Classifications

• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10B—ELECTRONIC MEMORY DEVICES
  • H10B12/00—Dynamic random access memory [DRAM] devices
• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10B—ELECTRONIC MEMORY DEVICES
  • H10B12/00—Dynamic random access memory [DRAM] devices
  • H10B12/01—Manufacture or treatment
  • H10B12/02—Manufacture or treatment for one transistor one-capacitor [1T-1C] memory cells
  • H10B12/03—Making the capacitor or connections thereto
  • H10B12/038—Making the capacitor or connections thereto the capacitor being in a trench in the substrate
  • H10B12/0383—Making the capacitor or connections thereto the capacitor being in a trench in the substrate wherein the transistor is vertical
• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10B—ELECTRONIC MEMORY DEVICES
  • H10B12/00—Dynamic random access memory [DRAM] devices
  • H10B12/01—Manufacture or treatment
  • H10B12/02—Manufacture or treatment for one transistor one-capacitor [1T-1C] memory cells
  • H10B12/03—Making the capacitor or connections thereto
  • H10B12/038—Making the capacitor or connections thereto the capacitor being in a trench in the substrate
  • H10B12/0385—Making a connection between the transistor and the capacitor, e.g. buried strap
• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10B—ELECTRONIC MEMORY DEVICES
  • H10B12/00—Dynamic random access memory [DRAM] devices
  • H10B12/01—Manufacture or treatment
  • H10B12/02—Manufacture or treatment for one transistor one-capacitor [1T-1C] memory cells
  • H10B12/05—Making the transistor
  • H10B12/053—Making the transistor the transistor being at least partially in a trench in the substrate

