KR20100084642A - Structure and fabrication of semiconductor architecture having field-effect transistors especially suitable for analog applications - Google Patents

Abstract

An insulated-gate field-effect transistor of its source/drain zones for reducing the parasitic capacitance along the pn junction between that source/drain zone and adjoining body material. In particular, the concentration of semiconductor dopant which defines the conductivity type of the body material increases by at least a factor of 10 in moving from that source/drain zone down to an underlying body-material location no more than 10 times deeper below the upper semiconductor surface than that source/drain zone. The body material preferably includes a more heavily doped pocket portion situated along the other source/drain zone. The combination of the hypoabrupt vertical dopant profile below the first-mentioned source/drain zone, normally serving as the drain, and the pocket portion along the second-mentioned source/drain zone, normally serving as the source, enables the resultant asymmetric transistor to be especially suitable for high-speed analog applications.

Description

STRUCTURE AND FABRICATION OF SEMICONDUCTOR ARCHITECTURE HAVING FIELD-EFFECT TRANSISTORS ESPECIALLY SUITABLE FOR ANALOG APPLICATIONS

TECHNICAL FIELD The present invention relates to semiconductor technology, and more particularly, to an insulated gate field-effect transistor ("FET"). All insulated gate FETs ("IGFETs") described below are surface-channel enhancement-mode IGFETs except where otherwise indicated.

An IGFET is a semiconductor device that electrically insulates a gate electrode from a channel zone in which a gate dielectric layer extends between a source zone and a drain zone. The channel zone in an enhanced-mode IGFET is a body region portion (often referred to as a substrate or substrate region) that forms a pn junction with the source and drain, respectively. In an enhanced-mode IGFET, the channel zone consists of all the semiconductor material between the source and the drain. During IGFET operation, charge carriers move from source to drain through the channel induced in the channel zone along the upper semiconductor surface. The threshold voltage is the value of the gate-source voltage that the IGFET switches between its on and off states for a given definition of an on / off state. The channel length is the distance between the source and the drain along the upper semiconductor surface.

IGFETs are used in integrated circuits ("ICs") to perform a variety of digital and analog functions. As IC operating performance has advanced over the years, IGFETs have been innovatively miniaturized, leading to a gradual reduction to the minimum channel length. IGFETs that operate in a manner defined by traditional models for IGFETs are often characterized as "long-channel" devices. An IGFET is described as a "short-channel" device when the channel length is reduced to the extent that the behavior of the IGFET significantly deviates from the typical IGFET model. Although both short channel and long channel IGFETs are used in ICs, the majority of ICs used for digital functions in very high density integrated circuit (VLSI) applications have the smallest channel length that can be easily produced through available lithography techniques. Ready

The depletion region extends along the junction between the source region and the body region. The other depletion region extends along the junction between the drain region and the body region. There is a high electric field in each depletion region. Under certain conditions, especially when the channel length is small, the drain depletion region can laterally extend into the source depletion region and integrate with the source depletion region below the upper semiconductor surface. This phenomenon is referred to as (bulk) punchthrough. If a punchthrough occurs, the operation of the IGFET cannot be controlled by its gate electrode. Punchthrough must be avoided.

As IGFET size is reduced, a variety of techniques have been used to improve the performance of IGFETs, including those operating in short channel systems. One performance improvement technique involves providing a two part drain to the IGFET to reduce hot-carrier injection (HCI). IGFETs are also generally provided with similarly configured two-part sources.

1 shows a conventional long n-channel IGFET 20 as described in US Pat. No. 6,548,842 B1 (Bulucea et al.). On the upper surface of the IGFET 20 is a recessed electrically isolated field-isolation region flanked by an active semiconductor island 24 having n-type source / drain (“S / D”) zones 26 and 28. (22) is provided. Each S / D zone 26 or 28 consists of a very heavily doped main portion 26M or 28M and a lateral extension 26E or 28E that is lightly doped but still heavily doped.

S / D zones 26 and 28 are channels of p-type body material 32 which consist of a lightly doped lower portion 34, a heavily doped middle well portion 36, and an upper portion 38. It is separated from each other by the zone 30. While most of the upper body-material portion 38 is suitably doped, the portion 38 is ion-implanted and heavily doped halo pocket portions 40 and extending along the S / D zones 26 and 28, respectively. 42). The IGFET 20 further includes a gate dielectric layer 44, a gate electrode 46 overlying it, electrically insulating gate sidewall spacers 48 and 50, and metal silicon compound layers 52, 54, and 56.

S / D zones 26 and 28 are generally mirror images of each other. In addition, the halo pocket portions 40 and 42 are also generally mirror images of one another such that the channel zone 30 is longitudinally classified symmetrically with respect to the channel dopant concentration. As a result, the IGFET 20 is a symmetric device. During an IGFET operation, one of the two S / D zones 26 or 28 can operate as a source and the other S / D zone 28 or 26 can operate as a drain. This is particularly suitable for the digital situation where each of the S / D zones 26 and 28 operate as a source and a drain respectively for a specific time period and respectively as a drain and a source for another time period.

2 shows how the net dopant concentration N _N varies as a function of the longitudinal distance x for the IGFET 20. Since IGFET 20 is a symmetric device, FIG. 2 presents only a half profile starting from the channel center. Curve segments 26M ^* , 26E ^* , 28M ^* , 28E ^* , 30 ^* , 40 ^* , and 42 ^* in FIG. 2 represent net dopant concentrations of regions 26M, 26E, 28M, 28E, 30, 40, and 42, respectively. . The dashed curve segment 40 "or 42" represents the total concentration of the p-type dopant forming the halo pocket 40 or 42, which is the S / D zone 26 or 28 between forming the pocket 40 or 42. And a p-type dopant introduced into position for.

