is then moved along the trajectory of the points “collectively” labeled “b”, “c” and “d” in FIG. 2 to shape the wire into two or three dimensions.

In the following description, for the sake of clarity, the movement of the capillary is described by instructing relative movement between the substrate and the capillary. Sometimes the movement in the x and y directions is made by moving the substrate (eg the x-y table on which the substrate is mounted), and the movement in the z direction is made by moving the capillary. In general, the capillary is typically directed in the z axis. However, it is within the scope of the present invention that capillaries having multiple degrees of freedom can be used to shape the shape of the wire stem.

In general, movement of the capillary 204 is performed by a positioning mechanism POSN 220 under the control of the microprocessor basic controller CONTROL 222, and the capillary 204 by any suitable link 224. Is linked to. As described in detail below, this allows point-to-point control of the position of the capillary to describe its trajectory.

As will be described in further detail below, vibration (eg ultrasonic) energy during passage of the trajectory is activated (eg on / off) under the control of the controller 222 and linked or directly mounted to the capillary 204. Can be supplied to the capillary 204 by a suitable transducer (USTR, 230).

As will be explained in further detail below, the wire 202 is cut upon reaching its final (solid line) position of the capillary. This is illustrated in FIG. 2 by an electrode 232 positioned adjacent to the wire 202 (typically in a fixed position relative to the capillary 204), the electrode 232 operating under the control of the controller 222. Electrical energy is received from the off (EFO) circuit 234.

As will be described in detail below, the cutting operation of the wire (e.g., by electron flame-off) is carried out from the appropriate light source 238 to the cutting position (position " d " By providing the directed ultraviolet light. As described in more detail below, the light source 238 can be directed at the cutting position d on the wire 202 or at the electrode 232 or between the electrode and the wire.

In the figures that follow, numerous exemplary forms of wire stems are described.

2A shows wire 204 having a free (near) end coupled to point “a” on surface 208a of substrate 208. The capillary 204 shown in the final position (compare with the final position in Fig. 2) is from “a” to “b”, from “b” to “c”, from “c” to “d”, “d Move along the trajectory of points from ”to“ e ”and from“ e ”to“ f ”. From point “a” to point “b” the capillary 204 moves in the vertical (z-axis) direction. From point “b” to point “c” the capillary 204 moves parallel to the surface 208a of the substrate 208 in the y-axis direction. Capillary tube 204 from point “c” to point “d” is in the vertical (z-axis) direction to point “d” which is displaced more vertically (higher) from surface 208a of substrate 208 than point “b”. Go to. Capillary tube 204 from point “d” to point “e” is the substrate 208 in the y direction opposite to the direction in which capillary 204 moved between points “b” and “c” (ie, in the negative y-axis direction). Move parallel to the surface 208a. This returns the capillary 204 to a point (“e”) that may or may not be directly above its starting point (“a”). From point “e” to point “f” the capillary 204 again moves in the vertical (z-axis) direction. Point “f” is displaced more vertically (higher) from surface 208a of substrate 208 than point “d”. As will be described in detail below, the wire is cut at point “f” to create a free upright wire stem whose proximal end is joined to the substrate and its distal end is displaced from the surface with the substrate.

This movement of the capillary 204 along the trajectories a → b → c → d → e → f applies a U-shape to the wire 202. This, of course, assumes that the wire 202 is sufficiently malleable (e.g. formable) to obtain a shape (e.g. U-shape) that substantially corresponds to the trajectory of the capillary. This U-shape is considered to illustrate the two-dimensional shape of the wire.

As described in more detail below, the preferred material of the wire represents a small but recognizable amount of springback that must be considered in the trajectory of the capillary 204 to apply the desired shape to the wire. Once the capillary 204 reaches its final position (“f”), as will be described in more detail below, the wire is cut at (or adjacent to) the capillary 204.

2B shows wire 202 with free end 202a coupled to point “a” on surface 208a of substrate 208. The capillary 204 shown in the final position (compare with the final position in Fig. 2) is from “a” to “b”, from “b” to “c”, from “c” to “d” and “d ”To“ e ”,“ e ”to“ f ”,“ f ”to“ g ”, and“ g ”to“ h ”along the trajectory of the points. From point “a” to point “b” the capillary 204 moves in the vertical (z-axis) direction. From point “b” to point “c” the capillary 204 moves parallel to the surface 208a of the substrate 208 in the y-axis direction. Capillary tube 204 from point “c” to point “d” is in the vertical (z-axis) direction to point “d” which is displaced more vertically (higher) from surface 208a of substrate 208 than point “b”. Go to. Capillary tube 204 from point “d” to point “e” is the substrate 208 in the y direction opposite to the direction in which capillary 204 moved between points “b” and “c” (ie, in the negative y-axis direction). Move parallel to the surface 208a. This returns the capillary 204 past the point “e”, which may (or may not) be above its starting point (“a”). From point “e” to point “f” the capillary 204 again moves in the vertical (z-axis) direction. Point “f” is displaced more vertically (higher) from surface 208a of substrate 208 than point “d”. Capillary tube 204 from point “f” to point “g” is the substrate 208 in the y direction opposite to the direction in which capillary 204 has moved between points “d” and “e” (ie in the positive y-axis direction). Move parallel to the surface 208a. From point “g” to point “h” the capillary 204 again moves in the vertical (z-axis) direction. Point “h” is displaced vertically more (higher) from surface 208a of substrate 208 than point “f”. As will be explained in detail below, the wire is cut at point “h” to create a free upright wire stem whose proximal end is joined to the substrate and its distal end is displaced from the surface with the substrate.

This movement of the capillary 204 along the trajectories a → b → c → d → e → f → g → h applies an S-shape to the wire 202. Here too, of course, it is assumed that the wire 202 is sufficiently malleable (e.g. formable) to obtain a shape (e.g. S-shape) that substantially corresponds to the trajectory of the capillary, as will be explained in more detail below. The good material of the wire represents a small but recognizable amount of springback that must be considered in the trajectory of the capillary 204 to apply the desired shape to the wire. Once the capillary 204 reaches its final position (“h”), as will be described in more detail below, the wire is cut at (or adjacent to) the capillary 204.

This S-shape of the wire 202 is considered to illustrate the two-dimensional shape of the outer ear if there is no x-axis displacement during the movement of the capillary along the trajectory. Obviously the capillary can move in the x-axis relative to one of the above-described movements from point to point (for example, moving from point “d” to point “e”), in which case the shape of the wire is more than a simple S-shape. It will become complicated and be considered as a three dimensional shape of the wire.

2C shows wire 204 having a free (near) end coupled to point “a” on surface 208a of substrate 208. The capillary 204 shown in the final position (compare with the final position in Fig. 2) is from “a” to “b”, from “b” to “c”, from “c” to “d” and “d Move along the trajectory of points from ”to“ e ”and from“ e ”to“ f ”. From point “a” to point “b” the capillary 204 moves in the vertical (z-axis) direction. From point “b” to point “c” the capillary 204 moves parallel to the surface 208a of the substrate 208 in the y-axis direction. From point “c” to point “d”, capillary 204 moves in the x-axis direction in a direction towards point “d” of y and z-axis coordinates that is substantially the same as point “c”. Capillary tube 204 from point “d” to point “e” is the substrate 208 in the y direction opposite to the direction in which capillary 204 moved between points “b” and “c” (ie, in the negative y-axis direction). Move parallel to the surface 208a. This returns capillary 204 to y-axis coordinates approximately equal to point “b” but to (offset) x-axis coordinates different from point “b”. From point “e” to point “f” the capillary 204 again moves in the vertical (z-axis) direction. Point “f” is displaced more vertically (higher) from the surface 208a of the substrate 208 than point “d”, and is approximately located at the same vertical (z-axis) coordinate as point “b”, and point “a” Or "b" is offset from the x-axis. As will be explained in more detail below, the wire is cut at point “f” to create a free upright wire stem whose proximal end is joined to the substrate and its distal end is displaced from the surface with the substrate.

This movement of the capillary 204 along the trajectories a → b → c → d → e → f applies a three-dimensional shape to the wire 202. It is also assumed here that the wire 202 is sufficiently malleable (e.g. formable) to obtain a shape that substantially corresponds to the trajectory of the capillary, and as described in more detail below, a good material of the wire Represents a small but recognizable amount of springback that must be considered (compensated) in the trajectory of capillary 204 to apply the desired shape. Once the capillary 204 reaches its final position (“f”), as will be described in more detail below, the wire is cut at (or adjacent to) the capillary 204.

FIG. 2D generally shows that the final position of the capillary 204 need not be above the substrate 208, but may be beyond the edge and below the surface 208a of the substrate. In FIG. 2D wire 202 has its free end 202a coupled to point “a” on surface 208a of substrate 208. The capillary 204 shown in the final position (compare with the final position in Fig. 2) is the trajectory of the points from “a” to “b”, from “b” to “c” and from “c” to “d”. Move along. As described in more detail below, the wire is cut at point “d” to create a free upright wire stem whose proximal end is coupled to the substrate and its distal end is displaced from the surface of the substrate. In this case the point “d” is below the surface of the substrate (not on the surface as in the embodiment of FIGS. 2A, 2B and 2C).

From point “a” to point “b” capillary 204 is vertically upwards (in positive z direction) and horizontally away from point “a” (positive y direction) from surface 208a of substrate 208. ) Move along arc trajectory. Point “b” represents the maximum vertical (z-axis) displacement of the capillary 204 (in the positive z direction) over the surface 208a of the substrate 208. From point “b” to point “c” the capillary 204 is directed vertically downward (in the negative z direction) towards the surface 208a of the substrate 208 and horizontally away from the point “a” (positive y direction). To the arc trajectory. Point "c" is nominally at the same z-axis position as point "a". Note, however, that point “c” is off the edge of the substrate. From point “c” to point “d” the capillary 208 extends vertically downward (in the negative z direction) beyond the surface 208a of the substrate 208 and further horizontally away from the point “a” (positive y Direction along the arc trajectory. The point “d” is therefore at a lower (z-axis) position than the surface 208a of the substrate 208, clearly out of the edge of the substrate.

This movement of the capillary 204 along the trajectories a → b → c → d applies a two-dimensional shape to the wire 202. However, within the scope of the present invention, a three-dimensional shape (eg, between points “a” and “b”) can be applied to the wire 202 by applying some x-axis movement to the capillary during the traversal of the trajectory. have. It is also assumed here that the wire 202 is sufficiently malleable (e.g. formable) to obtain a shape that substantially corresponds to the trajectory of the capillary, and as will be described in more detail below, a good material of the wire Represents a small but recognizable amount of springback that must be considered in the trajectory of capillary 204 to apply the desired shape. Once the capillary 204 reaches its final position (“d”), as described in more detail below, the wire is cut at (or adjacent to) the capillary 204.

As shown in FIG. 2D, the capillary 204 changes its direction, which is almost vertical at position “a,” to almost horizontal at position “b”. This is typically beyond the capabilities currently available in most wire couplers that hold the capillary in the fixed, vertical direction and move the capillary only in the z-axis direction while the substrate is moved in the x and y directions. As described in detail below, maintaining the vertical direction of the capillary through the traversal of the trajectory (for example, the trajectories shown in FIGS. 2A, 2B, and 2C) indicates that the wire “lags behind” as it exits the capillary. Cause This problem is solved by the features of the present invention described below.

In any of the trajectories described herein, it is proposed that the capillary 204 only moves in the z-axis direction and the x and y movements are applied by the x-y table on which the substrate is mounted. This obviously exacerbates the stiction of the wire exiting the capillary during the molding operation as described with respect to FIG. 2D.

2E, 2F, and 2G show that wire 202 is molded, with wire 202 coupled to exemplary terminal 212 at proximal end 202a and distal end 202b of the wire having an increased diameter zone in the form of a ball. It shows the other possible shapes that can be. These drawings are not intended to limit the number of shapes that can be made illustrative and easy.

For example, in FIG. 2E the wire is shaped to have an S shape. As with the embodiment of Figures 2A, 2B, 2C, the wire stems that are made begin in a first direction (i.e. from point "a" to point "b") away from the surface of the substrate and generally in the same direction (i.e., point " c ”to“ d ”).

In FIG. 2F, a simple loop is formed, and both the proximal end 202a and the distal end 202b of the wire 202 are coupled in the bonding region 210. Such loops are described in more detail below and are generally intended to have the function of bump contacts rather than as elastic contact structures.

In FIG. 2G a looped elastic contact structure is shown. The wire is three-dimensionally shaped between points a → b → c → d → e → f → g and both the proximal end (point “a”) and the distal end (point “g”) of the wire 202 are shown in FIG. 2F. In a manner similar to the example). In this case the free upright structure is fashioned and the middle part of the wire stem (between points “d” and “e”) is intended to make electrical connections to other electronic components.

2H shows wire 202 having its free end 202a coupled to point “a” on surface 208a of substrate 208. The capillary 204 shown in its final position (compared with the final position in FIG. 2) moves along the trajectory from point “a” to “b”. From point “a” to point “b” the capillary 204 is moved in a straight line to a position displaced from point “a” (“b”) in the vertical (positive z-axis) direction. Point “b” is shown to have the same y-axis coordinates as point “a” but this is not necessary. Moreover, during its movement from point “a” to point “b” the capillary 204 also moves in the positive or negative x-axis direction. In either case, the wire 202 moving from point “a” to “b” is straight and may be perpendicular (ie 90 degrees) or at an angle other than perpendicular to the surface 208a of the substrate 208.

Movement of the capillary 204 along the trajectory a → b applies a two-dimensional shape to the wire 202. Spring back of the wire is not a fundamental problem in this straight forming of the wire. As will be explained in further detail below, once the capillary tube 204 reaches its final position (“b”), the wire is cut at (or adjacent to) the capillary tube 204. Obviously, forming straight wires is a “minor” case of molding, and although included in the overall technique of fashioning a wire into a predetermined shape by the present invention, it does not limit the other lesser “minor” embodiments described above. Obviously straight wire stems as shown in FIG. 2H do not exhibit springable shapes as is known. The axially directed force along the straight wire stem only causes the straight wire stem to buckle. By elastic overcoating, the resulting elastic contact structure can exhibit some restoring force. A free upright straight wire stem, on the other hand, will act as a “cantilever” in response to a force directed off axis in the straight wire stem (either the wire stem itself is inclined or the force itself is off axis).

Exemplary trajectories of capillary 204 have been shown and described for artificial small numbers of points (a, b, c, d, etc.) for clarity of explanation, and in practice, multiple points along the trajectory are controlled by controller 222. It should be understood that it is programmable.

In all changes in the direction of the capillary 204, such as a change from the z-axis movement from “a” to “b” and from the y-axis movement from “b” to “c” (see any of FIGS. 2A, 2B and 2C). It is to be understood that the bending that is made of the wire has a radius that is not a sharp right angle bend.

The fashion shapes shown in Figures 2A, 2B, 2C, 2E, and 2G all include at least two bends. In general, even L or U-shaped (horizontal) opposite bends need to terminate the wire therefrom in the same amount of Z direction in which it was generated (for example, wire # in FIG. starting from the positive z-axis direction from “a” to “b” and completing its progression in the same positive z-axis direction from point “d” to “e”). And each shape includes at least two bends that include an S-shaped wire stem.

S-shape as shown in FIG. 2E is merely illustrative and is not intended to be limiting and should not be so interpreted. Many shapes of wire stems serve as the basis of an elastic contact structure by the present invention. However, note that, for example , in contrast to Form S shown in US Pat. No. 5,317,479, described above, Form S (for example in FIG. 2E) is sideways. Corresponding leads plated in this patent are in contact with the pads on the electronic component by the surface of their parallel “legs”. In the present invention, the elastic contact structure is preferably in contact with the pad on the electronic component by its distal tip. (In a similar manner, the proximal end is joined to the contact area by a similar tip of the elastic contact structure.)

In general, according to the present invention, the shape of the elastic contact structure is modeled for each application for a given stress, spacing, compliance and electrical performance limits.

As a general suggestion any desired shape for the wire stem can be fashioned using a wire coupler. As described below (see FIGS. 51A and 51B), devices other than wire couplers are suitably used for the non-metallic wire stem to form a stem having a springable shape.

Type of wire coupler

The present invention can be carried out using a conventional wire bonding machine (wire coupler). Preferably, however, certain improvements can be made to these machines to facilitate the formation of the elastic contact structure of the present invention.

Conventional wire bonders that can be used to carry out the invention use any combination of ultrasonic energy, thermal energy and compressive force to form a ball, wedge or other suitable bond between the free end of the wire and the substrate, Including but not limited to wire couplers.

(a) ultrasonic wire coupler

(b) thermosonic wire couplers

(c) thermocompression wire couplers

According to one aspect of the present invention, the function of the conventional wire coupler is preferably reinforced (enhanced) to carry out the techniques of the present invention. This modified wire coupler is hereinafter referred to as the "spring machine". In general, modifications required for wire couplers fall into two classes.

The first class for improving wire couplers is a. Ultrasonic molding hardware and software, b. Software for controlling the shaping of the wire stem (e.g., compensation for springback, etc.).

As detailed below, the second grade of improvement is a. Spark stabilization hardware, b. Haieraki Part Architecture Software, c. Software to control the cutting sequence cycle.

As an auxiliary theorem that it is desirable to provide improved functionality to a conventional wire coupler to shape the wire into a springable shape, the functional enhancement disclosed herein is applicable to situations where it is not necessary to shape the wire to have a springable shape. It is within the scope of the present invention. In other words, certain teachings of the present invention are applicable to the enhancement of "conventional" wirebond applications.

In general, wire couplers are known. In most of the elastic contact structures disclosed herein, the wire coupler is "incident" and any suitable wire coupler may be used.

Ultrasonic molding

FIG. 2 joins the free ends 102a, 202a of the wires 102, 202, and as described above, and shapes the wires into various exemplary (non-limiting) shapes described with respect to FIGS. 2A-2G. A wire coupling system (wire coupler) is shown. The improvement over the wire coupler (making the wire coupler into a spring machine) is the addition of a transducer 226 (FIG. 2) that vibrates the capillary 204.

According to one aspect of the invention, ultrasonic energy is applied through the capillary during the shaping of the wire (moving the capillary along the trajectory to apply a two-dimensional or three-dimensional shape to the wire stem). In general, this facilitates bending of the wire and facilitates the shaping of 3D (three-dimensional) shapes that are otherwise difficult. The use of ultrasonic energy during molding also improves the predictability of loop formation of conventional wire bonds.

This aspect of the invention also relates to a "conventional" wire bonding process (eg bonding pads) involving moving the capillary in the z direction while the table (with the interconnected electronic components mounted thereto) is moved in the x and y directions. Connecting wires to the leadframe fingers). It was observed that during this movement the wires tend to “hung up” or stick to the capillary, which introduces an undesirable degree of uncertainty into the process. According to the present invention, by using ultrasonic energy while moving the capillary (and / or table), the wire overcomes this “sticking” and is fed more freely from the capillary.

In general, as described above, making the wire stem includes engaging the free end of the wire, moving the capillary upward (z direction), and moving the table in the x or y direction.

In general, conventional wire bonding techniques involve coupling a free end of a supply wire to a substrate, moving the capillary upward (z direction), and moving the substrate in the x or y direction (xy table or other suitable positioning mechanism). Moving), moving the capillary downward again, and joining and cutting the wire to obtain arc-shaped wires joined at both ends.

In any wire bonding operation, the lagging that the wire exhibits when playing out from the capillary, in particular in the process of forming the wire stem, according to the invention, should be avoided.

According to the present invention, the application of ultrasonic energy during the process of forming the wire stem generally eliminates the tendency for the wire to lag behind as it plays out from the capillary. Moreover, the application of ultrasonic energy breaks the grain boundary of the wire-machining the wire during fashioning the wire to have a springable shape.

Ultrasonic wire combiners are already equipped with transducers that provide ultrasonic energy to perform the coupling. According to the present invention, the same ultrasonic transducer can be turned on (or kept on) during shaping the wire stem to facilitate shaping of the desired shape. The turn on and turn off of the ultrasonic transducer is easily included in the software.

Obviously, if the capillary is directed to various positions along its trajectory so that the wire is always fed straight from it (see eg FIG. 2D), sticking of the wire to the capillary will not be an important problem. However, this approach requires very large but possible modifications to existing wire couplers.

The use of ultrasonic energy in wire couplers is generally known for example to reduce sticking. Preferably, however, ultrasonic energy is applied to soften the wire for shaping and reduce residual springback during shaping the wire stem. In general, low ultrasonic frequencies are preferred.

Another embodiment of the present invention is to soften the wires during the forming step (e.g., in other ways using ultrasonic energy to make the wires easier to process (and to reduce sticking when fed from the capillary). In an annealing) wire coupler, heat may be applied to the wire (eg through the capillary) when fed from the capillary tube by an electron flame off (EFO) function or the like.

Another way to use the ultrasonic energy to overcome the sticking of the wire when supplied from the capillary is to provide an inert gas or forming gas through the capillary. This may be included in conjunction with the aforementioned ultrasonic energy application, or as a standalone function, as described in detail below (see FIG. 52).

Compensation of spring back of wire

Another improvement in making the wire coupler a spring machine is to program the controller 222 (FIG. 2) such that the trajectory taken by the relative movement of the capillary tube (facing the substrate) will produce the desired shape of the wire.

As mentioned above, the trajectory (relative movement) of the capillary is preferably controlled on a point-to-point basis to apply the shape to the wire and the springback of the wire causes the final shape to deviate from the capillary trajectory. 3A illustrates this “problem”. The dotted line 302 represents the trajectory along the path a → b → c → d → e → f of the capillary (e.g., 204 of FIG. 2A), and the solid line 304 shows the path a '→ b' → c '→ d The shape made of the wire (e.g. 202 of Figure 2A) along '→ e' → f 'is shown. Obviously point “a '” does not vary from point “a” because this is the point where the proximal end of the wire (compared to the substrate) is fixed (combined to the substrate).

A "problem" is one where the wire once cut takes the shape of "before" or "behind" the capillary trajectory at a particular portion along the length of the wire. For example, as shown in Figure 3A, from point a 'to b', the wire lags behind the path a → b of the capillary (back or left, depending on the view), and from point e 'to f' the wire Leading the path e → f.

Figure 3B shows a solution to the problem shown in Figure 3A. The solution is generally to move the capillary along the “virtual” path 312 offset (previously, lagging) from the desired path 314 of the wire (the wire path 314 defines the wire shape that is made). Include.

As shown in Fig. 3B, the capillary is to be moved (crossed) along a trajectory (path, 312) (shown in dashed lines) from point r → s → t → u → v → w, and once cut, The wire stem is located on path 314 (shown in solid line) along point r '→ s' → t '→ u' → v '→ w'. The wire “sets” itself in this path, offset from the trajectory of the capillary, even if this elasticity is very small due to the elasticity of the wire (unique springback). (Generally, the elasticity of the wire itself is insufficient according to the present invention.) In the example of FIG. 3B, the trajectory of the capillary between points r → s indicates the desired passage of the wire between points r 'and s'. Ahead (to the right in the figure), the trajectory of the capillary between points v → w lags behind the desired passage between points v 'and w' (to the left in the figure).

This result (offset the trajectory of the capillary from the desired wire path) preferably inputs the desired wire path 314 (eg on a computer workstation) initially which may be obtained from a design program already running on the computer. By inputting the physical properties of the wire (e.g. thickness, yield strength, etc.), and by calculating the capillary trajectory 312 made from a wire having a shape corresponding to the desired wire path 314 once cut. Preferably it may be made. Factors such as “lag behind” in the capillary (sticking when feeding the wire from the capillary) and the vibrating effect of the capillary (described above) when making (molding) the wire also produce the desired wire shape. It should be considered in determining the trajectory.

In general, the offset is due to two factors. (1) the lagging of the wire as it exits the capillary and (2) the inherent springback of the wire material. Complex algorithms can be developed that identify exactly how the path of a given wire is offset from a given trajectory, but generally use an experimental approach for each desired wire shape (from a given wire material). An estimate is made of the capillary trajectory that will occur at the desired wire shape, the resulting wire shape is analyzed, and the capillary trajectory is corrected until the desired wire shape is obtained.

As mentioned above, the heating of the wire when fed from the capillary facilitates the wire to be formed into a wire stem having a springable shape. Another advantage of applying heat to the wire when fed from the capillary is that heating of the wire (eg annealing the wire from 300 ° C. to 400 ° C.) causes the wire to spring back (eg, cut off). Reduce the tendency to

The tendency for the wires to spring back (eg when cut) can be advantageously used during the cutting process, as described in detail below. In general, the tip of the wire stem positioned to spring back when the wire is cut moves away from the capillary (and away from the EFO electrode), which facilitates control of the formation of the ball at the end of the feed spool.

It is within the scope of the present invention that a plurality of capillaries can be used simultaneously to produce a plurality of wire stems. For example, two capillaries acting at once (eg in series) can join and shape twice as many wire stems as “single” capillaries. Ten capillaries will reduce the manufacturing time to one tenth. This is useful for linear or rectangular arrangements where a plurality of similar wire stems need to be made in a repeatable pattern on the surface of the electronic component.

In the prior art, compensation of springback is known in the sense that a certain amount of overtravel is intentionally caused to form a conventional wire-bond loop.

Cutting of wire

As described above, once the proximal (free) end 202a of the wire 202 is coupled to the substrate 208 and made into the desired shape, the wire 202 is formed by an electron flame off (EFO) electrode 232. It is cut by the generated arc or the like. Once cut, the wire has a second (end) end, is fashioned (shaped) to a certain shape and is considered to be a “wire stem”.

According to one aspect of the present invention, providing light emission at the EFO electrode (cathode) stabilizes arc / plasma formation and provides more reliable and predictable wire cutting behavior. This technique can be used in conjunction with a negative EFO or a positive EFO and is called "Photo-Assisted Spark Stabilization (PASS)."

As shown in FIG. 2, the light source 238 directs light to the place where the wire 202 is cut, just below the capillary 204. Although light is shown focused on wire 202, it is preferably focused on electrode 232, which may be an anode or a cathode.

Preferably, the light is ultraviolet with a wavelength of 184 nm or 254 nm, but cutting enhancement can be obtained with light of other wavelengths as well.

As shown in detail in FIG. 4A, (light source 238 has been omitted for clarity of illustration). Wire 402 coupled to terminal 412 (compare 112) on substrate 408 (compare 108). (Compared to 102) is cut by a spark (not shown) generated by an electrode 432 (compare 232) located directly below the capillary 404 (compare 204). This leads the wire 402 to (a) a bottom, wire stem portion 430 coupled to the substrate, and (b) an upper feed portion extending through the capillary 404 from a spool (not shown, see 206 in FIG. 2) ( 431) into two parts.

As shown in Fig. 4B, similar to Fig. 7B of Case 1, this forms a ball 434 (of an increased diameter region) at the distal tip of the wire stem 430, and the feed material of the feed wire 431 At the end, a similar ball 436 is formed. These balls are important for the following reasons:

(a) Optionally, the ball 434 provides a “profile” at the tip of the wire stem 430, which is advantageous for connecting to the electronic component at the tip of the wire stem.

(b) The ball 436 is well suited for making a subsequent bond between the free end of the supply wire 431 and the substrate to form another (continuous) wire stem on the substrate.

The formation of balls (see (b) directly above) of the feed side of the wire fed from the capillary tube is not only essential for the subsequent joining of the wire but also presents a significant problem for the output. In conventional wire bonding the process was intended to perform in a highly automated manner. Checks for the presence of balls to establish subsequent engagements are known, and if they do not have balls, the process stops and requires manual intervention.

According to the present invention it has been found that the use of light emission associated with electron flame off cutting of wires improves wire ball formation and ball size distribution and reduces the loss of balls. Such supplemental spark wire cutting is useful in any (typical) wire bonding and can be used with various EFO circuit modifications, such as performing the process at low peak voltage conditions.

In general, from a technical point of view, the light assist spark stabilization (PASS) technique of the present invention makes field-related light emission more stable for breakdown (arc cutting of wires) and more controllable of the cutoff height (Z-axis coordinate of the cutoff). Use The latter advantage of precisely controlling the cutoff height of the wire is described in detail below.

As described above, the light source 238 (not shown in FIG. 4) may be focused at a point on the wire where it is desired to cut the wire. This is good for wire couplers that use an amount of EFO where the spark starts at the wire. Alternatively, light may flood-illuminate an area containing the electrode. This is good for wire couplers using negative EFOs where sparks start at the electrode. In either case, the use of ultraviolet light causes a breakdown of the ambient gas component and makes it easier for electrons to move from the electrode to the wire or vice versa. In general, flood lights (eg with respect to negative EFOs) tend to be self-selective and are more reliable than spotlights (eg with respect to positive EFOs). In flood lighting, the sharp tip of the electrode “selects” the point where the cut occurs.

In conventional continuous supply ball coupling, a high voltage arc (or EFO) is used to cut the wire in the middle of each surface bonding event. The wire cutting step of the continuous ball bond typically involves shearing the wire after forming the second surface bond. In general, the completed length of the sheared wire is not critical, and therefore the ability of the EFO to generate a uniform height of the cut wire is not critical.

In contrast to this situation, the uniformity of the wire height is very important in the continuous feed elastic molding process of the present invention. The ability of the EFO's cutting height to be controlled directly affects the quality of the final result, since the uniformity and planarity of the arrangement of the contacts is a direct function of this ability.

According to the present invention, ultraviolet light is used to stabilize wire cutting uniformity and spark breakdown when high voltage arcs are used for wire cutting. The formation of a high voltage arc in the gas between the two electrodes is a process of avalanche in which a cascade of electrons continues to increase until more current carrying plasma is formed between the cathode and anode of the discharge, creating more and more ionized gas molecules. to be. Initiation of the arc typically requires that field emission at the cathode electrode supply a few electrons that initiate breakdown.

According to the invention an ultraviolet (UV) lamp is used to illuminate the cathode to stimulate photoelectron generation at the cathode element of the EFO discharge. This lowers the critical electric field required for emission by the use of 3-5 eV (electron volts) UV (ultraviolet) photons to stimulate the generation of free electrons at the cathode under applied high field conditions. The role of the cathode may be performed by a cutting electrode (preferably a negative EFO device) or a continuous supply wire (positive EFO alternative). Flood or focus UV illuminators may be used. Focus illumination of the wire for positive EFO facilitates wire height control by localizing the electron emission point on the wire and controlling the point where the cutting plasma is initially formed. Flood illumination also acts as a ball height stabilizer because it stabilizes the arc formation between the cathode and anode.

It is believed that the improvement in ball formation by providing ultraviolet light during electron flame off is due to the following. In a conventional wire coupler the voltage at the flame off electrode (eg 232) is controlled by a circuit (eg 234). The voltage of the electrode is increased from 0 (e.g., 0, and is significantly reduced during the ignition period (the spark acts somewhat short circuit). The length of the ignition period is monitored and if it exceeds the predetermined length [(e.g. Ball is not considered to be formed, as set by a “watchdog” type timer in the control circuit of step 222. The process is inherently slightly intermediate, with a plurality of “trials” statistically from peak to peak. It can be shown by conventional statistical (eg bell-shaped) curves with fluctuations: The provision of ultraviolet light in the EFO increases the slope of the graph (in time) to the right of the peak so that it exceeds the desired duration. It is believed to reduce the incidence and increase the likelihood of ball formation within a given period of time, thus providing ultraviolet light during electron flame-off will reduce the onset time of the spark Significantly reduces the time-out indicating the formation failed to see, as well as chukhal.

Although the present invention has been described in connection with cutting wire stems under the tip of a capillary using electron flame off (EFO) technology, the wire may be cut using other means, such as mechanical means, or cut at other locations, such as capillaries. What can be included is within the scope of the present invention (see, eg, US Pat. No. 4,955,523, above, wherein the wire is cut in a capillary (bonding head)). However, if it is desired to make the distal end of the wire stem into a ball shape, a secondary step (ball formation separation) will be required to be performed.

According to one aspect of the invention, wire melting and ball formation consist of two separate sequential steps (rather than being mixed in one step). The wire stem is first melted (cut) and then a ball is formed at its ends (tips). This makes the wire height distribution more tight, thereby ensuring that the tips of the plurality of wire stems, which is a feature of the invention, described in more detail below, are on the same plane.

4C shows a wire stem cut in an arc from an electrode 432 having a stem 430 extending from the surface of the substrate 408 and the supply 431. In this case, the strength of the arc is adjusted (minimized) to sufficiently cut the wire while preventing balls of significant size (see 434 and 436 in Figure 4B) from forming. Balls that can be compared to the balls 434 and 436 shown in FIG. 4B can be formed in sequential processing steps using suitable ball forming techniques. It is within the scope of the present invention that the ball is formed on the tip of the supply (eg 431) without forming a ball on the tip of the stem portion of the wire (eg 430).

4D illustrates a method of using a spring back of the wire stem which serves as an advantage in cutting in controlling the formation of the ball on the tip of the supply 431 of the wire. Before sparking the electrode 432, the capillary 404 is moved approximately 0.5 mm in the X or Y direction. (More precisely, it will be the substrate moved by the X-Y table.) The EFO electrode 432 is moved with the capillary 404. In FIG. 4D, the capillary and the EFO are shown to be shifted left relative to the stem portion 430 of the wire. This lateral displacement of the capillary / EFO causes the wire stem portion 430 to snap to the right (as shown) upon cutting of the wire. This has a number of advantages, including:

(a) By moving the capillary before sparking, the wire is preloaded and “poised” at the minimum magnetic pole (eg, spark) to cut itself. This is analogous to stretching the strings tightly before cutting them with a knife.

(b) The wire stem portion 430, which generally represents the shortest passage to ground, causes the wire stem portion to snap from the EFO electrode 432 immediately when the wire is cut, and the supply portion 431 of the shortest passage to ground. It is a source. In general, the spark provided by the EFO electrode will require the shortest path to ground. In this way, ball formation at the tip of the supply of the wire can be more easily adjusted.

As described herein, failure to form a ball at the tip of the supply of the wire for secondary engagement will cause a stop of the process. A more subtle problem is that balls that do not have a definite size (eg, 436) may cause non-deterministic z-axis adjustments in the wire bonding process even though they are suitable for initial bonding. Balls are placed at the ends of the supply wires and reliable and repeatable sizes are expected. The technique of the present invention not only ensures reliable ball formation but also ensures uniformity in the size of the balls formed during wire cutting.

Coating of wire stem

According to the present invention, when the wire is bonded to the substrate and cut to have a wire stem made in a predetermined shape and having a distal end, the wire stem thereby is coated.

Generally, the coating applied to the wire stem is a conductive metal material such as nickel, cobalt, copper and alloys thereof and the like, which will be described in more detail below.

Overcoated wire stems are “elastic” and / or “compliance” contact structures, generally such structures are spring (elastic) properties (eg, plastic deformation) from the mechanical properties of the overcoated material and the elastic shape of the wire stem. Without compliance). The overall resilience of the resilient contact structure derives from these collectively "organized" properties.

Before continuing with this discussion, it may be helpful to clarify the following specific terms used in the text.

"Flexible" refers to a property that an object can be easily bent or twisted without being damaged, and is particularly applicable to the wire stem of the present invention.

"Elasticity" refers to the property of returning to the original shape after the object is compressed or bent, and is particularly applicable to the overcoated contact structure of the present invention.

"Elastic" refers to the property that an object resists bending (compression or tensile) forces and tries to return to its original shape after the compression force is removed. Normal springs exhibit elastic action. Elasticity is similar in elasticity to that.

"Plasticity" refers to the property of things to be deformed without damage, and is similar to flexibility.