Definitions

• the invention relates to a memory cell according to the preamble of claim 1 and memory cell arrangements composed of such memory cells.
• the invention further relates to a method for producing a memory cell and a memory cell arrangement.
• a one-transistor memory cell comprises a read-out or counting transistor and a storage capacitor.
• the information is stored in the storage capacitor in the form of an electrical charge, which represents a logical variable, 0 or 1.
• the storage capacitor must have a minimum capacitance for safe storage of the charge and simultaneous differentiability of the information read out. The lower limit for the capacitance of the storage capacitor is currently seen at 25 fF.
• the required area of the single-transistor memory cell must be reduced from generation to generation. At the same time, the minimum capacitance of the storage capacitor must be maintained.
• both the read-out transistor and the storage capacitor were implemented as planar components. From the 4 Mbit memory generation onwards, the area of the memory cell was further reduced by a three-dimensional arrangement of readout transistor and storage capacitor achieved.
• One possibility is to implement the capacitor in a trench (see, for example, K. Yamada et al., Proc. Intern. Electronic Devices and Materials IEDM 85, p. 702 ff).
• the electrodes of the storage capacitor act as a diffusion region adjacent to the wall of the trench and a doped polysilicon filling which is located in the trench.
• the electrodes of the storage capacitor are thus arranged along the surface of the trench.
• the effective area of the storage capacitor, on which the capacitance depends is increased compared to the space requirement for the storage capacitor on the surface of the substrate, which corresponds to the cross section of the trench.
• the upper capacitor electrode is the
• Trench capacitor with a horizontal selection transistor which is arranged above the trench capacitor and laterally offset, electrically connected via a line bridge made of polycrystalline silicon.
• Such an arrangement requires a relatively large amount of space due to the horizontally oriented selection transistor, so that a further increase in the packing density of the memory cells is only possible to a limited extent.
• EP 1 077 487 A2 describes a DRAM memory cell with a trench capacitor, in which a capacitor is formed in a lower section of a trench formed in a semiconductor substrate and a selection transistor is formed in an upper section of the trench.
• the channel region of the selection transistor extends along the side wall of the trench between a first source / drain connection coupled to the upper capacitor electrode of the trench capacitor. End and a second source / drain terminal located near the substrate surface.
• the gate electrode is located in the trench above the capacitor and a gate oxide layer is located at the interface with the channel region.
• the disadvantage of this arrangement is the relatively long channel length of the selection transistor and the resulting poor controllability and slow response time of the selection transistor.
• US-A-6, 137, 128 discloses a memory cell with a trench capacitor, which has a trench formed in a semiconductor substrate, in the lower trench region of which a lower capacitor electrode adjoins a wall of the trench and in which a storage dielectric and an electrically conductive trench filling are formed polycrystalline silicon are introduced as the upper capacitor electrode.
• a vertical MOSFET is arranged above the trench capacitor as a selection transistor, the gate as well as the source, drain and the channel region being formed as regions encircling a central insulation layer.
• the generic EP 0 905 772 A2 also describes a DRAM memory cell and a method for its production, in which a vertical MOSFET is formed as a selection transistor above a trench capacitor which is filled with polycrystalline silicon as the upper capacitor electrode.
• the MOSFET has a first n-doped source / drain region, a p-doped channel region and a second n-doped source / drain region, which are essentially deposited by epitaxy on the polycrystalline silicon of the trench filling.
• This layer sequence is structured vertically in such a way that it has a square cross section and a gate electrode layer covering the four sides and surrounding the layer sequence is deposited at the level of the channel region.
• a disadvantage of this method is that the MOSFET is grown essentially epitaxially on the poly-silicon of the trench filling, as a result of which defects and grain boundaries are introduced into the MOSFET.
• Another disadvantage is that prior to the manufacture of the MOSFET, vertical etching is carried out in a highly doped poly-silicon layer covering the substrate with the trench capacitor directly above the trench capacitor and then the gate oxide is applied directly to the etched side walls, so that the gate oxide not of optimal quality.
• Another disadvantage is to be seen in the fact that the shape of the gate electrode of the vertical MOSFET surrounding the channel region does not allow an optimal potential penetration of the gate electrode potential through the channel region.
• Another disadvantage is that the memory cell arrangement shown and the arrangement of the memory cells along the word line formed by the gate electrodes does not allow a high packing density.
• a first aspect of the present invention relates to a memory cell with a trench capacitor which has a trench formed in a semiconductor substrate, in whose lower trench region a lower capacitor electrode adjoins a wall of the trench and in which a storage dielectric and an electrically conductive trench filling are used
• Upper capacitor electrode are introduced, and a vertical field effect transistor arranged above the trench capacitor as a selection transistor.
• the upper capacitor electrode is connected to a first source / drain region of the selection transistor by a contact region which at least partially encloses its upper end section u on the start side.
• the connection of the trench capacitor and the selection transistor through the peripheral contact area makes it possible to produce the selection transistor with good crystalline quality and good electrical performance properties.