The presence of halo pockets 40 and 42 in IGFET 20 not only helps alleviate undesirable roll off of the threshold voltage at short channel lengths, but also reduces the net p− in channel zone 30. Form dopant concentration is caused to increase along each S / D zone 26 or 28, especially the lateral extensions 26E or 28E. Thus, the onset of punchthrough is alleviated because the thickness of the channel zone portion of the depletion region extending along the source-operated S / D zone 26 or 28 is reduced.

Body material 30 is provided with additional doping properties to further mitigate punchthrough. Based on the information shown in US Pat. No. 6,548,842 B1, FIG. 3A shows that the absolute concentration N _T of the p-type and n-type dopants extend through the main S / D portion 26M or 28M as a result of additional doping properties. It shows roughly how it changes as a function of depth y along the vertical line. Curve segment 26M ″ or 28M ″ in FIG. 3A represents the total concentration of n-type dopant that defines main S / D portion 26M or 28M. Curve segments 34 ", 36", 38 ", 40" and 42 "together represent the total concentration of p-type dopants that define respective regions 34, 36, 38, 40 and 42.

Additional doping characteristics are p-type anti-punchthrough ("APT; anti-punchthrouch") dopants that reach a maximum concentration at depths greater than 0.1 μm below the top semiconductor surface but below 0.4 μm below the top surface. Is achieved by ion implanting the -shaped upper body-material portion 38. In the situation shown in FIG. 3A where the main S / D portions 26M and 28M extend approximately 0.2 [mu] m below the top surface, the p-type APT dopant reaches its highest concentration at a depth of approximately 0.2 [mu] m. By placing the p-type APT dopant in this manner, the thickness of the channel-zone portion of the depletion region extending along the pn junction of the source-operated S / D zone 26 or 28 is further reduced to further mitigate the punchthrough. Is reduced.

Well region 36 is defined by ion implanting IGFET 20 into a p-type well dopant that reaches the highest concentration at a depth below the highest concentration of the p-type APT dopant. Although the highest concentration of the p-type well dopant is slightly higher than the highest concentration of the p-type APT dopant, the vertical profile of the entire p-type dopant is relative to the main S / D portion (26M or 28M) at the location of the highest well-dopant concentration. Flat. In particular, the N _T concentration of the total p-type dopant decreases significantly less than five times as it proceeds to the main S / D portion (26M or 28M) at the location of the highest well-dopant concentration.

US Pat. No. 6,548,842 B1 discloses that the p-type dopant profile along the vertical line described above through the main S / D zone (26M or 28M) reaches the highest concentration at a depth between the depths of the highest concentrations of APT and well dopant. It is disclosed that it can be smoother by injecting additional p-type dopants. This situation is shown in Figure 3b for the variation of this IGFET 20 where curve segment 58 "represents the variation caused by the additional p-type dopant. In Figure 3b, the highest concentration of the additional p-type dopant is APT. Between the highest concentrations of and well dopants, therefore, the N _T concentration of the total p-type dopant is significantly less than five times as it moves from the position of the highest well-dopant concentration to the main S / D portion (26M or 28M). Decreases.

In particular, for many analog applications where current flows only one way through the IGFET during device operation, a symmetrical IGFET structure is not needed. As disclosed in US Pat. No. 6,548,842 B1, the halo pocket portion can be removed at the drain side. Thus, IGFET 20 becomes long N-channel IGFET 60 as shown in FIG. 4A. Since channel zone 30 is asymmetrically longitudinally doped, the IGFET 60 is an asymmetric device. S / D zones 26 and 28 in IGFET 60 function as a source and a drain, respectively. 4B shows an asymmetric short n-channel IGFET 70 corresponding to long channel IGFET 60. In the IGFET 70, the source side halo pocket 40 approaches the drain 28. Net dopant concentration N _N as a function of longitudinal distance x along the upper semiconductor surface for IGFETs 60 and 70 is shown in FIGS. 5A and 5B, respectively.

Asymmetric IGFETs 60 and 70 receive the same APT and well implants as symmetric IGFET 20. Thus, along the vertical lines extending through the source 26 and the drain 28, the IGFETs 60 and 70 pass through the drain 28 due to the absence of the halo pocket 42 where the dashed curve segment 62 ″ is formed. It has the dopant distribution shown in Figure 3A, except that it exhibits a vertical dopant distribution, if the IGFET structure is provided with additional well implants for further flattening the vertical dopant profile, Figure 3B shows the dopant distribution through drain 28. It shows the inevitable vertical dopant distribution to redo the curve segment 62 "