“Compliance” as used in this text refers to the property of things exhibiting both elasticity and plasticity. In this respect, “compliance” is a broader term than “elasticity”. Contact structures of the present invention having a flexible (plastic) wire stem and an elastic (elastic) overcoat can be considered to have compliance.

For example, compliance contact structures formed in accordance with the techniques of the present invention may exhibit elastic deformation (pure elasticity) and plastic deformation (pure flexibility) such as 3 mils of elasticity and 7 mils of plasticity for a total bending of 10 mils. have. In general, this ratio of plasticity and elasticity can be easily adjusted by the composition of the overcoating layer (ie for certain wire stems). For example, pure ductile nickel exhibits a relatively large amount of plasticity compared to elasticity.

Suitable mechanical properties of the coating layer include thickness, yield strength and modulus of elasticity. In general, the elastic contact structure has a greater elasticity by having a larger thickness, yield strength and modulus of elasticity of the overcoating layer.

In general, the wire stem itself can simply maintain the shape of the overcoating layer without compromising the mechanical action of the elastic contact structure. As described herein, this is similar to the purpose of the poleswalk, even though the wire stem remains in place after being overcoated.

In Figure 5, similar to Figure 5 of Case-2, an exemplary wire stem 530 (compare 202, 303) is coupled with a terminal 512 (compare 112) on a substrate 508 (compare 108, 208). A wire 502 having a proximal base end 502a (compare 202a), consisting of a predetermined shape (similar to the shape shown in FIG. 2B for illustrative purposes) and a ball 534 at its distal end. (Compared to 434).

The example wire shape shown in FIG. 5 may be elastic, may act as a spring, and may react to forces generated axially (typically z-axis) below the wire stem (denoted by “F”). It is obvious. However, as mentioned above, for the wire stem to be elastic, the wire stem itself must have a significant content of elastic material, and gold (which is an optional conductor for semiconductor interconnection) is not elastic (gold reacts to applied forces Easily plastically deformed). As further described herein, the ability to form elastic contact structures is desirable for a wide range of applications. Obviously, if the wire itself is elastic (can function as a spring) and if the wire is dependent on a particular shape, a vast amount of force is readily applied to the joint at the base end of the wire stem (coupling 502a). This results in breakage of the joints and damage of the interconnects caused by the wire stems between the electronic components. The main feature of the present invention is to avoid this problem.

According to the invention, one or more layers are coated by plating on the wire stem 530, in which case the outermost (top) layer is a conductor material. The main results obtained by coating the wire stems are as follows:

(a) imparts elasticity to the contact structure comprising overcoated wire stems, especially if the wire stems themselves are essentially resistive materials (eg gold).

(b) improve the fixability of the wire stem to the substrate.

5 shows a wire stem 530 having multiple layers (two coating layers), the wire stem surrounding the wire 502 and covering the wire 502 with a first inner coating layer 520 and a first layer 520. ) And a second outer (top) coating layer 522. The first layer 520 covers the terminal 512 (compared to 112 in FIG. 1A) to which the base end 502a of the terminal 502 is coupled, and secures the wire there (ie, the bonding fixation of the wire). Increase significantly). The second layer 522 covers the first layer 520 in the region of the terminal 512 and increases the fixing of the wire 502 to the terminal.

When two layers 520, 522 overcoating the wire stem are opposed, one (or two) layers provide elasticity to the wire stem (otherwise inelastic) and an outer layer (or two) Layer) is electrically conductive.

For example:

Wire 502 is a soft gold material and has a diameter of 0.0007 to 0.0020 inch (0.00178 to 0.00508 cm).

The inner coating layer 520 is a copper “strike” and has a thickness of 12.7 to 25.4 micro centimeters (5 to 10 micro inches).

The outer coating layer 522 is nickel and has a thickness of 0.00508 cm (0.0020 inch).

In general, two reasons why a coating layer, such as copper, is chosen are:

(i) To improve the plating performance of embedded wire stems (as is known, certain materials, for example nickel, are quite difficult to plate).

(ii) to ensure good current conductivity to the overcoated wire stem (copper is known to have good electrical conductivity).

In general, coating layers such as nickel or alloys thereof are selected due to their mechanical properties, which are characterized by high yield strength and elastic contact structures that can respond elastically to applied forces (eg, terminals). Has the ability to lock firmly on

In some cases, it will be desirable for the top (eg, third) overcoat layer to provide solderability and the like and be compatible with the material of the contact area by electroplating. In such a case, for example, a thin top overcoat hard gold having a thickness of approximately 0.000254 cm (100 microinches) will be provided.

In high frequency applications, current tends to be distributed along the outermost layer of the coated wire stem. In such a case, gold is preferred as the outer layer of the plurality of overcoats.

FIG. 5A shows only one layer 540 (in contrast to the plurality of overcoating layers 520/522 of FIG. 5) that overcoats and surrounds the wire stem (and completely covers) and surrounds the terminal 512 on the substrate 508. The wire stem 530 with only is shown. As in the example shown in FIG. 5, the base end 502a of the wire 502 is coupled to the terminal 512, and a ball 534 is provided at the distal end of the wire 502. In this case, the coating layer 540 is elastic and conductive.

For example:

Wire 502 is a gold material and has a diameter of 0.00178 to 0.00508 cm (0.0007 to 0.0020 inch).

The outer coating layer 540 is nickel and has a thickness of 0.0005 to 0.0030 inch (0.00127 to 0.00762 cm).

5B shows a wire stem 530 with only one layer 544 overcoating and jacketing (partially covering) the wire stem and surrounding the terminal 512. In this case (jacketing), the overcoat layer 544 extends only partially from the base end 502a of the wire stem towards the distal end 534 of the wire stem. Preferably, as shown, the overcoat 544 surrounds the bent portion of the wire stem. In general, the vertical ends (not shown) of the wire stems do not contribute to the overall elasticity of the elastic contact structure. In this case, it is important that the wire stem 530 itself should be electrically conductive, because the wire stem protrudes from the coating 544 and the electrical to the contact pad (not shown) on the electronic component (not shown). This is because it makes contact.

For example:

Wire 502 is a gold material and has a diameter of 0.00178 to 0.00508 cm (0.0007 to 0.0020 inch).

The outer coating layer 544 is nickel and has a thickness of 0.0005 to 0.0030 inch (0.00127 to 0.00762 cm).

5C shows a wire stem 530 overcoated with a single layer 548 in a similar manner to that of FIG. 5A (compare 540). In this case, layer 548 is provided with a plurality of fine protrusions on the outer surface spaced longitudinally along the length of the wire stem. Such “jagged” coatings are sometimes referred to as “dendritic”. Such protrusions or irregular surface portions can be made in a number of ways, for example, by adjusting the processing conditions in the plating electrolytic cell (as described herein) causing sharp nodules to form in layer 548. This is exemplary for the outer conductive layer of the multilayer coating provided with fine protrusions.

For example:

Wire 502 is a gold material and has a diameter of 0.00178 to 0.00508 cm (0.0007 to 0.0020 inch).

The outer coating layer 540 is nickel and has a thickness of 0.00127 to 0.00762 cm (0.0005 to 0.0030 inch) using a common precipitation method known in the art. In addition, “foreign” particles (other than nickel) by the co-precipitation method are suitably silicon carbide, alumina, diamond and the like and have a diameter of approximately 3 microns. Such co-precipitation provides serrated peaks on the outer surface of the coating, which peaks have an average peak height of about 0.00127 cm (0.0005 inch). Co-precipitation can also be performed with particles of the same material as the ions in solution (eg nickel particles in a nickel plated electrolytic cell).

Another technique for forming a “jagged” coating (eg, 502) (ie, a technique other than the above described co-precipitation method) has been described as “abnormal” in plating solutions with respect to electroplating. abnormally) ”using high currents and“ abnormally ”low metal concentrations. As is known, this will result in a “microtreeing” effect in which nodules are formed in the plating.

5D shows a wire stem 530 overcoated in multiple layers including an inner layer 552 (compare 520 in FIG. 5) provided with microprotrusions (as described with respect to FIG. 5C). The “normal” outer registration layer 556 located on top of the inner layer 552 can be positioned in a “normal” manner and coincides with the micro-projection topography of the inner layer 552 to represent the micro-projection. . (For example, one strike of gold can be applied on a nickel layer with fine protrusions.) This is obviously easier when (i) multilayer coatings are desired and (ii) fine protrusions in the inner layer. It is preferable to form it.

5E illustrates one embodiment of the present invention in which wire 502 is bonded to terminal 512 on substrate 508 at its base end 502a and overcoated with an exemplary single layer conductive coating 562. have. In this case, the entire lower portion of the wire stem 530 including the area surrounding the terminal 512 is embedded in the mass 564. Mass 564 is an electrically conductive polymer mass, preferably a silicone rubber filled with silver particles, which serves to reduce parasitic inductance exhibited by the wire stem. Suitable electrical properties of mass 564 are conductive in the range of 10-2 to 10-6 ohm centimeters. The material of the mass 564 is selected such that the (limiting) action (elastic, compliance) of the wire stem 530 is not "creep" or significantly adversely affected.

For example:

Wire 502 is a gold material and has a diameter of 0.00178 to 0.00508 cm (0.0007 to 0.0020 inch).

The outer coating layer 544 is nickel and has a thickness of 0.0005 to 0.0030 inch (0.00127 to 0.00762 cm).

Polymer mass 564 has a Shore hardness between 10 and 60 (such as 30-50) and is filled with silver particles to have the desired conductivity. In general, the polymer mass is chosen so as not to significantly impair the compliance of the contact structure. It is within the scope of the present invention that conductive polymers that can be usefully employed can be used for the polymer mass 564.

The main purpose of the elastomeric mass 564 is to provide a conductive passage between the wire stem base end and the distal tip shorter than the full length of the wire stem [more precisely the closest position on the terminal 512 relative to the distal tip of the wire stem. ] Is provided. The shortest passage between the two positions consists of straight lines, and it is within the scope of the present invention that the polymer mass is provided in the area indicated by the dotted line.

FIG. 5F illustrates an exemplary overcoated freestanding elastic contact structure 530 (compare with forming wire 202 of FIG. 2E), wherein the elastic contact structure is from a terminal 512 (compared to 212) on the substrate. It is extended. In this description, the distal end (tip) 530b (indicated by the bold line and no force applied) of the contact structure 530 is located at position “A” on the surface of the substrate. The force indicated by the arrow “F” and exerted axially (toward the substrate 508) through the distal end of the elastic contact structure 530 is determined by the position “B” on the surface of the substrate (by the force applied to it). Cause the elastic contact structure 530 to bend as indicated by the dashed line to be positioned at ”. Bending force “F” occurs by pushing the substrate 508, which may be an electronic component against another electronic component, to provide an interconnection between the two electronic components through an elastic contact structure or vice versa.

Position “B” is closer to the surface of the substrate 508 than position “A”. Between these two positions A and B, the position "C" is indicated. In use (eg, when interconnecting two electronic components), applying a force F to the elastic contact structure 530 may first plastically deform from position “A” to position “C” If more force "F" is applied, it will elastically (elastically) deform from position "C" to position "B".

This “combination” of plastic and elastic deformations is meant “compliance” (as opposed to “elasticity”), which is not necessarily wrong, but rather a printed circuit in which a plurality of contact structures are not perfectly planar in surface. When not in constant contact with an electronic component such as a substrate (PCB), there can be a significant advantage in ensuring a constant contact force against contact. The contact structure that strikes the surface of the printed circuit board first is more plastically deformed than the contact structure that touches the surface of the printed circuit board last, but the compressive force of the entire contact structure will be fairly constant.

Exemplary elastic contact structures according to the present invention (e.g., 530 of Figure 5F) preferably have a total height of 0.0254 to 0.3048 cm (0.0100 to 0.1200 inch) (i.e. the contact structure from the surface of the substrate in the z-axis). Distance between distal ends) and will have a spring constant (“k”) (in most applications) between 0.1 and 20.0 g / mil (grams per 0.001 inch), preferably between 0.5 and 5.0 g / mil.

A person skilled in the art will understand that it is difficult to set a "preferred" spring constant because the elasticity and / or flexibility of the contact structure is specified under a particular application. By manipulating the thickness, shape and material of the wire stem and the material and thickness of the overcoat in accordance with the techniques described herein, a substantially possible blend of elasticity and plasticity can be achieved. However, for the purpose of this description, a spring constant (“k”) of 3.0 g / mil is preferred for electronic devices using the contact structure of the present invention to connect semiconductor die (which may be mounted directly to a die or probe). Do. The overall compliance of the contact structure is up to 15 mils (0.015 inch) for plastic deformation (eg, the distance between positions “A” and “C”) and 3 mils (0.003 inch) for elasticity (eg Distance between positions “C” and “B”). The relative contributions of elasticity and plasticity will be readily made in the individual devices. For example, the contact structure on the interposer can be fabricated to exhibit a strain of 10 mils overall allowing for 5 mils of plastic strain and 5 mils of elastic strain. Contact structures may be fabricated that exhibit substantially “pure” elasticity.

Coating technology, material and thickness

The possibility of coating the wire stem by plating is mentioned in the text. In accordance with the present invention, a number of different techniques can be used to overcoat the wire stems to provide the mechanical and chemical properties of the elastic contact structure.

The main feature of the coating applied to the wire stem is continuity, ie the formation of homogeneous materials along the entire length of the wire stem. Avoiding this discontinuity in the composition or metallurgy of the coating minimizes the possibility of local stress locations occurring due to repeated bending of the elastic contact structure, which may cause breakage.

In general, the material of the coating material is significantly stronger than the material of the wire and provides the desired "spring property" to the contact structure.

Coatings can be applied to a number of readily available techniques, including but not limited to:

(a) wet electrochemical techniques including electrolytic or non-electrolytic solution plating of metals;

(b) Electroplating method of electroplating nickel from nickel and its alloys or electroplating nickel from standard nickel sulfonic acid solution. This method makes it possible to attach a controlled thickness of coating onto a wire stem having a tensile strength of at least 80,000 pounds per inch (psi), such as at least 200,000 psi (14,060 kg / cm 2).

(c) chemical vapor precipitation (CVD), microwave elevated CVD (MECVD), physical vapor precipitation (PVD), and the like, including processes including evaporation and sputtering.

(d) A number of processes that lead to the precipitation of materials (conductive materials) through decomposition of gas, liquid or solid precursors (CVD is one such process).

(e) Wave soldering or electrolytically-adherent soldering for attaching the solder to the wire stem.

The above-described wet electrochemical and electroplating processes are carried out in accordance with "standard" processes and can be carried out using molten salts, ionic solutions and the like.

For the purposes of the present application, the process of providing a coating on a wire stem by one of the techniques described above is collectively named "overcoating" (of wire stem).

As a useful comparison in describing the elastic contact structure of the present invention described in Case-1, the wire stem is conceptualized as a "skeleton" and the coating on the wire stem as a "muscle". This is because the wire stem defines the shape of the contact structure and the coating on the wire stem provides the wire stem with significant mechanical properties (eg, elasticity). As discussed above, as a more useful comparison, the formed wire may be considered as a "fall walk" or "mandrel" which serves to provide the shape of the overcoat.

In general, the importance of overcoating (eg plating) in the process is inevitable and should not be overlooked because the overcoat functions as the main component of the contact structure.

Conventionally, plating has generally been treated as a process in which the properties of the underlying structure can be improved. This is because plating can avoid corrosion of the underlying structure, alter the appearance of the underlying structure and provide desirable properties (eg, firmness) to the surface of the underlying structure. In other words, plating was not seen as a "end" means. The plated structure depends on the properties of the underlying structure itself to perform its intended function. For example, an automobile bumper would function very well if it had not been plated by providing the bumper with an aesthetic appearance as well as protection from corrosion through plating.

In conventional electronic interconnects as described above in US Pat. No. 5,317,479, plating is used to improve the rigidity of the bottom lead structure. If the lower lead structure does not remain intact, the concept of flexible leads is disrupted.

In contrast to conventional plating “mind sets”, the present invention relies on plating (ie, overcoat material) to provide substantially all of the desired functionality. The text describes how the wire stem is removed after overcoating and the result is still a fully functional elastic contact structure. (However, plating processes on mandrels of ceramics, waxes, plastics, etc. have been used for phonograph recording control and automotive radiators.)

In making the flexible electrical contact structure according to the invention, the coating (e. G. Plating) is again described.

Suitable materials for coating (ie, coating layers in multiple layers of coating on the wire stem) include, but are not limited to:

Nickel, copper (especially thin “strikes” as a middle layer of a plurality of coatings on gold wire stems or as a current conducting layer), cobalt, iron, and Kovar (tm) (eg, 29% nickel; 17% cobalt); 0.3% manganese; balance iron) and “alloy 42” (42% nickel; 0.1% carbon; balance iron) or similar materials containing iron / nickel / cobalt (eg, 42% nickel, 40.7% iron) , Alloys including 17% cobalt and 0.3% manganese) or low expansion alloys (eg nickel-cobalt);

Gold (especially hard gold) and silver. Both have excellent current conducting capabilities;

Elements of the platinum group;

Precious or semi-precious metals;

Tungsten and molybdenum (both can be suitably attached by a CVD process);

cobalt;

Zinc (especially on aluminum wire stems);

Tin (particularly in forming the eutectic described herein);

Solders (mainly in producing raised solid phase contacts);

Platinum, rhodium, lussendium, other elements of the platinum group, a group consisting of copper, and semi-noble metals selected from gold-silver-copper alloys and their alloys (especially for the top coating layer).

Of course, the techniques for applying such coating materials to the various wire stem materials described herein will vary from application to application. For example, gold is one of the preferred materials for wire stems. However, due to the good electrical properties, plating on gold is somewhat undesirable. Moreover, gold is not a good "initiator" for plating. Thus, according to one aspect of the invention, when plating nickel onto gold (especially non-electrolytic plating), it is desirable to first apply a thin layer of copper "strike" onto the gold wire stem.

In freestanding elastic contact structures, nickel is suitable as an overcoat material, in particular as an inner material (compare 520) of a plurality of layers of coating material. Such nickel inner layers, which tend to be oxidized, are overcoated with a precious metal or semi-noble metal top coating layer 522 such as gold, silver, platinum group elements and alloys thereof.

In the use of contact structures that function as raised solder contacts (described in more detail below), nickel, cobalt and their alloys are suitable as overcoat materials, for example 0.000076 to 0.0127 cm (0.00003 to 0.00500 inch), preferably It may have a thickness in the range of 0.000127 to 0.00762 cm (0.00005 to 0.00300 inch).

Solders typically containing lead and tin elements are suitable for overcoat and can be applied to the wire stem by conventional wave soldering devices. Embodiments of the present invention utilizing solder overcoats will be described in more detail below.

An important advantage of forming multiple layers of coating on the wire stem is that the material of each individual layer can be selected to provide the physical properties of the elastic contact for a given application. These physical properties include tensile strength and yield strength and the like.

Another advantage of forming multiple layers of coating on the wire stem is due to the fact that nickel has a relatively strong contact force in order to break through the oxide formed by having strong oxidative properties. (Low contact force due to good electrical connection is generally preferred.) For this reason, nickel tends to be not the best element as a single layer coating or as a top layer of multiple layers of coating on a wire stem.

Similarly, although soldering top coats are highly desirable in certain applications, typical solders include tin and gold (ie, gold wire stem) is very reactive to tin. To prevent such undesirable reactions, an intermediate “block” layer can be applied between the wire stem and the top layer of the coating. In the case of a gold wire stem and a soldered top layer (with tin), the barrier between the nickel alloy will prevent the reaction between the (wire) and the tin (solder overcoat). For example, a first coating layer of nickel alloy having a thickness of 254 to 2540 micrometers (100 to 1,000 microinches) may be applied between the gold wire and the top coating layer of the soldered portion.

As the elastic contact structure functions as a “spring”, certain advantages will occur when an internal compressive stress is introduced into the coating (at least one layer of the coating) during application of the coating to the wire stem. For example, certain additives may be added to the plating solution, such as plated on the wire stem, to introduce internal stresses in the coating. For example, in the nickel sulfonate plating solution, the following additives may be added to the plating solution: saccharin (sodium salt), sulfur urea, butindiol, meta-benzene disulfonic acid (sodium salt), 1,3, 6 naphthalene trisulphonic acid (trisodium salt). Boric acid may also be added to the plating solution for pH adjustment. The internal stress of the plating portion can be controlled by the overall temperature of the plating portion, the plating rate, the pH of the plating solution, the nickel concentration, and other plating parameters. In general, the additives described in the text are the main contributors to controlling stress in the plated portion, including compressive stress, tensile strength and “theta” (θ) stress.

If the wire stem itself has some elasticity that is “adjusted” with respect to the elasticity of the coating, the formed wire stem (or a plurality of formed wire stems) may be formed during overcoating (eg, plating) of the stem (shown in FIG. 5F). Under stress pre-pressurized by the applied force “F”. Such “mechanical” means for prepressurizing the contact structure may be desirable in certain applications.

According to the present invention, the overcoated wire stem functions as a spring, and the spring constant "K" calculated for a given configuration is based on the known properties of the material. In general, as described herein, the overcoat is applied at a somewhat constant thickness at least along the bend of the wire stem along the length of the wire stem, thereby facilitating the calculation. By using a process such as electroplating, it has a constant thickness along the length of the wire.

According to one aspect of the invention, the coating is applied to have a thickness that is carefully considered and not constant along the length of the wire stem.

5G shows a wire stem 530 plated with an overcoat 568 while applying “localized” heat (indicated by arrow “H”) to the substrate. This is readily accomplished by attaching the substrate 508 on a heat strip (not shown). In this way, the substrate and the wire stem become hotter than the surrounding plating bath.

Although the gold wire 502 is a good thermal conductor, the base end 502a of the wire stem becomes hotter than the distal end 502b of the wire stem, and the entire wire stem 530 is placed in the plating bath during the plating operation. Immersion is due to the notable temperature component along the length of the wire stem. This is due to the coating having a larger thickness at the base end of the wire stem than at the distal end of the wire stem. (This change in thickness is exaggerated in Figure 5G for clarity. See Figure 5B.)

By making the coating 568 thicker than at the freestanding end of the wire stem when the wire stem is secured to the substrate, the following advantages are involved:

The anchoring of the wire stem to the substrate is high for a given “average” coating thickness;

The compliance of the elastic contact structure is greater. The thickness of the coating to match the desired average coating thickness at the elastically manipulated end of the wire stem becomes smaller;

Thus, the stress imposed by the elastic force works well at the base end 502a of the wire stem. As a result, the tendency for the elastic contact structure to rupture from the substrate is reduced.

As an example, when raising the temperature of the substrate to 80 ± 10 ° C. (when raising the overall temperature of the plating bath to room temperature), the coating portion which is uniformly deposited along the length of the wire stem is 1.5: 1 to 5: Can be introduced to represent a thickness element of one. This aspect of the invention is applicable to both electroplating and electroless plating.

In contrast, cooling of the substrate (for bath), for example using a thermal coupling device, can be used to fix the coating thickness.

Within the scope of the present invention, any suitable technique for fixing the thickness of the plated member may be used by imposing a temperature component on the plated member.

In accordance with the present invention, the local heat applied during the plating process (ie, over the entire heat of the plating bath) can be usefully used to “fix” the thickness of the coating at multiple contents. As described herein, thickening the coating can be used to bring the heat source closer. Other non-uniform coatings of constant thickness may be used. For example, if the coating (eg nickel) tends to be thicker at the distal end of the wire stem without applying heat, the coating thickness may be constant or along the length of the wire stem. As described, the coating may be “reinforced” to become thicker at the base end of the wire stem. In addition, where it is desirable to attach a coating of constant thickness across the surface of a planar substrate, heat may be used to “react” to the tendency of the plating to be inconsistent (ie, thinner at the spot), thereby providing a substrate (i.e., , Silicon wafers, to ensure a constant coating thickness across the surface. (Spots or areas on the surface of the substrate with thin coatings can be locally heated to make the coating thicker in these areas.) This is also within the scope of the present invention.

FIG. 5H shows one embodiment of fixing a coating thickness similar to that shown in FIG. 5G except for coating a straight wire stem (eg, a “pin”). Straight wire stem 502 is coupled by a base end 502a to a terminal on the substrate. Heat (“H”) is applied to the substrate during plating. The overcoat 578 will be thinner towards the distal end 502b of the wire stem 502 than at the base end 502a of the wire stem.

In many of the embodiments described herein, nickel is well suited for overcoat material (eg, at least one of the overcoat layers) and is deposited on the wire stem by plating. Although such processes are generally well known and can be readily understood by those skilled in the art, Roger Brugger, Robert Brugger, 1970, of the relevant arts incorporated herein by reference. In Nickel Plating by Robert Draper Ltd. (UK), a clear explanation of nickel plating can be provided.

Characteristics of contact area

The present invention describes coupling a free end of a wire to a contact area (eg, 100 in FIG. 1) which may be a terminal (see 112 in FIG. 1A) on a substrate (eg, an electronic component). . The contact region 110 is required to be formed of a material that is metal and includes: It is not limited to these materials:

(a) gold and its alloys;

(b) aluminum and its alloys;

(c) silver and its alloys;

(d) copper and its alloys;

(e) metals of the platinum group.

In many embodiments described herein, the contact region is terminal 112. Such contact areas are not limited to terminals or bonding pads, but there are a plurality of contact areas on a single continuous layer substantially shaped by photoresist and etching (see, eg, FIG. 1B) to have a plurality of contact areas. Should be understood.

The contact area (such as a terminal) may comprise a plurality of layers, the top layer of which is a material such as gold or aluminum. In general, the material of the contact region of the substrate is within the range known as the material of such contact region.

Material properties

A number of materials have been described that are suitable for wires, coatings, and for limited areas in which the wires are bonded. Although not forming part of the invention, a brief description of the salient features of these many materials will be taken in conjunction with the description of the practicality of the invention. This brief description is for the purpose of description and is in no way intended to be limiting. An easier “shopping list” of metal materials and their related properties can be found in Harper and Sampson, McGraw-Hill, Inc., Second Edition, 1994, in the Electronic Materials and Process Handbook. Chapter 5 (“Metal Pieces”) may be provided in lines 5.1 to 5.69.

Aluminum-Due to its good electrical and mechanical properties, aluminum is an important contact material. However, aluminum as a contact material is not suitable because it is easily oxidized. If aluminum is used in contact junctions, it should be plated or covered with copper, silver or tin.

Beryllium-This material exhibits high stiffness and strength with respect to density ratio, high thermal conductivity and low thermal expansion.

Copper-Copper is widely used because of its high electrical and thermal conductivity, low cost and ease of manufacture. The main disadvantage of copper contacts is their low resistance to oxidation and corrosion. Pure copper is relatively soft, annealed at low temperatures, and lacks spring properties.

Epoxy-Epoxy is a resin that is generally nonconductive and cures at room temperature or at high temperatures. When used as an electrical connection (eg, instead of soldering), conductive epoxy filled with silver or gold particles is used. Such conductive epoxy typically exhibits a volume resistivity of 1 mΩ / cm or less. Excess epoxy is selected in the application in which the epoxy is to be used.

Eutectic-A "eutectic" is a compound of metals such as gold and tin that exhibits a lower melting temperature than the component metals. Examples include 80% gold-20% tin and 63% tin-37% lead. The properties of the eutectic that expands upon melting are used in the examples described herein.

Gold-Pure gold has the best resistance to oxidation and sulfated. However, their low melting point and their susceptibility to corrosion limit their use in electrical connections for low currents. Gold has very low contact resistance when in contact with palladium and rhodium.

Nickel-Nickel and its alloys are typically characterized by good strength and corrosion resistance. Nickel metal has a relatively good electrical conductivity.

Platinum Group Metals-Platinum and palladium are two important elements of the platinum group. Such metals have a high resistance to rust, thereby providing contact covers for relays and other devices with low contact forces.

Precious Metals The following materials generally exhibit good corrosion resistance (except for silver under sulfide environments). : Gold, iridium, osmium, palladium, platinum, rhodium, ruthenium and silver.

Silver-Pure or alloyed silver is a widely used material for the manufacture and breakage of contacts. The mechanical properties and hardness of pure silver are improved by alloys, but their thermal and electrical conductivity produce the opposite effect.

Soldering-Soldering is an alloy of the group which is widely used in the electronics and electrical industries. Its composition is mainly based on tin and lead and finely contains silver and antimony. The solder is a low strength material at room temperature and quickly decreases in strength at the high temperatures encountered by electronic components. The relative ratio of tin and lead in soldering is usually 50:50.

Tungsten and Molybdenum-Most tungsten and molybdenum contacts are in the form of composites with other components such as silver or copper.

Self-planarizing features

In most applications having contact structures extending from the surface of the electronic component (such as pins extending from semiconductor packages), it is highly desirable that the contact structures are located on the same plane. For projecting contact structures (eg fins) oriented from a plane, this means that the fins are at a constant height. Often, manufacturing tolerances of electronic components are provided to ensure the flatness of the components due to packaging or the like that may occur in the thermal assembly and processing steps. Thus, by definition, the tip of a pin mounted with the same length to the wrapped substrate will not be the same plane.

According to the invention, the vertical (z-axis) position of the tip of the wire stem can be easily adjusted from the cutting operation (described in the text). This ensures the predictable location (preferably coplanar) of the tips of the plurality of wire stems with the relative ease with which the coating is evenly applied to the plurality of wire stems. This feature of the invention is particularly applicable when the wire stem is located at different levels in one electronic component and terminates on the plane of another electronic component (including the same plane of two or more electronic components) or vice versa. Do. Related to this example are when the wire stem is oriented from the electronic component and the electronic device mounted thereon and terminates on the plane of another electronic component. With regard to Figs. 4A and 4B, the guarantee of the precision in the cutting operation is described.

6A illustrates an application 600 of the present invention, in which the first and second plurality of protruding contact structures 602, 604 (one of the plurality of contact structures) are shown (shown in FIG. ) Extending from the first electronic component 606 (orienting), each having two different levels of terminals (different from the z-axis coordination) on the top surface, and mounted to the top surface of the first electronic component 606. And a third plurality of protruding contact structures 614 (two of which are shown) extending from the top surface (shown) of the electronic component 618. In this embodiment, the protruding (elastic) contact structure is shown S-shaped (in a manner similar to the S-shaped wire stem of FIG. 5A), and the overcoat is omitted for clarity.

As shown in the figure, all of the first, second, and third groups of protruding contacts are oriented at different levels (z-axis coordination), and of the third electronic component 622, which is an exemplary plane, They are all terminated (at their tips) on the bottom surface (shown). (Contact pads and terminals that the tips of the elastic contact structures come into contact with are omitted for clarity.)

Due to the different distances at which the contact structures 602, 604, 614 are traversed (results oriented at different levels and terminated in the same plane with each other), some of the protruding contacts must be “extended”. For example, this can be done simply by no longer making (or shortening) the distal end of the contact structure or by modifying the bend of the contact structure as required. This is shown and is desirable to provide various contact structures having the same elastic properties.

In an exemplary embodiment 600 of the invention, component 606 may be a printed circuit board, component 618 may be a decoupling capacitor or resistor (ie, a passive electronic component), and component 622 may be It may be an unpacked semiconductor device (die). The invention is not limited to specific electronic components. Among the advantages of fabricating interconnects in this manner, there is an advantage that the parasitic inductance in the parts can be minimized, which is very important when handling high frequency operated semiconductor devices (eg 622). . This illustration (Fig. 6A) is similar to Fig. 12 of Case-1 showing two straight contact structures per pad 92. As shown in Figs.

FIG. 6B shows an example of the opposite situation for FIG. 6A in the application 620. Contact structures 602, 604, 614 are oriented on third component 622 and terminate on first component 606 and second component 618.

6A and 6B illustrate the degree of adjustment to the height of the contact structure that may be implemented by the present invention. In the first case (Figure 6A), the contact structures are oriented from different levels and terminated at the same (coplanar) level. In the second case (Figure 6B), the contact structures are oriented from the same level and terminated at different levels. In either of the cases shown, the positions of the tips of the elastic contact structures can be precisely adjusted during the cutting operation as described in Figures 4A-4D.

Within the scope of the present invention, a configuration as shown in FIGS. 6A and 6B can be used to mount and connect one die on the other die. Furthermore, referring to Figure 6A, it is unnecessary to have a recess in the surface of the component 606 to use advanced techniques to produce contact structures that are oriented from different levels and having all of the tips in a given plane.

6C shows another configuration 650 that utilizes the “self-planning” feature of the present invention. In this embodiment, the first representative contact structure 652 is oriented from a terminal on the top surface (shown) of the first electronic component 656, and the second representative contact structure 658 is the first electronic component 656. Is oriented from the terminals on the top surface of the substrate. Contact structures 652 and 658 (shown with only one bend for clarity) are terminated from the edge of component 656 with their respective tips (ends) and the surface of another electronic component 664. In contact with Notably, however, the tips 652b and 658b are on the same plane that lies in the x-z plane (compare FIG. 2), while their base ends lie in the x-y plane of the surface of the electronic component 656. This illustrates that the base end and the distal end of the contact structure need not be placed in a parallel plane, and the contact structure is not separated from any arbitrary positions (any plane of the surface of the first electronic component) from any other set of positions ( Which may be formed to extend in any plane of the surface of the second electronic component, provided that this is not limited by the relative orientation of the two (or more) electronic components, each being distinct. That is, as long as a “free” (unblocked) passageway exists between two (or more) electronic components requiring interconnection, the contact structure can be formed extending between the two electronic components, and the contact The structure can be made to exhibit some and all of the advantages (eg, elasticity) described in the text.

Within the scope of the present invention, all of the first group of elastic contact structures are terminated on one plane (whether or not the elastic contact structures are on the same plane with the oriented substrate) and are oriented from the same plane of the same substrate. The elastic contact structures of the group are terminated in different planes that are not coplanar with the substrate or plane on which the first group of elastic contact structures are terminated.

Also within the scope of the present invention, each of the elastic contact structures oriented from the surface of the substrate terminates at different (specific, individual) heights on the surface of the substrate.

The "self-planarizing" nature of the present invention (ie, all of the elastic contact structures oriented from different levels can be terminated in a common plane) provides a number of exhalations with non-current interconnect technology. The “common contact tip plane” feature allows additional components in the electronic component to have elastic contact structures (or connections of elastic contact structures on the substrate) for improved functional integration, wherein the additional components may include resistors and resistors as well as active devices. Capacitors (eg, decoupling capacitors). This provides a desirable interconnect of very low inductance and capacitance.

Although not explicitly shown, it should be understood that the elastic contact structures of the present invention need not be oriented from "level" terminals, and that the terminals may be asymmetrical (at one angle to other terminals on the substrate). In general, this is due to the fact that the base end of the wire stem (ie, the end of the wire first joined to the terminal, etc.) is essentially in one position and the configuration of the wire stem is generally independent of the orientation of the terminal. This is important in that additional components (e.g., 618 shown in Figure 6A) are incorporated, and the terminals are not coplanar with each other. This is also important when one terminal is asymmetrically coupled to the other terminal. Although the process of joining the wire to the terminal is best performed when the wire is bonded to the surface of the terminal substantially normal (eg 90 ° ± 20-30 °), some degree of obliqueness can be accommodated and The degree of obliqueness is generally limited only if the terminal (ie the contact area) is too oblique to interfere with the movement of the capillary (ie the edge of the capillary hits the terminal and mechanically interferes with the joining operation).

Electronic parts

Coupling the free end of the wire to a “substrate”, configuring a flexible wire stem to have a springable shape, and two (or more) electronic components, one of which has elastic contacts formed thereon The invention has been described in connection with overcoating wires with spring material to form an elastic contact structure for interconnecting them. Suitable substrates include, but are not limited to the following, many of which are described in more detail below.