• the contacting area is preferably produced from essentially monocrystalline doped semiconductor material, the selection transistor can be deposited thereon by epitaxial growth and thus be made from a semiconductor material with good crystalline quality.
• the contacting area can advantageously be such that it encompasses the upper end section of the upper capacitor electrode, that is to say the electrically conductive trench filling, in an annular or tubular manner, that is to say completely surrounds the upper end section on the circumference. If the contacting region is tubular, it can enclose the upper end section of the upper capacitor electrode with its lower section and the source / drain region of the selection transistor in the form of a tube or flange with its upper section.
• ring or tube does not in any way define a specific, approximately circular tube cross-section. Rather, the trench can have any conceivable cross section in this section.
• the upper capacitor electrode is preferably not in direct mechanical contact with the source / drain region of the selection transistor.
• the trench filling forming the upper capacitor electrode is mostly provided by doped polycrystalline silicon.
• the epitaxial growth should only take place on the monocrystalline contact region, but not additionally on the poly-silicon of the trench filling.
• An insulation layer preferably a silicon oxide layer (TTO, trench top oxide), is therefore deposited on the surface of the upper end section of the poly-silicon trench filling.
• an intermediate layer can be arranged between the contacting region made of monocrystalline semiconductor material and the polysilicon of the upper end section of the upper capacitor electrode in order to avoid that crystal defects from the polysilicon continue into the contacting region.
• This intermediate layer may but does not hinder the electrical connection between these areas, for example, it is formed as an ultra-thin tunnel contact layer, such as silicon nitride.
• An insulation layer for example an oxide layer, is preferably inserted between the contacting area and the semiconductor substrate in order to electrically isolate the contacting area from the substrate on the one hand and to prevent the contacting area from diffusing out into the substrate on the other hand.
• the selection transistor of the memory cell is a field-effect transistor, which has a layer structure comprising the first source / drain layer, a channel layer and a second source / drain layer adjoining the annular contacting region.
• the layer sequence of the field effect transistor can be structured vertically in such a way that it has an essentially oval cross section, the layer sequence being surrounded at least in the region of the channel layer by a gate electrode layer with an essentially oval circumferential profile and by a gate oxide layer located between them.
• the advantage of a MOSFET transistor shaped in this way is that, on the one hand, a good potential penetration of the gate electrode potential through the channel region is provided, and on the other hand, the oval shape of the gate electrodes can be used to enable the memory cells that are joined together in one direction in a memory cell arrangement. you can arrange them tightly.
• Such a MOSFET transistor can thus advantageously be used as a selection transistor in a memory cell according to the invention.
• the layer sequence of the MOSFET transistor can be deposited, for example, on the TTO insulation layer and epitaxially on the contact area.
• the trench capacitor is formed in the semiconductor substrate and an insulation layer is deposited on the semiconductor substrate, within which the gate electrode layer is formed.
• the contacting region can, for example, extend to the interface between the semiconductor substrate and the insulation layer his.
• a plurality of memory cells each containing field-effect transistors as selection transistors as described above, can be combined in a memory cell arrangement in which a number of
• Memory cells are arranged along a direction in which the gate electrode layers of adjacent memory cells are joined together in such a way that their oval circumferential courses overlap in one section and the gate electrode layers coincide in this section.
• This arrangement enables a tight integration of memory cells along the mentioned direction.
• Such an arrangement can be developed by arranging a number of memory cells in the form of a matrix and are controllable by word and bit lines, and the word line is formed by gate electrode regions of memory cells arranged along one of two orthogonal directions, and the bit line runs in the other of the two orthogonal directions via the selection transistors and in each case with their second source / drain Layers are electrically connected.
• the invention further relates to a method for producing a memory cell which has a trench capacitor and a vertical field-effect transistor arranged above the trench capacitor, the method comprising the steps:
• the upper capacitor electrode is often formed from polysilicon, a direct mechanical contact between the polysilicon and the crystalline silicon of the transistor and thus the propagation of crystal defects into it can be prevented.
• Fig.l a semiconductor substrate with applied etching mask layers
• 6 shows the memory cell after the epitaxial growth of the layer sequence of the MOSFET; 7 shows the memory cell after the formation of the gate electrode layer;
• FIG. 9 shows a plan view of a matrix-shaped memory cell arrangement with word and bit lines.
• FIGS. 1 to 8 each show a cross section through a memory cell to be processed along a plane symmetrically dividing the trench.
• an approximately 200 nm thick STI (shallow trench isolation) insulation layer 2 made of SiO 2 is first deposited on a monocrystalline silicon substrate 1, which serves to isolate the finished memory cells from one another, as will be seen later.
• An approximately 100 nm thick first mask layer 3 made of SiN and then a second mask layer 4 made of SiO 2 are deposited on this. These layers serve as hard mask layers in the subsequent etching processes.