US Pat. Nos. 6,078,082 and 6,127,700 (both Bulucea) describe IGFETs with asymmetric channel zones but with different vertical dopant properties than those used in the novel IGFETs of US Pat. No. 6,548,842 B1. In addition, an IGFET having an asymmetric channel zone is (a) "Asymmetrical Halo Source GOLD drain (HS-GOLD) Deep Sub-half Micron n-MOSFET Design for Reliability and Performance" IEDM by Buti et al. Tech . Dig . , December 3-6, 1989, pages 26.2.1-26.2.4, (b) "A Cost-Effective 0.25 μm L _eff BiCMOS Technology Featuring Graded-Channel CMOS (GCMOS) and a Quasi-Self- by Chai et al. Aligned (QSA) NPN for RF Wireless Applications ", Procs . 2000 Bipolar / BiCMOS Circs . and Tech . Meeting , 24-24 September 2000, pages 110-113, (c) "Channel Engineering for High Speed Sub-1.0 V Power Supply Deep Submicron CMOS" by Cheng et al . , 1999 Symp . VLSI Tech ., Dig . Tech . Paps . , June 14-16, 1999, pages 69 and 70, (d) "Channel Engineering for Analog Device Design in Deep Submicron CMOS Technology for System on Chip Applications" by Deshpande et al., IEEE Trans . Elec . Devs . , September 2002, pages 1558-1565, (e) "A High Performance 0.1 μm MOSFET with Asymmetric Channel Profile" by Hiroki, IEDM Tech . Dig . , December 1995, pages 17.7.1-17.7.4, (f) "Improving Manufacturability of an RF Graded Channel CMOS Process for Wireless Applications" by Lamey et al., SPIE Conf. Microelec. Dev. Tech. II, September 1998, pp. 147-155, (g) Ma, et al., "Graded-Channel MOSFETs (GCMOSFET) for High Performance, Low Voltage DSP Applications", IEEE Trans. VLSI Systs . Dig . , December 1997, pages 352-358, (h) "Laterally-Doped Channel (LDC) Structure for Sub-Quarter Micron MOSFET" by Matsuki et al . , 1991 Symp. VLSI Tech ., Dig . Tech . Paps . , Pp. 28-30, 1991, pages 113 and 114, and (i) Su, et al., "A High-Performance Scalable Submicron MOSFET for Mixed Analog / Digital Applications," IEDM. Tech . Dig . , December 1991, pages 367-370.

The term "mixed signal" refers to an IC that includes both digital and analog circuit blocks. Typically, digital circuits use the most aggressively scaled n-channel and p-channel IGFETs to obtain the maximum potential digital speed at a given current leakage specification. Analog circuits use IGFETs and / or bipolar transistors that are subject to different performance requirements than digital IGFETs. In general, the requirements for analog IGFETs include high linear voltage gain, good small and large signal frequency response at high frequencies, good parameter matching, low input noise, well controlled electrical parameters for active and passive components, and reduction. Parasitic devices (especially reduced parasitic capacitance). Using the same transistors for analog and digital blocks may be economically beneficial, but this typically results in weak analog performance. Many of the requirements imposed on analog IGFET performance conflict with the results of digital scaling.

Among other things, the electrical parameters of analog IGFETs are subject to more precise specifications than IGFETs in digital blocks. In analog IGFETs used as amplifiers, the output resistance of the IGFET needs to be maximized to maximize its intrinsic gain. The output resistance is also important in setting the high frequency performance of the analog IGFET. In contrast, output resistance in digital circuits is of relatively less importance. As long as the digital circuit can distinguish the logic states of logic "0" and logic "1", for example, higher current driving and consequently reduced values of output resistance in the digital circuit instead of higher digital switching speeds are acceptable. Can be.

The shape of the electrical signal passing through the analog transistors is important for circuit performance and should be kept as reasonable as possible, usually free of harmonic distortion and noise. Harmonic distortion is mainly caused by the nonlinearity of transistor gain and transistor capacitance. Thus, the degree of linearity requirement for analog transistors is very high. Parasitic capacitances at pn junctions have inherent voltage nonlinearities that must be reduced in analog blocks. In contrast, signal linearity in digital circuits typically has secondary importance.

The small signal analog speed performance of IGFETs used in analog amplifiers is determined at the small signal frequency limits and involves small signal gain and parasitic capacitance along the pn junction to the source and drain. Similarly, the large signal analog speed performance of the analog amplifier IGFET is determined at the large signal frequency limit and involves nonlinearity of the IGFET characteristics.

The digital speed of the logic gate is defined in terms of the large signal switching time of the transistor / load combination, thus involving the drive current and the output capacitance. Thus, analog speed performance is determined differently than digital speed performance. Optimizations for analog speed and digital speed can be different, which will require different transistor parameter requirements.

Digital circuit blocks mainly use tiny IGFETs that can be manufactured. As a result, the dimensional spread is inherently large, so parameter matching in digital circuits is often relatively poor. In contrast, in analog circuits, good parameter matching is generally needed to achieve the required performance. It is typically required that analog transistors be manufactured with dimensions larger than digital IGFETs by keeping the analog IGFET as short as possible in order to have as low a source-drain propagation delay as possible.

In view of the foregoing, it is desirable to have a semiconductor architecture that provides IGFETs with good analog characteristics. Analog IGFETs must have a high small signal rate with high intrinsic gain, high output resistance, and reduced parasitic capacitance (especially reduced parasitic capacitance along the source and drain junctions). It is also desirable that this architecture be able to provide high performance digital IGFETs.

The present invention provides such an architecture. In accordance with the present invention, the semiconductor structure includes a primary IGFET having a relatively low parasitic capacitance along the pn junction of at least one of the pn junctions forming the source / drain boundary. Although available for digital applications, the main IGFETs are particularly suitable for analog applications and can achieve good analog performance.

The semiconductor structure of the present invention may include additional IGFETs similar to the main IGFET but configured with opposite polarity. Accordingly, the two IGFETs form a complementary-IGFET architecture that is particularly useful for analog circuits. The semiconductor structure of the present invention may also include additional IGFETs, or two additional counter-polar IGFETs that are particularly suitable for digital circuits. Thus, the overall architecture can be used in mixed signal ICs.