(a) interconnect and interposer substrates;

(b) semiconductor wafers and dies made of any suitable semiconductor material such as silicon (Si) or gallium arsenide (GaAs) or the like;

(c) product interconnect sockets;

(d) test sockets;

(e) sacrificial members, elements, and substrates described herein;

(f) a semiconductor package (“package”) and chip carrier comprising the ceramic and plastic packages described herein; And

(g) connector

Semiconductor package

The present invention is well suited for providing an elastic contact structure (including flexibility, deformation and flexibility) on an outer surface of a semiconductor package.

Semiconductor packages are used to (1) receive (protect) semiconductor (IC) dies in several kinds of package bodies and (2) provide external connections for connecting the packaged dies to external systems. Typically, the external connections provided by the semiconductor package are spaced farther from each other than the connections to the semiconductor die. In this sense, the semiconductor package acts as a “space transformer”. This paragraph describes the applicability of the contact structure of the present invention in connection with interconnecting from a packaged die to parts of an external system.

Many techniques for packaging semiconductor equipment are known, including plastic molding, ceramic packaging, and PCB substrate packaging.

Generally, prior art ceramic packaging techniques include mounting a die in a cavity of a packaged body made up of one or more conductive trace (conductive line) layers sandwiched between ceramic layers. The die is connected to the inner end of the conductive trace extending into the cavity by conventional wire bonding techniques or the like. The outer end of the trace is connected in the ceramic to an external pin, lead or pad on the outer surface of the ceramic package body.

Generally, prior art plastic molding techniques include mounting a die on a paddle of a relatively rigid leadframe, where the leadframe has a patterned layer of conductive lead wires (conductive lines). . The die is connected to the inner end of the conductive leadframe lead wire by conventional wire bonding techniques or the like. The interior of the die and leadframe is encapsulated by a plastic molding composite. The outer end of the conductive lead wire extends out of the molded plastic body to connect to the external component.

In general, PCB substrate packaging techniques include mounting the die to a printed circuit board (PCB) substrate having conductive traces. The die is connected to the inner end of the conductive trace extending to an area adjacent to the die, for example by conventional wire bonding techniques. After being connected to the inner end of the trace, the die and the inner end of the trace can be encapsulated in plastic or resin. The outer end of the conductive trace is connected in the PCB to an external pin, lead or pad on the outer surface of the ceramic package body.

7A shows a ceramic package 700 having a plurality of elastic contact structures 730 mounted to an exterior (bottom) surface of the package body 702. The contact structure 730 in this example has an S shape (as shown in FIG. 2E). The package body 702 is provided with a cavity 704 extending into the package body from its upper surface (shown). Semiconductor die 706 is disposed in a cavity sealed by lid 708. The semiconductor die is connected to the end of the conductive trace (not shown) extending into the cavity 704 by the coupling wire 710. These traces are connected to conductive pads (not shown) on the bottom surface of the package body 704 by conductive vias (not shown), additional patterned layers of conductive traces (not shown), and the like. The elastic contact structure 702 is mounted one-to-one to the conductive pad. In such “cavity-up” packages, contact structures arranged in an array of rows and columns may be “populated” on the entire bottom surface of the package body. In some other figures presented herein, it will be appreciated that the elastic contact structures (eg, 730, 750) are shown simply in bold lines, and that these contact structures are overcoated.]

FIG. 7B shows an exemplary plastic semiconductor package 740 with a plurality of elastic contact structures 730a, 730b used as external interconnects. Contact structures 730a and 730b in this example have an S-shape (as shown in FIG. 2E). The state in which the conductive leadframe elements (fingers) 744a and 744b extend out of the package body 742 to the outside of the package body is shown. Elastic contact structures 730a and 730b are mounted to the exterior of each leadframe finger 744a and 744b extending out of the package body. Likewise, semiconductor die 746 may optionally be connected to the inner end of leadframe finger 744 by elastic contact structure 750 (as shown).

As shown in FIG. 7B, leadframe fingers 744a and 744b may alternately extend (eg, short-long-long, etc.) from the semiconductor package body 742 to two different distances. In this way, the effective pitch of the remote ends of the elastic contact structures 730a, 730b can be increased. Further discussion of such pitch spreading can be found in the description of FIG. 15A. In general, the ability to increase the effective pitch from the proximal end to the remote end (front end) of the contact structure allows for easier connection of electronic components to the printed circuit board. In a sense, this is true for all semiconductor packages, providing an external interconnect pitch that is wider than the spacing of the bond pads on the semiconductor die. In the prior art, the peripheral lead wire package has a pitch between the lead wires limited to about 20 mils due to restraint when soldering a component connected by lead wires to a printed circuit board. By alternately arranging the peripheral lead wires 744a and 744b as shown in Fig. 7B, a larger density can be easily achieved while maintaining these constraints.

FIG. 7C shows a PCB-type package 780 having a plurality of elastic contact structures 730 mounted to an exterior (bottom) surface of a printed circuit board (PCB) 782. The contact structure 730 in this example has an S shape (as shown in FIG. 2E). Semiconductor die 786 is mounted to the top surface (shown) of PCB 782 and is connected to a conductive trace (not shown) on top of PCB 782 by a coupling wire 790. Contact structure 730 is mounted to a conductive pad (not shown) on the bottom surface (shown) of PCB 782. Conductive pads on the bottom surface of PCB 782 are connected to conductive traces on the top surface in PCB 782 via plating through holes (not shown) and additional wiring layers (not shown) in PCB 782. A small amount of epoxy 792 is applied over die 786 and bonding wire 786 to encapsulate die 786. Alternatively, PCB 782 may be overmolded with plastic.

7D shows another package 790, where semiconductor die 791 is coupled to the inner end of leadframe finger 793 in a conventional manner by coupling wire 792. In FIG. A passivation layer 794 with openings for coupling to the die is superimposed on the die. This is quite common. According to the present invention, the full integrity of the bond (i.e. wire bond) is achieved by immersing a die bonded to the leadframe in an appropriate plating bath to plate (i.e., overcoat) the bond wire (and the inner end of the leadframe finger). The plating (e.g. nickel) serves to harden the bond wire in the manner described above and to securely bond the bond wire to the bond pads on the die (and optionally to the leadframe fingers). do. Such hardening of the bonding wires helps to minimize problems associated with wire wash (movement) and the like (e.g., shorting of wires to wires), forming an insulator 795 on the die [glob topping ( glob-topping) helps to minimize problems associated with thermal expansion (stress on joints) in subsequent processing steps. In addition, overcoating of the joining wires helps support the die when the back of the die is open as shown.

As shown in this example, the die (and leadframe) may be partially overmolded (but not completely) leaving the backside of the exposed die. This accommodates a heatsink structure (described below in connection with, for example, FIGS. 24D and 24E) for mounting a heatsink (not shown) on the back of the die or on the back of the die. It is advantageous to. In general, this is best achieved by mounting the die and leadframe on top of a support layer 797 (e.g., polyimide film) having an opening aligned with the center portion of the die back (where the edges of the die are placed on the support layer). Supported by]. In this figure, the external connection to the package has been omitted so as not to confuse the figure. Within the scope of the present invention, any suitable external connection may be employed including the elastic contact structure of the present invention.

Sacrificial elements

According to one aspect of the invention, the elastic contact structure can be formed on or from the electronic component to the sacrificial element.

The use of a sacrificial element in connection with forming an elastic contact structure is described in case-1 of FIGS. 6A-6C, similar to FIGS. 8A-8C.

8A is shaped into a U-shaped loop in a wire 802 with a proximal end 802a coupled to a first terminal 812 on a substrate 808 and does not cut the remote end 802b of the wire 802. And the remote end 802b of the wire can be coupled to the second terminal 820 by a suitable wedge-bond or the like.

The resulting looped wire stem 830 is overcoated with one or multi-layer coating 840 surrounding the entire wire stem 830 and the terminals 812, 820 as shown in FIG. 8B. The second terminal 820 serves as an electrical contact for the electroplating process (if such a process is used to overcoat the wire stem) and also differs (higher than the two ends of the wire stems 802a, 802b). ) is suitably positioned on top of the sacrificial layer serving to provide z-axis coordinates.

As shown in FIG. 8C, after overcoating the wire stem, the sacrificial layer 822 leaves a gap 824 between the end 802b and the surface of the substrate 808 (selective etching, etc.). Can be removed by The “suspension” of the end 802b may be used to pre-test or test an electronic component (such as a semiconductor die) (described in detail below) or to provide a removable electrical interconnect to the electronic component, It is particularly important to form spring contacts of controlled geometry that can elastically engage mating terminals on a component or substrate. The gap 824 allows its z-axis deflection (movement) when the tip 802b of the resulting contact structure is forced.

As shown in Figure 8C, the contact structure 830 is for contacting at a point along its length, not at its remote end 802b. This is illustrated by the downward pointing arrow labeled "C". [Similar results can be obtained for the contact structures 1640 and 1642 shown in FIG. 16C.]

FIG. 8D shows a state in which the procedure just described can be rearranged so that the sacrificial member (reference numeral 822 of FIG. 8A) can be removed before overcoating the wire stem (see, eg, FIG. 8B). .

As used herein, a “sacrifice element” such as element 822 described above is generally an element (such as a layer) on an electronic component 808 on which an elastic contact structure is formed.

Sacrificial Member and Probe Embodiment

The use of a sacrificial member in connection with forming an elastic contact structure useful as a probe is shown in Example-2 in FIGS. 14 and 15 similar to FIGS. 9A and 9B.

9A illustrates an embodiment 900 for the use of a sacrificial member 902 (shown in dashed lines) in connection with forming an elastic contact structure 930 suitable for use as a probe. In this example, the sacrificial member is suitably formed of aluminum.

A plurality of recesses 904 (one of which is shown) are formed at the top surface 902a of the sacrificial member 902 by etching, engraving, stamping, or the like. The bottom surface (shown) of the recess 904 has an irregular topography, such as in the form of an inverted pyramid ending at the vertex. A thin layer 906 of conductive material, such as gold or rhodium (alternatively, tin or solder as when contacting the solder terminal), is deposited in the recess in any known manner. The recess 904 is then substantially filled with the conductive material 908 in any known manner. Then, a layer 910 of conductive material such as gold is deposited over the fill material 918 in any known manner. Such sandwich structures of gold 906, nickel 908 and gold 910 form suitable tip structures (“contact pads”) for the probe.

The wire 912 is configured to couple to the surface of the layer 910 at its proximal end 912a and extend past the edge of the electronic component 920 where the wire diverges, and at the remote end 912b the electronic component ( 920 on terminal 922. The configuration shape of the wire is similar to that shown in FIG. 2D (ie, the wire protrudes beyond the edge of the component 920).

The wire is then overcoated by a conductive material such as nickel 914, or by the multilayer coating described in connection with FIG. 5, which also overcoats the terminal 922 on the electronic component as described above. In order to allow the overcoat to cover only the desired area on the sacrificial member, the entire surface of the sacrificial member except for the recess 904 may be masked by a suitable masking material such as a photoresist (not shown). . (This masking may be the “rest” remaining after forming and filling the recess to make the contact pad.)

As shown, the sacrificial member 902 is held in place with respect to the electronic component 920 by a suitable spacer element 916 (shown in dashed lines), which spacer element is simply a photoresist material. Can be.

After completion, the spacer element 916 and the sacrificial member 902 are removed leaving the elastic contact structure 930 extending from the electronic component 920, each of which has a controlled geometry at its end. Has a contact pad in shape. For example, the vertex of the reverse pyramid of the contact pad performs a reliable electrical connection to the terminal (pad) of another electronic component (not shown) to be detected (eg, pretest, test, etc.) in connection with the detection. Useful for doing If the total force involved is relatively small, the point (vertice) partially penetrates the terminal of the electronic component being detected. Generally, in this case, the electronic component 920 may be a test card (printed circuit board) having a plurality of probe structures 930 extending into the area for introducing the electronic component being detected. . The test card may suitably be in the form of a ring in which the probe 930 extends from the inner edge of the ring to the bottom of the ring.

The order of events described above within the scope of the present invention,

(a) First, wire 912 is coupled to terminal 922 of electronic component 920 and / or

(b) The wire 912 is rearranged to overcoat 914 after the sacrificial member 902 is removed.

9B illustrates an embodiment 940 of a completed probe 942 similar to the probe 930 of embodiment 900 described above with the following differences. In this case, the end of the probe 942 (compare 930) is wedge-coupled to the contact pad 944 having a single projecting nub 946 rather than a plurality of points, and the end of the probe 942. 948 (compared to 912b) is ball coupled to the electronic component 950 (compared to 920).

As shown in Figure 9C, a useful (e.g., good) contact tip for the probe may be formed on the sacrificial member 960 of thin aluminum (foil) in (or on) the sacrificial member in the following manner. have.

The foil is provided with a temporary backing 962 such as a plastic sheet to increase the structural integrity of the foil.

The face of the foil is patterned with a thin [approximately 3 mil] layer of photoresist 964 leaving the opening at the location where the contact tip is to be formed.

Within the opening of the photoresist, a thin (approximately 100 μ ″) layer 966 of hard gold is deposited (by plating or the like) on the foil.

A very thin (about 5-10 μ ") layer (strike) of copper 968 is deposited (by plating, etc.) on the layer of hard gold. [Copper strikes are somewhat selective and mainly It is to be understood that the gold layer 966 is provided to help later plating of the layer.]

A thick (about 2 mil) layer 970 of nickel is deposited (by plating or the like) on the copper strike.

A thin (about 100 μ ″) layer 972 of soft gold is deposited on nickel (by plating or the like).

This allows the gold wire (not shown) to be easily bonded (to the soft gold layer), the hard gold surface 966 for contacting the electronic component, the nickel layer 970 to provide strength, and the soft gold layer 972 to be easily bonded. Four-layer contact tip with () is formed. As mentioned above, after joining the wire to the sacrificial member 960, the wire is plated (eg with nickel) and the sacrificial member is removed (and vice versa).

Additional Probe Embodiments

The ability to form an elastic contact structure using a sacrificial element on a substrate (as in FIG. 8A) or a sacrificial member away from the substrate (as in FIG. 9A) has been described above generally.

As described above, the elastic contact structure described in Figures 9A and 9B is suitably employed as a probe that contacts the electronic component to pre-test, test, and exercise the functionality of the detected element.

According to the invention, the elastic contact structure can be incorporated into a chip probe card.

10A-10I illustrate a manufacturing method for making a chip probe card using a sacrificial member. Generally, openings are formed in the photoresist on the sacrificial member (eg, aluminum or copper), and optional topography in the sacrificial member in the resist opening (typically by an additional etching step or shaped tool). After photography is generated, at least one conductive layer is plated or deposited inside the opening, at which point the sacrificial member is mounted to a chip probe card (e.g., the bottom) (this mounting may be the first step performed). have). The wire is then bonded to the sacrificial member and the card and overcoated with a spring material (or any other material). Finally, the sacrificial layer is removed without adversely impacting the contacts (probes).

In the probe embodiments described herein where the contact pads typically have contact pads formed at the tip of the wire, suitable topologies for temporarily connecting to the equipment to be tested (detected), the material that coats the shaped wire may be It should be noted that the elasticity sought in the application does not need to be conductive as long as the wire itself provides the necessary conductive path between the probe card (eg) and the device to be detected.

A first phase (phase-1) of processing flow comprising processing the sacrificial member is shown in FIGS. 10A-10C.

FIG. 10A illustrates a patterned photosensitive resist 1002 of a patterned charge having an opening 1004 (one of which is shown) defined therein applied over a sacrificial material substrate 1008, such as an aluminum or copper sheet. One step is shown.

FIG. 10B shows the next step where the generally flat planar substrate material in opening 1004 is embossed to exhibit topography by ultramachining or lithography (eg, including etching) or the like. This is illustrated by an embossing tool 1010 that pushes the surface of the substrate in the opening 1004, creating a plurality of recesses 1012 extending into the surface of the substrate 2008.

10C shows a thin layer 1020 of a material suitable for contacting the surface of the substrate in the opening 1004 with other electronic components, such as rhodium, after the tool 1010 has been removed (or the surface of the substrate has been embossed by etching or the like). Plating). Thick layer 1022 of other material, such as nickel, is applied over thin layer 1020. This layer 1022 substantially fills the recess 1012 formed by the embossing process, and preferably overfills the recess 1012 as shown. Next, a thin layer 1024 of bondable (eg soft) gold is applied over the rear layer 1022 in the opening 1014. Layered structure 1026 (see FIG. 10D) in opening 1014 defines the contact pad to which the wire is to be bonded.

After performing the steps shown in FIGS. 10A-10C, the sacrificial member 1008 is ready for contact mounting.

A second phase (phase-2) of the process flow, including mounting the contact retaining substrate to the sacrificial member 1008 and manufacturing the elastic probe contact structure, is shown in FIGS. 10D-10C.

FIG. 10D shows a state in which a sacrificial member 1008 provided per phase-1 phases has spacer elements 1030 mounted to its surface in its peripheral region. The spacer element 1030 is a ring (eg, square ring) having a central zone 1032 corresponding to the zone where the (eg) semiconductor die is introduced in contact with the probe (described later) after completion of the process. It is suitable to be formed of a thick layer of photoresist (alternatively a polymer, or alternatively a metal shim) patterned to have a shape. A ring shaped substrate 1040 having a central open area aligned with and corresponding to the open area 1032 is disposed on top of the spacer element 1030. (This substrate 1040 is, for example, a multilayer PCB.) The upper surface (shown) of the ring-shaped substrate 1040 is a plurality of contact regions 1042 that are preferably defined near the opening of the substrate 1040 [ Comparison with reference numeral 110 in FIG. The substrate 1040 is illustrated as a multilayer circuit board (eg, a PCB) in which an insulating material layer and a conductive material layer are alternated, and separate electronic components (such as a test circuit device) may be mounted thereon.

Within the scope of the present invention, substrate 1040 may simply be an insert mounted on a heterogeneous separate probe card (not shown).

FIG. 10E illustrates that wire 1050 may electrically connect them between the contact areas 1042 on PCB 1040 at their ends 1050a and 1050b and the materials 1022 in the openings on the sacrificial substrate 1008 [described above]. In a similar manner to loop 802]. For example, one end 1050a of wire 1050 is ball coupled to layer 1022 in zone 1012 and the other end 1050b is wedge coupled to contact zone 1020. Wire 2050 is shaped (as described above) into a shape suitable for acting as an elastic probe contact structure (as described below after being overcoated). In essence, it does not matter that the ends of the wires 1050 are joined first or the ends of the wires are joined second.

10F shows that after wire 1050 is bonded between the substrates 1008 and 1040 (and shaped to some shape after the first bond and before the second bond), the wire 1050 completely terminates the wire 1050. Covering and covering the contact area on the substrate, and overcoated with a conductive material 1052 completely covering the area in the opening 1012 of the sacrificial substrate 1008. The method and manner of overcoating the wire with a material that gives elastic properties to the resulting contact (test) structure can be any of the techniques described above.

10G shows a state in which the sacrificial substrate 1008 can be removed by chemical etching or the like after the wire 1050 is overcoated. As shown, removing the sacrificial substrate includes not only removing the layer 1002 but also removing the spacer element 1030. From this, a plurality of contact regions 1042 (two of which are shown) are provided for elastic contact to the corresponding contact regions 1062 of the corresponding plurality (two of which are shown) on the semiconductor equipment 1060. With a plurality of probe contact structures 1054 (each of which includes overcoated 1052) extending from below the interior of the opening of the card substrate 1040, two of which many are shown. The card substrate 1040 can be obtained.

As clearly shown in FIG. 10G, each elastic contact probe 1054 has a tip formed by layers 1022 and 1020 that are “mirror images” of the embossed pattern provided by zone 1012. 1054b).

In this manner, as described in more detail below, a device under test is important for performing pretests and testing of semiconductor equipment prior to packaging the equipment or for producing fully validated equipment. An elastic and temporary connection can be made between the DUT) and the test card 1040.

10H and 10I show a second, final and end step in phase-2, wherein the sacrificial substrate 1008 [prior to overcoating (see FIG. 10I) the molded wire 1050 [FIG. And spacer element 1030 are removed (FIG. 10H).

10J and 10K show other embodiments of an elastic contact structure that functions as a probe, in accordance with the present invention. Figure 10J shows an embodiment of a probe similar to Figure 11 of Example-2, and Figure 10K shows a similar embodiment to Figure 12 of Example-2.

10J illustrates another embodiment 1070 of a probe type contact structure. As in the above embodiment, the wire stem 1072 (flexible elongate element) has one end 1072a coupled to the contact region 1076 on the substrate 1076. The other end 1072b of the wire stem 1074 is coupled (shown wedge-coupled) to the preformed contact tip 1078 on the sacrificial member 1080 in any suitable manner, and the resulting probe It is overcoated (eg with nickel material 1079) to provide the desired deflection characteristics. Appropriate spacer elements 1083 (compare 1030) are suitably employed. Ultimately, both the spacer element 1083 and the sacrificial substrate 1080 are removed.

10K illustrates another embodiment 1084 of a probe type contact structure. As in the above embodiment, the wire stem 1086 (flexible elongate element) has one end 1086a coupled to the contact region 1074 on the substrate 1076. The other end 1086b of the wire stem is coupled (shown wedge coupled) to the preformed topographically shaped contact pad 1088 on the sacrificial member 1090 in any suitable manner. Wire stem 1086 is provided with a multilayer coating in the following manner. A layer 1092 of conductive material (such as nickel) is deposited (eg, plated) over the wire stem 1082 to impart elasticity to the wire stem and secure the wire stem to the contact pad 1092. A layer 1094 of dielectric material (such as silicon dioxide) is applied over nickel layer 1092. Dielectric material 1094 may surround contact pad 1088 (not shown) to help secure the wire stem to the contact pad. Next, the wire stem can be masked by immersing the tip of the wire stem to which the contact pad is attached into a suitable masking material such as a photoresist (not shown). Finally, another layer 1096 of conductive material (such as gold) is deposited over the wire stem. This forms a coaxial (shielded) conductor, the outer layer 1096 of which can be grounded (as indicated by ground symbol “G”) in any suitable manner to control the impedance of the probe structure. Appropriate spacer elements 1093 (compare 1030) are suitably employed. Within the scope of the present invention, the wire stem may be overcoated (as described above) prior to attaching the contact pad 1088 to it (eg by brazing). In this example, the contact pad is shown topographically (compare 1026 in FIG. 10D), the characteristic of which is particularly useful for detecting, in particular for detecting solder ridges (see for example FIG. 12E).

Fabrication of contacts on the sacrificial member

In the main part, the end of the wire is joined to an electronic component, the wire is formed to be a wire stem having an elastic shape, the overcoated wire is made of an elastic material, and optionally the other end of the wire is a sacrificial element or sacrificial material. Techniques for making elastic contact structures by selectively coupling to members have been described. In this way, the elastic contact structure is mounted to the electronic component.

According to the present invention, a plurality of elastic contact structures do not mount the elastic contact structure to the electronic component, but are separately heterogeneous for mounting to an electronic component (after brazing, etc.) later (after manufacturing the elastic contact structure). It is manufactured as a structure. In other words, the supply portion of the contact structure (eg “bucket”) can be manufactured and stored for later attachment (mounting) to an electronic component. This is similar to the prior art, which manufactures a plurality of pins and later solders them to the package body.

FIG. 11A shows a first step in fabricating a separate elastic contact structure, where the photoresist 1102 of the patterned layer is applied to the surface (top shown) of the sacrificial substrate 1104. Doing. This photoresist 1102 is provided with a plurality (three of many) openings 1106a, 1106b, 1106c (eg, using conventional screening techniques).

FIG. 11B shows the next step of (optionally) generating topography in the sacrificial substrate 1104 in the resist opening using the formed (embossed) tool 1110. (The use of the tool is an alternative to the method of etching the topography on the surface of the sacrificial substrate). As shown in this figure, the topography was generated in the sacrificial substrate 1104 in the opening 1106a, in the sacrificial substrate 1104 in the opening 1106b, and in the sacrificial substrate 1104 in the opening 1106c. It should be created in the future.

11C-11E show subsequent steps of the process of manufacturing a plurality of separate elastic increasingly structures, for an exemplary one of the plurality of contact structures (in an exemplary corresponding opening of the openings 1106a); have.

11C plated a first conductive layer (e.g., of hard gold) on a sacrificial substrate 1104 in opening 1106a, and (eg, made of nickel) on top of first layer 1122. 2 Contact layer 1120 in opening 1106a (compared to reference numeral 990) by plating a conductive layer and plating a third conductive layer (e.g. of hard gold) on top of second layer 1124 It is shown forming a. (It should be understood that the three-layer contact pad shown in this set of drawings is merely exemplary and that at least one layer is required.) In general, the bottom layer 1122 should be made of a material that is easy to contact the electronic component, 1124 should be made of a material that is easy to bond to the wire stem. As described above, a four-layer structure can be formed that in turn comprises a thin layer of hard gold (for contacting an electronic component), very thin layer of copper, and a thin layer of soft gold (for bonding to a gold wire stem). It should be understood that the thin copper layer is optional.

FIG. 11D illustrates a method of joining the wire to the contact tip 1120 and shaping the wire to be the wire stem 1130 in any suitable manner described above (eg, as compared to FIG. 2A).

FIG. 11E illustrates overcoating the wire stem with an elastic (eg nickel) material 1132 in any suitable manner described above (eg, as compared to FIG. 5A).

11F shows the final step, where the sacrificial substrate 1104 and photoresist 1102 are removed by any suitable process, such as how to clean the photoresist and dissolve the sacrificial substrate. This completes the manufacture of the elastic contact structure 1130 having the contact tip 1120.

In this way, a plurality of individual elastic contact structures can be formed. For example, the contact structures formed by the techniques shown in FIGS. 11A-11F may be individually mounted one-to-one of pads on the outer surface of the semiconductor package (by brazing, soldering, epoxy adhesion, etc.). As just described above, in many applications it is desirable that the plurality of contact structures be delivered at one time (“gang transfer”) to the outer surface of the semiconductor package.

Contact bunch propagation

As described above with respect to Figures 11A-11F, a plurality of separate "unmounted" elastic contact structures may be formed for later mounting to an electronic component.

In accordance with the present invention, a plurality of elastic contact structures can be fabricated on a sacrificial substrate (eg in the manner in FIGS. 11A-11E) and then mounted (bundled) in a bunch on an electronic component. (Generally, this is accomplished by omitting the step of detaching the contact structure from the sacrificial substrate or by omitting the step shown in FIG. 11F).

According to the present invention, a plurality of preformed elastic contact structures mounted on a sacrificial carrier (sacrificial substrate) can be delivered in a single step to an electronic component, such as a fully assembled ceramic package. (The ceramic package having a plurality of elastic contact structures on its outer surface has been described above with reference to FIG. 7A as a whole.)

12A shows a package body 1252 (compare 702) with a central cavity 1254 (compare 704) on which semiconductor equipment (die) 1256 (compare 706) is mounted. A fully assembled ceramic package 1250 is shown. As shown, the die 1256 has a plurality of engagement pads (two of which are shown) disposed around the perimeter region of its front face, and the package body has a corresponding perimeter disposed around the perimeter of the cavity. It has a plurality of terminals (of which many are shown) (inner ends of the conductive traces). The bond pads of the die 1256 are connected one-to-one to the terminals of the package body by a bond wire 1260 (compare 710) in a conventional manner, and the cavity 1254 is a lid 1258 (reference 708). In comparison with). This is an example of the prior art "up cavity" packaging technique, where the entire bottom surface of the package body may be "fully-populated" in an array of conductive pads 1280, the semiconductor pad having a semiconductor Pins, ball ridges, etc. are mounted to make the connection between the package and external equipment or systems. In the case of “cavity-down” packaging, the lid occupies a central area outside the package, and the pad 1280 “partially-fills” the surface of the package body outside of the central area of the lid. populate) ”. The present invention is applicable to upstream or downstream cavity packaging techniques, and to fully or partially compacted arrangements.

Generally, in the prior art, contact structures such as pins are individually delivered to electronic components such as ceramic packages using automated equipment. In general, special molds or fixtures are needed that can typically be used only for certain individual package (and pin) arrangements. This represents tooling steps that delay product development cycles and tooling costs that are difficult to depreciate. Moreover, control over the process is incomplete, causing one or more pins to typically need to be removed or replaced by hand. This is not cost effective, as illustrated by the example of a package containing expensive chips (eg $ 1000 semiconductor equipment). Remanipulating the package to replace a defective pin can result in chip damage.

12A illustrates a plurality of elastic contact structures 1230 (compare 1130) fabricated on a sacrificial substrate (compare 1104). This corresponds to FIG. 11E described above in which the photoresist 1102 is cleaned from the sacrificial substrate 1104, but it is not necessary to clean the photoresist in this processing step. As shown in the figure, the plurality of contact structures 1230 mounted on the carrier 1204 may have any position of the package 1252 with the tip 1230b of the contact structure aligned with the pad 1280 on the package. To be mounted. This is easily accomplished by an automated part handling device. As also shown in this figure, a large amount of solder 1282 (alternatively bronze material, conductive epoxy, etc.) is deposited on each contact pad 1280 before introducing the contact structure 1230.

12B shows a state in which the contacts are carried in a batch with their tip embedded in solder 1282 on the corresponding contact pad 1280. This is accomplished by performing the above steps in the furnace (if soldering is used) (to reflow the soldering). As shown in this figure, it is not essential for the tip 1203b of the contact structure to be in physical contact with the contact pad 1280, as the solder reflows around the tip of the contact structure 1203 so that the contact structure and the contact are made. This is because a reliable electrical connection is made between the pads.

12C shows the final step, where the sacrificial substrate 1204 is removed (compare FIG. 11F).

In the manner described above, a plurality of elastic contact structures can be preformed and subsequently (later) mounted to the surface of the electronic component (eg package body). Within the scope of the present invention, forming contact structures in advance can be accomplished by, for example, molding techniques using ceramic or graphite molds. Additionally, within the scope of the present invention, the contact structures may be mounted to any electronic component or silicon die or the like, such as the interconnect substrate of the assembly embodiment described above.

The advantage of the bunch transfer technique of the present invention is that the tooling (and the costs associated therewith) is minimized, and for any given package shape, a template for a given pattern (eg, an arrangement) of the elastic contact structure (described above) Should not be confused with a molded wire stem that serves as a template as described herein]. (In general, the desired pin layout is already available in CAD software, and a simple “macro” can be used that converts the alignment points into a masking pattern to form contact areas on the carrier.)

Within the scope of the present invention, a plurality of elastic contact structures may be unreasonably delivered to an electronic component (see, for example, FIGS. 12A-12C), and then the mounted elastic contact structures (on the component) may be plated or further plated. Can be. In other words, (i) a plurality of non-overcoated wire stems may be overcoated to the electronic component and then overcoated, or (ii) a plurality of wire stems to which one or more overcoating layers have been applied to the electronic component. And then overcoated by an additional layer of material.

In addition, within the scope of the present invention, the technique of excessively transferring a plurality of contact structures to electronic components may also be applied to inelastic contact structures such as package pins. In such a case, the plurality of package pins are fabricated or mounted on a sacrificial substrate for later delivery (and typically brazing) to electronic components.

An exemplary process of forcing a plurality of contact structures to an electronic component is described with reference to FIGS. 12A-12C. Within the scope of the present invention, the process is applicable not only to the elastic contact structures of the present invention but also to "conventional" contact structures such as package pins. However, the bunch transfer process is particularly advantageous in connection with the present invention for the following reasons.

When coupling wires to terminals (eg, on a package), and when employing an electronic flame off (as in the case of constructing the elastic contact structures directly on the package pad), packaging There is a possibility of generating a spark that adversely affects the semiconductor device. In many cases, semiconductor equipment is mounted in these packages (eg, ceramic packages) before mounting the external interconnects in the package. There is a costly situation to be avoided as much as possible if there is even slight damage to the semiconductor equipment in the "late" step of this process. The herd delivery technique of the present invention has improved yield and solves this problem in a good manner.

An additional advantage involved in fabricating a plurality of contact structures on a sacrificial substrate and then forcing the structures into a package (or any electronic component) is that the spacing between the contact pads can be controlled. The contact structures may begin at a first interval (pitch) on the sacrificial member and may be formed to have a second interval at both ends thereof. Within the scope of the present invention, the contact structures may be initiated at a sparse pitch on the sacrificial member, connected to the package (or any electronic component) at a fine pitch, or vice versa.

Another advantage entailed from the above-described techniques is that ultimately leading to the package pins (ie contact structures that are unreasonably delivered to the package body or any electronic component) is not only at their spacing (described above), but also ( In the manner of manufacturing the probe contact pads, which will be described in more detail later), they can be well controlled in terms of their shape and topography. By reliably controlling the spacing and placement of the tip of the contact structure on the electronic component, the requirements for manufacturing conductive traces, pads, and the like on a substrate (e.g., a printed circuit board) on which the electronic product is to be mounted may be somewhat relaxed. .

The bunch transfer technique described herein is particularly advantageous for mounting a plurality of contact structures on a semiconductor die, whether separated or not. Within the scope of the present invention, the contact structures can be fabricated on a sacrificial substrate that is sized enough to force a plurality of contact structures to a plurality of non-separated semiconductor dies on a wafer with one fell swoop. Alternatively, it is also possible to fabricate contact structures on a sacrificial substrate that is the size of one non-separated semiconductor die, and to force these contact structures only to selected “good” dies on the wafer. In either case, contact structures can be mounted to the die on the wafer prior to dicing (separation).

In another embodiment of the flock transfer technique of the present invention shown in Figure 12D, a plurality of elastic contact structures are inserted into holes in the package, such as in a plastic package.

FIG. 12D illustrates bunch transfer of contact structures 1230 (two are shown for clarity and sacrificial substrates are omitted) into recesses 1220 of the surface of electronic component 1222, such as a plastic package body. . Recess 1220 may be filled with conductive material 1224, which may be, for example, a solder that may be reflowed, or a conductive epoxy material.

Within the scope of the present invention, the contact structures can be individually delivered (mounted) one-to-one with respect to the package without being unreasonably delivered.

Within the scope of the present invention, the herd delivery technique described above may be used to form a plurality of probe structures on a probe card.

The overall advantages of the bunch transfer technology of the present invention are evident in contrast to certain prior art techniques for mounting contact structures on a semiconductor package. These prior arts include preparing a package, manufacturing a pin, feeding pins one by one from a fin source (“bucket”), and mounting these pins to the package body by a reheating process. According to the technique of the present invention, all of these similar steps are integrated into one.

Figure 12E illustrates a probe card (substrate) 1252, preferably with a plurality of hermeticly conveyed elastic contact structures 1254 (compare with reference numeral 1230). [Sacrifice herd delivery substrate 1204 is omitted for the sake of clarity of illustration] Contact structures can be custom made according to their shape and overcoat material to allow varying levels of overtravel and elasticity. The tip of the contact structure 1254 has a topographical shape to allow for controlled penetration of the solder ridge 1256 on the electronic component 1258, the solder ridge being (eg) by a “C-4” process. Is formed.

As shown in Figure 12F, after fabricating the plurality of elastic contact structures 1230, the contact structures are further overcoated in the manner described above with respect to Figure 5G, resulting in a tapered outer layer of the coating ( 1232 is obtained, wherein the outer layer is thicker at the mounted (near) end of the contact structure to increase the fixation of the contact structure. This increased fixation characteristic is desirable with respect to the probe embodiment, where the contact structure is subjected to repeated bending.

Temporary / Permanent Method

Known processes in integrated circuit (chip) manufacturing methods include pre-testing and functional testing of the chip. These techniques are typically performed after packaging the chip.