• trenches 5 are then defined using conventional photolithography and resist technology in which trenches 5 are to be etched.
• a trench is produced in these regions using a first etching process, extends through the STI insulation layer 2 and produces a first trench section 5A in the substrate, which has a depth of approximately 200 nm below the main surface of the substrate 1. This depth defines the length of the contact section to be created later.
• the trenches 5 have an elongated shape in cross section, for example a rectangle with an edge length of approximately 100 nm ⁇ 250 nm or an oval with corresponding longitudinal and transverse dimensions.
• rectangles of the specified dimensions are provided on the mask side, with in practice due to the small dimensions result in oval trenches.
• a matrix-shaped arrangement of trenches 5 of this type is produced which are spaced apart from one another in both orthogonal directions by approximately 100 nm.
• the first trench section 5A is then widened by a few 10 n in all sides by isotropic etching, as a result of which the width of the ring-shaped contacting section is defined.
• an intermediate layer 6 (“liner”) made of SiO 2 is deposited, which in the trench section 5A will have the function of both electrically isolating the contact area to be produced from the substrate 1 and also diffusing it out of the contact. to prevent the clocking area into the substrate 1.
• the anisotropic etching process removes the intermediate layer 6 in the base area again on a section which corresponds approximately to the original dimensions of the trench 5. The result of these method steps is shown in FIG.
• FIG. 3 shows how a silicon filling 7 is subsequently deposited in the trench section 5A by means of selective epitaxy, for example by means of a CVD process or the like, so that it completely fills it.
• the silicon filling 7 is produced by epitaxy on the silicon of the substrate 1, it can be produced with a good crystalline quality.
• an in-situ doping can be carried out in order to provide sufficient electrical conductivity of the contact area to be formed from the silicon filling 7.
• the silicon can also initially be deposited in a nominally undoped manner and the doping can be made up in a later process step, as will be seen later. The result of these process steps is shown in Fig.3.
• the epitaxially grown silicon filling 7 is then restored in one area by an anisotropic etching process removed, which corresponds to the cross section of the original trench 5, so that an annular contact region 7.1 of doped, essentially monocrystalline silicon remains.
• the ring-shaped contacting region has an oval circumferential course, since - as already indicated - the trench 5 has an oval cross section.
• an approximately 5 nm thick oxidation layer (SiO 2 ) 8 is provided (deposited or generated by thermal oxidation) on the inner wall of the ring-shaped contact area 7.1, which serves on the one hand as protection against etching and also has the function of the ring-shaped contact area 7.1 to be shielded in the event of a doping to be carried out later in order to maintain an already preset doping.
• the oxidation layer 8 is produced in the entire trench, but is only shown in the trench section 5A. The result of these process steps is shown in Fig.4.
• the trench 5 is deepened by an anisotropic etching process, thereby producing a second trench section 5B in which the trench capacitor is to be produced.
• the lower capacitor electrode (not shown) can first be provided in a manner known per se. If desired, as already indicated above, the contacting region 7.1 can be doped simultaneously with this doping, so that in FIG. 3 the silicon filling 7 does not have to be doped in situ and the oxidation layer 8 can be made thinner or can be omitted entirely.
• the lower capacitor electrode can also be provided by depositing a metallic layer in the second trench section 5B, as was described, for example, in DE 199 44 012. Preferably too in this case, a previous doping was carried out in order to make an ohmic contact with the silicon substrate 1.
• a dielectric layer 9 with a thickness of approximately 5 nm, the SiO 2 and / or Si 3 N 4 , and optionally silicon oxynitride or also A1 2, is deposited in the second trench section 5B as a capacitor dielectric 0 3 , Ta0 5 or Ti0 2 (possibly with admixtures of hafnium and / or zirconium).
• the upper capacitor electrode 10 is introduced in a manner known per se.
• this is doped polysilicon, with which the trench is first completely filled and then etched back again, so that it has a filling height in the trench up to approximately the middle region of the contacting region 7.1.
• the dielectric layer 9 is then removed above this. The result of these process steps is shown in Fig.5.
• the SiO 2 layer 4, the oxidation layer 6 located in the upper trench section on the layers 2 and 3 and the oxidation layer 8 are then removed by an etching step.
• the SiN mask layer 3 is then - as shown in FIG. 6 - evenly withdrawn from the edge of the trench by isotropic etching and thinned at the same time. As will be seen later, this already serves to prepare the gate electrodes that are still to be produced.
• this must first be connected in an electrically conductive manner to the poly-silicon of the upper capacitor electrode 10, since a gap has formed between them due to the removal of the oxidation layer 8 located between them.
• Tunnel contact layer 11 made of SiN on the poly-silicon and then the remaining space between the contacting area 7.1 and the upper capacitor electrode 10 is filled by depositing a polycrystalline silicon layer and subsequent removal outside of this space.
• the tunnel contact layer 11 With the tunnel contact layer 11, a direct mechanical contact between the polysilicon of the upper capacitor electrode and the monocrystalline silicon of the contacting region 7.1 and thus the spread of crystal defects are to be avoided.
• the tunnel contact layer 11 if the tunnel contact layer 11 is not necessary, it can also be omitted and instead the connection of the contacting region 7.1 to the upper capacitor electrode 10 can be carried out by a reflow method, in which poly-silicon flows into the intermediate space through a temperature treatment step.