Returning to the main IGFET, it includes a channel zone, a pair of source / drain (“S / D”) zones, a gate dielectric layer overlying the channel zone, and a gate electrode overlying the gate dielectric layer over the channel zone. The main IGFET is produced from a semiconductor body having a body material doped with a semiconductor dopant of the first conductivity type to become the first conductivity type. The channel zone is part of the body material and thus is of the first conductivity type. The S / D zone is located in the semiconductor body along the upper surface of the semiconductor body and is laterally separated by the channel zone. Each S / D zone is of a second conductivity type opposite to the first conductivity type to form a pn junction with the body material. Body material extends laterally under the S / D zone.

Importantly, the dopant of the first conductivity type in the body material is a lower body that is 10 times or less deep, preferably 5 times or less deeper than a particular S / D zone of the S / D zones below the upper semiconductor surface. Have a concentration decreasing at least 1/10, preferably at least 1/20 when moving upward from the material position to that particular S / D zone. In other words, the concentration of the dopant of the first conductivity type in the body material moves downward to the body-material position at a depth of 10 times or less, preferably 5 times or less, below the upper semiconductor surface than the particular S / D zone. When increased by at least 10 times, preferably at least 20 times. Usually, these subsurface body-material positions lie primarily below the channel zone and the entire S / D zone, respectively. By providing this "hypoabrupt" dopant distribution to the body material, the parasitic capacitance along the pn junction between the body material and the particular S / D zone is relatively low. Thus, major IGFETs can achieve high analog performance.

Typically, the primary IGFET is an asymmetric device in which the channel zones are asymmetrically doped dopant longitudinally. Specifically, the concentration of the dopant of the first conductivity type in the body material is such that where the channel zone meets a particular S / D zone along the upper semiconductor surface and where the channel zone is along the top surface the remaining S / D of the S / D zones. Lower than where John meets. Thus, while a particular S / D zone constitutes a drain during IGFET operation, the remaining S / D zones constitute a source. The concentration of the dopant of the first conductivity type in the body material is typically lower, preferably at least 1/10, where the channel zone meets the drain along the top surface than where the channel zone meets the source along the top surface. Preferably at least 1/20 lower. In other words, the concentration of the dopant of the first conductivity type in the body material is typically at least 10 times higher and preferably where the channel zone meets the source along the channel surface than where the channel zone meets the drain along the upper surface. Is at least 20 times higher.

The electric line of force from the drain is located in the channel zone near the source and is located in the channel zone near the source instead of removing the ionized dopant atoms in the depletion region along the source to adversely lower the absolute value of the potential barrier for most charge carriers coming from the source. Because it removes ionized dopant atoms that provide higher channel-zone dopant concentrations in close proximity, high dopant concentrations along the source side of the channel zone protect the source from relatively high electric fields in the drain. This alleviates punchthrough. Accordingly, the combination of the aforementioned hypobolite vertical dopant profile under a particular S / D zone (ie, drain here) and increased channel-zone dopant concentration at the source side can achieve high analog performance without punchthrough failure.

The hypobolite vertical dopant profile under a particular S / D zone can be implemented in a variety of ways. In one implementation, the concentration of the dopant of the first conductivity type in the body material reaches a local maximum at the aforementioned substrate body-material location underneath a particular S / D zone. And the concentration of the dopant of the first conductivity type in the body material typically decreases sharply as it moves upward from its body-material position to a particular S / D zone.

In the fabrication of the aforementioned implementation of the main IGFET according to the present invention, a semiconductor well dopant of the first conductivity type is typically introduced by ion implantation into the semiconductor body to define the well portion of the first conductivity type. The use of ion implantation in performing well-doping allows the well dopant to reach its maximum concentration at the aforementioned subsurface body-material locations. The gate electrode is provided on top of the semiconductor material intended to be a channel zone, and is separated from the semiconductor material by the gate dielectric material. A second conductive semiconductor source / drain dopant is introduced into the semiconductor body to form an S / D zone.

Additional processing is performed to complete the fabrication of the aforementioned implementation of the main IGFET. The well-doping step and further processing are performed under conditions that cause the vertical dopant profile below the particular S / D region to be hypoblotted. In particular, the concentration of well dopant is reduced by at least 1/10 when moving from the aforementioned subsurface body-material position to a particular S / D zone.

At least at the end of IGFET fabrication, the body material is of a first conductivity type. The semiconductor material constituting the body material and the S / D zone at the end of IGFET fabrication may initially be of the second conductivity type. If so, the well-doping step converts the lower portion of this material to the first conductivity type. In one version of the fabrication process, complementary doping is performed through a semiconductor dopant of the first conductivity type to convert the remaining upper portion of the material to the first conductivity type. In another version of the fabrication process, the portion of the well dopant diffuses upwards to the upper portion of this material during further processing and is not particularly performed by other doping of the first conductivity type or second conductivity type following the well-doping step. Induce all of the upper portions of the material to be virtually converted to the first conductivity type.

In another implementation of the main IGFET, the concentration of the dopant of the first conductivity type in the body material substantially experiences a step decrease as it moves upward from the subsurface body-material location to the particular S / D zone. For example, the body material may comprise a subsurface (embedded) body-material portion and a sub-adjacent body-material portion extending directly to the upper semiconductor surface and overlying the subsurface body-material portion including S / D. It may also include. The subsurface body-material portion lies underneath the S / D zone and, at its closest, at a depth of 10 times or less, preferably 5 times or less, below the upper semiconductor surface than the S / D zone. The subsurface body-material portion may for example be predominantly uniformly doped. Next, the concentration of the dopant of the first conductivity type in the body material is intersected from the sub-surface body-material portion to the surface-adjacent body-material portion, and through the surface-adjacent body-material portion, through the specific S / D zone. And substantially experience step reduction at least 1/10 when moving further upwards, and remain at least 10 times smaller than in the subsurface body-material portion.