Today's integrated circuits typically produce several, typically identical, integrated circuit dies (typically as square or rectangular die sites) on a single (usually round) semiconductor wafer, and then the dies (chips) to each other. Manufactured by scribing and slicing a wafer for separation (unification, dicing) from the wafer. An orthogonal grating in the “scribe line (cuff)” region extends between adjacent dies and sometimes includes test structures to evaluate the manufacturing process. These scribe line regions and any contained within these regions will be destroyed after the die is separated from the wafer. The separated dies are finally packaged individually, such as wire bonded connections between the bond pads on the die and the conductive traces in the package body (see, eg, FIGS. 7A-7C above).

"Pretest" is a process that simply powers a chip (die) ("static" pretest), or powers up and exercises the chip's functionality to a significant extent ("dynamic" pretest). In both cases, preliminary testing is typically performed by “temporarily” (or removably) connecting to the chip at high temperature, the purpose of which is to identify the defective chip prior to packaging the chip. Preliminary testing is typically performed one-to-one after the dies have been separated (diced) from the wafer, but it is also known to perform the pretest before separating the dies. Typically, a temporary connection to the die is made by a test probe with a “flying wire”.

In addition, a functional test can be performed by temporarily connecting to the die. In some cases, each die is provided with a built-in self test (self-start, signal generation) circuit arrangement that exercises some of the functionality of the chip. In some cases, test jigs may be made for each die, with the probe pins aligned precisely with the bond pads on the particular die requiring the test. These test jigs are relatively expensive and require excessive manufacturing time.

As a general suggestion, package leads are optimized for pre-testing (or not functional testing) assembly. Prior test substrates of the prior art are expensive and often experience thousands of cycles (ie, typically one cycle per die to be tested). Moreover, different dies require different pretest substrates. Pretest boards increase the overall manufacturing cost and are expensive, which can only be depreciated for large scale operation of a particular piece of equipment.

If the die was partially tested before the die was packaged, the die is packaged so that the packaged die can be connected to external system components. As mentioned above, packaging typically involves some kind of “permanent” connection to the die, such as by a joining wire. (Sometimes such a permanent connection may or may not be performed, but this is generally not desirable.)

As a result, the temporary connections required for the pre-test and / or pre-packaging test of the die are often different from the permanent connections required to package the die.

It is an object of the present invention to provide a technique for temporarily or permanently connecting electronic components, such as semiconductor dies, using the same interconnect structure.

It is yet another object of the present invention to provide a technique for temporarily interconnecting dies to perform pretests or tests of the dies prior to or after detaching the die from the wafer.

It is another object of the present invention to provide an improved technique for temporarily interconnecting dies, whether or not the same interconnect structure is used to permanently connect the dies.

In accordance with the present invention, elastic contact structures can perform a dual task as temporary and permanent connections to electronic components such as semiconductor dies.

According to the present invention, the elastic contact structures can be mounted directly to the semiconductor die, and the elastic contact structures can serve multiple purposes.

(a) The elastic contact structures can reliably and temporarily contact the test board, which is as simple and concise as a conventional printed circuit board.

(b) It is possible to reliably and permanently contact the circuit board when the same elastic contact structures are held in place by spring clips or the like.

(c) The same elastic contact structures can be reliably and permanently connected to the circuit board by soldering.

13A illustrates a technique for performing a "socketless" test and a pretest of a package electronic component having a plurality of elastic contact structures 1330 disposed on an outer surface of the package. Such parts may be fabricated according to the process described in connection with FIGS. 12A-12C.

To perform testing, pre-testing, etc. of the electronic component 1302, a simple (eg, easy and cheaply manufactured) printed circuit having a plurality of pads 1312 patterned on the top surface (as shown). It is held on a test card (board) 1310 which can serve as a substrate (PCB). Electronic component 1302 is aligned with card 1310 using any suitable alignment means (such as positioning pins, not shown) such that each elastic contact structure 1330 is retained on corresponding pad 1312. Can be. As a result, an elastic, “temporary” connection can be made between the card 1310 and the electronic component 1302. The card 1310 is provided with an edge connector (not shown) and the like, and optionally, a pre-test circuit (not shown) so that testing and pre-testing of the electronic component can be easily performed. Among the advantages of this technique is that no "special" sockets need to be fabricated to perform these tasks.

13B shows the same elastic contact structure 1330 as used for the socketless and pretest of the package electronic component (FIG. 13A) between the electronic component 1302 and the interconnect substrate (system board) 1320, and the like. It can be well adopted to make a permanent connection. The substrate 1320 is provided with tips of the elastic contact structure 1330 on the package component 1302 and a plurality of contact pads 1322 aligned one-to-one. The permanent connection between the component 1302 and the substrate 1320 may (i) apply a “permanent” pressure to the component 1302 using spring clips or the like (not shown) to deflect the component relative to the substrate. Or by (ii) soldering the component 1302 to the substrate 1320.

As shown, the elastic contact structures 1330 are soldered to the pads 1322 on the board 1320. This can be easily done by preparing a predetermined amount of solder (eg, solder paste) 1324 on each pad, pressurizing the part against the board, and running the assembly through the furnace to reflow (thermal) the solder. Can be.

The inventive technique of using the same elastic contact structures 1330 for both temporary and permanent connection to electronic components is an elastic contact structure mounted on active semiconductor devices (ie, bare, unpacked dies). Especially beneficial in their relationship. (In this case, the package shown in Figs. 13A and 13B, together with the elastic contact structures 1330 mounted thereon, can be simply replaced by a "bare" semiconductor device.)

It is within the scope of the present invention that tip structures such as contact tip 1120 of FIG. 12F can be used for the tips of the elastic contact structure 1330.

It is also within the scope of the present invention that the substrates 1320 and 1310 may be one and the same rather than separate entities (as shown and described).

Chip-Level Mounting Process

Mounting the elastic contact structure of the present invention directly on semiconductor dies clearly falls within the scope of the present invention. This is particularly important in light of the prior art of wire bonding to dies disposed in any kind of packages that require external connection structures (eg pins, leads, etc.). In general, semiconductor dies do not withstand significant imposition of heat, as is typically required when soldering fins to packages, since a significant amount of heat further diffuses the laid-out diffusion regions within the die. can not do it. This is of greater concern when the geometry of the device is reduced (eg, with submicron geometry). As a general suggestion, in some fabrication processes (e.g. CMOS), there is a thermal "budget" and the impact (e.g., reflow glass) of all the processing steps that the die is subjected to heat is carefully considered and calculated. Should be.

In general, the present invention provides a technique for mounting contact structures directly on semiconductor dies without significantly heating the die. In general, the adhesion of the wire stem to the die and subsequent overcoating of the wire stem (eg, plating) can be compared to a device fabrication process (eg, plasma etching, reflow glass) that applies a temperature of several hundred degrees Celsius to the die. When performed at a relatively “trivial” temperature. For example, the adhesion of gold wires is typically at 140 ° C to 175 ° C. Adhesion of aluminum wires can be done even at much lower temperatures, such as room temperature. Plating temperatures depend on the process but generally do not include temperatures above 100 ° C.

14A-14E illustrate a process of placing elastic contact structures on silicon chips (dies) prior to positioning on silicon chips or separating from semiconductor wafers. An important feature of this process is to provide an important shorting layer for overcoating the molded wire stems of the elastic contact structures by electroplating (described above). As long as electroplating involves attaching material from solution in the presence of an electric field, the electric field will not only damage sensitive semiconductor devices, but it will certainly ensure that the electric arc (such as the electrical flame-off technique of severely handling the wire as described above). It has the potential to damage semiconductor devices, and during processing, a short circuit layer provides electrical protection of such sensitive electrical components. Optionally, the shorting layer can also be grounded.

14A shows a semiconductor substrate 1402 having a plurality (only two of which are shown) adhesive pads 1404. Adhesive pads 1404 are covered with a passivation layer (usually silicon nitride) 1406 having openings on each of the adhesive pads 1404. Typically, the holes in the passivation layer 1406 cause the adhesive wire to adhere to the adhesive pad to wire bond the substrate (eg, die) to a lead frame or the like. For all intents and purposes, the openings in the passivation layer may be the size of the adhesive pad 1404 regardless of the fact that metallization of the adhesive pad may extend beyond the openings in the passivation layer 1406. It limits (area). (Usually, the adhesive pad itself simply occupies one position in the pattern of conductors in the metallization layer.) This is known in the art of semiconductor fabrication and has additional conductivity, insulation and semiconductor between the adhesive pads (upper metallization layer). Material layers are omitted for clarity. Typically, but not necessarily, the adhesive pads are all at the same level on the semiconductor substrate (apparatus) (eg, when the previous layer is planarized), and the adhesive pads are or are not in common plane or are not important for the purposes of the present invention. .

14A shows conventional processes on the entire surface of the substrate 1402 (over the passivation layer 1406 and into the openings in the passivation layer) to make electrical contacts with the adhesive pads 1404. It is shorted together by aluminum, Cu (titanium-tungsten-copper), Cr-Cu (chromium-copper), or the like applied by. A patterned resist (photoresist) layer 1412 is applied over the shorting layer 1410 and patterned to have openings 1414 aligned directly over the adhesive pad 1404. In particular, the openings 1414 of the resist layer 1412 can have any size and are preferably “virtual” (defined by the openings 1414 through the resist 1412 to the shorting layer 1410). It is desirable that the adhesive pad be larger than the openings in the passivation layer 1406 so that it has a larger area than the “actual” adhesive pad 1404. According to one aspect of the invention, the area of the virtual adhesive pad is 10%, 20%, 30%, 40%, 50%, 60, 70% more than the actual adhesive pads (as defined by the openings in the passivation layer). Much larger, such as up to 80%, 90% or 100%. Typically, the adhesive pads (and their openings) are square (as described above). However, certain types of adhesive pads are not particularly suitable for the present invention applicable to adhesive pads having a rectangular, circular or elliptical or the like.

14B shows the next step in the process of mounting the elastic contact structure on the substrate 1402. The wires 1420 are formed such that the distal end 1420a is bonded to the shorting layer of the opening 1414 and has a shape suitable for functioning as an elastic contact structure upon overcoating. In general, any of the techniques described above for forming wire stem shapes can be used at that stage. In this example, wire 1420 is formed into a wire stem having the same shape as that described in FIG. 2A.

14C shows a next step in the process of mounting elastic contact structures on a substrate 1402 where the wire stem (molded wire 1420) is overcoated with one (or more) layers 1422 of conductive material. Is shown. (In the examples above, only the top layer of the multilayer coatings are required to be conductive. In addition, any of the above-described processes and materials for overcoating molded wire stems can be used in this step. In this example, The wire 1420 is electroplated (overcoated) with nickel, as in the above example, overcoating is to determine the elasticity of the final contact structure and also greatly promote the adhesion of the contact structure to the substrate. In this example, the entire substrate is immersed in the electroplating bath and nickel is selectively plated uniquely on the wire stems and in the openings 1414 of the resist 1412 (nickel is not electroplated on the resist material). In this way, elastic contact structures 1430 are provided.

14D shows the next step in the process of mounting the elastic contact structures on the substrate 1402 where the wire stems 1420 are overcoated to form the elastic contact structures 1430. The resist layer 1412, which was evident in the last three steps, was excluded. In this step, the virtual adhesive pads are simply the contact area (comparative 110) on the continuous shorting layer 1410.

Figure 14E illustrates the final stage of the process of mounting the elastic contact structure on the substrate. In this step, the shorting layer 1410 is excluded at all locations except under the overcoat 1422. For short-circuit layers formed of a material that is easily etched (ie, without etching overcoating material 1422 or passivation material 1406), this may be achieved by selective wet etching (ie, by selecting the appropriate etchant). Can be performed. In this embodiment, the only “basic” requirement for performing selective etching is that the material of layer 1410 is different from the material of coating 1422 and one layer 1422 without dissolving the other material 1422. There is a reactant that will dissolve). This is within the understanding of those skilled in the art to which the present invention most closely belongs.

An obvious advantage of the process of the present invention is that a wider "virtual" contact area is generated than is present (ie, at the opening to the passivation layer). In addition, although the die substrate may have square (or rectangular, or circular) actual contact pads, the process of the present invention provides for openings in the resist 1412 of virtual contact pads (such as rectangular, circular, elliptical, etc.). Cause). In addition, it is only required that the virtual contact pads overlap with the actual contact pads. In other words, the center of the virtual contact pad can be biased from the center of the actual contact pad. This allows to “stagger” the tips (ends) of the elastic contacts, and another feature is required to form at least two different wires or orientations (when directly adhering to the linear arrangement of the actual contact pads). .

As described above, the process of mounting the elastic contact structures 1430 on the substrate may be performed on an already separated die or on a die (die site) before being separated from the semiconductor.

In addition, the steps described above can be performed on semiconductor dies that are not separated from the wafer. (See Figure 15 for a discussion of mounting contact structures on the dies prior to detaching the dies from the wafer.) Figures 14F and 14G are the same processes as those of Figures 14A-14E, as will be described later below. However, contact structures are applied to the dies before separating from the wafer.

14F shows a post finishing step in which elastic contact structures 1430 are mounted to a plurality of die sites (only two of which are shown) on a semiconductor wafer. An appropriate scribing or cupping tool 1450 (such as a saw) is to be held on the wafer between adjacent die sites, forming a plurality of separating dies, each die having elastic contact structures mounted thereon. .

FIG. 14G illustrates another optional post-finish step that may be performed before (ie, independently) before and after the post-finish step shown in FIG. 14F. In this step, an appropriate hermetic (eg, polymer) coating 1460 is applied to the surface of the substrate, so that not only the proximal ends 1430a of the elastic contact structure 1430 but also the entire surface of the substrate (not shown) Applied to the edges. Typically, (ie preferably) such coatings are insulating material and covering the distal end (tip) 1430b of the elastic contact structure 1430 should be avoided (as shown). If unavoidable, the insulating material 1460 covering the tip 1430 of the elastic contact structure should be excluded. Additionally, since the insulating material can change the elastic (spring) properties of the contact structure 1430 (largely) applied there by the overcoating 1422, the insulating material 1460 is used to determine the length of the elastic contact structure. Coatings other than minute (very small) parts should be strictly avoided. This step represents an important feature of the invention in that semiconductor dies, in particular aluminum adhesive pads, can be hermetically sealed from the surroundings (atmosphere). Such hermetic sealing of the die allows a less hermetic (and less expensive) package to be used. For example, ceramic packages are very hermetic (moistureproof) and very expensive. PCB board-type packages tend to be much less airtight and are comparable in price to plastic packages.

Wafer-level mounting of elastic contact structures

The foregoing discussion has generally emphasized mounting the elastic contact structures of the present invention on discrete substrates comprising semiconductor dies. The present invention has a broader scope and is particularly advantageous for mounting the elastic contact structures of the present invention on a die before separating (cutting) the die from the wafer. This provides the opportunity to perform pre-testing and testing of undivided dies before detaching from the wafer, using the elastic interconnect technology of the present invention. Mounting the contact structures on the undivided dies was briefly described with reference to FIGS. 14F and 14G.

In general, in the prior art, testing dies that are not separated at the wafer level is complicated by electrical (eg, die selection mechanisms installed in the wafer and / or dies) that tend to be complex and lead to significant increases in manufacturing costs. Or some type of die selection techniques, either mechanical or mechanical (eg, probe, flying wire, etc.). The opportunity, according to the invention, for manufacturing “final” contact structures on undivided dies and for using these structures to test and permanently connect the dies, makes these intermediate steps unnecessary and after testing It tends to be more economical compared to separation methodologies.

In addition, during fabrication of dies on a wafer, incompleteness in the wafer is often found prior to wafer processing. Some dies fabricated at such incomplete die sites must be discarded immediately (after cutting), without even “effort” to test these dies.

FIG. 15 shows a portion 1502 of a semiconductor wafer, showing a plurality of die sites 1504a,... 1504o defined by a grid of cuff (scribe) lines 1506. Elastic contact structures 1530 were mounted to adhesive pads (not shown) at each of the die sites 1504a... 1504d and 1504f .. 1504o. The resilient contact structures 1530 are not mounted to the die site 1504e (determined to be defective before mounting to the resilient contact structures). As shown in this figure, all the elastic contact structures on the die site are “oriented” such that any portion of the elastic contact structures occupies a position directly above the cuff line 1506.

After separating the dies from the wafer, the dies are covered (or encapsulated) with a suitable insulating material, and the tips of the elastic contact structures can be left exposed for subsequent interconnection to the board or card.

In general, the ability to directly fabricate elastic contact structures on semiconductor dies prior to separating the semiconductor dies from the wafer shows tremendous advantages in the overall process of fabricating semiconductor devices. This can be illustrated as follows.

In the conventional process sequence of the prior art, the dies are searched while on the wafer, then separated from the wafer, then wire bonded to the fingers of the lead frame, and then assembly of the die and lead frame The encapsulation die is inserted into the mold for this encapsulation, so that the packaged die is removed and cut from the mold (eg, “flash” and forming (parts of the lead frame fingers extending from the package body are formed into appropriate gull wing shapes) ).

In the usual process sequence of the present invention, the dies are searched while on the wafer, the elastic contact structures are mounted on “good” (passed) dies, the dies are separated from the wafer, and then coated or encapsulated. As a general suggestion, searching for dies in the manner described above is preferably limited to dies having less than one hundred adhesive pads that must be searched, such as a memory device. Nevertheless, searching dies at wafer level (prior to separation), especially for preliminary testing purposes, is generally greatly facilitated by the disclosed process.

In Figure 15, the elastic contact structures 1530 are shown as being disposed on both sides of the die, and the elastic structures on any one side of the die are all shaped identically and orientated in the same direction. This establishes a “pitch” or spacing between the tips of adjacent elastic contact structures having the same pitch as the pitch of the adhesive pads on which the elastic contact structures are mounted, as is apparent.

This illustrates the advantage that elastic contact structures suitable for direct connection to printed circuit boards and the like can be mounted directly to semiconductor (eg silicon) devices to form a “chip size package”. Such a device with directly mounted elastic contact structures is ready for testing and pre-testing, and is ready for direct interconnection to a card or board, for example, as discussed in connection with FIGS. 13A and 13B.

For the purposes of this discussion, certain semiconductor devices have a lower limit on how adhesive pads, in particular a row of adhesive pads, can be placed in close proximity, and this lower limit is the “pin-out” of the device. Assume that the pitch referred to herein is set. (It should be understood that the term “pin-out” is commonly used to describe the signal transfer of adhesive pads rather than the physical spacing of the adhesive pads. This pin-out pitch is used to amplify (deploy) the pin-out pitch. In the context of packaging dies, there is a tendency to be quite fine (small) compared to pad spacing that can be practically achieved on a printed circuit board, with special consideration given to the general acceptance of using adhesive wires, lead frames, and the like.

In general, the critical limitation in board design is that the contact (solder) pads should be spaced far enough apart so that, in some cases, the conductive traces can pass between them to validate a "complex" interconnection scheme. Also, as a general suggestion, the larger the solder pads, the "acceptable" more solder can make the more reliable solder connections.

In accordance with a feature of the present invention, elastic contact structures having various shapes and orientations may be mounted to a substrate (eg, semiconductor dies), useful for increasing the effective pitch of the device pin-out.

In addition, when mounting elastic contact structures on separate dies, it can be done in a relatively straightforward manner to shape the contacts to extend beyond the outer periphery of the die. In general, when mounting elastic contact structures to electrical components, the shape and size of the (overcoated) wire stem is not substantially limited, and relatively relatively, such as on a printed circuit board from a relatively small gap, such as on a die. Easily allows fan-out).

However, it is within the scope of the present invention that contact structures extending beyond the outer periphery of the die may be mounted on undivided dies on the wafer. This requires cutting the wafer from the opposite side, for example, as such contact structures will be placed on the cuff line.

Another advantage of the present invention is that when the wire stem is plated (overcoated), the overcoated material can be formed in areas of the electronic component not specifically intended to make an interconnect. For example, the edges of the electronic component may be plated while plating the wire stem mounted on the front side of the electronic component, or opposite sides of the electronic component may be plated while plating the wire stem. In general, any area on the unmasked electronic component may be plated. (In many of the embodiments described above, the contact area (eg, 110) to which the wire stem is adhered to the electronic component is defined by openings or the like in the photoresist.

Figure 15A illustrates an embodiment of the present invention, in which the orientation of the contact structure is shaken to increase the effective density and is the same as that of Figure 24 in Case-2. The figure shows a semiconductor die 1520 equipped with a plurality of different contact structures, in accordance with the techniques described above. The first portion 1522 of the contact structures is formed (molded, curved) to have a relatively large deviation (ie, distal end from proximal end). The second portion 1524 of the contact structures is formed (molded, curved) to have a relatively small excursion (ie, a distal end from the proximal end). In this manner, as shown, the spacing between proximal ends of adjacent contact structures 1522 and 1524 is “m” and the spacing between distal ends of adjacent contact structures is “n”. As shown in the figure, linear contact structures 1528 extending perpendicular to the surface of the electronic component 1520 may be formed on the electronic component. These contact structures 1528 are intended to function as alignment pins that engage with corresponding alignment elements (such as holes) on other electronic components, such as a printed circuit board (PCB). Preferably, these alignment pins 1528 are not elastic, but can certainly be fabricated in the same process steps as the elastic contact structures 1522, 1524.

Optionally, a capsule is disposed on the surface of the substrate to enclose the lower portions (shown) of the contact structures and mechanically reinforce the attachment of the elastic contact structures to the surface of the substrate.

The swinging of the tips of the contact structures, according to the invention, allows the designer to relax the "ground rule" (design principle) of the board on which the electronic component is to be mounted, the contact pads being different and / or Or from each larger solder pad.

In use, temporary connections may be made to the electronic component 1520 via the contact structures 1522, 1524, 1526, and subsequent permanent connections may be made in the same manner as described above with reference to FIGS. 13A and 13B. The electronic component 1520 may be made via the 1522, 1524, and 1526. This preferably facilitates wafer-level pre-testing of dies not separated on the wafer, the feature of which is particularly good for (but not limited to) semiconductor memory devices. It is within the scope of the present invention that the contact structures 1522, 1524, 1526, 1528 are collectively moved to the wafer (or chip) 1520 in the manner described above. The collective movement technique generally eliminates the need to form a shorting layer (comparative 126) on the electronic component since the contact structures are fabricated “off-line” (ie on a sacrificial substrate).

Short layer not required

In many of the embodiments described above, the use of a shorting layer has been described (see, for example, conductive layer 126 in FIGS. 1C-1E). The shorting layer is particularly suitable for overcoating wire stems by an electroplating process. The use of a conductive sacrificial structure to which all wire stems are connected also facilitates electroplating by simply shorting (electrically connecting together) a plurality of wire stems as well.

16A shows a first step in a process, in which a sacrificial structure 1602 is used in connection with shaping and overcoating a plurality of wire stems 1630 and 1632 mounted (bonded) to a semiconductor die 1612 and FIG. have.

The sacrificial structure 1602 is formed as a cage-like structure from a conductive (and easily removed at the end of the process) material, such as aluminum, and has an outer ring 1604 defining the area in which the die 1612 is disposed, and of the ring A crossbar 1606 spanning from one side (not shown) to the opposite side of the ring 1604 (not shown in cross-sectional perspective view). This causes their openings to run parallel from one side of the ring to the opposite side of the ring, parallel to (and with each other) crossbar 1606.

In general, the sacrificial structure (cage) is disposed above the semiconductor die 1612 so that the openings 1608, 1610, respectively, on the die 1612 before mounting the wire stems 1630, 1632 to the die 1612. Are aligned with the adhesive pads in parallel rows of.

As shown, the wire stems of each row of adhesive pads along each side of the die alternately extend to the outer ring 1604 and the inner crossbar 1606 and are bonded to the sacrificial structure by wedge bonding their distal ends. In this way, the sacrificial structure 1602 shorts all wire stems together and can be easily connected (not shown) for subsequent plating of the wire stems.

16B shows the next step in the process, which is plated, in the manner described above, such that the wire stems 1630 and 1632 function as the elastic contact structures 1640 and 1652, respectively.

In the next step, it is necessary to separate (remove) the sacrificial structure, and in general there are two possibilities: (i) distal ends of the elastic contact structures can be cut from the sacrificial structure, or (ii) the sacrificial structure is elastic It may be dissolved (eg, etched) without cutting the tips of the contact structure.

16C shows that sacrificial structure 1602 is melted, leaving a die on which elastic contact structures 1640 and 1642 are mounted. In most of the previous embodiments, the elastic contact structures generally make contact with other parts, whereas in this embodiment, the elastic contact structures 1640 and 1642 are each different parts (not shown). Is molded to make a contact (as indicated by arrow “C”).

In general, by changing the orientation of the contact structure 1640 (facing the inside of the die) and the contact structure 1644 (facing the outside of the die), the effective pitch of the contact structures is greater than the pin-out pitch of the die. May be larger (see Figure 15A). Facing outwardly facing structure 1641, the tips 1644b deviate from the edge of die 1612 and do not cause such an apparent problem (ie, the tip of the contact structure depends on the surface of the die, depending on the contact force). Contact).

16A-16C, die 1612 is shown having a passivation layer 1614 on the top surface (shown) in the manner described above.

16D shows a sequence of other cases where the sacrificial structure 1602 is removed before overcoating the wire stems 1630 and 1632. The first step, described with respect to Figure 16A, remains the same and the final structure is as shown in Figure 16C.

The technique described above with reference to FIGS. 16A-16C provides for a thinner sacrificial structure 1602 that simply lies above the wafer (rather than extending below the side edges of each die, as shown in FIGS. 16A-16C). Can be performed at the wafer level by simply providing

It is within the scope of the present invention that the electronic component 1612 is “free” from the sacrificial structure 1602 by simply cutting the contact structures (eg, FIG. 16) or wire stems (eg, FIG. 16D).

A general advantage of using a sacrificial structure (eg, 1602) is that it does not require any flame-off of the electronic component and the electronic component 1612 is subjected to very high and potentially damaging voltages (eg 2000 volts). .

In addition, the contact structures (or wire stems) can be stabilized like hard wax material and subjected to polishing (polishing) in a plane parallel to the plane of the electronic component such that the contact portion (eg, 1642C) is (eg, contact structure). Or by fully grinding through the wire stem) to the free end of the contact structure. This will be described later by way of example with reference to Figs. 53C and 53D.

16E and 16F illustrate a technique 1650 of fabricating an elastic contact structure in a manner suitable for stacking chips (semiconductor dies), one by one. The sacrificial structure 1652 (see 1602) is disposed above the first electronic component 1662 (see 1612). The wire 1658 is configured such that one end 1658a is adhered to a pad on the first electronic component 1662, and has an elastically deformable form (in the same manner as in FIG. 16A), and an intermediate portion of the wire 1658. 1658c is bonded to the sacrificial structure 1652 (without cutting). As shown, the sacrificial structure 1652 may be provided with a contact tip 1654 (see 1026 in FIG. 10C) to which the middle portion of the wire is bonded. The wire is further shaped to extend from the sacrificial structure 1652 in an elastically deformable form (eg, see S-shape in FIG. 2E) and cut to have a free end 1658b. The molded wire stem may be plated to be an elastic contact structure, before or after removing the sacrificial structure 1652 (see FIG. 16B) or (FIG. 16D), and the topographical contact applied to its free end 1658b. (See 1026).

After the sacrificial structure 1652 is removed, a second electronic component 1672 is disposed between the first electronic component 1662 and the intermediate portion 1658c of the elastic contact structures (overcoated wire stems), thereby removing the first electronic component 1662. It results in an interconnection between the terminals 1674 of the first electronic component 1662 and the second electronic component 1672. The advantage of this technique is that the interconnect (overcoated portion of the wire stem between 1658c and 1658b) also extends from the second electronic component 1672 for connection with an external system (other electronic components). For example, the first electronic component 1662 is a microprocessor, and the second electronic component 1672 is a memory device.

Interposer

Figures 17A-17D of the present application are similar to Figures 6, 7, 13, and 22 of Case-2, respectively, and show contact structures mounted on a printed circuit board (PCB) for use as interposers. . As shown herein, an “interposer” is a generally flat substrate having contact structures disposed on both sides thereof, the contacts on one side of which are in contact with the other sides in the interposer. Is electrically connected. Such interposers are disposed between two electronic components to be interconnected. In general, it is used when it is not desirable to mount contact structures directly to interconnected electronic components.

17A shows a plurality (only one of which) fashioned wire stems 1702 are disposed on the top surface 1704a (shown) of the printed circuit board 1704, and a plurality (only one of which is shown). Illustrated an embodiment 1700 of an interposer with fashioned wire stems 1706 disposed on the bottom surface 1704b (shown) of the printed circuit board 1704. In particular, the printed circuit board 1704 is provided with a plurality (only one of which) plated through holes 1708 having one or more layers of conductive material 1710 fabricated in a conventional manner. The wire stems 1702 are mounted to the plating 1710 (bonded in any conventional manner as described above) to protrude upwardly (shown) from the printed circuit board 1704, and the wire stems 1706 are printed circuit boards. It is mounted to the plating 1710 so as to project downward (shown) from the substrate 1704 (adhered in any conventional manner as described above). Wire stems 1702 and 1706 are overcoated (in any conventional manner described above) with one or more (only one shown for illustration purposes) layers 1712 of conductive material that give elasticity to the wire stems. In addition, the coating 1712 firmly secures the wire stems to the through hole plating 1710 and covers the entire surface of the through hole plating 1710 as shown. The electrical path is provided by overcoat 1712 from the contact structures 1702 to the contact structures 1706. In this manner, elastic contact structures are formed on both sides of the interposer substrate and a terminal, pad on a first electronic component (not shown) with an upper “set” of contact structures 1702 disposed above the interposer. Make contacts with the back, and make contacts with terminals, pads, etc. on the second electronic component (not shown) with a lower "set" of contact structures 1706 disposed below the interposer-both sets of contact structures are flexible (elastic) To make electrical connections to the individual electronic components.

Thus, in FIG. 17A, one can easily observe that the interposer is provided with elastic contact structures on both sides of the “circuited” (metalized) electrical substrate. For this embodiment 1700, any suitable means of connecting the upper springs 1702 to the lower springs 1706, which may or may not include the through hole 1708, meets the intended purpose of the interposer. Let's go. The contact structures 1702 and 1706 of this embodiment may be elastic (pure spring) or flexible (showing a combination of elasticity and plasticity).

When a flexible (or elastic) electrical connection is required only on one side of an interposer, an interposer embodiment 1720 as shown in FIG. 17B can be fabricated and used. In this embodiment, a plurality (only one of them) fashioned wire stems 1722 are disposed on the top surface 1724a (shown) of the printed circuit board 1724. When covered with a conductive elastic sheath 1732, these wire stems 1722 form elastic contact structures for making contacts, pads, and the like (not shown) disposed above the interposer.

As in the previous embodiment 1700, the printed circuit board 1724 is provided with one or more layers of conductive material 1730 fabricated in a conventional manner. The wire stems 1722 are mounted to the plating 1730 so as to protrude upwardly (shown) from the printed circuit board 1724 (bonded in any suitable manner as described above).

In this embodiment 1720, inelastic contact structures 1706 adhere one end 1726a of wire 1726 to a location on through plated portion 1730, and the other end 1726b of wire 1726. The other, preferably on through-hole plating 1730 in a manner similar to forming a traditional wire contact loop (the wire has two ends bonded between the two points and has a fixation shape between the two points). Is formed on the bottom surface 1724b (shown) of the printed circuit board 1724 by adhering to the radially opposite position.

The wire stems 1722 and 1726 then firmly secure the wire stems 1722 and 1726 to the through hole plating 1730 and, as shown, the entire surface of the through hole plating 1730. Overcoated (in any suitable manner described above) with one or more layers 1732 of conductive material covering (only one shown for purposes of illustration). In addition, overcoating gives elasticity to the wire stems. The electrical passage is provided by overcoating from the contact structures 1732 to the contact structure 1736. In this manner, elastic contact structures are formed on both sides of the interposer substrate, and a first electronic component (not shown) in which an upper "set" of contact structures 1722 is disposed above (shown) of the interposer. Contacts, terminals, pads, etc. are made on the second electronic component (not shown) with a lower “set” of contact structures 1726 disposed below the interposer. In this way, only the upper set of contact structures 1722 makes elastic contacts (on the first electrical component). For this embodiment 1720, the contact structures 1726 can be selected as freestanding, pinned contact structures (as described above) because it is expected that no elastic connection between the interposer and the second electrical component is required. have.

Clearly, for this embodiment 1720, only the top contact structures 1722 are elastic (or flexible), and the bottom contact structures 1726 are rigid and soldered to other components disposed below the interposer 1720. It is suitable for making.

17C illustrates another embodiment 1740 of an interposer. As in the foregoing embodiment, a printed circuit board 1744 having a top surface 1744a (shown) and a bottom surface 1744b (shown) is plated with layers of one or more conductive materials 1750. Through holes 1748 of (only one of them are shown) are provided. This embodiment inserts a capillary (not shown) through the through hole 1748, adheres the free end 1742a of the wire 1742 to the sacrificial member 1743, and connects the capillary through the through hole 1748. After moving upward, it is appropriately fabricated by fashioning the elastic wire stem 1746 protruding upward from the top surface 1744a of the printed circuit board 1744. When overcoated with conductive material 1552 (in any suitable manner described above), the elastic contact structures protrude upward (shown) from the top surface of the printed circuit board 1744 and the bottom surface 1744b of the printed circuit board. Elastic contact structures are formed that protrude downwards (shown). Clearly, the diameter of the through hole 1748 must be larger than the diameter of the capillary tube to enable this technique. Capillaries that feed wires having a diameter of approximately 0.001 inches are known to have diameters as small as approximately 0.006 inches. Therefore, a through hole having a diameter of approximately 0.010 inches will provide a large gap for the capillaries through it. (Alternatively, as described above, the sacrificial member can be removed before overcoating the wire stems.)

Clearly, for this embodiment 1740, the plated through holes providing conductive regions on both sides of the substrate are not shown, but only one side of the interposer substrate 1744 needs to have conductive traces. In addition, the wire stem portion 1742 does not need to extend through the through hole. The same scheme can be easily accomplished by bonding the stem portions 1742, 1746 to a conductive pad near the edge of the interposer substrate 1744.

For the interposer embodiments described above, it is apparent that the "top set" of contact structures may be different from the "bottom set" of contact structures, and one or two sets are elastic contact structures. In general, it is desirable that all of the contact structures of a given (ie, top or bottom) set be identical to one another. However, they differ from one another within the scope of the present invention.

The elastic contact structures formed (fashioned and overcoated) according to the invention are flexible. Depending on the intended use, it is desirable to limit their compliance. For this purpose, "stops" can be easily manufactured to limit the compliance (spring deflection) of the elastic contacts.

17D shows an embodiment 1760 of an interposer incorporating a "stop". This embodiment incorporates the materials of the interposer embodiment 1700 of FIG. 17A as indicated by the dotted lines. In this embodiment, in addition to fashioning the wire stems 1702 and 1706, a compliance limiting stop (standoff) 1779 is a simple loop on the flat through hole 1778 of the printed circuit board 1764 (see FIG. 17B). (See loop 1726). As mentioned above, such a loop 1779 becomes elastic when overcoated. To limit the compliance of the resilient contacts 1702 on the top surface of the printed circuit board 1764, the loop extends to a position labeled "B", which is the vertical range of the resilient contact structure 1702 (label "A"). Down) (as shown), at any desired distance, such as 0.003 inches. In this manner, the deflection of the elastic contact structure 1702 (which results from the force exerted by the electronic component held on the interposer) can be controlled (ie, limited).