• an insulation layer 12 (TTO, Trench Top Oxide) is applied to the surface of the upper capacitor electrode 10, for example by HDP oxide deposition (high density plasma) and subsequent isotropic etching back.
• This insulation layer 12 is advantageous because it separates the poly-silicon 5B of the trench filling from the crystalline silicon of the transistor which subsequently has to be grown epitaxially and thus prevents crystal defects from spreading out of the polysilicon into the crystalline silicon.
• the insulation layer 12 can, if appropriate, also be omitted if there is no risk of the crystalline silicon located above being impaired. This can be the case, for example, if the trench filling in the upper region is not formed from polysilicon but from another electrically conductive material.
• the upper region of the trench is then filled by means of selective silicon epitaxy and the MOSFET selection transistor 20 of the memory cell is produced in the process.
• the remaining portion of the Si0 2 mask layer 4 and the SiN mask layer 3 is then removed. The result of these process steps is shown in Fig. 6.
• Such a selection transistor 20 is also referred to as a floating body transistor, since it is not built up on a substrate with a constant electrical potential.
• the layer sequence of the selection transistor 20 is formed in the upper trench section located within the STI insulation layer 2, it has the same cross-sectional structure as that of the etched trench 5, that is to say preferably an oval-shaped structure.
• anisotropic etching first removes the section of the STI insulation layer 2 which was exposed by the previous withdrawal of the SiN layer 3, and thus space for the gate electrode 24 is created, as is shown in the result in FIG.
• This anisotropic etching is a self-aligned process by means of which a trench is produced in the STI insulation layer 2 around the layer sequence of the selection transistor 20, the course of which is adapted to the cross section of the initially etched trench 5 and thus now to the cross section of the exposed layer sequence of the selection transistor 20.
• the circumferential profile of this trench produced by the anisotropic etching is preferably of an elongated, in particular oval, shape.
• the gate electrode layer 24 to be subsequently produced in the trench will be produced with an oval circumferential profile, which is equally spaced from the layer sequence of the selection transistor 20 at every point. Since the STI etching - as already mentioned - is carried out by a self-aligned process, the lithography step usually provided at this point by means of the so-called AA mask is omitted.
• a gate oxide layer 25 is produced on the exposed layer sequence of the selection transistor 20, for example by thermal oxidation.
• the circumferential course of this gate oxide layer 25 is also matched to the cross section of the etched-off layer sequence and thus preferably has an oval circumferential course.
• the horizontal layer section of the gate oxide layer 25 on the source / drain layer 23 is then removed again by an etching step.
• the trenches are also etched in a matrix-shaped memory arrangement in such a way that overlap regions 24.3 result between the oval circumferential courses of memory cells joined together along an orthogonal direction. Because the oval trenches of the individual memory cells are joined to one another in such an overlapping manner, a dense sequence and thus a high level of integration of the memory cells on the chip can be generated. This overlap then also creates lines that can later be used as a word line for memory addressing. The word lines weave around the transistors from one trench area to the next, so to speak. 9 thus shows two word lines 24.1 and 24.2 from the entire memory arrangement in the vertical direction, which are crossed by three bit lines in the horizontal direction. The memory cells are located below the crossing points.
• word and bit lines are exactly orthogonal to one another and the memory cells of adjacent word lines 24.1, 24.2 are exactly open arranged at the same height next to each other.
• An alternative arrangement with an even higher packing density provides that the memory cells are offset in height from one another by exactly half a distance, so that the height of a memory cell of one word line comes to lie between two memory cells of the adjacent word line. As a result of this offset, it is necessary that the bit lines no longer run orthogonally but at an oblique angle to the word lines.
• the gate oxide 25 is produced and then the gate electrode layer 24 is deposited in the trenches.
• a pure polysilicon gate or a layer sequence composed of a metal and polysilicon can optionally be used as the gate electrode layer 24.
• the gate electrode material on the STI insulation layer 2 is removed by means of chemical mechanical polishing.
• the gate electrode 24 can then be etched back slightly to ensure that it does not overlap with the upper source / drain layer 23 of the selection transistor. The result of these process steps is shown in FIG. 7.
• bit lines 15 can then be applied thereon in a direction orthogonal to the word lines and contacted with the through contacts 14. Three bit lines 15 are shown in FIG.

Landscapes

• Engineering & Computer Science (AREA)
• Manufacturing & Machinery (AREA)
• Semiconductor Memories (AREA)
• Jib Cranes (AREA)

Applications Claiming Priority (3)

Publications (1)

Family

ID=7695362

Family Applications (1)

Country Status (7)

Families Citing this family (27)

Family Cites Families (19)

• 2001
  • 2001-08-14 DE DE10139827A patent/DE10139827A1/de not_active Ceased
• 2002
  • 2002-02-13 US US10/473,330 patent/US20040104192A1/en not_active Abandoned
  • 2002-07-12 KR KR1020047002145A patent/KR100613927B1/ko not_active IP Right Cessation
  • 2002-07-12 WO PCT/DE2002/002559 patent/WO2003017331A2/de not_active Application Discontinuation
  • 2002-07-12 JP JP2003522141A patent/JP4050230B2/ja not_active Expired - Fee Related
  • 2002-07-12 EP EP02762218A patent/EP1417707A2/de not_active Withdrawn
  • 2002-07-12 US US10/486,758 patent/US7268381B2/en not_active Expired - Fee Related
  • 2002-07-23 TW TW091116370A patent/TWI223439B/zh not_active IP Right Cessation

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events