In other words, the present invention is particularly suitable for analog circuits and meets a semiconductor architecture having an IGFET, or a pair of anti-polar IGFETs. The architecture may include additional IGFETs, or a pair of counter-polar additional IGFETs, which are particularly suitable for digital circuits. The resulting architecture can handle mixed-signal applications very well. Thus, the present invention provides a significant advance over the prior art.

1 is a front cross-sectional view of a prior art symmetric long n-channel IGFET.
FIG. 2 is a graph of net dopant concentration along the upper semiconductor surface as a function of longitudinal distance from the channel center for the IGFET of FIG. 1.
3A and 3B are graphs of absolute dopant concentration as a function of depth along the vertical line through the source / drain zone at two different well-doping conditions for the IGFETs of FIGS. 1, 4A and 4B.
4A and 4B are front cross-sectional views of each prior art asymmetric long n-channel IGFET and prior art asymmetric short n-channel IGFET.
5A and 5B are graphs of net dopant concentration along the upper semiconductor surface as a function of longitudinal distance from the channel center for each IGFET of FIGS. 4A and 4B.
6 is a front cross-sectional view of an asymmetric long n-channel IGFET constructed in accordance with the present invention to have a semiconductor well portion of the same conductivity type as the underlying semiconductor material.
7A-7C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of longitudinal distance along the upper semiconductor surface for the IGFET of FIGS. 6, 18A, 68A, or 68B.
8A-8C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along the vertical line through the source of the IGFET of FIG. 6, 11, or 13.
9A-9C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along a pair of vertical lines through the channel zone of the IGFET of FIGS. 6, 11, 13, or 15; to be.
10A-10C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along a vertical line through the drain of the IGFET of FIGS. 6, 11, 13, 18A, or 18B. .
11 is a front cross-sectional view of an asymmetrical short n-channel IGFET constructed in accordance with the present invention to have a semiconductor well portion of the same conductivity type as the underlying semiconductor material.
12A-12C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of longitudinal distance along the upper semiconductor surface for the IGFET of FIG. 11.
13 is a front cross-sectional view of another asymmetric long n-channel IGFET configured in accordance with the present invention to have a semiconductor well portion of the same conductivity type as the underlying semiconductor material.
14A-14C are respective graphs of individual dopant concentrations, absolute dopant concentrations and net dopant concentrations as a function of longitudinal distance along the upper semiconductor surface for FIGS. 13, 15, 18B or 18C.
15 is a front cross-sectional view of an additional asymmetric long n-channel IGFET configured in accordance with the present invention to have a semiconductor well portion of the same conductivity type as the underlying semiconductor material.
16A-16C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along a vertical line extending through the drain of the IGFET of FIG. 15.
17A-17C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along a vertical line extending through the drain of the IGFET of FIG. 15 or 18C.
18A-18C are front cross-sectional views of three separate long n-channel IGFETs constructed in accordance with the present invention, each having a semiconductor well portion of the same conductivity type as the underlying semiconductor material.
19A-19C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along a vertical line extending through the source of the IGFET of FIG. 18A or 18B.
20A-20C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along a vertical line extending through the source of the IGFET of FIG. 18C.
21 is a front cross-sectional view of an asymmetric long n-channel IGFET constructed in accordance with the present invention having a semiconductor well portion of a conductivity type opposite to the underlying semiconductor material.
22A-22C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of longitudinal distance along the upper semiconductor surface for the IGFET of FIG. 21 or 27A.
23A-23C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along a vertical line through the source of the IGFET of FIG. 21 or 25.
24A-C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along the vertical line through the drain of FIGS. 21, 25, 27A, or 27B.
25 is a front cross-sectional view of another asymmetric long n-channel IGFET configured in accordance with the present invention having a semiconductor well portion of a conductivity type opposite to the underlying semiconductor material.
26A-26C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of longitudinal distance along the upper semiconductor surface for the IGFET of FIG. 25 or 27B.
27A and 27B are front cross-sectional views of two separate long n-channel IGFETs constructed in accordance with the present invention, each having a semiconductor well portion of a conductivity type opposite to the underlying semiconductor material.
28A-28C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along a vertical line extending through the source of the IGFET of FIG. 27A or 27B.
29.1 and 29.2 of FIG. 29 are front cross-sectional views of two portions of a complementary-IGFET semiconductor structure constructed in accordance with the present invention.
30.1 and 30.2 of FIG. 30 are front cross-sectional views of two portions of another complementary-IGFET semiconductor structure constructed in accordance with the present invention.
31A to 31O, 31P.1 to 31R.1, and 31P.2 to 31R.2 are front cross-sectional views illustrating steps of fabricating the complementary-IGFET semiconductor structure of FIGS. 29.1 and 29.2. The steps of FIGS. 31A-31O apply to the structural portions shown in both FIGS. 29.1 and 29.2. 31p.1-31r.1 show additional steps leading to the structural part of FIG. 29.1. 31p.2 to 31r.2 show additional steps leading to the structural part of FIG. 29.2.
32A-32C are alternatives to the steps of FIG. 31E for fabricating changes in the complementary-IGFET semiconductor structure of FIGS. 29.1 and 29.2 starting with the structure of FIG. 31D repeated as FIG. 32A in accordance with the present invention. Is a front cross-sectional view showing the typical steps.
33A-33F illustrate the steps of FIGS. 31C-31F to fabricate a change in the complementary-IGFET semiconductor structure of FIGS. 29.1 and 29.2 starting through the structure of FIG. 31B repeated as FIG. 33A in accordance with the present invention. A front cross-sectional view showing other alternative steps for.
34 is configured in accordance with the present invention to have a semiconductor well portion of a conductivity type opposite to the semiconductor material immediately underlying and without complementary n-type dopant implantation into the semiconductor material over the well portion as defined earlier. A front cross sectional view of an asymmetric long p-type IGFET made in accordance with the present invention. Asymmetry prepared according to the process of FIGS. 31A-31O, 31P.1-31R.1 and 31P.2-31R.2 using alternative steps of FIGS. 32A-32C or 33A-33F. The p-channel IGFET is an implementation of the p-channel IGFET of FIG. 34.
35A-35C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of longitudinal distance along the upper semiconductor surface for the IGFET of FIG. 34.
36A-36C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along a vertical line through the source of the IGFET of FIG. 34.
37A-37C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along a pair of vertical lines through the channel zone of the IGFET of FIG. 34.