A notable feature of the embodiment 1760 of FIG. 17D can be fabricated in the same process as the elastic contact structures 1702 and 1706.

17E illustrates another embodiment 1780 of an interposer. In this case, interposer substrate 1762 has a plurality of cascaded plated through holes 1784 extending from its (shown) top surface to its bottom surface (shown). In particular, the substrate 1762 is preferably a thermoplastic substrate in which excursion holes are molded, as shown in the figure. When metallized, when metallized, it forms a "step" (or "platus") in the body of the substrate suitable for mounting the contact structures. A plurality of (only two of them shown) contact structures 1786 are mounted in the through holes 1784 to extend beyond the (shown) top surface of the substrate 1762 and a plurality of (only two of them shown) elastic contacts The structures 1788 are mounted in the through holes to extend beyond the bottom surface (shown) of the substrate 1762. In this way, the deflection of the elastic contact structures 1786, 1788 according to the compressive force is “stop” (Fig. Inherently limited by the substrate 1742, which acts as the above-described deflection limiting stop 1779 of 17D. In this manner, the sufficiently deflected elastic contact structures 1786 and 1788 are fully embedded in the through holes. It is within the scope of the present invention that the holes are filled with a soft conductive elastic mass in the manner described above with respect to FIG. 5E for the same reason.

Additional Interposer Embodiments

The use of the present invention in the context of an interposer has been briefly described with reference to Figures 17A-17E.

The following are within the scope of the invention:

(a) the same elastic contact structures may be formed on both sides of the interposer;

(b) other elastic contact structures may be formed on each side of the interposer;

(c) the spacing between adjacent elastic contact structures may vary depending on the side of the interposer;

(d) one or more elastic contact structures may be formed at each “point” on a given side of the interposer;

(e) Standoff and / or alignment and / or spring clip elements may be formed in connection with the elastic contact structures (see, eg, element 1528 of FIG. 15A, and element 1780 of FIG. 17D).

18A shows an interposer 1800 suitable for placement between the spaced opposing surfaces of two electronic components (1840 and 1850, shown in FIG. 18B), such as a semiconductor die 1840 and a printed circuit board 1850. An embodiment is shown.

Interposer 1800 is made of a substrate 1802, such as a laminated printed circuit board, through which a plurality of (only two of which are shown) holes 1804, 1805 extend. The substrate 1802 is covered with a conductor 1806, such as copper, covering both sides and extending through the through holes 1804, 1805. This is known.

A resist layer 1812 is applied to the top side (shown) of the PCB 1802, extending almost to the through holes 1804. The same resist layer 1814 is applied to the bottom side (shown) of the PCB 1802, extending almost to the through holes 1804, 1805. The through holes 1830 are plated with a nickel layer 1808, which is overcoated with the gold layer 1810. These layers 1808 and 1810 cover copper 1806 in the through holes 1804 and 1805, and holes 1804 and 1805 onto some of the surfaces of the PCB 1802 that are not covered with resist 1812 and 1814. Extends past the edges of the. This provides plated through holes 1804 and 1805 suitable for mounting the elastic contact structures in accordance with the present invention. PCBs with cladding 1806 and through holes 1804 and 1805 coated with multilayer (eg, nickel 1808, gold 1810) are known and commercially available.

The first wire stem 1820 is bonded to the top surface (shown) of the PCB 1802 at the through hole 1804 and, as described above, into a skeleton (a form suitable for functioning as a spring, at the time of covering). It becomes fashion. The second wire stem 1821 is adhered to the top surface (shown) of the PCB at the through hole 1804, preferably at the circumferential position of the through hole 1804 near the wire stem 1820, as described above. As described above, it is fashioned into a skeleton (a form suitable for functioning as a spring at the time of coating). The wire stems 1820 and 1821 extend upwards (shown) in the same direction to each other, towards the first terminal of a first electronic component (not shown) disposed above (shown) of the interposer 1800. do.

The third wire stem 1822 is bonded to the bottom surface (shown) of the PCB 1802 at the through hole 1804 and, as described above, into a skeleton (a form suitable for functioning as a spring, once coated). It becomes fashion. The fourth wire stem 1823 is bonded to the bottom surface (shown) of the PCB 1802 at the through hole 1804 at the circumferential position of the through hole 1804 near the wire stem 1822, and described above. As described above, it is fashioned into a skeleton (a form suitable for functioning as a spring at the time of coating). The wire stems 1822, 1823 extend downwards (shown) in the same direction to each other toward the first terminal of a second electronic component (not shown) disposed below (shown) of the interposer 1800. do.

The fifth wire stem 1824 is bonded to the top surface (shown) of the PCB 1802 at the through hole 1805 and, as described above, into a skeleton (a form suitable for functioning as a spring, once covered). It becomes fashion. Sixth wire stem 1825 is preferably bonded to the top surface (shown) of the PCB at through hole 1805 at the circumferential position of through hole 1804 near wire stem 1824 and described above. As described above, it is fashioned into a skeleton (a form suitable for functioning as a spring at the time of coating). The wire stems 1824 and 1825 extend upwards (shown) in the same direction to each other, toward the second terminal of the first electronic component (not shown) disposed above (shown) of the interposer 1800. do.

The seventh wire stem 1826 is bonded to the bottom surface (shown) of the PCB 1802 at the through hole 1805 and, as described above, into a skeleton (a form suitable for functioning as a spring, once coated). It becomes fashion. The eighth wire stem 1827 is bonded to the bottom surface (shown) of the PCB 1802 at the through hole 1805 at the circumferential position of the through hole 1805 near the wire stem 1826, and described above. As described above, it is fashioned into a skeleton (a form suitable for functioning as a spring at the time of coating). Wire stems 1826 and 1827 extend downwards (shown) in the same direction to each other toward the second terminal of a second electronic component (not shown) disposed below (shown) of interposer 1800. do.

Once the wire stems 1820,..., 1827 are mounted to the substrate 1802 and fashioned in an elastic form, the wire stems are overcoated with a spring material 1830, such as nickel, in the manner described above. Overcoating 1830 covers the fully exposed (not covered with resist 1812 or 1814) of through holes 1804 and 1805 to provide firm fixation to the wire stems and provide elasticity to the wire stems Wrap the wire stems to During the coating 1830 process, the copper layer 1806 acts as a “short” layer (shorting all through holes together, thus shorting both wire stems together) to facilitate electroplating.

Once the wire stems 1820,..., 1827 are overcoated, the resists 1812, 1814 are removed and the copper cladding 1806 between the through holes 1804, 1805 is removed by selective etching (ie, nickel). For overcoating 1830, chemically) is removed to the copper cladding 1806.

This is a pair of upwardly extending elastic contact structures (shown) for connecting to the first electrical component 1840 and extending downwards (shown) for connecting to the second electrical component 1850 (shown). It has pairs of elastic contact structures, resulting in an interposer making a one-to-one (pad to pad) connection between two electrical components.

As is apparent in this embodiment, the wire stems are provided in pairs (eg, pairs 1820 and 1821, pairs 1822 and 1823, pairs 1824 and 1825, and pairs 1826 and 1827). In general, the stems in each pair of stems (eg, stems 1820 and 1821) extend at the same height as the surface on which they are mounted (although shown in a slightly perspective view). A connection is made to a corresponding terminal (or pad, or conductive region) A pair of contact structures 1820/1821 is connected to a terminal 1842 (shown in dashed lines) on the first electronic component 1840 (shown in dashed lines). A pair of contact structures 1822/1823 connects to a terminal 1852 (shown in dashed line) on the second electronic component 1850 (shown in dashed line) and connects to the interposer 1800. It is connected to the pair 1820/1821 by a through hole 1804. The pair of contact structures 1824/1825 are connected to a terminal (shown in dashed line) on the first electronic component 1840 (shown in dashed line). And a pair of contact structures 1826/1827 (shown in dashed lines) on the second electronic component 1850 (shown in dashed lines). Connected to the) terminal (1854), and by a through hole 1805 in the interposer 1800 are connected to the pair (1824/1825).

The advantage of making a connection with pairs of elastic contact structures (eg, 1820, 1821) is a reduction in inductance (ie, using only one elastic contact structure per connection).

Another advantage of making a connection with a pair of elastic contact structures (eg, 1820, 1821) is that it provides a margin simply, under the general premise that two contacts are better than one.

The interposer embodiment 1800 of FIGS. 18A and 18B is shown having the same contact structures at each interconnection point and on both sides of the interposer. The contact structures are preferably all the same on a given side of the interposer (eg, wire stems 1820, 1821, 1824, and 1825 are generally fashioned in the same height and shape), but the contact structures on the other side of the interposer (Eg, wire stems 1822, 1823, 1826, and 1827) may be fashioned to a different shape and height than the contact structures on opposite sides of the interposer.

In the description of the interposers below, the plated through-holes are shown as having one layer for clarity of illustration, but a multilayer through-hole may be suitable as described above. Also, for the sake of clarity, only one of the plurality of contact structures formed on the interposer is shown.

FIG. 19A shows an interposer 1900 suitable for being disposed between the spaced apart opposing faces 1940 and 1950 (shown in FIG. 19B) of two electronic components, such as semiconductor die 1840 and printed circuit board 1850. FIG. One embodiment is shown.

Interposer 1900 is made of a substrate 1902, such as a fibrious printed circuit board, having a plurality of through-extending holes 1904 (only one of which is shown). Substrate 1902 has a conductor 1906 plated in a known manner so that the through hole 1804 is plated through the hole in a patterned manner. (The steps involving the filling of the resist to perform patterning for clarity are omitted.)

A removable member 1970 is disposed below the bottom surface 1902b of the substrate and is attached to the substrate with a suitable removable temporary adhesive 1972, such as a conventional photoresist. The removable member 1970 is provided with a recess 1974 to form a topological (topographical) contact tip 1976 on the wire stem, as in the manner described above with respect to FIGS. 10D and 11D. do.

The free end 1920a of the first wire stem 1920 is coupled to the through-hole conductor 1906 on one face 1902a of the interposer substrate 1902 and extends upwardly above the face 1902a in a spring shape. Molded.

The free end 1922a of the second wire stem 1922 is coupled to the topological tip 1976. This may be accomplished by moving the capillary down through the aperture 1904. Typically, the diameter of the capillary tube is about 0.15 mm (0.006 inch) and the diameter of the hole is 0.25 mm (0.010 inch) or more, so that the capillary tube can be traversed into the hole. Thereafter, the wire stem 1922 extends upwardly through the through hole 1904 to have a spring shape, and preferably adjacent to the engaging end 1920a of the first wire stem 1920, the interposer substrate ( The second end 1922b is coupled to the through-hole conductor 1906 on one side 1902a of 1902.

The wire stems 1920, 1922 are then overcoated with a nickel layer 1930, for example, to form a resilient contact structure (and the joined ends of the wire stems are fixed).

Finally, as shown in FIG. 19B, the removable member 1970 is removed (eg, by dissolving the removable member and the adhesive 1972 with a suitable solvent). In this manner, terminals 1942 (shown in hidden lines) on the first electronic component 1940 (shown in hidden lines) and terminals (shown in hidden lines) on the second electronic component 1950 (shown in hidden lines) An interconnect is made between (1952).

As in the case of the above-described embodiment using a removable substrate, the removable member 1970 functions as a shorting member to facilitate electroplating of the coating 1930 on the wire stem. Also, as in the embodiment described above, the removable member 1970 may be removed before, but not after, the overcoat 1930 as shown.

For example, interposer 1900 may be disposed between a package (electronic component 1940) and a printed circuit board (electronic component 1950), in which case the contact tip 1976 (on the printed circuit board). It has the advantage that it is connected to a tin or solder layer to connect to the pad and the tip of the stem 1920 is connected by soldering (to connect the pad on the package).

As described above with respect to interposer 1800 of FIGS. 18A and 18B, in interposer 1900 of FIGS. 19A and 19B, contact structure 1920 on one side 1902a of interposer 1900 It may be made to have a different shape and different height than the contact structure 1922 on the other side 1920b of the interposer 1900. Further, the tip shape of contact 1920 on one side 1902a may be different from the tip shape of contact 1922 on other side 1902b. In particular, the material of the wire stem 1920 on one side 1902a of the interposer 1900 is the same as the material of the wire stem 1922 on the other side 1902b of the interposer 1900, although the same material is good. May be different. In general, it is also within the scope of the present invention that the overcoating 1930 is different for the wire stems on each side of the interposer, but if the overcoating is the same (material and process) for the contact structure on both sides of the interposer, Since discontinuity of the is prevented, overcoating is the same (material and process) for the contact structure on both sides of the interposer. Such “variations on the theme” are generally applicable to all interposer embodiments described herein.

Due to the fact that the electrical connection between the two parts 1940, 1950 does not depend on the through hole, the inter-hole, not the plated through hole 1904 shown (i.e., shown in the embodiment of Figures 18A and 18B) is not interposed. It is within the scope of the present invention that the pads are provided only on the top surface 1902 of the poser substrate 1902.

In addition, the end 1922a of the wire stem 1922 is coupled to the removable member 1970 and the capillary tube is extended upward through the through hole 1904 (at the same position as shown by reference numeral 1922b in FIG. 19A). Two wires, which are easily visualized by coupling the middle of the wire 1922 to the terminal, and then continuing to move the capillary to form the wire stem 1920 as a “continuation” of the wire stem 1922. It is within the scope of the present invention that the stems 1920, 1922 are formed of one wire stem. In this manner, one wire connection directly between the terminal 1942 of the first electronic component 1940 and the terminal 1952 of the second electronic component 1950 without depending on the electrical conductivity of the through-hole material 1906. This will be done.

When attempting to form wire stems 1920 and 1922 with two separate wire stems (as shown), the lower wire stems (extending through through holes 1904) may be formed prior to forming the upper wire stems 1920. It is preferred to form 1922 first. Otherwise, wire stem 1920 will be “in the way” during lower wire stem 1922 formation. In this way, it will generally be appreciated that all embodiments described in this application can change the order of events for a particular application as appropriate.

Figure 20A illustrates an interposer similar to the embodiment 1900 of Figure 19A, except that one wire is made to interconnect between two electronic components 2040 and 2050 (shown in Figure 20B). 2000) is shown. In addition, as with the interposer embodiment, the interposer 1900 is disposed between the spaced and opposing faces of two electronic components 1940 and 1950, such as the semiconductor die 1840 and the printed circuit board 1850. However, it is not limited to these two specific electronic components.

The interposer 2000 is made of a substrate 2002 such as a printed circuit board having a plurality of holes 2004 extending therethrough (only one of which is shown). As suggested in the description of the embodiment 1900 described above, a plurality of pads 2006 (shown here as one layer) disposed on the top surface 2002a (shown) are shown as a single layer, but multilayer pads such as gold over nickel on the copper Only).

This embodiment represents a feature that may be mounted to other interposer embodiments described herein, that is, a ground or power face 2060 may be installed in the printed circuit board 2002. In particular, the upper surface of the substrate 2002 (shown) is provided with a signal surface on the ground or power surface 2062 (which is the power surface when the surface 2060 is a ground plane) and the lower surface 2002b (shown), The opposite may be true. It is also within the scope of the present invention that substrates such as 2002 may be provided with resistors, capacitors, capacitive layers, planar resistors, and the like to enhance functionality.

In the same manner as the embodiment 1900 described above, a removable substrate 2070 is disposed below the bottom surface 1902b of the substrate and is typically attached to the substrate with a suitable removable temporary adhesive 2072, such as a photoresist. . Removable substrate 1970 is provided with recess 2074 to form a topological tip 2076 on the wire stem, as in the manner described above with respect to FIGS. 10D and 11D.

The free end 2020a of the wire 2020 is coupled to the conductive pad 2006 on the surface 2002a of the substrate 2002 and the removable substrate 2070 to which the middle portion 2020c of the wire is coupled to the tip 2076. ) Is formed into a spring shape extending downward through the through-hole (2004). These wires extending from the pads 2006 to the tips 2076 constitute the first portion 2022 of the molded wire stem and have a suitable shape to form a first conductive contact when overcoated in the manner described above.

After joining the middle portion 2020c of the wire 2020 to the tip 2076, the wire 2020 is further shaped by moving the capillary upwards again through the hole, thus the upper surface (shown) of the substrate 2002 (2002a) forms a second portion 2024 of the wire stem suitable for forming a second conductive contact when connected over and overcoated in the manner described above.

Wire stem 2020 (ie, both portions 2022, 2024 of wire stem) is then overcoated with nickel layer 2030, for example, over top surface 2002a (shown) of substrate 2002. An elastic contact structure is formed having an extended end and an end extending below the bottom surface 2002b (shown) of the substrate 2002.

Finally, as shown in Figure 20B, the removable substrate 1970 is removed by dissolving the removable substrate and adhesive (e.g. photoresist) 2072 with a suitable solvent, for example. In this manner, between the terminal 2042 (shown in hidden line) on the first electronic component 2040 (shown in hidden line) and the terminal 2052 (shown in hidden line) between the second electronic component 2050 (shown in hidden line) An interconnect can be made.

As in the case of the above-described embodiment, which generally uses a removable member, the removable member 2070 functions as a shorting member that facilitates electroplating of the coating 2030 on the wire stem portions 2022, 2024. Also, as in the embodiment described above, the removable member 2070 may be removed before, but not after, overcoat 2030 (as shown).

20A and 20B, in this embodiment 2000, a spacer 2080 is preferably provided on the top surface 2002a (shown) of the substrate. The first portion 2024 of the wire stem extends over the top surface (shown) of the spacer 2080 by, for example, 0.13 mm (0.005 inch). As in other embodiments described herein, the wire stem protrudes above about 0.13 mm (0.005 inch) other structures (e.g., spacer 2080) and is about 0.076 cm (0.003 inch) in response to a "normal" loading force. It is usually sufficient to be designed to have a flexibility of). For example, the wire stem of the present embodiment is shaped such that the lower (shown) top of the wire stem (ie, the topological tip of the wire stem) extends 0.13 mm (0.005 inch) beyond the bottom surface 2002b of the substrate.

It is understood that the total wire length (electric path) between terminal 2042 and terminal 2052 is typically shorter than the total wire length between terminal 1942 and terminal 1952 (see previous embodiment 1900) of this embodiment. This is an advantage of the interposer 2000. In this case, the total wire length is determined by the path pointed as a wire between points “A” and “C”. In the previous embodiment 1900, the total wire length is applied to the wire stem 1922 from point “B” to “C” plus the path pointed by the wire stem 1920 from point “A” to “B”. Determined by the path pointed to. (Ie AC <AB + BC.) These shorter paths are generally good in terms of inductance.

In this embodiment 2000, the interposer 2002 primarily functions as a support for the wire. However, as described above, the interposer substrate 2002 may be provided with ground, power, and signal planes. Having a ground and signal side in the substrate (including one side of the substrate) is applicable to any other PCB substrate embodiment described herein and is beneficial for stray capacitances.

Figure 21 illustrates an embodiment 2100 of an interposer capable of making "engineering changes", i.e., redirecting a signal to a particular pin (i.e., contact pad) of an electronic component. The multilayer PCB-type substrate 2102 is provided with a number of plated through holes 2104 and 2106 (only two of which are shown). A patterned conductive layer 2108 is provided on the top surface (shown) of the substrate 2102. The bottom surface (shown) of the substrate 2102 is provided with a patterned conductive layer 2110. Elastic contact structures 2112 and 2114 are shown mounted to molded contact pads of upper conductive layer 2108. Elastic contact structures 2116 and 2118 are shown mounted to molded contact pads of lower conductive layer 2110.

As shown, the pattern of conductive layers 2108, 2110 is such that the electrical interconnect path is between the elastic contact structure 2116 and the elastic contact structure 2112, rather than the contact structure 2112 directly above the elastic contact structure 2116. To be formed. In this example, the substrate is shown as a ground plane 2120 and a power plane 2122 formed of layers in a multilayer substrate.

The ability to make technology changes (e.g., rearrangement of pin-outs) with the interposer as described with respect to FIG. 21 has great flexibility since the interposer itself is technology changed. The technology change is simply performed by re-routing the conductive traces on the interposer substrate again.

22A shows an embodiment 2200 of a technique for forming an elastic contact structure in an interposer structure having dimensional stability. In this case, a metal substrate (stiffener) 2202, such as copper, is used and the metal substrate is provided with a coating of dielectric material 2204. The substrate 2202 has a plurality of openings 2206 (only one of which is shown) extending through the substrate on which the elastic contact structure is to be formed. Conductive removable substrate 2210 (see 1104 of FIG. 11A) is disposed below substrate 2202 (shown) and is spaced from the substrate by a suitable removable spacer 2212. Wire stem 2220 is coupled to removable substrate 2210 and molded to have a springable shape, and then overcoated to be elastic in the manner described above to become an elastic contact structure. Thereafter, the through hole 2206 may be filled with a conductive elastomeric material 2208 (see 564 in FIG. 5E). Thereafter, the removable substrate 2210 and the spacer 2212 are removed (eg, by dissolving) in the manner described above. In this manner, the upper portion (shown) of the elastic contact structure 2230 is suitable for contacting the second electronic component and the lower portion (shown) of the elastic contact structure is suitable for contacting the second electronic component, and the interposer The interconnection is made with one elastic contact structure 2230 extending from both sides of the. It is also within the scope of the present invention that the substrate 2202 (similar as described above) may be made of plastic, or reinforced plastic.

As shown in FIG. 22B, the elastic contact structure 2230 may be soldered to a pad 2232 on the electronic component 2234.

For example, the structure shown in FIGS. 22A and 22B has the (great) advantage that the interconnect structure remains loose in the stiffener 2202 with only an elastomeric “bolb”.

It is within the scope of the present invention that the stiffeners 2202 are, for example, two layered structures such as two semi-rigid sheets (each with a series of holes) with a gel between the two sheets. In any case (ie, the stiffener is monolayer or multilayer), the contact structure remains loose (suspended) in the opening of the stiffener. This allows the contact structures to align themselves to abut the pads on the external parts that are not on the same plane as each other.

It is also within the scope of the present invention that the elastomeric material can cover a contact structure that is “punched” through the elastomeric material to be compressed to contact the external component.

The elastomeric material serves two purposes: (1) to hold (suspend) the contact structure in the stiffener and (2) to optionally short the contact structure to provide a series electrical path from one end to the other.

Figure 22C illustrates an embodiment 2240 of a technique for forming a support structure around the contact structure after mounting the contact structure on the electronic component. In general, a number (only one of which) contact structures 2242 are contact pads of electronic component 2246 in any suitable manner, such as the gang-transfer technique described above with respect to FIGS. 12A-12C. (2244). The component is then inverted and a flexible dielectric molding compound 2250 is poured while exposing the tip 2248 of the contact structure.

For the interposer, it can be seen from the above description that the contact structure extending from the two sides of the interposer can be manufactured according to various techniques.

22D-22F illustrate another interposer embodiment 2250. Preferably, an elastic contact structure 2252 having a morphological tip 2254 is fabricated on a removable substrate 2256 (see FIG. 11E) such as aluminum. Spacers or the like of the photoresist 2258 are disposed on the top surface (shown) of the removable substrate 2256. A printed circuit board 2260 having a series of plated through holes 2262 is aligned over removable substrate 2256 such that elastic contact structure 2252 is line-up with through holes 2262. Preferably, the plated through hole has a copper layer 2264 covered by a tin-lead (eg, solder) layer 2266.

FIG. 22E shows a printed circuit board (20) with an elastic contact structure 2252 having an intermediate portion in contact with the tin-lead layer of the corresponding through hole and having two ends 2252a and 2252b extending from the through hole 2262. 2260 is located.

Heat is then applied to re-flow the tin-lead (solder) layer so that the middle portion of the elastic contact structure is permanently mounted to the printed circuit board 2260. Thereafter, the removable structure 2256 and the spacer 2258 are removed (as shown in FIG. 22F) and extended from the through-holes 2262 and the two ends of the elastic contact structure suitable for contacting each other electronic component ( 2252a, 2252b). In Figure 22F, the reflowed solder 2266 is shown as a black portion.

Another semiconductor package embodiment

A number of "straightforward" designs have been described for using elastic contact structures on the outer surface of "conventional" semiconductor packages. Such semiconductor packages typically have a number of internal interconnects, such as conductive biases between layers (ie, between ceramic package layers). Each of these interconnects represents a level of complexity in the design, layout, and performance of the package, each producing its own package from each other, depending on the required interconnects. This means additional cost and additional manufacturing time per semiconductor package, which is a desired goal.

In accordance with a feature of the present invention, a "via-less" package is provided which avoids the problems associated with the inner layer interconnections and the related problems.

FIG. 23A shows an embodiment 2300 of a semiconductor package based on a multilayer substrate 2310, such as a PCB substrate.

Multilayer substrate 2310 is shown having two (or more than two) insulating layers 2312, 2114 and layer 2312 is disposed over layer 2114 (as shown). Bottom surfaces of layers 2312 and 2314 are provided with conductive traces 2316 and 2318 respectively patterned in a known manner.

Upper insulating layer 2312 (shown) is formed from a square ring having a center opening 2320, and lower insulating layer 2314 (shown) has a center opening 2322 aligned with center opening 2320. It is also formed as a square ring. As shown, top layer 2312 extends past the outer edge of bottom layer 2314 (to the left or to the right as shown), and opening 2320 is smaller than opening 2322. In this manner, the conductive layer (conductor) 2316 of the top layer 2312 is exposed.

Electronic component 2330, such as a semiconductor device, may be attached to die attach material 2323 to plate 2324 (eg, silver-filled adhesive or silver-glass composite). Is mounted by The plate is preferably a metal plate and is larger than the opening 2320. As shown, plate 2324 is mounted to layer 2312 on the side of the layer opposite conductive conductor 2316 with a suitable adhesive (not shown). In this manner, semiconductor die 2330 is disposed in opening 2320.

Coupling wires 2326 (only one of which is shown) are provided between the front (bottom as shown) surface of semiconductor die 2330 and the exposed inner portion of conductive lead 2316. Similarly, coupling wires 2328 (only one of which is shown) are provided between the front side of semiconductor die 2330 and the inner portion of conductive lead 2318. This is easily accomplished with "standard" wire-bonding equipment.

Elastic contact structures 2302 and 2304 are mounted to the exposed outer portion of conductive lead 2316. Similarly, elastic contact structures 2306 and 2308 are mounted to the outer portion of conductive lead 2318. Each contact structure 2302, 2304, 2306, 2308 is coupled to a respective one of a plurality of conductive leads 2316, 2318. In this manner, each conductive lead is electrically interconnected to any of the bond pads on the face of the semiconductor die through bond wires 2326 and 2328.

In this manner, the contact structures 2302, 2304, 2306, 2308 preferably start from different heights of the PCB 2310 but extend to the same height (downward in view). In particular, the contact structure can be molded (and overcoated) to be elastic.

In the manner described above, external connections suitable for mounting on semiconductor die 2330, motherboard, etc., without the need for packages with complex patterns of bias (layer to layer conductive paths) that require relatively expensive process steps. Complex interconnection between 2302, 2304, 2306, and 2308 may be performed.

FIG. 23B illustrates another implementation of a package assembly in which a multilayer substrate is formed having three exemplary insulating layers 2351, 2352, and 2353, each having conductive wires disposed on top surfaces 2354, 2355, and 2356 (shown). An example 2350 is shown. The upper two insulating layers 2351 and 2352 are each a ring-shaped structure with a central opening. The opening in layer 2351 is larger than the opening in layer 2352, and the inner portion of conductive lead 2355 is exposed. Layer 2353 need not be ring shaped. The conductive layer 2353 is exposed in the opening of the second insulating layer 2352.

Electronic components 2360, such as semiconductor dies, are disposed on top of the openings, and are connected to conductive layers 2354 and 2355 having elastic contact structures 2361 and 2362, respectively. The resilient contact structures 2361 and 2362 preferably start (mounted in) from the conductive layers 2354 and 2355 and end in the same plane as described above.

Individual electronic devices 2370, such as decoupling capacitors, are mounted and connected to conductive leads on the conductive layer 2356 within the openings of the insulating layer 2352. Electronic component 2360 is connected to decoupling capacitor 2370 with elastic contact structure 2363 in the manner described above.

External connections to the package are provided in the same manner as the previous embodiment 2300. However, in the present embodiment 2350, elastic contact structures 2364, 2365, and 2366 are mounted to various levels of bottom surface of the package at through holes connected to the conductive layers 2354, 2355, and 2356, respectively.

In this embodiment 2350, a reduced interconnect (bias) between the layers (eg, 2351, 2352, 2353) can be seen. These through layer connections (eg, biases) are relatively straight and may typically be slightly reduced in number compared to prior art designs.

Another feature shown in FIG. 23B is that the contact structures 2361, 2362, 2363 all have different spring formations. In this way, a wide variety of applications are possible, and this principle can be extended to many other embodiments described herein.

Figure 23C shows another embodiment 2380 of a package assembly. In this embodiment, the individual electronic devices (see decoupling capacitor 2370 of previous embodiment 2350) have been omitted for clarity of illustration.

This embodiment 2380 is generally similar to the previous embodiment 2350 and similarly prevents interconnection between independent inner thin layers of the multilayer substrate 2338.

As shown in this embodiment, the semiconductor die 2382 is formed by a plurality of elastic contact structures (flip-chip contacts) starting from different levels on the multilayer substrate 2384. Is connected to. Also in the present embodiment, a bottom surface (shown) of the substrate 2384 is provided with a number of inelastic straight pinned contact structures 2388 that may be formed by the techniques described above. It is within the scope of the present invention that any suitable design may be used that provides an external connection 2388 to a substrate 2384 that includes a resilient or flexible contact structure or solder column (bump) rather than the pin 2388 shown. Is in.

As an option, substrate 2384 may be surrounded by a suitable insulating material 2390 such as epoxy to protect the package from ambient conditions (such as moisture), for example.

This embodiment 2380 is an example of a flip-chip “ready” PCB based substrate that does not require drilling.

A general advantage of the embodiments shown in Figures 23A, 23B and 23C is that these embodiments are "self-planarizing" and are essentially "via-less" semiconductor packaging techniques.

Loop embodiment

In the above, the elastic contact structure is shaped so as to be a freestanding wire stem having a suitable shape so as to be overcoated, and the function of the wire which becomes an independent contact structure having elasticity (and improved fixing to a substrate) by overcoating. A method has been described in which adjacent ends are bonded to a substrate.

It has also been described that the distal end of the wire can be coupled within the contact region of the substrate (see, eg, FIG. 2F). This structure is particularly suitable for overcoating the loop with solder such that it is a solder contact having a controlled geometry and a high aspect as described in Examples 1 and 2. This solder overcoated loop embodiment of the present invention has been briefly described herein because it is generally in a different category than the "elastic contact structure".

In many electronic applications, multiple raised solder contacts (or patterns, such as patterns or arrays on electronic components) for mounting solder-bumped components to other electronic components, such as printed circuit boards with bonded pattern pads. , “Solder bumps”). In the past, attempts have been made to achieve an alternative height or shape ratio (height: width) of such solder bumps. This is due to the tendency of the solder bumps to be the same height as the width (shape ratio of 1: 1) and the surface tension in the molten state (see surface tension of water). In general, higher is preferred. In particular, the ability to control the height and shape of the solder contacts is a secondary concern.

One example of an attempt to connect using solder bumps on the outer surface of a package is a product wherein a semiconductor die is mounted in a central region on the front side of a printed circuit board and a plurality of conductive leads extending from the vicinity of the periphery to the central region on the front side of the PCB This is found in the supplied Overmolded Plastic Pad Array Carrier (OMPAC). The die is connected to the inner end of the conductive wire having a conventional bonding wire. Near the periphery of the PCB, there is a plated (conductive) through hole (bias) extending to the back of the PCB connected to each lead on the front of the PCB. The back side of the PCB is provided with conductive conductors each having an end connected to each via. In this way, signals (and power) to and from the die are connected through the coupling wires through the front leads and via the bias to the bottom leads. The plastic molded body is formed over the die and partially covers the front side of the PCB. Each bottom lead is connected to a "site" (specific location) on the back of the PCB. These sites are arranged in a square array of evenly spaced columns and columns. Each site is provided with solder bumps (ball bumps) that collectively constitute external connections for packaging (assembly). In this way, the entire assembly may be surface mounted to a corresponding “motherboard”, ie, a circuit board that includes other circuit elements, etc. for mounting the assembly to the electronic system. This kind of package is called a "ball-bump grid array" (BGA) for other package types known as "pin grid arrays" (PGAs) or for leaded packages. do.

In US Pat. No. 5,014,111 (Cuda et al., 5/91, US Classification 357/68), wherein the invention is named a package provided with electrical contact bumps and electrical contact bumps, conductive terminals are provided on a substrate and electrode pads are provided on another substrate Is disclosed. A two-stepped electronic contact bump is formed on an electrode pad that includes a first riser and a second riser formed on the first riser, respectively.

It is an object of the present invention to provide an improved bump grid array package.

The invention is not limited to forming solder bumped connections in a package, but is also useful for forming “flip-chip” connections between dies. In general, as described by way of example in US Pat. No. 4,067,104, described above, flip-chip coupling techniques involve stacking thin pads or metal bumps on one side of a silicon chip and a suitable substrate with interconnect circuitry. do. Thereafter, the chip is “flipped” over the top of the substrate, and the chip and the corresponding pads on the substrate are bonded to each other.

It is an object of the present invention to provide an improved technique for connecting electronic components in flip-chip form.

According to the present invention, solder bumps of controlled shape exhibiting a relatively high shape ratio can be formed on the electronic component. Solder bumps with high shape ratios are commonly referred to as “solder columns”.

Like FIG. 16 of Embodiment 2, FIG. 24A shows an embodiment 2400 of a raised bump electrical contact formed on a terminal 2412 on a substrate of a substrate 2408. As shown in FIG. Coupling the first end 2402a of wire 2402 at a first location on terminal 2412 (by ball coupling as shown) and shaping wire 2402 in a loop (in the manner of FIG. 2F) to (wedge) Bump-type contacts are formed by coupling the second end 2402b to a second location on the terminal 242 (such as coupling). The wire stem shaped (eg, U-shaped) is a conductive material 2420 positioned over the entire length of the wire 2402 and over the terminal 2412 (including over the edge of the terminal and covering the joint of the end of the wire). ) Is overcoated in the manner described above. In this way, it is possible to provide a relatively rigid contact structure protruding from the face of the terminal having a controllable shape and having a relatively high shape ratio (height: width). Overcoating material 2420 is preferably solder, which would be a relatively inflexible contact structure. However, overcoating material 2420 may be a springable conductive material such as nickel so that the contact structure has some elasticity. Overcoat 2420 may be a multilayer overcoat in the manner of FIG. 5.

If desired, the same to provide two or more "redundant" raised bump electrical contacts on terminals 2412, each having an overcoat 2420 surrounded by the same contact structure of the type described with respect to Figure 24A. It can be formed on the terminal.

FIG. 24B, similar to FIG. 17 of Embodiment 2, shows another contact structure 2450. In this embodiment, two wire stems 2430 and 2432 are each formed in the manner of contact structure 2400 (FIG. 24A). In this case, the wire stem is overcoated with a material that prevents the reaction of gold with the tin component of the solder (as described below).

As shown, the two contact structures 2430 and 2432 are spaced apart on the terminals 2442 and 2412 on the substrate 2458 (see 2408) and may be parallel to each other. It is within the scope of the present invention that any number of contact structures in any direction relative to each other may be mounted to terminals 2242 on substrate 2458.

In the description of Figure 24B, a wire stem (shown as "bridging" (filling the space between) two separate contact structures 2430, 2432 to provide one "unitary" solder bump ( 2430 and 2432 are overcoated with solder 2455. However, if desired, the two contact structures 2430 and 2432 are sufficiently spaced apart so that the solder 2455 does not connect two (or more) contact structures, thus providing two (or more) solder bumps per terminal. It becomes a contact. However, in most cases it is desirable that the solder be filled in the space between the wire and the terminal.

In a preferred embodiment, the solder bump contacts comprise at least 70% by volume, for example at least 80% solder, the remainder of the bump contacts being the wire itself and a (almost negligible) barrier layer (eg nickel alloy). Cover this gold wire.

Thermal path embodiment

Heat is necessarily generated during operation of the semiconductor device, and if left unreduced, the device may be destroyed. Thus, it is known to provide some kinds of heat sinks for such devices. In general, heat sinks have one of two forms. These may be integrated with the device package or may be connected outside the device package. In either case, heat sinks generally include thermal masses of intimate thermal conductivity with semiconductor devices, and may include convection or forced air cooling of the thermal masses. Providing an effective thermal path between the die and the heat sink has long been a challenge.

It is an object of the present invention to provide an improved technique for providing an effective heat path between a heat generating die and a heat mass (heat sink).

According to the invention, thermal interconnections are formed on the terminals of the substrate. In general, thermal interconnection functions to conduct heat from a substrate, which may be a semiconductor die. In general, it may be within the scope of the present invention that the terminal where the thermal interconnection is formed may be an “active” terminal of the semiconductor die, but the terminal where the thermal interconnection is formed is an “active” electrical terminal of the semiconductor die. There is no need. In particular, thermal interconnect terminals may be fabricated as other “active” terminals (eg, bond pads) on the same side of the die.

FIG. 24C shows an embodiment 2470 of a technique for forming thermal interconnection on a substrate 2482. An “oversize” terminal 2268 (ie, relative to the electrical connection terminal) is formed on the side of the substrate 2478. Wire 2472 has a free end 2472a coupled to the terminal near its perimeter. Using the technique described above, the wire is formed into a loop and is cut into another loop that is rejoined to the terminal without being cut and then rejoined to the terminal, thereby forming a fence structure of multiple loops extending around the outer periphery of the terminal as shown. Repeatedly combined and formed to form. In the final step, the final loop is cut at a position corresponding to the origin 2472a of the first loop. The multiple loops forming the fence structure preferably extend over a terminal to a common (same) height.

FIG. 24D shows the fence structure of FIG. 24C to provide bump or freestanding solder contacts that can function as slugs or heat sinks that conduct heat from the substrate 2478 to a heat sink (not shown). using any suitable means, such as wave soldering). Solder 2475 surrounds the entire terminal 2482 preferably in the manner of solder 2455 in FIG. 24B. In general, the fence outline will match the outline of the terminal (usually square) and the outline of the solder mass will match the outline of the fence.

Figure 24E illustrates another embodiment 2450 that is adapted to dissipate heat through the back of the semiconductor die. In this embodiment, the semiconductor die 2454 has a number of elastic contact structures 2454 mounted to the front face. The back of the die is covered with a conductive layer 2456 such as gold. Multiple heat dissipation structures 2458 are coupled to layer 2456 behind the die. Such heat dissipating structures are not intended to perform interconnection between the die and other electronic components (although it is within the scope of the present invention that they may perform a ground connection as a whole). Rather, this heat dissipation structure 2458 is adapted to dissipate heat from the die during operation. Multiple heat dissipating structures increase the total effective area behind the die (ie, the total area of the heat dissipating structure plus the rear area of the die) to improve convective cooling. If the back of the die is pressed against a heat sinkable structure (eg a housing, etc.), the conductive thermal path is also performed by the heat dissipating structure. Preferably, the heat dissipation structure 2358 is made of copper wire stem (copper is a good thermal conductor), and the same process step by overcoating the elastic contact structure 2454 (which may be a gold wire stem, for example). It can also be overcoated in.

Assembly of electronic components

In particular, the packaged electronic component (i) connects two (or more) electronic components mounted in the interposer structure to the elastic contact structure, and (ii) connects the plurality of elastic contact structures to the electronic component (eg individually or The purpose of the mounting) has been described above. In the following figures, exemplary assembly of electronic components has been described.

25, similar to FIG. 25 of Embodiment 2, is provided with a plurality of elastic contact structures 2504 and 2506 on a surface 2502a on a first electronic component 2502. The other electronic component 2508 is provided with a plurality of surfaces 2508a (two of which are shown) elastic contact structures 2510, 2512. The faces 2502a and 2508a of the two electronic components 2502 and 2508 face each other as shown, spaced apart from each other. The elastic contact structures 2504, 2506, 2510, 2512 are shown coupled to the engagement zone 110 in a manner similar to the technique shown in FIGS. 1C-1E. However, it is within the scope of the present invention that the elastic contact structure may be combined (ie, wire stem) and overcoated in any of the exemplary ways described above.

The printed circuit board (PCB) 2520 ("circuited"-i.e., with conductive wires / pads on both sides) is provided on both sides 2502a and 2508a of the two electronic components 2502 and 2508, respectively, as shown. ) Is placed between. The upper surface 2520a of the PCB 2520 is provided with a number of contact zones 2522 and 2524 (only two of which are shown), which may be terminal pads or the like, and the lower surface 2520b of the interposer structure is provided with terminals, Multiple (only two of which are shown) contact areas 2526 and 2528 are provided, which may be pads or the like. Other electronic devices (not shown) may be mounted on both sides of the printed circuit board.

As shown, the tip of the elastic contact structure 2504 abuts the pad 2522, the tip of the elastic contact structure 2506 abuts the pad 2425, and the tip of the elastic contact structure 2510 is the pad 2526. A tip of the elastic contact structure 2512 abuts the pad 2528. In this way, both electronic components 2502 and 2508 are electrically connected to the PCB 2520.

The PCB has conductive leads (not shown) on both sides 2520a and 2520b. In this way, each lead is connected on the edge of the PCB 2520 at an edge connector terminal (edge connector pad). Two such terminals 2532, 2534 are shown (among others). Terminal 2532 is shown on top surface 2520a (shown) of PCB 2520 and may be connected via wire to (for example) terminal 2522, terminal 2534 is PCB 2520 Is shown on the bottom surface 2520b (shown) and may be connected via a lead to (for example) terminal 2526.

Electronic components 2502 and 2508 are supported on interposer substrate 2520 with suitable clamps (not shown) such that their respective elastic contact structures 2504, 2506, 2510, and 2512 are angled on PCB 2520. The pads 2522, 2534, 2526, and 2528 are in contact. Preferably, contact pads 2522, 2524, 2526, and 2528 are provided with a solder mass 2540, which may be reflowed, for example by passing the assembly through an oven (ro), so that the solder carried by the contact pads Solder joints are formed with the distal ends of the respective elastic contact structures 2504, 2506, 2510, and 2512. In the figure, the solder is shown to be reflowed. In general, embodiments of the electronic assembly described herein may use conductive epoxy masses instead of solder.

Alternatively, in the final step, the compressed assembly may be an appropriate insulator 2530, which may be an adhesive or polymer material filled with suitable particles, such as, for example, thermally conductive particles, to improve the heat dissipation properties of the assembly (as shown). Is surrounded internally. In addition, certain suitable fillers may be useful for continuing to reduce thermal expansion of the polymer encapsulant.

The compressive force may be removed after reflow soldering or after applying insulation 2530.

Such a structure is suitable for, for example, mounting and electrically connecting a plurality of electronic components (for example, bare memory chips) to each side of a printed circuit board. For example, a memory (e.g., RAM) module may be formed by this technique of edge-mounting multiple memory chips on one or both sides of a PCB as shown in Figures 36A-36C. Therefore, SIMM (Single In-Line Memory Modules) modules (boards, cards) fabricated in this way have much more storage capacity than “standard” SIMM modules using the same chip, and the packaged chips are assembled by soldering to the PCB. It can be manufactured at a lower cost (excluding the price of the chip) than a lower capacity SIMM module.

In general, by fabricating an electronic assembly in accordance with the techniques of the present invention, more semiconductor chips can be mounted in a given PCB area more effectively using the “real estate” on the PCB. In particular, because the die mounted to the PCB is not packaged (eg plastic packaging), the overall thickness of the PCB to which the die is attached will be smaller than the conventional assembly of the packaged die.

Although the semiconductor die is not entirely contained within the package body in the "packaging" in the conventional sense, the electronic component assembly of Figure 25 and the electronic component (semiconductor die) assembly described below on a printed circuit board may be considered as a packaging technique.

A method different from that described in FIG. 25 is used to perform double-sided flip-chip attachment of two silicon dies 2502 and 2508 to an interconnect substrate having its own elastic contact structure (electronic components (as shown in FIG. 25). Elastic contact structures 2504, 2506, 2510, and 2512 originate from the substrate 2520 (rather than from 2502, 2508).

In practice, the elastic contact structure may be overcoated (pre-wetted), for example, with a large amount of solder, mainly at the tip of the elastic contact structure. When the parts are heated together, the solder reflows to form a “hourglass” joint with most of the solder distributed near and at the ends of the elastic contact structure, so that a minimum amount of solder bends (springs) in the elastic contact structure. Remains in injury. This will be true regardless of whether the elastic increasingly structure is mounted to a substrate (eg 2520) or the electronic component (eg 2503, 2508) itself. Alternatively, a small amount of solder paste may be applied to the pad (eg, 2522) to secure the tip of the (reflow furnace) contact structure.

The importance of this and other electronic assemblies described above is that the flexible (including pure elasticity) contact structure (eg 2504) can be mounted directly to the silicon device (eg 2502) without damaging the device. Such contacts are well suited for evaluating (testing, pre-testing electronic components before final packaging). In particular, these flexible contact structures can be used to determine the rate of thermal expansion between electronic components (eg 2502) and components (eg 2520). The difference can be accommodated, thereby improving the reliability of the interconnection (see, for example, the description of FIG. 44).

It is understood from the prior art search that in the past there have been limited attempts to place contact structures (not solder bumped contacts) directly on silicon. For example, in US Pat. No. 4,955,523 (“Laychem”) described above, wire 7 is coupled to a contact on an electronic component 1, such as an integrated circuit chip. The wires are cut to approximately the same length from each other. Insulating material 17 is applied to the layer of the component to surround the bond between the wire and the contact. The insulating material is described in the Leychem patent as a relatively flexible polymer, such as partially crystallized. The need for this layer will probably help to increase the integrity of the bond in the adjustment of the wire to the contacts. On the other hand, as described above, according to the techniques of the present invention, the overcoating material is a conductive metal material (in most embodiments) that firmly secures the wire stem to the terminal to impart some elasticity / flexibility to the contact structure. The fastening function of the overcoating is particularly severe in terms of elastic (or flexible) contact structures that react to compressive forces. As described below, in most embodiments described herein, the wire stem is essentially sufficient once it is overcoated. The technique of the Raychem patent is clearly not the same regardless of any similarity shown in the present invention.

The subtle advantages of the embodiment as shown in FIG. 25 include the "part" of the "card" pre-assembled to the substrate (e.g., 2520), where "card" includes electronic components (e.g. 2502). Assembly ”may be prefabricated (tested) and later incorporated into a more complex assembly (shown in FIG. 37 to be described later).

In Figures 25-32, an assembly of electronic components (e.g. 2502) with spring (elastic or flexible) contacts is shown interconnected to an interconnect substrate (e.g. 2502). It is also within the scope of the present invention that the contacts are “inverted” and originated from the interconnect substrate, in which case the interconnect substrate will be an interposer. However, this is not desirable (ie, using an interconnect substrate as a contact mounted interposer), and a more effective interposer structure is described above (FIGS. 17A-22F). In the embodiment of the electronic assembly shown in Figures 25-32, once the assembly is assembled, the electronic component (e.g. 2502) solders the contact pad (e.g. 2522) on the interconnect substrate (e.g. 2520). Will be soldered to the interconnect substrate (eg 2520) by reflowing the solder paste.

As with FIG. 26 of Embodiment 2, FIG. 26 shows an electronic component having elastic contact structures 2604, 2606, 2610, 2612 extending from faces 2602a, 2608a in an exemplary manner similar to the assembly 2500 of FIG. Another embodiment 2600 of the assembly of 2602, 2608 is shown. On the other hand, in the previous embodiment 2500, a compressive force (“F”) is used to yieldly compress the electronic components 2502 and 2508 against the PCB 2520, and the electronic components 2502 and 2508 are the solder itself ( Held in position by 2540 or by insulating material 2530, in this embodiment, electronic components 2602 and 2608 are specially disposed on (mounted on) faces 2620a and 2620b of PCB 2620. It is held in position relative to the PCB 2620 by molded stems 2630, 2632, 2634, 2636, acting as spring clips (holding springs). All four of these specially shaped wire stems (spring clips) 2630, 2632, 2634, 2636 are overcoated to be appropriately elastic in the manner of elastic contact structures that are electrically connected but need not be electrically conductive.

In general, the spring clips will be placed in multiple locations on the PCB and will cooperate with each other to secure each electronic component to the PCB. For example, four spring clips may be used to secure the square electronic components to the PCB to join opposite sides of the paired electronic components. For the sake of understanding, only two spring clips are shown per electronic component.

As shown, the representative spring clip 2630 is shaped to compress (press) the elastic contact structure of the electronic component 2602 with respect to the interposer structure 2620. In particular, the spring clip 2630 extends generally straight in a straight upward direction from the top surface 2620a (shown) of the PCB 2620 and bends inward (toward the spring clip 2632), the electronic component being spring clipd. It is more bent outwardly (spaced from the spring clip 2632) so that it can be inserted between the tips of 2630 and 2632. The spring clip 2632 is a “mirror image” of the wire stem 2630. In this manner, the electronic component 2602 is inserted between the tips of the spring clips 2630 and 3632 as indicated by the initial position of the electronic component 2602 shown in hidden line, so that the electronic component 2603 is inserted. The interposer until the bends of the tips of the spring clips 2630 and 2632 “snap” back towards each other until a final position (shown in solid line) is reached to engage the rear surface 2602b of the electronic component 2602. Toward structure 2620 (by means not shown) The tip of the spring clips 2630 and 2332 can be deformed outwardly (spaced from each other as indicated by the arrows marked “OUT”). Preferably, the electronic component 2602 is a spring clip 2630. After spreading between the tips of 2632, wire stems 2604 and 2606 begin to contact (and thus compress) each pad on PCB 2620. The electronic component is positioned (indicated by a solid line). When “snap” to, the wire stems 2604 and 2606 will be compressed. The spring clips 2634 and 2636 may be shaped to operate in the same manner as the wire stems 2630 and 2632. Electronic component 2608 If this electronic component 2602 is the same and each of the elastic contact structures 2604, 2606, 2610, 1612 have the same system, the spring clips 2634, 2636 are preferably the same as the spring clips 2630, 2632. Connection to the rear side of the part grounds the “body” of the electronic part (or While it may be useful as to bring it to another predetermined potential, the spring clips 2630, 2632, 2634, 2634 are primarily PCB 2620, like any electrical connection made between these spring clips (i.e., bends). The electronic components 2602 and 2608 are held in position with respect to the back surface and the rear surfaces 2602b and 2608b of the electronic components 2602 and 2608 are incidental.

In the present embodiment 2600, the edge connectors (see 2352 and 2534 in Fig. 25) are omitted for simplicity as necessary. As in the previous embodiment 2500, prior to overcoating with solder 2640 and resolding the solder 2640 prior to inserting the PCB contact pad onto the PCB to ensure a soldered connection between the distal end of the elastic contact structure and the PCB contact pad. The assembly can be passed through the furnace to flow. The amount of solder placed on the PCB contact pads is sufficient to secure the distal ends of the elastic contact structure but the solder does not degrade the elastic contact structure (which may occur if solder flows over the bends of the elastic contact structure). You'll find that you're adjusting so badly.

Compared to any of the embodiments for the assembly 2600 of FIG. 26, or another assembly disclosed herein, including spring clips, etc., it is within the scope of the present invention that the electronic component is not soldered to the PCB. This makes it easy to remove the electronic component for replacement or upgrade of the electronic component by pressing the end of the spring clip outward (indicated by the arrow marked “OUT”) by hand to release the electronic component from the junction with the PCB. .

Elastic contact structures 2604, 2606, 2610, 2612 originate from the substrate 2620 where electronic components 2602, 2608, such as semiconductors with compatible contact pads (solderable metallization) on the surface, snap in place. It is also within the scope of the present invention that it is coupled in such a way that the contact pad is in contact with the tip of the elastic contact structure.

Clearly, when spring clips (eg 2630, 2632, 2634, 2636) are used, solder 2640 (eg conductive epoxy) and the like are not essential and optional. This is also the case in other embodiments such as the embodiment 2700 of FIG. 27 using the "holding means", which will be described immediately below.

Figure 27, similar to Figure 27 of Case-2, illustrates an embodiment 2700 in which an electronic component 2702 (one of the several shown) is mounted only on one side 2720a of a printed circuit board 2720 and electrically connected thereto. As shown, it substantially shows half (top) of the device 2600 of FIG.

In this embodiment 2700, as in the previous embodiment 2600, the elastic contact structures 2704 and 2706 are attached to the front surface 2702a of the electronic component 2702 in any suitable manner (eg, soldering, brazing). Mount it. It should also be noted that such a device can be reversed so that a contact can be mounted to (extending from) the substrate 2720.

Any of the assemblies in which the electronic component (such as 2702) is pressed against the interconnect plate (such as 2720) and soldering (such as epoxy treatment) is optional are included within the scope of the present invention. Pressurization is sufficient to ensure reliable electrical interconnection and to replace electronic components (e.g. for upgrade or repair) (e.g., to disperse clips 2730 and 2732, It is clear that removing 2702, replacing with other similar components, may be somewhat simpler as it relies solely on pressurization.

In this embodiment 2700, as in the previous embodiment 2600, the electronic component 2702 is attached to the spring clips 2730 and 2732 mounted to the upper surface 2720a (as shown) of the PCB 2720. Mechanically mounted to the PCB (2720).

Solder 2540, 2640 is not shown in this figure. This indicates that the electronic component 2702 has been carefully assembled to the PCB 2702 so that it can be removed (by displacing the spring clip).

It is within the scope of the present invention that external spring elements, clamps, and the like may be used in place of the exemplary spring clips 2730 and 2732 shown.

25, 26, and 27 all refer to printed circuit boards 2520, 2620, and 2720, the printed circuit board elements may consist of any suitable substrate having a terminal formed thereon, which includes electronic components. It should be noted that the resilient contact structure mounted on the is electrically connected.

As with the embodiment 2600 of FIG. 26, the elastic contact structures 2704 and 2706 are derived from the substrate 2720, in which case the electronic component 2702 with compatible contact pads on the surface snaps into place. It is also within the scope of the present invention that it is simply coupled in such a way that the contact pads of the electronic component come into contact with the tip of the elastic contact structure of the substrate.

In general, the embodiments of FIGS. 25-30 all attempt to mount the elastic contact structure to an electronic component rather than a PCB, where the PCB is optional.

Figure 28, similar to Figure 28 of Case-2, shows an electronic component 2802 (such as a semiconductor die, etc.) for facilitating removal of the electronic component 2802, like the embodiment 2700 of Figure 27. The embodiment 2800 of the assembly assembled to FIG.

Electronic component 2802 has a number of elastic contact structures 2804 and 2806 that are mounted to and protrude from the surface in a suitable manner as described below. In this case, the elastic contact structures 2804 and 2806 are configured to have a shape such that the elastic contact structures 2804 and 2806 fit snugly into corresponding plated through holes 2822 and 2824 in the printed circuit board 2820. do. The advantage of such a device is that the electronic component 2802 can be easily disconnected (disconnected) from the printed circuit board 2820 and replaced (eg, replacing a defective component or upgrading a component). Is the point. Conductive traces on one or more surfaces (two surfaces) of the printed circuit board 2820 have been omitted for simplicity.

In more detail, the distal regions of the elastic contact structures 2804 and 2806 are at a distance d (crossing the ground) that is about 0.0076 to 0.0381 mm (0.0003 to 0.0015 inch) less than the inner diameter D of the through holes 2822 and 2824. It is formed into a shape having a bent portion extending laterally. This configuration allows the elastic contact structure to fit snugly in the plated through hole (elastically slightly interference fit). This assembly 2800 may be soldered, of course, but soldering does not allow the electronic component 2802 to be easily separated from the PCB 2820.

As described above, according to the present invention, a preferred technique for mounting an elastic contact structure to a semiconductor component is provided. 25-29 and subsequent embodiments of the assemblies described herein illustrate the technique described in FIGS. 1C-1E, where the elastic contact structures 2804, 2806 (compare 130) are formed of a protective layer ( 2808 (compared to 124) to conductive pads 2810 (compared to 132).

The embodiment of Fig. 28 is excellent in simplicity. The electrical contact structures 2804 and 2806 can be easily mounted to the silicon electronic component 2802 and thus plugged into a conventional plated through hole in the printed circuit board 2820. In this way, the spring on the electronic component (ie the contact structure) is spring biased against the wall of the through hole (plated hole). The electronic component is selectively soldered in place by applying a certain amount of solder paste into the through hole and heating the assembly to reflow the solder paste.

Figure 29, similar to Figure 29 of Case-2, illustrates an assembly 2900 of two electronic components 2902 and 2908 assembled to a printed circuit board (PCB) 2920 in a manner similar to the embodiment 2600 of Figure 26. Another embodiment is shown. In this case, the spring clips 2930, 2932, 2934, 2936 are mounted to the electronic components 2902, 2908 rather than the PCB 2920, and function in a similar manner to mechanically couple the components 2902, 2908 to the PCB 2920. Coupling Two spring clips (latch springs) 2930 and 2932 are shown extending from the front face 2902a of the electronic component 2902. Two spring clips (latch springs) 2934 and 2936 are shown extending from the front face 2908a of the electronic component 2908.

As shown, the representative spring clip 2630 is formed in a shape such that the elastic contact structure of the electronic component 2602 is pressed against the intermediate structure 2620. More specifically, the spring clip 2630 extends generally straight upwards from the front face 2902a of the electronic component 2902 and is bent outwardly (from the spring clip 2932) and then (spring clip 2932). The bent portion of the distal end of the spring clip 2930 is positioned to be sized and shaped to be inserted and mounted in a corresponding through hole 2922 extending through the PCB 2920. 2932 is of the same size and shape as spring clip 2930 and is positioned to extend into a corresponding through hole 2924 extending through PCB 2920. These through holes 2922 and 2924 are plated. However, the through holes can be plated through holes, and if plated, the electrical connection is similar to that of the embodiment 2800 of FIG. It is within the scope of the present invention that it is made by a spring clip 2930 between electronic component 2902 and PCB 2920. Electronic component 2902 is indicated by arrow F with respect to PCB 2920. When pressurized as shown, the spring clips 2930 and 2932 enter the holes 2922 and 2924 and are displaced and spring-backed so that the spring clips hold the opposite side 2920b of the PCB 2920. 2902 is mechanically secured to PCB 2920 with elastic contact structures 2904 and 2906 being pressed against corresponding pads 2922 and 2924 on surface 2920a of PCB 2920. To remove 2902, the distal tip of the spring clips 2930 and 2932 must be manually manipulated to retract through the through holes 2922 and 2924. The spring clips 2934 and 2936 are each through hole in the PCB 2920. Insert the electronic component 2908 by inserting it through (2926, 2928). It is operated in the same manner as the spring clips 2930 and 2932 by fixing in the CB 2920.

Generally, the elastic contact structure is not soldered to the PCB in order to be able to break the interconnect between the electronic component and the PCB. However, soldering an elastic contact structure to a PCB in the same manner as described for the embodiment 2500 of FIG. 25, as an example, is within the scope of the present invention.

If no solder is used and the electrical connection between the electronic components 2902 and 2908 and the substrate 2920 is dependent only on mechanical contact, the elastic contact structure may be provided with a fine protrusion (as described above) and a top of the embodiment of the probe (as described above). Can be overcoated to have a topological tip. In either case, electronic components 2902 and 2908 are simply “clicked” into place on substrate 2920 and held therein by latch springs 2930, 2932, 2934, and 2936.

In a manner similar to the embodiment 2700 of FIG. 27, which is substantially half of the embodiment 2600 of FIG. 26, the embodiment 2900 of FIG. 29 provides a “pressing contact” interconnect (if solder is not used). Similarly, it can be "half-type" which makes it easy to mount one electronic component (for example, 2902) on the substrate 2920.

Generally, in an embodiment of the electronic assembly disclosed herein having means for pressing the electronic component against the PCB (eg, spring clips), soldering the distal end of the elastic contact structure is optional, In the case of using soldering, a contact structure (for example, a straight fin structure) in which elasticity is weakened can be used. However, as will be described in more detail below (see FIG. 44 for example), it is preferable as long as the shape-shaped contact structure easily accommodates effects such as thermal expansion.

Figure 30, similar to Figure 30 of Case-2, illustrates another embodiment 3000 in which two electronic components 3002, 3008 are mounted on a printed circuit board (PCB) 3020. In the manner as described above (for example, description of the embodiment 2500 of FIG. 25), the elastic contact structures 3004 and 3006 extend from the front 3002a of the electronic component 3002 and the PCB 3020. Contacting the corresponding pads 3022 and 3024 on the surface 3020a of the elastic contact structures 3010 and 3012 extend from the front 3008a of the electronic component 3008 and onto the surface 3020b of the PCB 3020. Contact with corresponding pads 3026, 3028. Both electronic components 3002 and 3008 are pressed against PCB 3020 as indicated by arrow F, which illustrates two techniques for securing electronic components 3002 and 3008 to PCB 3020. .

The electronic component 3002 is provided with straight alignment pins 3030 and 3032 formed in the manner described above with respect to FIG. 2E. These straight alignment pins 3030, 3032 need not be elastic and need not be conductive, but these pins are easily mounted to the electronic component 3002 and are elastic contacts extending from the front surface 3002a of the electronic component 3002. The shape (linear) is formed and overcoated in the same process steps used to form the structures 3004 and 3006. The size and location of the pins 3030 and 3032 allow for insert mounting through respective through holes 2922 and 2924 extending through the PCB 3020. These pins 3030 and 3032 allow the electronic component 3002 to be simply aligned with respect to the PCB 3020 so that the distal ends of the elastic contact structures 3004 and 3006 have respective pads 3022 on the surface 3020a of the PCB 3020. , 3024). The electronic component 3022 is pressed against the PCB 3020 in the direction of the arrow F and secured thereto by an appropriate adhesive 3070 (shown in partial cross section, compared to 2530 in FIG. 25).

The electronic component 3008 is provided with initial straight alignment pins 3034 and 3036 which are formed in the manner described above with respect to FIG. 2E. These alignment pins 3034, 3036 need not be elastic and need not be conductive, but these pins are easily mounted to the electronic component 3008 and are elastic contacts extending from the front surface 3008a of the electronic component 3008. The shape (straight) is formed and overcoated in the same process steps used to form the structures 3010 and 3012. The size and location of the pins 3034 and 3038 allow for insert mounting through respective through holes 2926 and 2928 extending through the PCB 3020. These pins 3034, 3036 allow the electronic component 3002 to be preferentially aligned with respect to the PCB 3020 so that the distal ends of the elastic contact structures 3010, 3012 have respective pads on the surface 3020a of the PCB 3020. 3026, 3028). Once the electronic component 3028 is pressed against the PCB 3020, it is secured there and supported against the surface 3020a of the PCB 3020 by bending the distal ends of the pins 3034 and 3036 as shown. This allows the electronic component 3080 to “fasten” in place facing the printed circuit board 3020.

In this embodiment, the fins 3030, 3032 and the fins 3034, 3036 (with their proximal ends) are conductive pads 3050, 3052, 3054, 3056 on the surfaces 3002a, 3008a of the electronic component 3002, 3008. It is seen that these pads (ie, 3050, 3052) overlie the passivation layer 3060. These pads 3050 and 3052 are suitably formed in the same materials and steps as the wire stems of the elastic contact structures 3004 and 3006 to form bonded bond pads (photosensitive and metal layers 126 of FIGS. 1C-1E). Compared to resist 128).

In this embodiment, the elastic contact structures 3004 and 3006 are shown soldered to respective pads 3022 and 3024 on the PCB 3020 in the same manner as the embodiment 2500 of FIG. As described above, conductive epoxy may be used instead of solder. It is also shown to maintain intimate contact with the respective pads 3026, 3028 on the PCB 3020 by bending the alignment pins 3034, 3036 without soldering the elastic contact structures 3010, 3012.

Although two electronic components 3002 and 3008 are shown fixed to the PCB in two different ways, this is merely illustrative. Two electronic components may be secured to the PCB in the same way (for example, both with alignment pins and adhesives, and both with soldering).

In practice, the lower electronic component 3008 (as shown) is first placed in position, the pins 3034 and 3036 are bent, and then the upper electronic component 3002 is mounted on the printed circuit board. do. As in the embodiment described above, the use of solder to electrically connect the electronic component (especially the electronic component 3008 with the bent alignment pin) and the printed circuit board 3020 is optional, and the user The connection may also rely on the tip of the resilient contact structure to actively connect the pad onto the printed circuit board (and to easily disconnect the electronic components). However, droplets of solder paste (omitted for simplicity in the drawings and illustrated that the solder / epoxy is optional) were applied to pads 3026 and 3028 during the assembly process and droplets of solder paste on pads 3022 and 3024 were applied. It is therefore generally preferred to reflow. Generally, pin 3034 to prevent electronic component 3008 from exiting printed circuit board 3020 during subsequent processing steps (eg, mounting another electronic component 3002 and refluxing the assembly). , 3036). In the prior art, a technique such as glue bonding (gluing the lower electronic component 3008 to the printed circuit board 3020) has been used.

The use of adhesives (e.g., 2530 and 3070) has been described, where adhesives typically have electronic components (e.g., 2502, 3002) in the absence of fastening structures (e.g., 3034 and 3036). For example, it serves to fix the 2520, 3020. It is preferable that an adhesive consists of a material of the kind which shows shrinkage at the time of hardening.

It is within the scope of the present invention to provide alignment pins on both electronic components (eg, 3002, 3008) and to fix them to the printed circuit board 3020 without bending the pins. In this case, it is preferable to apply an adhesive between each of the electronic components and the printed circuit board, and this arrangement is similar to the upper half of Fig. 30, which is mirror symmetrical (both sides are the same).

Assembly containing electronic device

25 to 30 illustrate how an electronic component (such as 2502a) is secured to one side 2520a of a printed circuit board type substrate (such as 2520) and another electronic component (such as 2508 is described to be fixed to another side of the printed circuit board type substrate 2520 (for example, 2520b).

Figures 31 and 32 are similar to Figures 31 and 32 of Case-2, and as described below, an electronic device assembly, such as a separation capacitor (or resistor, or any other kind of electronic component), may be used. The method of coupling to a printed circuit board is illustrated.

The use of separate capacitors in an assembly of electronic components is generally known to have an advantage in the performance of the semiconductor device in the assembly. As a general suggestion, the closer the isolation capacitor is to the semiconductor device, the greater the effect. In general, traditional packaging techniques are basically inadequate for this purpose.

FIG. 31 shows an embodiment 3100 of two electronic components 3102 and 3108 mounted on each of two sides 3120a and 3120b of a printed circuit board 3120. FIG.

A large hole (through hole) 3150 is provided through the PCB 3120 at a position where it is desired to be below the electronic component 3102, 3108. The hole 3150 has a depth dimension in units such as the thickness of the PCB 3120 (vertical dimension when viewed in the drawing) and enters the ground (when viewed in the drawing) an area that is circular or rectangular (or any other shape). Area). For the purpose as described herein, the hole 3150 may be circular and have a specific diameter.

The hole 3150 may be provided with a plating material in a manner somewhat similar to making a plated through hole. In this illustrative example, the first plating portion 3152 is disposed on the left side (as seen in the figure) of the hole 3150 and extends only partially (less than 180 ° around the circumference of the hole 3150). . A second plated portion 3154 (which may of course be manufactured in the same process steps as the first plated portion 3152) is disposed on the right side (as seen in the drawing) of the hole 3150 and the circumference of the hole 3150. It extends partially (less than 180 °) in the circumference. Plating 3154 is shown connected to a trace 3156 on the upper side 3120a of the PCB.

The elastic contact structures 3104 and 3106 are mounted to the front side (lower side when viewed in the drawing) of the electronic component 3102, and the elastic contact structures 3110 and 3112 are moved to the front of the electronic component 3108 in the above manner. It is mounted on the side (upper side, when seen in the drawing).

An advantage of the electronic assembly of this embodiment is that the short circuit path that can be made between the electronic device 3160 and the electronic components 3102 and 3108 is very short. In one example, the elastic contact structure 31006 is obviously coupled with one of the power supplies (ie, power or ground, or VSS or VDD) “pin outs” of the electronic component 3102.

In the present embodiment, the electronic device 3160, which is a capacitor, is disposed in the hole 3150 and has two lead portions (tabs or plates) 3162 and 3164 extending from the device. As shown, one lead portion 3322 is connected to the plating portion 3152 and the other lead portion 3164 is connected to the plating portion 3154 (plating portions 3152 and 3154 are not shorted to each other). Although not shown directly in the figure, the plating portion 3152 is coupled (connected) with a trace terminating in the pad, where the elastic contact structure (not shown) of the electronic component 3102 is pressed.

An example of a component such as a capacitor (eg, 3160) underneath other components such as a packaged semiconductor device (eg, 3102, but not an unpacked 3102), the component 3160 may be a printed circuit board ( As an example, it is generally known to place by placing it in the pocket (recess) of 3120. The previously described embodiment 3100 is basically in close contact with the component 3160 to other components 3012, except that the benefits of the elastic contact structure and the valuable fixed area on the printed circuit board 3120 are consumed. It relates to a combination having the same performance that can be arranged.

32 shows another embodiment 3200 of two electronic components 3202 and 3208 mounted on two sides 3220a and 3220b of the printed circuit board 3220, respectively.

Elastic contact structures 3204 and 3206 are mounted to the front side (bottom side) in the electronic component 3202, and the elastic contact structures 3210 and 3212 are front side of the electronic component 3208 (the top in the figure). Cotton).

In this embodiment, a known manner such as wiring a trace to one of a plurality of conductive pads 3222, 3224, 3226, 3228 on the top surface 3220a and the bottom surface 3220b of the PCB 3220. The conductive pads 3252, 3254, 3256, and 3258 connected by the same are provided.

The first electronic device 3260 having two leads 3326 and 3264, such as a capacitor, is mounted on the top surface 3220a of the PCB 3220, and each of the leads 3326, 3264 is a conductive pad 3252. , 3254).

The second electronic device 3270 having two leads 3327 and 3274 such as a capacitor or the like is mounted on the bottom surface 3220b of the PCB 3220, and each of the leads 3327 and 3274 has a conductive pad 3256. , 3258).

As described above with respect to the embodiment 3100 of FIG. 31, the general concept of placing components under components is not entirely new. However, as well illustrated by the present embodiment 3200, the distinct advantage is that component parts (e.g., 3260) are different from solder bumps, except that recesses and the like are provided in the printed circuit board 3220. The ability to manufacture very large (long) contact structures to be placed in these other component parts (eg 3202) arises.

The elastic contact structure is mounted on the electronic component, the modified embodiment is easily realized, and in the electronic component which is a semiconductor device, it is generally preferable that the elastic contact structure is directly mounted on the semiconductor device, but the elastic contact structure is mounted on the PCB. Any assembly that comes into contact with a selected contact region (eg, an adhesive pad) on an electronic component is within the scope of the present invention. Moreover, PCBs (interconnect substrates) on which elastic contact structures are mounted are generally classified as interposers.