38A-38C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along the vertical line through the drain of the IGFET of FIG. 34.
39 and 40 show net dopant concentrations as a function of depth and longitudinal distance for each computer simulation of (i) an asymmetrical stage n-channel IGFET constructed in accordance with the present invention and (ii) a reference symmetrical stage n-channel IGFET. 3d graph.
41 and 42 are graphs showing dopant contours as a function of depth and longitudinal distance from the source location for each of the computer-simulated IGFETs of FIGS. 39 and 40.
FIG. 43 is a graph of net dopant concentration as a function of longitudinal distance from the source location for the computer-simulated IGFETs of FIGS. 39 and 40.
44A and 44B are respective graphs of absolute dopant concentration and net dopant concentration as a function of depth along a pair of vertical lines through each source and drain for the computer-simulated IGFETs of FIGS. 39 and 40. .
45A and 45B are graphs of direct transconductance and direct drain current as a function of gate-source voltage at respective threshold and saturation conditions for the computer-simulated IGFETs of FIGS. 39 and 40.
46A and 46B show (i) a novel asymmetric long n-channel IGFET generally corresponding to the novel short channel IGFET of FIG. 39 and (ii) a reference symmetric field generally corresponding to the reference short channel IGFET of FIG. 40. It is a graph of direct transconductance and direct current drain current as a function of gate-source voltage at threshold and saturation conditions, respectively, for computer simulation of an n-channel IGFET.
FIG. 47 shows gates for computer simulations of (i) the novel IGFET of FIG. 39, (ii) the reference IGFET of FIG. 40, and (iii) an additional reference symmetrical stage n-channel IGFET lacking anti-punchthrough injection. It is a graph of direct current drain current density as a function of source voltage.
FIG. 48 is a graph of linear drain current as a function of drain-source voltage for the computer-simulated IGFETs of FIGS. 39 and 40.
49 is a circuit diagram of an n-channel IGFET and associated parasitic capacitances.
50 is a circuit diagram of a small signal model of the n-channel IGFET and associated parasitic capacitance of FIG. 49.
51A-51C are circuit diagrams of single-IGFET amplifiers arranged in a common-source, common-gate, and common drain configuration, respectively.
52 is a circuit diagram of a single-IGFET amplifier arranged in a common-source short-output configuration.
53 is a circuit diagram of a small signal model of the amplifier of FIG. 52.
54 is a graph of net dopant concentration as a function of distance from pn junctions for models of three different p-type dopant distributions.
FIG. 55 is a graph of depletion-layer capacitance as a function of reverse voltage for the models of the three dopant distributions of FIG. 54.
FIG. 56 is a graph of net body dopant concentration as a function of distance from the pn junction for a model of junction capacitor where the more lightly doped side experiences a step change in dopant concentration.
FIG. 57 is a graph of junction capacitance of a region as a function of reverse voltage for the junction capacitor modeled in FIG. 56.
58A and 58B show composite front cross-sectional views of dopant contours as a function of depth and longitudinal distance from the center of the channel for computer simulation of each asymmetric short n-channel IGFET and asymmetric long n-channel IGFET constructed in accordance with the present invention. It is a graph.
FIG. 59 is a graph of direct drain-body capacitance as a function of drain-body voltage for the computer-simulated IGFETs of FIGS. 39 and 40.
FIG. 60 is a graph of direct source-body capacitance as a function of source-body voltage for the computer-simulated IGFETs of FIGS. 39 and 40.
FIG. 61 is a graph of cut-off frequency as a function of series drain current for the computer-simulated IGFETs of FIGS. 39 and 40 and the additional new IGFET of FIG. 63.
62 shows (i) a novel asymmetric long n-channel IGFET corresponding to the new short channel IGFET of FIG. 39, (ii) a reference symmetric long n-channel IGFET corresponding to the reference short channel IGFET of FIG. 40, and ( iii) A graph of the cut-off frequency as a function of the series drain current for computer simulation of an additional novel asymmetric long n-channel IGFET corresponding to the additional novel short channel IGFET of FIG. 63.
63 is a front sectional view of another computer-simulated asymmetric short n-channel IGFET constructed in accordance with the present invention.
FIG. 64 is a graph of net dopant concentration as a function of longitudinal distance from the source location for the computer-simulated IGFETs of FIGS. 39 and 63.
65 shows (i) an asymmetric n-channel IGFET constructed in accordance with the present invention, (ii) a reference symmetric n-channel IGFET with a halo pocket portion along each source / drain zone, and (iii) each source / drain Is a graph of the threshold voltage as a function of channel length for a reference symmetric n-channel IGFET lacking halo pockets along the zone.
66 is a front sectional view of a further complementary-IGFET semiconductor structure constructed in accordance with the present invention.
67 is a graph of absolute dopant concentration as a function of depth for (i) two asymmetric n-channel IGFETs constructed in accordance with the present invention and (ii) a reference symmetric n-channel IGFET.
68A and 68B are front cross-sectional views of two separate additional asymmetric long n-channel IGFETs constructed in accordance with the present invention.
69A-69C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along the vertical line through the source of the IGFET of FIG. 68A or 68B.
70A-C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along a pair of vertical lines through the channel zone of FIG. 68A or 68B.
71A-71C are respective graphs of individual dopant concentrations, absolute dopant concentrations, and net dopant concentrations as a function of depth along a vertical line extending through the drain of the IGFET of FIG. 68A or 68B.
72A-72D are front cross-sectional views of four additional discrete complementary-IGFETs constructed in accordance with the present invention.
Similar reference signs are used in the description and the drawings of the preferred embodiments to represent the same or very similar items or items. Numerical portions of reference symbols with single prime ('), double prime ("), asterisk ( ^* ), and pound ( ^# ) signs in the figures containing graphs each indicate similarly numbered regions or zones in the other figures. "X" in the cross-sectional view of the IGFET provided with the semiconductor well dopant indicates the position of the maximum concentration of the well dopant, The electrical isolation spacer (not shown) is based on the manufacturing method of these IGFETs, Figure 13, Figure 15, Figure 18B. 18C, 25, 27B and 34 may be located along sidewalls of the gate electrode of the IGFET.
In the dopant-distribution graph, "individual" dopant concentration means individual concentrations of n-type dopants introduced separately and p-type dopants introduced separately, and "absolute" dopant concentrations refer to the total n-type dopant concentration and the total concentration. p-type dopant concentration. The "net" dopant concentration in the dopant-distribution graph is the difference between the absolute (or total) n-type dopant concentration and the absolute (or total) p-type dopant concentration. The net dopant concentration is represented as a net "n-type" when the absolute n-type dopant concentration exceeds the absolute p-type dopant concentration, and the net when the absolute p-type dopant concentration exceeds the absolute n-type dopant concentration. It is represented as "p-type".