In general, mounting an elastic (or compliance) contact structure to a semiconductor device is sometimes referred to herein as &quot; springs on silicon. &Quot;

Carrier assembly

Figures 33-35 show another embodiment of an assembly of electronic components similar to Figures 33-35 of Case-2, which is represented by the term "carrier assembly" for description herein.

33 illustrates an embodiment 3300 of a carrier assembly. The multilayer substrate 3320 may be provided with a plurality of elastic contact structures extending from an upper side (as seen in the drawing) side 3320a to be connected to the electronic component 3302, and an electronic component such as a motherboard (not shown). It has a plurality of elastic contact structures extending from the bottom (as seen in the figure) side 3320b to be connected.

A first group (set) 3322 of elastic contact structures is mounted at a first height on the front surface 3320a of the PCB 3320 and extends therefrom to a predetermined height above the PCB. An electronic device 3370, such as a separation capacitor, is also mounted to the front surface 3320a of the PCB 3320. A second group (set) 3324 of elastic contact structures is mounted at the height of the electronic device 3370 and extends from the same height as the elastic contact structures 3322 of the first group. In this manner, two groups of elastic contact structures 3322 and 3324 can be connected to the planar surface 3302a of the electronic component 3302. This is similar to that described above in connection with Figure 6A.

The bottom surface 3320b of the PCB is stepped as shown, and has contact pads (terminals) 3324, 3326, 3328 disposed at a first height, contact pads 3334 disposed at a second height, and a second one. And contact pads 3342 (only one shown) disposed at three heights. The first height is higher than the second height (away from the front surface 3320a) and the second height is higher than the third height. It is shown that conductors (traces) and vias are provided in the multilayer PCB 3320 in a known manner. It is known to use conductive materials such as molybdenum and tungsten for multilayer substrates 3320 made of ceramic (rather than as printed circuit board materials). The detailed interconnections performed within the body of the substrate 3320 are omitted for simplicity.

A first group of elastic contact structures 3380 (three are shown) are mounted to the contact pads 3324, 3326, 3328 at a first height on the bottom surface 3320b of the PCB, all of which are represented by dotted line A. Same as (i.e. extend the same distance from the bottom of the PCB). These elastic contact structures can be configured in any shape (such as the shape shown in FIG. 3B) as described above and incorporate the desired desired physical properties (such as the plasticity and elasticity described with respect to FIG. 5F) as described above. You can have it.

A second group of elastic contact structures 3302 (two are shown) are mounted to the contact pads 3330 and 3332 at a second height on the bottom surface 3320b of the PCB, all of which are the same (ie, PCB). Extends the same distance from the bottom surface of the plane). The second group of elastic contact structures 3302 is longer than the first group of elastic contact structures 3380 and extends at a distance from the PCB 3320 as indicated by dashed line A. FIG. In this way, the tip (end) of the elastic contact structure 3302 is coplanar with the tip of the elastic contact structure 3380 (see the description of coplanarity described above with respect to FIGS. 6A-6C).

In a similar manner, a third group of elastic contact structures 3338 (one shown) are mounted to the contact pads 3334 at a third height on the bottom surface 3320b of the PCB, all of which are the same (ie , Extend the same distance from the bottom surface of the PCB). The third group of elastic contact structures 3338 is longer than the second group of elastic contact structures 3338 and extend at a distance from the PCB 3320 as indicated by dashed line A. FIG. In this manner, the tip (end) of the elastic contact structure 3338 is coplanar with the tips of the elastic contact structures 3380 and 3382.

Preferably, the elastic contact structures 3380, 3382, 3384 all have the same bends in the middle of their entire length so that they can exhibit the same elasticity to each other. The additional length (eg, the additional length of the contact structure 3338 facing the contact structure 3380) simply extends the end portions (legs) of these elastic contact structures to an appropriate length without changing the elastic properties of the contact structure. It is easily accommodated by making it the predetermined length, without changing the shape of a bending part.

As described above, the ability to include isolation capacitors (or multiple isolation capacitors) in an assembly is a very important consideration that affects the performance of packaging certain electronic components such as microprocessors. In general (ie, approximated), the shorter the path from the semiconductor device to the isolation capacitor, the better (e.g., the inductance is lower). The assembly shown here provides a very short path, resulting in low inductance coupling between the printed circuit board 3320 and an electronic component 3302, such as a high speed microprocessor. Moreover, it is within the scope of the present invention that the electronic component 3302 is an exposed die (eg, an unpacked microprocessor) and this shortens the electrical path (elastic contact structure 3324).

Alternatively, the electronic component 3302 may be packaged by providing a lid (cover, partially shown) 3290 surrounding the electronic component and placing it against the top surface 3320a of the printed circuit board 3320. . Alternatively, a capsule material (eg, epoxy) may be used without the use of a lid to package the electronic component 3302.

The advantages of the configuration as shown in Figure 33 have a planar or stepped bottom surface (as shown), optionally include a separation capacitor 3370, etc., and in the absence of a separation capacitor the top surface of the substrate On each side of this planar ceramic or PCB substrate 3320 a number of elastic contact structures prepared as an interconnect semiconductor die 3302 in the form of flip chips can be provided.

The general advantage of the assembly as shown in FIG. 33 is that the number of vias required (compared to conventional semiconductor packaging techniques) is reduced and the utilization space for the conductors (for example, to implement interconnection schemes and Space to improve routing is wider and costs are reduced.

It is within the scope of the present invention to mount multiple electronic components 3302 to a common (single) substrate 3320. In addition, a recess is provided in the printed circuit board which extends from the upper surface (as seen in the drawing) (or the lower height of the multilayer board not covered by the higher height of the substrate in the selected area). It is within the scope of the present invention that the elastic contact structure derived from the pad in the set has a leg longer than the rest of the elastic contact structure, and the elastic contact structure 3324 is made shorter than the elastic contact structure 3322.

The elastic contact structures 3380, 3382, 3384 are formed in the same manner as the elastic contact structures 2804, 2804 of FIG. 28 so that the printed circuit board 3320 is another printed circuit board (compared to the printed circuit board 2820). It is also within the scope of the present invention that can be plugged in.

By using spring (elastic or compliance) contacts as external interconnects 3380, 3382, 3384, electronic components packaged in this manner are easily tested and / or pre-tested and subsequently socketed. The same interconnect can be used to make a connection to the motherboard or the like, for example by surface mount. It is generally the same in other embodiments disclosed herein.

FIG. 34 shows a low inductance coupling in the manner as described with respect to FIG. 33 with a composite assembly of electronic components 3402 mounted to a printed circuit board 3420 with a separate capacitor 3470 disposed therebetween. An embodiment 3400 is provided that provides a short path. The assembly of the electronic component 3402, the PCB 3420, and the isolation capacitor 3470 constitute the first auxiliary assembly 3425 of the composite assembly 3400.

In this embodiment, the bottom surface 3420b of the PCB 3420 is shown flat rather than stepped (compare FIG. 33) for simplicity of illustration. The entire assembly of electronic component 3402, PCB 3420, and isolation capacitor 3470 is mounted to another printed circuit board 3450, which may be a motherboard, integrated substrate, or the like. In the manner as described above, the bottom surface of the PCB 3420 (compare 3320) has a plurality (only six of which are shown) extending from this bottom surface 3420b to the common plane (compare A in FIG. 33). Elastic contact structures 341 to 3426 are provided.

Motherboard 3450 is provided with a plurality of (only six of which are shown) contact surfaces 3401-3456, which are convenient extensions (on the top surface 3450a of motherboard 3450) of plated through holes. The tip of each of the elastic contact structures 341-3426 is in contact with and soldered to each of these contact regions 3651-3456 and indicated by solder 3440. As described above, when the assembly 3440 is clamped to the motherboard 3450 by clamping means (not shown), it is necessary to solder the elastic contact structures 341 to 3426 to the contact areas 3451 to 3456. There is no.

It should be noted that in any of the embodiments described herein, solder is used to connect between the tip of the contact structure and the conductive pad or region, and an electrically conductive epoxy is used in place of the solder to complete the assembly.

It should also be noted that in any of the embodiments described herein, one electronic component may be assembled to a printed circuit board or the like and a plurality of electronic components may be assembled to the printed circuit board in a similar manner.

The embodiment 3300 of FIG. 33 is also mounted on top of the component part 3450 of FIG. 34 in the manner shown in FIG. 34 (eg, by replacing the assembly 3300 with the assembly 3425). Belongs to the scope of.

35 shows a short path showing low inductance coupling in the same manner as described with respect to FIG. 33 with an electronic component 3502 mounted to a printed circuit board 3520 with a separate capacitor 3570 disposed therebetween. Another embodiment 3500 is provided. The assembly of the electronic component 3502, the PCB 3520, and the isolation capacitor 3570 constitute a first auxiliary assembly 3525 (compare 3425) of the composite assembly 3500.

In this embodiment, an interposer substrate 3560 similar to the interposer 1720 of FIG. 17B is disposed between the assembly 3525 and another printed circuit board 3550, such as a motherboard, an integrated substrate, and the like. A number of elastic contact structures 3601 to 3566 (only six are shown) extend upwards (as shown) from the top surface 3560a of the interposer substrate 3560 and their tips extend to the bottom of the assembly 3525. Each of the contact pads 3251 to 3526 on the face (ie, bottom 3520b of PCB 3520) is in contact, in which case there is no solder. (Elastic contact structures 3601 to 3566 are shown in detail without the details of the wire stems overcoated for simplicity.)

The motherboard 3550 is provided with a number of contact structures 3551 to 3556 (only six are shown), which engage with plated through holes in the motherboard 3550 as in the case of the motherboard 3450 of FIG. 34. .

Multiple (only six shown) bumped (generally inelastic) contact structures 3568 (only one shown in the drawing) extend from the bottom surface 3560b of the interposer 3560 and the contact area on the PCB 3550 Are placed in contact with 3535-3556. These contact structures 3568 may or may not be soldered to the contact regions 3551-3556 (as shown).

As shown, each of the PCB 3520, the interposer 3560, and the PCB 3550 is provided with a plurality of (only two) holes aligned with each other, through which the plurality of threaded studs 3530 are provided. 3532). The nuts on the ends of the studs 3530 and 3532 cause the entire assembly 3550 to be pressurized so that the elastic contact structures 3551 to 3566 on the top surface 3560a of the interposer 3560 are the bottom surface 3520b of the PCB 3520. It is possible to substantially contact the contact pads (3521 to 3526) on the (). If the bottom contact structure 3568 of the interposer 3560 is preferentially soldered to the contact areas 3551-3556 on the PCB 3550, there is no need to provide holes in the interposer 3560. In either case, this secures the assembly 3525 to the PCB 3550 by mechanical means (eg, studs and nuts as described above, optionally nuts and bolts, cams, levers, spring clips, etc.). Represent various technologies. The advantage of this embodiment is that the pads 3251 to 3526 on the bottom surface 3520b of the PCB 3520 are different from the soldered connections shown on the bottom surface 3420b of the PCB 3420 in the previous embodiment 3400. No solder is used to make contact with the furnace.

It is also within the scope of the present invention to use other means than the threaded studs 3530, 3532 to secure the assemblies 3520, 3560, 3550 with clips, latch engagement springs, and the like.

PCB 3550 (and optionally interposer 3560) constitutes a test jig, such as for conducting tests and pretests on electronic component 3502, and then moves assembly 3525 to the motherboard of the notebook computer. Mounting on another PCB 3550 such as is also within the scope of the present invention.

It is within the scope of the present invention to use another type of interposer (ie, another form other than FIG. 17B) in the assembly 3550.

As described with respect to the embodiment 3300 of FIG. 33, the lead 3390 may be provided in any of the embodiments described herein in which the electronic component (eg, 3502) is mounted to the substrate 3520. Alternatively, the adhesive can be added to package electronic components.

36A, similar to FIG. 36 of Case-2, illustrates another technique for assembling electronic components (embodiment 3600), where multiple (only two), such as “exposed” (unpacked) semiconductor devices are shown. ) Electronic components 3602, 3603 are on one side 3620a of interconnect substrate 3620, and many (only two are shown) electronic components 3608, 3609, such as exposed (unpacked) semiconductor devices. Another side 3620b of this interconnect substrate 3620 is mounted in the manner described in FIG. 25. (This indicates that the silicon chips are packaged side by side on the card / substrate.) That is, for example, the electronic components 3602, 3603, 3608, 3609 are optionally by suitable insulating adhesive 3630 (compare 2530). May be secured to the interconnect substrate 3630. (Optionally, the electronic component may be fixed to the interconnection substrate by pressing a spring clip or the like as described above.)

Each of the electronic components 3602, 3603, 3608, 3609 is provided with a plurality of elastic contact structures 3610 extending from the front face, and the interconnect substrate 3620 has a corresponding plurality of corresponding ends for receiving the distal ends of the elastic contact structures. Contact pads are provided. As shown, each contact pad is suitably provided with some conductive epoxy 3612 to accommodate the distal end of the corresponding elastic contact structure 3610.

In this manner, a large number of electronic components can be interconnected (eg, PCB) using the available space on the interconnection board more effectively than packaging semiconductor devices (eg, plastic packaging, leaded). Can be mounted on

36B and 36C illustrate exemplary applications of the techniques shown in FIGS. 25 and 36A, respectively.

FIG. 36B shows that a number of exposed semiconductor devices 3650a through 3650r are mounted on the surface of the interconnect substrate 3652 in two rows and the semiconductor devices are effectively placed in close contact with each other in close proximity to each other. A number of edge connector pads 3654 are shown as being at the edge of the interconnect substrate 3652, wherein the edge connector pads are placed on a pad that accommodates the distal end of the elastic contact structure mounted to the semiconductor devices 3650a to 3650r. Connected by conductive traces (not shown).

FIG. 36C shows that a number of exposed semiconductor devices 3660a through 3660i are mounted on the surface of the interconnect substrate 3662 in a row and the semiconductor devices are effectively placed in close contact with each other in close proximity to each other. A number of edge connector pads 3664 are shown as being at the edge of the interconnect substrate 3662, which edge pads receive (shown in) a pad that accommodates the distal end of the elastic contact structure mounted to the semiconductor devices 3660a to 3660i. Is connected by conductive traces.

As shown in Figure 36A, the interconnect substrate 3620 is raised or oblique (although not preferably raised) with respect to the electronic component. That is, it is not parallel to the electronic components 3602, 3603, 3608, 3609. This figure shows one of the many advantages of using an elastic contact structure (eg, an electronic component), that is, the elastic contact structure inherently compensates for small deviations in planarity. As described above, in order to accommodate nonplanarity, it is sometimes desirable for the elastic contact structure to have a plastic deformation of a certain size (as opposed to elastic deformation). In general, contact structures exhibiting elasticity and plasticity are referred to as "compliance" contact structures.

In this embodiment and other embodiments described herein, the final assembly is intended to protect electronic components from environmental damage, including mechanical (such as processing assemblies) and environmental (such as dust and moisture). Can be completely covered with a suitable insulating material (eg glob top epoxy). This allows for a final, fully packed assembly. It is also within the scope of the present invention that a suitable heat sink is coupled to such a final assembly in any suitable manner (eg, by allowing the end of the heat sink to be embedded very close to the electronic component within the overlapped glow top epoxy). Included in

Another embodiment 3700 of the FIG. 37 electronic component, similar to that of FIG. 36 of Case-2, wherein two assemblies (“previous assemblies”), each of which is compared to the assembly 3600 described in FIG. 3710 and 3712 are stacked.

As shown, the interconnect substrate 3720 of the assembly 3710 is provided with a plurality of (only one) contact pads 3722 on its bottom surface 3720b (as shown). The interconnect substrate 3740 (compared to 3620) of the assembly 3712 is provided with a plurality (only one) of elastic contact structures 3702 on its upper surface 3730b (as shown).

In the process of integrally covering assembly 3710 and assembly 3712 with a suitable insulating material 3730 (compare 2530), the material 3730 fills the gap 3732 between the two assemblies 3701 and 3712. Fill it. With regard to this integral covering, the assemblies are pressed against each other in the same manner as described above so that the tips of the elastic contact structures 3702 are in contact with the contact pads 3722.

26, 27, 29, 30 or the like, such as a spring clip as described with respect to FIG. 35 or other mechanical means such as described with respect to FIG. 35, may be used instead of the adhesive 3730 to hold the assemblies 3710, 3712 together. It is included in the scope of the present invention.

In the drawings of this embodiment, for simplicity, the interconnection substrate is not raised (compared to the interconnection substrate 3620 of FIG. 36).

Most of the above has described a number of electronic assemblies in which electronic components (eg, xx02) are assembled to the interconnect substrate xx20 in a flip chip manner. In other words, the electronic components were made parallel to the interconnect substrate. (In the case of using the reference numeral xx02 or the like throughout the description of the present specification, any suitable number such as “25” is used instead of the letter “xx” (while maintaining the position), that is, “xx02”. Can be represented as “2502”, “2602”, “2702”, etc.)

According to one aspect of the present invention, an electronic component can be mounted (and connected) to the interconnect substrate in an edge direction.

As described above (for example, as described with respect to FIG. 25), a plurality of cards 3710 and 3712 are manufactured in advance, a pre-test is optional, a list is made, and then shown in FIG. Assemble into more complex assemblies as shown. Such auxiliary assemblies (ie, 3710 and 3712) can be easily stacked as shown, simply by providing contacts on the card and providing a card-to-card interconnect structure (e.g., 3742) between the card and the card. It becomes possible. As a general suggestion, this technique is superior to "3D" lamination of individual silicon components (electronic components) and may be applied to multiple chip modules or other electronic components of memory components.

FIG. 38, similar to FIG. 38 of Case-2, illustrates an embodiment 3800 of an electronic assembly, where a number of electronic components 3802, 3803, 3804, 3805, 3806 are along an edge to an interconnect substrate 3820. Is mounted.

Each of the electronic components 3802-3806 is provided with a plurality of (only one) elastic contact structures 3842-3846, of which representative elastic contact structures 3822 are shown in more detail in FIG. 38A. The upper surface 3820a (as shown) of the interconnect structure 3820 is provided with a plurality of (only five) contact pads 3822 to 3826. In practical use, the tips (eg, 3842b) of the elastic contact structures 3832 to 3846 contact each of the elastic contact pads 3822 to 3826 in a manner similar to that shown in FIG.

A plurality of pads (386 shown), generally larger than the number of electronic components 3802-3806, interconnect boards 3820 at locations between the outer plates of the electronic components 3802-3806. It is provided on the upper surface 3820a of the. Spring clips 3831 to 3836 are mounted to the pads 3861 to 3866, respectively, and their shape is molded as follows.

An exemplary spring clip 3832 extends upwardly from the top surface 3820a of the interconnect substrate 3830, the shape of which extends to the left (as shown) and to the right (as shown). It is formed to have a bent portion. All spring clips are preferably formed in the same way, with the bends facing away. The bend extending in the left direction of the spring clip 3832 cooperates with the bend extending in the right direction of the adjacent spring contact 3831 (to the left as shown) to form a gap somewhat smaller than the thickness of the electronic component 3802. Form. In this manner, the electronic component 3802 fits snugly between the spring clips 3831 and 3832 on either side so that they can be held in place along the edge with respect to the surface 3820a of the interconnect substrate 3820. do. Electronic component 3803 is similarly held in place along the edge relative to interconnect substrate 3820 by spring clips 3832 and 3833. Electronic component 3804 is similarly held in place along the edge with respect to interconnect substrate 3820 by spring clips 3835 and 3834. Electronic component 3805 is similarly held in place along the edge relative to interconnect substrate 3820 by spring clips 3834 and 3835. Electronic component 3806 is similarly held in place along the edge with respect to interconnect substrate 3820 by spring clips 3835 and 3836.

As with the spring clips as described above (e.g., 2630 and 2632), the spring clips 3831 to 3836 are manufactured in the same manner as the elastic contact structures, and are optional, but not essential, as the spring clips are connected to the interconnect substrate. A spring clip forms an electrically conductive path between each electronic component held in place.

For example, as described with respect to FIG. 26, the pads 3822 to 3826 are pretreated with solder so that once the electronic components are pressed into place (between each spring clip), the solder is reflowed to bring the pads 3822 to 3826. ) And the entire assembly can be passed through the furnace such that a solder connection is made between each tip of the elastic contact structures 3832 to 3846.

As shown in Fig. 38A, the elastic contact structure 3842 of the present embodiment is a pressing force F for holding the electronic component 3802 on which the elastic contact structure 3842 is mounted in a plane (not an orthogonal plane) in place. It is formed into a shape to react to.

It is also within the scope of the present invention for the elastic contact structure to have a shape that is formed to react to a pressing force that is directed at substantially any angle with respect to the electronic component on which the elastic contact structure is mounted. As an example, in the embodiment of the “probe” illustrated by FIGS. 9A and 9B, the pressing force is applied from below the substrate on which the elastic contact structure is mounted. Although described herein as having a specific exemplary shape, the scope of the present invention is not limited to any of the illustrated shapes.

It is also within the scope of the present invention that the electronic components 3802, 3803, etc. do not form a right angle to the substrate 3620.

Dust test for interface unit

The use of elastic contact structures in probe cards for pre-testing and performing functional tests on electronic components has already been described above with respect to FIGS. 10A-10G as an example.

In some applications, it is important to pinpoint and analyze failures of complex integrated circuit devices. In one example, a partially packaged device during a test (DUT) has a PCB with conductive areas contacted by “pogo pins” (straight pins with coil springs) that extend from the test interface substrate through the spacers. Can be plugged into the socket on the top. The assembly of the test interface substrate, spacer, PCB, socket and DUT is inserted into the opening in the vacuum chamber of the E-beam probe device as an example so that the DUT can be probed by the E-beam under vacuum atmosphere. This is only one application for the DUT test for the interface device according to the present invention.

39 illustrates a test interface device 3900 in accordance with the present invention. Semiconductor die 3902 has a plurality of solder bumps 3904 arranged in one example on a bottom surface in a row (as shown). The “space transducer” substrate 3910 has a corresponding plurality of elastic contact structures 3912 disposed on its top surface (as shown). The distal ends of the elastic contact structures 3912 are arranged at the same pitch (spacing) as the solder bumps 3904. Spatial converter 3910 is a substrate formed of alternating layers of insulating material and patterned conductive material (eg, foil) in the same manner as a multilayer PCB. The bottom surface (as shown) of the space transducer 3910 is provided with a plurality of contact structures 3914 arranged at greater intervals (at more spaced distance from each other) than the elastic contact structures 3912. The space transducer 3910 presses the die against the distal end of the elastic contact structure 3912 in the same manner as described above (see, eg, FIG. 12B for the gang-transfer of the contact structure). The distal end can be assembled to the semiconductor die 3902 by soldering (or brazing) to the semiconductor 3902.

Interposer substrate 3920 is provided with a plurality of elastic contact structures 3922 on its upper surface (as shown) and its lower surface (as shown) in a manner similar to any of the interposer substrates described above. A plurality of elastic contact structures 3924 are provided in the.

The probe card substrate 3930 is provided with a plurality of contact pads 3932 on its upper surface (as shown). Means are provided for securing the probe card 3930 to the space transducer 3910, between which penetrates a hole in the probe card 3930, optionally an interposer 3920, to form a screw in the space transducer 3910. It is clamped by an interposer such as a bolt extending into the hole. In this manner, multiple semiconductor assemblies 3902, 3910 can be tested by a test circuit (not shown) coupled with the probe card 3930.

The technique of the present invention is simpler and more economical than conventional techniques for performing a "known good die" test. Conventional probe cards of the prior art include contacts mounted therein, which generally means that a unique (for example, expensive) probe card is needed for the particular device to be tested. The probe card 3930 of the present invention has greater “replacement” and can be easily applied to unique devices by the space converter substrate 3910 produced easily and inexpensively.

Several additional embodiments

Hereinafter, specific modifications to the various techniques described above will be described, which are included in the scope of the present invention.

Non-conductive stem

As described above, the wire (eg 102) basically serves as a medium (fallwalk, mandrel) overcoated (top structure) formed (plated) thereon. Thus, in many cases, as long as the mandrel is mounted to the substrate, it is generally not important whether the mandrel is electrically conductive, nor is it generally important whether the mandrel is shaped and the mandrel is overcoated ( As a general suggestion, it is preferable to make the outermost layer plating part electrically conductive).

In view of this, it is also included in the scope of the present invention to form the wire stem (long flexible member) from an insulating material such as plastic (for example, nylon), polymer, or the like. This is evident within the scope of those skilled in the art in which almost all of the invention relates to plating conductive metal materials (eg nickel) on the mandrel. In one example, the surface of the insulating mandrel may be first activated to be ion water soluble, followed by non-electrode plating with a first layer of conductive material, followed by electroplating with a second layer of conductive material.

It is known that various various plastic materials (such as polymers) liquefy to high pressure, for example, to form the epidermis under atmospheric pressure.

40A illustrates a technique for supplying polymer material under high pressure through nozzle 4002. The nozzle is placed in close proximity with the terminal 4012 on the electronic component 4008 (compared to 108) and the liquefied polymer flows out of the nozzle to contact the terminal. The nozzle is then moved in the z direction and in any of the x and y directions to form stems 4002 (compare 102). Once a stem is produced in this manner, the pressure is turned off and the nozzle is moved away, leaving its distal end. In this manner, the function of the nozzle 4004 in forming the wire stem is very similar to the function of the wire bonding capillaries 104 and 204 as described above. When the stem is cured, it is easily overcoated with a conductive material in the manner described above.

As described above, certain advantages also arise from the use of multiple capillaries to simultaneously form multiple wire stems on electronic components. In many cases in the same way, various nozzles 4004 can be used to form various non-conductive stems on electronic components.

40B illustrates a nozzle head 4020 having various (only two shown) orifices 4022 that allow non-conductive material to penetrate and be supplied to the surface of the substrate 4028 to form various stems 4026. . (Terminals and the like have been omitted from this figure for the sake of simplicity.) It is clear that the stems formed in this manner all have the same shape, which is indicated by dashed line 4024 terminated by an arrow. In practice, the nozzle head 4020 is a perforated plate having a plurality of orifices 4022 arranged in a suitable manner, such as a rectangular arrangement.

It is not currently suggested that the technique described with respect to Figures 40A and 40B is "preferred." These are primarily provided to illustrate that the stem 4002 that is shaped to be overcoated does not need to be a conductive wire. However, any suitable insulator material may be used for the stem within the gist of the present invention.

Stem Attached Without Coupling

As fully described herein, the wire coupler can be used to mount (couple) the free end of the supply wire to the contact area on the substrate, the free end of the supply wire being derived (when coupled to the substrate). Becomes the proximal end of.

Considering that the overcoating (top structure) acts to secure the resulting contact structure to the substrate, within the gist of the present invention, the wire is not mounted corresponding to the substrate but rather mounted therein in an unbonded manner. In general, any technique of securing the free end of the supply wire to the substrate will be sufficient.

41A shows capillary head 4104 (compare 104), through which wire 4102 is fed. In this embodiment, the terminals 4112 (compare 112) are disposed over the substrate 4108 (compare 108) and overcoated with a material that can be penetrated by the tip of the wire. Bonding, such as ultrasonic coupling, is not required to secure the proximal end of the resulting wire stem to the electronic component. Rather, the wire 4102 is not sufficiently taut and the material 4114 is not sufficiently penetrating (ie, plastically deformable) so that the resulting wire stem (see FIG. 41B) is in place during the performance of overcoating the wire stem. maintain. In this embodiment, the wire stem is shown as a rigid wire stem, and for the sake of clarity the wire stem can be formed into any elastic shape described herein (ie, see FIGS. 2A-2E). Material 4114 may be a solder, conductive epoxy, or any other suitable “penetrate” material.

Within the gist of the present invention, the above-described techniques for inserting (rather than coupling) wires to terminals (or pads, etc.) on electronic components can be used to mount the fin-like structure to electronic components such as semiconductor packages. Wire materials include beryllium-copper, cobar (tm), elastic steel, and the like (relatively strong materials). Optionally, the resulting finned structure can be overcoated to securely secure the pin to the terminal (ie, by plating with gold).

Within the gist of the present invention, wire stem 4012 is overcoated with conductive material 4116 in the manner described above, as shown in FIG. 41C.

Within the gist of the present invention, wire 4102 is made of any suitable material, such as high elastic carbon fibers, for insertion into mass 4114 (such as solder, or conductive epoxy).

Manually shape wire stem

As described above, the wire stem is shaped to have an elastic shape by adjusting the relative movement of the capillary (ie, 204) facing the substrate (ie, 208).

In the gist of the present invention the shape of the wire stem is increased or completely satisfied by external means rather than by the drive of the capillary.

42 illustrates the free end of the supply wire 4202 (compare 202) extending from the end of the capillary 4204 (compare 204). As indicated by the dashed line, capillary 4204 is a “mechanical” member (such as a plate provided with at least one through hole for coupling the tip of the wire to the terminal 4212; 112 on the substrate 4208; 4220 passes through opening 4202.

Once the free end of the supply wire is coupled to the terminal, the drive of the capillary and instrument member is controlled to set the desired spring shape for the resulting wire stem. Such driving includes capillary and block driving in the same x and y directions. (The instrument member also drives in the z direction, as with the capillary.) If the instrument member is simply held in place, the wire stem can be manually held at any position along its longitudinal direction during the shaping process. After removing the wire stem to have a distal end (tip), the block can simply be raised to separate the wire stem.

Within the gist of the present invention, two or more instrument members each having a corresponding hole and through which the capillary traverses to join the wire to the terminal can be used in a similar manner to augment the shaping process. Also, with respect to the instrument block, the capillary tube (after joining the free end of the supply wire to the terminal) can draw the wire to a sufficient length, and the overall stem shaping process can be performed by controlling the driving of the two blocks.

As will be evident, the use of a mechanism (ie 4220) outside of the wire coupler to shape the wire not only creates a tight bend (twist) but also the heavy amount of displacement of the capillary tube (usually related to any shape of the bond wire (to a large extent). Allow the wire to be shaped into a shape that avoids).

FIG. 42A takes the concept of FIG. 42 further advanced, and is particularly beneficial in situations where there is an array of wire stems, and it is difficult for the capillary to be adjusted without bumping into adjacent wire stems. As shown in FIG. 42A, a plurality (only two are shown) of wire stems 4232, 4234 are coupled to terminals 4236, 4238 on the surface of electronic component 4240, respectively. The wire stem is formed so as to protrude vertically upwards from the surface of the electronic component (for example, compare FIG. 4B). Three outer mechanisms 4242, 4244, 4246, each formed as a rigid plate member with multiple holes, are introduced over the protruding wire stem. The instruments, as shown, are introduced over the wire stem and spaced apart from each other. As shown in Figure 42B, by securing the instruments 4242 and 4246, and by moving the instruments 4244 (towards the right, as shown), the wire stem can be "gang shaped". After shaping the wire stem, the instrument is removed. Any wire stem shape prevents removal of the instrument. To avoid this problem, the mechanism is preferably formed as a honeycomb structure, as shown in Fig. 42C.

FIG. 42C shows another embodiment of a gang shape that includes a plurality of (4 × 4 arrangement shown) protruding wire stems and uses an external instrument (one of which is shown) formed as a honeycomb structure. In this embodiment, the function of the two honeycomb plates 4252, 4254 is moved to the array (see arrows in the figure) so that the space between the individual branches of the honeycomb plate is separated from the individual wire stems [instruments 4242, 4244, 4246]. In a manner similar to that of a hole]. While two honeycomb plates 4252 and 4254 are shown per vertical position (along the wire stem), for example, stabilizing the wire stem from moving towards the right side (as shown) or the wire stem is turned to the right side. It is sufficient to use only one honeycomb 4252 to twist. In addition, within the gist of the present invention, the top honeycomb pairs can form their shape and then deform the wire stem-to several normal heights. The honeycomb plates discussed herein are suitably formed by an electrical discharge machining (EDM) process and can be made of silicon, iridium, ceramics, metals, tungsten carbide, and the like. Small spaces between the branches of honeycomb plates of less than 10 mils are easily performed.

Overcome sticking and preclean wire

As described herein above, applying ultrasonic energy to the capillary during shaping the wire stem is used to overcome sticking when the wire moves from the capillary.

Within the gist of the present invention, means other than ultrasonic energy may be used to reduce the sticking of the wire as it moves from the capillary, which means may be increased or may be used instead of the ultrasonic energy described above.

Figure 43A illustrates a capillary 4304 (compare 204) with a feed wire 4302 (compare 202) through. The capillary, as is known, has a bore extending axially through its body. The tail of the capillary (as shown in the figure, top) is frustoconical shaped by a housing (cover) 4320 having an opening 4322 and the tail of the capillary fits snugly into the opening and extends through the supply wire. The orifice 4326, which communicates with the opening 4324 and the internal cavity 4328, is closed by the housing. An inert gas or “forming” gas is supplied to the cavity 4328 through the orifice 4326 at a pressure (above atmospheric pressure) by the supply reservoir 4330. The gas is, for example, argon, hydrogen, carbon dioxide, or carbon monoxide.

During wire bonding, gas flows into the cavity 4328, and the gas moves from the cavity in two directions as indicated by the arrows. The gas flows downwardly (as shown in the figure) through the capillary to reduce the stiction of the wire as the wire acts from the capillary. The gas flows upwards (as shown in the figure) to preclean foreign matter in the supply wire (such as dust) before the wire enters the capillary.

Within the gist of the present invention, the gas injection orifice 4336 (compare 4326) is incorporated into the capillary itself (eg, the midpoint of the capillary) to eliminate the need for another housing. However, since the opening in the distal end of the capillary tube is typically too large (e.g., to facilitate insertion of the feed wire into the distal end of the capillary tube), a small opening (slightly larger than the diameter of the wire 4302) may cause It is arrayed over the distal end of capillary 4344 (comparative 4304) to regulate (eg, reduce) release. This is illustrated in Figure 43B.

Orientation of Contact Structure

 As mentioned above, multiple contact structures can be mounted to the electronic component in a rectangular array.

As discussed above, when an electronic component such as a semiconductor device operates, it generates heat that causes the semiconductor device to expand around the center of the semiconductor device. In many cases, packages (or PCBs, etc.) to which semiconductor devices are interconnected are less affected by heat (because they do not generate heat) and expand to a lesser extent than semiconductor devices. This difference in thermal expansion apparently stresses on the interconnects and consequently one end of the interconnect is displaced more than the other end of the interconnect. The same phenomenon occurs when a heat generating semiconductor package is connected to a printed circuit board (PCB).

According to the invention, the orientation of the plurality of shape contact structures mounted on the electronic component is arranged to accommodate the difference in coefficient of thermal expansion between the electronic component and other electronic components to which the electronic component is interconnected.

Figure 44 illustrates an electronic component having an array of elastic contact structures mounted on its surface. Each contact structure is represented by a dashed line (origin) in this example. Edge contact structures 4430a, 4430b, 4430c, 4430d are illustrated in perspective view. The expansion of electronic component 4402 is represented by arrow E, with all points away from the center of electronic component 4402.

In this example, each elastic contact structure has an orientation that can be arbitrarily defined as the plane in which the wire stem of the contact structure is bent. As is apparent from this figure, the orientation of the four corner contact structures 4430a, 4430b, 4430c, 4430d is arranged such that the thermal expansion (dotted line) intersects the plane of the contact structure at a normal (e.g., 90 °) angle. do.