Reference numerals used in the following and in the figures refer to the following meanings where the adjective "lineal" means per unit IGFET width, and the adjective "areal" means per unit side region. Has

A _I ≡ current gain

Depletion-Area Capacitance in C _da ≡ Region

C _d0a ≡ 0 Value of local depletion-area capacitance at reverse voltage

C _DB ≡ Drain-Body Capacitance

C _DBw ≡ direct drain-body capacitance

C _GB ≡ gate-body capacitance

C _GD ≡ Gate-Drain Capacitance

Gate dielectric capacitance in the C _GIa region

C _GS ≡ gate-source capacitance

C _L ≡ load capacitance

C _SB Shock Source-Body Capacitance

C _SBw ≡ Direct Source-Body Capacitance

f ≡ frequency

f _T ≡ cut-off frequency

f _{Tpeak 피크} Peak value of cut-off frequency

g _m 고유 Inherent transconductance of IGFET

g _mw 직 direct transconductance of IGFET

g _mb 트랜스 transconductance of the body electrode

g _{meff 유효} Effective transconductance of the IGFET in the presence of a source resistor

g _{msatw 직} saturation transconductance of IGFET at saturation

H _A ≡ amplifier transition function

I _D ≡ Drain Current

I _Dw ≡ series drain current

I _D0w ≡ 0 Leakage of direct drain current at gate-source voltage

i _i ≡ Small signal input current

i _o ≡ Small signal output current

K _S 상대 relative permittivity of semiconductor materials

k ≡ Boltzmann constant

L ≡ channel length

L _G ≡ gate electrode length

L _{GDoverlap 길이} Longitudinal distance at which the gate electrode overlaps (or overlies) the drain

L _{GSoverlap 길이} Longitudinal distance at which gate electrode overlaps source

N _A ≡ acceptor dopant concentration

N _B Net Dopant Concentration in Body Material

N _B0 , N _B0 '≡ pn The value of the net dopant concentration at the junction-adjacent constant-concentration portion of the material on the lighter doped side of the junction.

N _B1 , N _B1 '값 The value of the net dopant concentration at the junction-remote constant-concentration portion of the material on the lightly doped side of the pn junction

N _D ≡ Donor Dopant Concentration

The value of the net dopant concentration in the constant-concentration material on the more heavily doped side of the N _D0 ≡ pn junction

N _I ≡ individual dopant concentration

N _N ≡ Net Dopant Concentration

N _T ≡ absolute dopant concentration

n _i ≡ Intrinsic carrier concentration

q ≡ charge

R _D 직렬 Series Resistance at Drain of IGFET

Series resistance at the gate electrode of R _G ≡ IGFET

R _on 선형 linear-area on-resistance of IGFET

Series resistance at the source of R _s IGFET

s ≡ conversion variable

T temperature

t _d ≡ Depletion-area thickness

t _d0 ≡ 0 Depletion-zone thickness at reverse voltage

t _GI ≡ gate dielectric thickness

V _BI ≡ Built-in Voltage

V _BS ≡ DC Body-Source Voltage

V _DB ≡ DC Drain-Body Voltage

V _DD high high supply voltage

V _DS ≡ DC Drain-Source Voltage

V _GS ≡ DC Gate-Source Voltage

V _g ≡ gate voltage amplitude

V _in ≡ input voltage amplitude

V _out ≡ output voltage amplitude

V _R ≡ DC Reverse Voltage

V _Rmax 최대 Maximum DC Reverse Voltage

V _SB ≡ DC Source-Body Voltage

V _SS low supply voltage

V _T ≡ threshold voltage

v _gs ≡ small-signal gate-to-source voltage

v _nsat ≡ electron saturation rate

W ≡ channel width

x ≡ longitudinal distance

y ≡ depth, vertical distance, or distance from pn junction

y _d 값 The value of the distance from the pn junction to the remote boundary of the depletion region in the body material

y _d0 값 The value of the distance from the pn junction of the body material with the step change at a uniform net dopant concentration to the remote boundary of the junction-adjacent constant-concentration portion.