The straight forward approach aligns all of the contact structure in a square on the upper right side (as shown) of the electronic component parallel to the contact structure 4430a, and the upper right side of the electronic component parallel to the contact structure 4430a (as shown). Similarly) align all of the contact structure in a square, align all of the contact structure in a square on the lower right (as shown) of the electronic component parallel to the contact structure 4430b, and electrons parallel to the contact structure 4430c. Align all of the contact structure to the square of the lower left side (as shown) of the component, and align all of the contact structure to the upper left (as shown) square of the electronic component parallel to the contact structure 4430d. However, given that the expansion from the center of the electronic component generally occurs radially, within the gist of the present invention, each contact structure (only four of those illustrated) has a plane of the contact structure at which the expansion vectors are individually at close angles. It is arrayed to penetrate. In Figure 44, the contact structure is shown aligned in "45 '" for clarity of illustration.

Also within the gist of the present invention, the tip "leg" of the contact structure is deflected from the base "leg" of the contact structure in the direction of thermal expansion, thereby supporting accommodation of thermal expansion.

Curvature

As described above, as with respect to Figure 5F, the elastic contact structure of the present invention reacts with the compressive force upon compression. The range of spring coefficient k has been described.

Within the gist of the present invention, other bending mechanisms (eg, rather than compression) can be introduced to improve the elastic properties of the contact structure.

Figure 45 illustrates an elastic contact structure 4500 similar to the contact structure shown in Figure 2C. The proximal end of the contact structure is mounted to the substrate 4508 (compare 2508) at point a, and the contact structure has a remote tip (at point f). The upper “leg” of the contact structure (eg, the vertical portion of the contact structure between the points f and e) is the lower contact “leg” of the contact structure (eg, between the points a and b). Deflection (for example in the x direction)). When the compressive force F acts on the contact structure, the compressive force at least partially exerts a torsion as indicated by the arrow T. Torsion as a bending mechanism can advantageously be used to combine the elastic and / or composite properties of the contact structure. Preferably, the contact structure is such that torsion occurs symmetrically with respect to the “arm” of the contact structure (the horizontal part of the contact structure between points b and c and the horizontal part of the contact structure between points d and e). Is formed.

Within the gist of the present invention, the contact structure can be designed (eg shaped) to exert a compressive force through a number of bending mechanisms including compression, torsion, tension, bending, and the like. In addition, additional forces (eg, in addition to compression) can be accommodated with other bending mechanisms. For example, the contact structures 4430a, 4430b of the previous embodiment (FIG. 44) can be basically aligned to accommodate thermal expansion upon tensioning and exert a compressive force upon compression.

Inductance reduction

As discussed above, for example with respect to FIG. 5E, in many cases, it is appropriate to reduce the added inductance by the relatively long overall length of the contact structure in order to reduce the inductance effect. In the previously described embodiment of FIG. 5E, the elastomeric material may be provided with conductive paths between the remote tip and the proximal end of the wire stem shorter than the overall length of the wire stem (or, more precisely, the close point on the terminal at the remote tip of the wire stem). Can be used to provide.

In Euclidean geometry, the short passage between two points is a straight line. According to the invention, the separating and distinguishing wire is provided with a shaped wire stem so as to provide a small length of electrical passage in which the contact structure is mounted between the far end of the contact structure and the terminal (for example).

46A-46F illustrate a technique for minimizing the electrical passage between two points and the inductance derived therefrom in a similar manner as discussed for the elastomeric material 564 of FIG. 5E.

46A shows a “jumper” wire 4602 with one end connected to a terminal 4612 (compare 212) on a substrate 4808 (compare 208). Jumper wires 4602 are connected in the manner of the wire stems described above, but as will be apparent, are not intended to function in an elastic manner. Wire 4602 extends almost straight upward (downward as shown in the figure) from the surface of substrate 4608 but is preferably slightly twisted. Wire 4602 is a flexible wire, such as a 0.0008 inch gold wire.

46B illustrates a masking layer 4604, such as a photoresist, applied to the remote portion of wire 4602. FIG. This can be done simply by “immersing” the substrate into a container of liquid masking material (not shown).

46C illustrates a step where a small portion of masking material 4604 is removed from most of the tip of wire 4602. This exposes the material of the wire 4602 for subsequent overcoating (eg plating) as described below.

46D illustrates the step in which another wire 4604 is connected to terminal 4612. This wire 4606 is formed (e.g., bent) in the manner described above (e.g., see Figure 2A and below) to support the overcoating (i.e., overlapping structure) to form an elastic contact structure. Elastic shape). The far end (tip) of the resulting shaped wire stem is intimately directly above (far below, as shown in the figure) of the wire 4602.

46E shows that wires 4602 and 4606 are overcoated by nickel plating on the wires in the manner described above (eg, see FIG. 5 and below). Overcoat 4608 completely covers (surrounds) wire stem 4606 and covers terminal 4612 to securely fasten wire stem 4606 there. Overcoat 4608 connects wire 4602 and remote ends of wire stem 4606 to each other. In this manner, a substantially straight passage is formed between the terminal 4612 and the far end of the wire stem 4606. The resulting elastic contact structure 4630 exhibits low inductance regardless of the overall length of the elastic portion of the contact structure (eg, overcoated wire stem 4606) as a result of this nearly straight passage.

Another approach to the techniques described above, including masking jumper wires 4602 to prevent plating (and therefore maintaining its elasticity), is to use insulated wires as jumper wires. For example, a gold wire with an insulating coating can be used as the jumper wire. (See Fig. 46G showing an insulated wire 4650 having a conductive core 4652 and an insulating coating 4474). Such insulated wires are well known and one of them is described in the referenced US Pat. No. 4,860,433. Such insulated wires are useful in bonding to the leadframe of the semiconductor die to prevent the wires from shorting to each other in the process of plastic molding. For example, a gold wire having a diameter of 0.0008 inches and covered by a polymer (or plastic material such as urethane) can be used as a jumper wire. In general, the insulating coating on these wires does not contribute significantly to the mechanical properties of the wires.

In use, (ie, before joining the proximal end of the jumper wire to the terminal and withdrawing a predetermined length of the jumper wire), insulation at both ends of the insulated jumper wire will be removed during the process of cutting the supply wire, The exposed wire is left underneath to facilitate plating.

Since jumper wire 4602 provides a basic electrical path from component 4608 to other components (not shown), wire stem 4606 need not be electrically conductive within the spirit of the present invention.

Tapered wire stem

As described above with respect to Figures 5G and 5H, it is possible to "cut" the length of the overcoat in order to change the properties of the resulting contact structure.

According to the invention, it is possible to "cut" the length of the wire stem in order to change the properties of the resulting contact structure.

Figure 47 illustrates a wire stem 4702 mounted to a terminal 4712 on a substrate 4708 (e.g., an electronic component) by its proximal end. Depending on the material of the wire stem (eg aluminum), the cross section of the wire stem can be selectively reduced with a suitable solvent (eg potassium hydroxide). In this configuration, the wire stem is shown to have a smooth transition from the region of the smallest diameter at its far end to the region of the largest diameter at its proximal end. Within the gist of the present invention, the diameter can be “stepped” by immersing the proximal portion of the wire stem (eg, upper two thirds as shown) in a solvent. In this example the wire stem is shown in a straight line for clarity of illustration. The technique of the thin part of the wire stem can equally be applied to the wire stem having an elastic shape.

Wire stem removal

As described above, upon establishing the formation of the superimposed superstructure (eg, plated overcoating), the underlying polework (wire stem) is once overcoated, and the underlying wire stem is the overall mechanical of the resulting contact structure. Contributes slightly to (eg, structural) properties.

According to the invention, the wire stem can be removed after being overcoated.

48A illustrates a wire stem 4802 mounted to a terminal 4812 on a substrate (eg, an electronic component) by its proximal end. Masking material 4810, such as a photoresist, is applied over a small portion of the far end (tip) of the wire stem. Next, as shown in FIG. 48B, the wire stem is overcoated with at least one layer 4820 of the material in accordance with the techniques disclosed above. As shown in FIG. 48C, masking material 4810 is removed (by dissolving, washing, etc.). This results in an overcoated contact structure with the far end of the wire stem exposed through the overcoating material. Next, as shown in FIG. 48D, the wire stem is at least partially removed by immersing the contact structure in a solvent (eg, potassium cyanide for gold wire stem, potassium hydroxide for aluminum wire stem). This results in a hollow overlapping structure 4820 (overcoating). The process can end at this point. The resulting contact structures can then be interconnected to other electronic components (not shown). In the case of soldering the far end of the contact structure to another electronic component, the solder is introduced into the remaining cavity by capillary action by the removed wire stem to improve the electronic connectivity and structure integration of the interconnects.

Optionally, an additional overcoating layer 4824 is applied by plating, such as a thin layer of hard gold, which will fill the hollow opening in the overcoating 4820, as illustrated in FIG. 48E. . Again the wire stem of this example is shown in a straight line to illustrate the clarity.

Eutectic contact tip

In general, through the above-described embodiments, it is an overcoating material to form contacts in other electronic components, and the wire stem is “burned” in the overcoating. If the wire stem has better contact properties (eg, conductivity, corrosion resistance, etc.) than the overcoating material, it will be slightly counterintuitive to “burn” the wire stem interior of the overcoating.

According to the invention, the wire stem is allowed to be exposed at the far end (tip) of the contact structure.

49A illustrates a contact structure 4900 fabricated in the following manner. Wire stem 4902 is mounted to terminal 4912 on substrate 4902 by its proximal end and is made of a material such as gold wire. Thin film layer 4920 of a material such as tin is applied over the gold wire stem. A mass of masking material 4922, such as a photoresist, is applied to the far end of the tin overcoated wire stem. (Optionally, the masking material is applied over the wire stem 4902 before the tin overcoat 4920 is applied). A layer 4920 of spring material, such as nickel, is applied over the tin overcoated wire stem. Next, masking material 4922 is removed. This results in a contact structure similar to the contact structure shown in Fig. 48C.

The overall contact structure is then sufficiently heated so that the gold 4902 and tin 4920 form a gold-tin eutectic. Within the gist of the present invention, as materials capable of forming eutectic in addition to gold and tin, for example, lead and tin may be used.

Next, the contact structure is cooled. The property of eutectic materials is that they expand in the molten state. For example, gold tin eutectic expands to 4% (volume). Spring material (eg nickel) 4924 is not affected by the heating and cooling cycles. Thus, upon cooling, the tin overcoated wire stem expands and is partially forced from the far end of the contact structure, resulting in a shape as shown in Figure 49B, where the shortest portion of the contact structure 4900 is located on other electronic components. Gold tin eutectic material 4930, which is “braze ready” for mounting to contact pads. The braze ready tip is substantially part of the initial wire stem 4902 (and additional layer 4920) that “leaks” out of the open end overcoating. When the eutectic wire stem is cooled, it is the cavity remaining on the back, as indicated by the two cavities 4940 illustrated in FIG. 49C.

Within the gist of the present invention, an outer layer of gold added may be applied to the contact structure (eg, on a nickel layer) before and after heating / cooling of the contact structure.

As long as the wire stem is joined to the contact area at its proximal end and its distal end is exposed (to connect to other electronic components), within the gist of the present invention the overcoating is formed of an insulating (dielectric) material (eg , FIGS. 49B and 49C).

Geometry contact tip

As described above, with respect to FIGS. 10A-10I and 11A-11F, for example, the contact structure of the present invention can be manufactured with a contact tip portion for use in the category using the elastic contact structure of the present invention as a probe. .

50A has a number of elastic contact structures 5030 mounted to their outer surface by their proximal ends. Contact structures are formed in accordance with the techniques disclosed above. For example, the contact structure has a gold wire stem and nickel overcoating. As described above, the far end 5030a of the contact structure can be readily manufactured on the same plane. Generally, however, they will be gentle (eg, if simple nickel overcoating is applied to the wire stem). As illustrated in this figure, the far end 5030a of the wire stem 5030 is applied to surface grounding or EDM combustion or the like so that the far end of the contact structure is flat, and optionally, of the geometry (rough, knullated) A) has surface polishing.

50B illustrates another embodiment in which the semiconductor package 5052 has a number of elastic contact structures mounted to their outer surface by their proximal ends. In this example, prefabricated contact tip 5090 is applied to the far end of the contact structure. Contact tip 5090 may have any desired geometry and is attached to the end of the contact structure by soldering, brazing, or the like.

In general, the various ways to make the geometry of the contact tip may be achieved by (a) making the tip of the original geometry by adjusting the plating process as described above (see, eg, FIGS. 5C and 5D), or (b ) “Establish” the contact structure on the prefabricated tip structure (see, for example, contact tip 1026 in FIG. 10D), or stand up the contact tips individually, and then bridge (or) them to the contact structure. And the like).

Demi coating

In general, through the above-described embodiment, the overcoating (upper structure) has a cylindrical shape (circular cross section, surrounding the underlying wire stem). This is the case when the wire stem is removed as described with respect to Figures 48A-48E.

According to the invention, the superstructure can be formed to have an almost semicircular shape (cross section) and the underlying false work can be removed.

51A illustrates a wire stem 5102 mounted to a terminal 5112 on a substrate 5108 by its proximal end. A layer of masking material 5110, such as a photoresist, is applied over the entire length of the wire stem 5102. The wire stem covered with the photoresist is then exposed to light by being illuminated from the side (eg rather than from the top) by light (eg, light from actinic rays) as indicated by arrow L. "Develop" part of the resist. The remaining portion of the photoresist becomes the “shadow” of the wire stem and does not develop.

In the next step, illustrated in Figure 51B, the partially overcoated wire stem is layered with a layer of spring material 5120, such as nickel. This results in a partially extended overcoating around the wire stem (which prevents the residual amount from overcoating by the residual (photoresist). The photoresist is thus (optionally) removed and the process flies.

Generally, as shown in FIG. 51C, the residual photoresist is removed and the wire stem is melted (by applying heating to the contact structure) to form material on terminal 5112.

Within the gist of the present invention, the gold wire stem 5102 is first covered with a thin layer of tin as described above with respect to Figure 49A so that the residual material 5014 becomes eutectic.

Contact center of wire stem

As described with respect to the embodiment of FIG. 8C (for example), what is described herein indicates that the contact structure of the present invention contacts other electronic components by the center of the contact structure rather than by the free end of the contact structure. .

Examples of contact structures are made in other electronic components by overcoating materials (see, eg, FIG. 18B) or by wire stem materials (see, eg, FIG. 49B).

According to a feature of the invention, the electronic contacts between the contact structures mounted on the first electronic component may be made by the center of the wire stem being overcoated rather than the overcoating material.

52A illustrates a wire stem 5202 with one end 5202a coupled to the substrate 5208 (compare 208) and the other end 5502b coupled to the substrate 5208. Ends 5202a and 5202b are both coupled to the same contact area 5210 (compare 110) on substrate (electronic component) 5208.

If the two ends of the wire stem are joined (as opposed to only one end and the other end is free), the wire stem is shown to be similar to the loop embodiment of Figure 2F or similar to the three-dimensional loop embodiment of Figure 2G.

FIG. 52B is similar to the way in which the middle section of the wire stem is masked in the embodiment of FIG. 48C to prevent the tip portion of the wire stem 4802 from being continuously overcoated (eg, plated) of the masked portion of the wire stem. Illustrating the step of masking with a photoresist.

Figure 52C illustrates the next step where the masked wire stem is overcoated with a material such as nickel (5220; compare 4820).

52D illustrates the step in which masking 5212 is removed. This leaves the center portion 5202c of the exposed wire stem easy to contact other electronic components. In this category, gold is a good choice for the wire stem 5202 due to its excellent electrical contact properties and the overcoating material 5220 is electrically conductive (just to establish the spring property of the elastic contact structure). Does not matter.

Multiple Free Normal Wire Stem, Single Removal Step

In many of the embodiments provided above, wires (eg, gold wires) may be coupled to, shaped (including straight), and freely cut into contact areas on the electronic component. In this way, one end of the resulting wire stem is attached to the electronic component, and the other (free) end of the wire stem is suitably useful for making contact with the other electronic component. In general, this requires forming each free top wire stem individually by repeating the joining and cutting steps for each wire stem.

According to a feature of the invention, the plurality of free normal wire stems consists of a plurality of joining steps and a single cutting step.

This embodiment can be understood by referring to Figures 52A-52D described earlier. In this case, however, ends 5202a and 5202b of wire stem 5202 may be coupled to the same contact region 5210 or two separate contact regions 110 and 110 (not shown) on substrate 5208. .

In this embodiment, it is generally preferred that the gold wire stem 5202 be first overcoated with a thin layer of tin, ultimately forming gold-tin eutectic in the manner described above with respect to FIGS. 49A and 49B.

In this embodiment, after removing the masking 5212, the contact structure is heated to a temperature sufficient to reflow the eutectic wire stem, so that the exposed “bridge” (smooth bend) between the two “legs” of the contact structure. Two free normal contact structures 5230, as shown in FIG. 52E, each having a eutectic end (compare FIG. 49B), which is a “tip” (far end) suitable for making contacts to other electronic components, thereby “collapsing” 5202c. , 5232).

Within the spirit of the present invention, this principle can be applied to a continuous loop as shown in Fig. 24C, so that each free end of the wire stem is cut multiple times without requiring cutting (e.g., EFO as described above). Form a top contact structure.

Within the spirit of the present invention, the technique of “collapsing” a bridge of one or more loops to create two or more (eg, twice the number of loops) free normal contact structures (or wire stems before overcoating) is shown in FIG. It can be applied to a loop structure as illustrated in Figs. 15C, 16, and 24C.

In accordance with an embodiment of the present invention, a plurality of single bond wires can be looped between two electronic components and then cut, forming a plurality of free normal wire stems (or overcoated wire stems). For example, as illustrated in FIG. 52F, a single wire stem 5252 has a first end 5322a mounted to the first electronic component 5244 and a second end 5302b mounted to the second electronic component. Have FIG. 15 is intended to illustrate that two electronic components 5244 and 5246 may be adjacent non-separated semiconductor dies on a semiconductor wafer.

As described above, according to the present invention, a contact structure mounted on an unseparated semiconductor die performs an operation in which the edge of the die, that is, the area between two adjacent dies, cuts (dices) the die (saw) the die. It does not extend beyond the cuff region.

As shown in FIG. 52F, the “bridge” region of wire stem 5252 can simply be cut in the same operation of cutting the die with cupping saw blade 5250. 14F comparison.

The concept of creating multiple free top contact structures without cutting may be performed with a simple wire bond loop that extends from one terminal to another, or from a terminal on one die to a terminal on another die (compare FIG. 15). In addition, a series of loops as shown in FIG. 24C can be processed in such a way as to leave a number of free normal wire stems on the back and each mounted to a separate terminal on the electronic component.

Also, for example, the wire stems shown in Figure 16B may have their tops removed in any suitable manner, separating the frame from the die rather than dissolving the frame.

In general, within the gist of the present invention, loops can be formed (typically from terminal to terminal), and their gentle bends are removed in any suitable way to form two free normal wire stems for each loop. For example, the loops can be protected with a material such as wax and polished to separate the legs from each other. This can be done before or after overcoating. If performed after overcoating, the wire stem is exposed and the techniques and benefits described above with respect to FIGS. 49A-49C will be relevant.

For example, Figure 53A illustrates a number of loops 5302, 5304 (similar to loop 202 of Figure 2F) formed between terminals 5308, 5308, 5310, 5312 on the surface of electronic component 5314; Two are shown). 53B shows (e.g., potted) loops 5302, 5304 protected with removable material 5320, such as rigid wax. After potting in this manner, the polishing (polishing) mechanism 5322 pushes down the potted loop and the loop is removed through the potting material 5320 and the smooth bends of the loops 5302, 5304. (This is indicated by the dotted line P in the figure). Thereafter, the potted material is removed (by melting). This results in two free normal wire stems (not shown) in each loop. Within the gist of the present invention, the wire stem (loop) is overcoated before potting or after polishing (removing potting material). If the wire stem is overcoated before being potted, the wire stem is exposed to form a brazable tip in the manner described above with respect to FIG. 49C.

Within the gist of the present invention, a loop wire stem (eg, 5302) is connected from a terminal on one electronic component to a terminal on another electronic component (as illustrated, rather than two terminals present on the same electronic component). Extend.

By fabricating multiple wire stems from loops and the like, electronic components (such as semiconductor devices) on which the loop (and ultimately the free-normal contact structure) are mounted can produce high and potentially dangerous voltages associated with electronic flame-off techniques (described above) (e.g., For example, thousands of volts are dissociated during discharge.

53C and 53D illustrate another technique for making a free steady wire stem without electron flame off from the loop in accordance with the present invention. As illustrated, the wire stem 5332 extending from the terminal 5332 on the electronic component 5558 is formed in a loop and connected back to the terminal (or another terminal on the electronic component, or another terminal on the other electronic component). A substantial portion of one "branch" (leg) of the loop is covered with masking material 5354, such as a photoresist. The loop is then overcoated with material 5558 and the photoresist is removed, at which point the first masked branch of the loop can also be removed, forming a free normal overcoated wire stem as illustrated in FIG. 53D. .

How to Create Flat Tab Tips, and Interconnect

As described above, the far end (tip) of the contact structure may be provided as a contact pad of geometry or the like. For example, within the gist of the present invention, the tip portion of the contact structure may be provided as a flat plate tab (pressure plate). In this way, the interconnection to the external component is via an intermediate medium of what is known as a "z-axis conductive adhesive" which is a known material with conductive (eg gold) particles contained therein and which is conductive under pressure ( Easy to break, especially in fragile external parts).

Figure 54 illustrates an overcoated wire stem 5402, the distal end (tip) of which is provided with a flat tab (pad) in a manner similar to that described above with respect to Figure 10D (element 1026). The electrical interconnect is effective from the contact structure 5402 to the external electronic component 5406 by means of the z-axis conductive adhesive 5408 with the conductive particles 5410 maintained entirely therein. When the contact structure 5402 is forced against the electronic component (not shown) against the external component 5406 (see arrow C), the adhesive 5408 becomes compressed and conductive.

Advantages

As mentioned above, the prior art is filled with various interconnect shapes, and each shape can be applied to only one environment.

For example, a "conventional" wire bond may be a plurality of points (eg, bond pads) on a first electronic component (eg, a semiconductor die) and a corresponding number of points on a second electronic component (semiconductor package element). Effective by making electrical connections between (e.g., terminals), by coupling the wire to the first point, extending the wire slightly “loosely” to the second point, and coupling the wire to the second point do. However, this shape is not of particular value for bonding a semiconductor package to a printed circuit board.

For example, it is known to provide an outer surface of a semiconductor package with an array of pins, leads, solder bumps, etc. to make electrical connections between the package and the printed circuit board on which the package is mounted. However, packaging electronic components such as semiconductor dies in this manner require outsourcing packaging, which often involves the transfer of expensive semiconductor devices. In addition, once the semiconductor device is mounted therein, it is difficult to deduct the package, and as a result, the overall packaged semiconductor device is removed if the packaged semiconductor device is a test failure.

For example, the socket provides a means for making a temporary (removable) connection between the semiconductor package and the printed circuit board. However, sockets cannot be applied to make connections to bare (unpackaged) semiconductor components.

For example, contacts having a limited degree of elasticity are known for making connections between electronic components. Typically, these contacts are arrayed (in a supported state) by a carrier as highlighted in the above-mentioned US Pat. No. 4,705,205. Also, although these contacts may have a conductive shape (eg, an S shape) to act as a "spring", the material from which these contacts are made is not conductive to act as an elastic contact structure. In addition, any such contact supported on (rather than extending from) the carrier is not free by definition.

Some non-intuitive advantages include soldering (or epoxying) the tip (far end) of the contact structure that extends from the die (with the elastic contact structure mounted thereon) to the printed wiring board (see, eg, Figure 36A). It is to use a shape elastic contact structure such as mounted on a semiconductor die in an inelastic manner. In this flip chip type application of the invention, the degree of elasticity can be maintained (if desired). In addition, due to the “projection” nature of the high aspect ratio contact structure, the cleaning and inspection possibilities (eg, of the solder flux) are increased compared to conventional solder bump type flip chip surface mounting processes.

Another non-intuitive advantage of the present invention is the use of a process such as plating to produce contact structures suitable for microelectronic applications. A review of the prior art will show that plating is used to provide a protective coating over a substantial electronic component (such as a lead frame lead) rather than structurally. In addition, gold is widely accepted as an optional material for making electrical interconnects in many applications, so plating something (eg nickel) onto gold (eg gold wire stems) is the counter intuitive.

Various elastic contact structures, methods for manufacturing the same, applications suitable for the same have been described above.

Electroplating processes are preferred for overcoating wire stems, and certain parameters such as strength, ductility, tensile strength are required to be appropriate. The process of optimizing these parameters is intended to be empirical according to some general guidelines, as disclosed in the paper under the nature of electrodeposited metals and alloys by Iliam saprene.

Although the invention has been illustrated and described in detail in the drawings and above, it is to be understood that the same is intended to limit the understanding that only preferred embodiments have been shown and described, and that all changes and modifications are required to be protected within the spirit of the invention. no.

For example, the elastic contact structure can be formed as a separate unit so that it can be applied to electronic components. This is similar to the gang transfer embodiment described above with respect to Figure 12A, which illustrates that a number of elastic contact structures are formed on a removable substrate for continuous transfer to electronic components. However, within the gist of the present invention, an elastic contact structure is manufactured and then removed from a removable substrate and accumulated in an electronic component suitable for future applications (such as by automated equipment) on a separate basis. Such individual elastic contact structures can be formed by using two removable members, and after the elastic contact structures have been manufactured, the two removable members are removed.

Based on the description disclosed herein, those skilled in the art will understand how to make the present invention practical. Nevertheless, a useful “guide book” reference that provides a basic understanding of electronic contact technology is found in the Y & I Contact Manual, Kenneth E., whose related parts are incorporated herein by reference. Pitney, found in 1973. For example, as discussed herein, “ideal” contact materials are characterized by (1) high electrical conductivity (low resistance) such that compression and volume resistance are low, and (2) high heat such that joule heat is conducted quickly from the contact interface. High strength to provide conductivity and (3) softness to provide low compression resistance due to large spots, (4) high stiffness suitable for low mechanical wear, and (5) ability to act as contact and cantilever beams to provide low mechanical wear (6) high precious metals suitable for extended shelf life, low electrical noise and good reliability; (7) the ability to form a thin lubricating film, but not to provide an excessive friction polymer; and (8) a low cost.

It is clear that the present invention satisfies many of these criteria and that simultaneous arrival of all these properties is not possible.

As is further recognized in the Y & Y contact manual, there are many cases where cantilever beams play an important role in the contact structure. (A simple cantilever is a beam of uniform cross section in which its support is suspended (through its length) and a load acts on its suspended area. According to the present invention, straight (linear) wire stems arranged at an angle (not vertical) to the surface of the substrate overcoated and mounted as disclosed above will act substantially as a cantilever beam.

The present invention includes within its gist the case where the bond wire is bonded to an electronic component and overcoated (eg plate) with a material selected for its structural properties.

Based on the description disclosed herein, it implies that certain features of the selected embodiments described above may be incorporated in the other embodiments described above, as will be apparent to those skilled in the art to which this invention pertains.

For example, in any of the embodiments disclosed above, the contact structure can be substantially pure elasticity (eg, elastic) or can exhibit a composite (eg, complex) of plasticity and elasticity. Therefore, in some cases where "elasticity" is used, "complexity" can be replaced by the "special case" of elasticity. In addition, as used herein, any contact structure that is composite is elastic.

For example, rather than making the tip (far end) of the wire stem to allow it to be exposed after overcoating (see, for example, FIGS. 48A-48E or 49A-49C), the wire stem (without masking) ) And the tip of the wire stem is grounded (with a polishing or polishing instrument) to expose the wire stem.

For example, when the wire stem is reflowed to expose, as shown in Figures 49A-49C, an external mechanism, such as a mold, may be caused to remain on the far end 4930 of the contact structure to extend the tip portion. Give the desired shape. This includes a mold mechanism that causes the tip to have an engraved (wedge) shape.

For example, in some cases where the far end of the contact structure is “permanently” attached to another electronic component, this may be done with any suitable conductive material, such as soldering or conductive epoxy. Alternatively, the remote portion of the contact structure can be formed to bias itself in the plated holes of other electronic components, as shown in FIG.

For example, in most cases where contact structures are mounted to a semiconductor device, it is desirable that the at least adjacent portions and surrounding areas on the surface of the substrate (semiconductor device) are protected with a solidifying material such as a polymer. In many cases, the entire semiconductor device can be protected in this manner.

Undoubtedly, many other "variations" on the "topic" described above will occur to those skilled in the art to which this invention pertains, and such variations will be within the spirit of the invention, as disclosed herein.

For example, in the embodiments disclosed or suggested herein, the masking material (eg, photoresist) is patterned by exposure to light that is applied to the substrate, passes through the masking, and chemically removes a portion of the masking material. Masking by directing an appropriate integrated light beam (e.g., from an excimer laser) to a portion of the masking material (e.g., a blank reinforced photoresist) to be removed (i.e., conventional photosensitive techniques) and removed. Alternative techniques may be used that include removing these portions of the material, or guiding the light beam properly integrated into the reinforcing portion of the masking material (without using masking) and then chemically cleaning the unreinforced masking material. .

For example, in the main part described above, the stem (core, poles walk) is in the wire state, but within the gist of the present invention, the core may be an elongated planar member such as a conductive ribbon or tab. These tabs can be easily punched out of a relatively soft metal flat cardboard, and then overcoated to exhibit the desired elasticity and / or complexity in the manner described herein above.

It will be clearly understood that the elastic contact structure of the present invention can be formed on a removable substrate and then removed therefrom to provide a plurality of separate individual elastic contact structures suitable for mounting to an electronic component.

Additionally, the elastic contact structures disclosed herein that include a corer false walk overcoated by superstructure overcoating can be considered as a "composite interconnect element".

In this main part, the contact force F is applied longitudinally to the elastic contact structure, in other words, the contact force is applied to the terminal in which the elastic contact structure (shown otherwise) is mounted in the direction of the mounting base of the elastic contact structure. Discussion was directed towards. See, eg, FIG. 5F. The contact force may act on the elastic contact structure in any direction (including transverse direction) in which the elastic contact structure reacts to the force.

Examples have described how a plurality of elastic contact structures can be formed on a plurality of prefabricated contact tips on a removable substrate. See, eg, FIGS. 10A-10I. As shown in Figure 9C, it will be apparent that a number of prefabricated tip structure may be brazed at the tip of the resilient contact structure prefabricated on the terminal of the electronic component.

Claims (50)

Fabricating a plurality of separate contact structures not attached to the substrate, Providing a substrate on the surface comprising an array of electrical connections; Attaching the plurality of contact structures to the array of electrical connections. The method of claim 1 further comprising dividing the plurality of contact structures into one or more groups prior to the attaching step. 3. The method of claim 2, further comprising selecting one or more of the groups of the plurality of contact structures prior to the attaching step. 4. The method of claim 3, wherein the attaching step further comprises sequentially attaching the selected one or more groups of the groups of the plurality of contact structures. 4. The method of claim 3, wherein the attaching step further comprises attaching the selected one or more groups of the groups of the plurality of contact structures at once. The method of claim 1 wherein the electrical connection comprises a metal pad. The method of claim 1, wherein the contact structure is elastic. The method of claim 1, wherein the fabricating step includes forming the contact structure on the sacrificial substrate and removing the contact structure from the sacrificial substrate. The method of claim 8, wherein the forming step includes sequentially applying a plurality of materials to the sacrificial substrate. The method of claim 9, wherein at least one of the materials comprises a patterned material layer. The method of claim 10, wherein at least one of the materials comprises a photoresist. 10. The method of claim 9, wherein at least one of the materials constitutes the contact structure. The method of claim 12, wherein a plurality of the materials make up the contact structure. 10. The method of claim 9, wherein at least one of the materials constitutes a tip of the contact structure. The method of claim 14, wherein at least another of the materials constitutes a wire bonded to the tip. The method of claim 15, wherein at least another of the materials forms an overcoat that at least partially surrounds the wire. The method of claim 14, wherein the tip is patterned. The method of claim 1, wherein the fabricating step includes forming a patterned layer of material in the sacrificial substrate, wherein the patterned layer includes openings corresponding to the contact structure. 19. The method of claim 18, wherein the fabricating step further comprises forming a geometric pattern in the opening in the sacrificial substrate. 19. The method of claim 18, wherein the step of manufacturing further comprises forming a contact tip of the contact structure in the opening. 21. The method of claim 20, wherein forming the contact tip comprises depositing at least one material in the opening. 21. The method of claim 20, wherein the step of manufacturing further comprises wire bonding one end of a wire to the tip and spacing the second end of the wire away from the tip. 23. The method of claim 22, wherein the manufacturing step further comprises overcoating the wire. 21. The method of claim 20, wherein the fabrication step further comprises removing the sacrificial substrate. The method of claim 1, wherein the substrate comprises a probe card assembly. The method of claim 1, wherein the attaching step comprises contacting an end of the contact structure with the array of electrical connections on the substrate and securing the end of the contact structure to the array. The method of claim 1, wherein the attaching step includes inserting an end of the contact structure into a recess in the substrate and securing the end of the contact structure in the recess, wherein the recess is in the substrate. Corresponding to the array of electrical connections on the phase. The method of claim 1, wherein the attaching step comprises permanently attaching each of the plurality of contact structures to the array of electrical connections. The method of claim 1, wherein the attaching step comprises metal coupling the plurality of contact structures to the array of electrical connections, respectively. Providing a substrate on the surface, the substrate comprising an array of electrical connections; Manufacturing a plurality of elongated contact structures on a sacrificial substrate in a pattern corresponding to at least a portion of the array of electrical connections; Attaching the plurality of contact structures to the array of electrical connections, respectively; Removing the contact structure from the sacrificial substrate. 33. The method of claim 30, wherein the contact structure is elastic. 31. The method of claim 30, wherein the forming method comprises sequentially applying a plurality of materials to the sacrificial substrate. 33. The method of claim 32, wherein at least one of the materials comprises a patterned material layer. 34. The method of claim 33 wherein said at least one of said materials comprises a photoresist. 34. The method of claim 33, wherein at least one of the materials forms the contact structure. 36. The method of claim 35 wherein a plurality of said materials make up said contact structure. 33. The method of claim 32, wherein at least one of the materials constitutes a tip of the contact structure. 38. The method of claim 37, wherein at least one of the materials forms a wire bonded to the tip. The method of claim 38, wherein at least another of the materials forms an overcoat that at least partially surrounds the wire. 38. The method of claim 37, wherein the tip is patterned. 31. The method of claim 30, wherein the fabrication step further comprises forming a patterned layer of material in the sacrificial substrate, wherein the patterned layer includes openings corresponding to the contact structure. 42. The method of claim 41, wherein the manufacturing step further comprises forming a geometric pattern in the opening in the sacrificial substrate. 43. The method of claim 42, wherein the manufacturing step further comprises forming a contact tip of the contact structure in the opening. 44. The method of claim 43, wherein forming the contact tip comprises depositing one material in the opening. 45. The method of claim 44, wherein forming the contact tip comprises depositing a plurality of materials in the opening. 44. The method of claim 43, wherein the manufacturing step further comprises wire bonding one end of a wire to the tip and spaced apart from the tip of the second end of the wire. 47. The method of claim 46, wherein the manufacturing step further comprises overcoating the wire. 31. The method of claim 30, wherein the substrate comprises a probe card assembly. 31. The method of claim 30, wherein the attaching step comprises contacting an end of the contact structure with the array of electrical connections on the substrate and securing the end of the contact structure to the array. 31. The method of claim 30, wherein the attaching step includes inserting an end of the contact structure into a recess in the substrate and securing the end of the contact structure in the recess, the recess being in the substrate. Corresponding to the array of electrical connections on the phase.

Images