y _dmax 값 The value of the distance from the pn junction of the body material with the step change at a uniform net dopant concentration to the remote boundary of the junction-remote constant-concentration portion.

y _D 깊이 Depth value at drain bottom

y _S 깊이 Depth value at source bottom

y _ST 깊이 Depth value at the remote boundary of the junction-adjacent upper body-material portion

y _W ≡ Depth value at the location of the maximum concentration of the well dopant

ε ₀ 유전 permittivity of free space (vacuum)

μ _n ≡ electron mobility

ω ≡ each frequency

ω _in 값 Value of each frequency at the input pole

ω _out 값 Value of each frequency at the output pole

value of each frequency in ω _z ≡ 0

ω _p 값 Value of each frequency at the pole

In this specification, ie, both above and below, the long channel n-channel IGFET and the short channel n-channel IGFET are referred to as the long n-channel IGFET and the short n-channel IGFET, respectively. Similarly, each of the long channel p-channel IGFET and the short channel p-channel IGFET is referred to herein as the long p-channel IGFET and the short p-channel IGFET. As described below, the term “surface-adjacent” means adjacent to (or extending to) the upper semiconductor surface (ie, the upper surface of the semiconductor body consisting of a single crystal, or predominantly single crystal semiconductor material).

In general, no particular channel-length value separates the short channel scheme and the long channel scheme of the IGFET operation, or generally does not distinguish between the long channel IGFET and the short channel IGFET. Short-channel IGFETs, or IGFETs operating in short-channel systems, are simply IGFETs whose characteristics are significantly affected by the short-channel effects. Long channel IGFETs, or IGFETs operating in a long channel scheme, are the opposite of short channel IGFETs. A channel length value of approximately 0.4 mm roughly constitutes the boundary between the short channel and long channel schemes for the background art in US Pat. No. 6,548,642 B1, and includes various factors (eg, gate dielectric thickness, minimum printing Long channel / short channel boundaries can occur at higher or lower values of the channel length based on possible feature sizes, channel zone dopant concentrations, and source / drain-body junction depths.

IGFET when vertical body-material dopant profile under drain is hypobrupt due to subsurface maximum at well dopant concentration

6 shows an asymmetric long n-channel IGFET 100 constructed in accordance with the present invention to be particularly suitable for high speed analog applications. Long channel IGFET 100 has a single crystal silicon ("mono") in which a pair of highly heavily doped n-type source / drain (again, "S / D") zones 102 and 104 are located along the upper semiconductor surface. Silicon ") from the semiconductor body. Typically, S / D zones 102 and 104 are referred to below as source 102 and drain 104, respectively, because they typically function as source and drain, respectively, but not necessarily.

Usually, the drain 104 is doped slightly higher than the source 102. Usually, the maximum value of the net dopant concentration N _{N in} the source 102 along the upper semiconductor surface is at least 1 × 10 ²⁰ atoms / cm 3, typically 4 × 10 ²⁰ atoms / cm 3. The maximum value of the concentration N _{N in} the drain 104 along the top surface is usually at least 1 × 10 ²⁰ atoms / cm 3, typically 4 × 10 ²⁰ atoms, to slightly exceed the maximum top-surface N _N concentration in the source 102. Slightly larger than / cm 3. However, as discussed below in connection with the novel IGFET of FIG. 63, drain 104 is often more lightly doped than source 102. For example, the maximum value of the concentration N _{N in} the drain 104 along the upper surface may be 5 × 10 ¹⁹ atoms / cm 3, and the maximum top-surface N _N concentration in the source 102 is at least 1 × 10 ²⁰ atoms. In the case of / cm 3, it can be lowered to about 1 × 10 ¹⁹ atoms / cm 3.

Source 102 extends to distance y _S below the upper semiconductor surface. Drain 104 extends to depth y _D below the upper semiconductor surface. Source depth y _S is usually 0.1-0.2 μm, typically 0.15 μm. Drain depth y _D is usually 0.15-0.3 mu m, typically 0.2 mu m. Thus, drain depth y _D usually exceeds source depth y _S by 0.05-0.1 μm.

The source 102 and the drain 104 are (p) forming a source-body pn junction 110 with the source 102 and (b) a drain-body pn junction 112 with the drain 104. It is laterally separated by an asymmetric channel zone 106 of the mold body material 108. The p-type body material 108 is typically more than the lightly doped lower portion 114, the heavily doped middle well portion 116, and the source 102 and drain 104 below the upper semiconductor surface. It consists of an upper portion 118 extending deeply. Thus, the upper body-material portion 118 typically includes the entirety of the channel zone 106. In this case, the p− lower body-material portion 114 and the p + well portion 116 extend laterally under the source 102 and the drain 104.

P + well portion 116 is defined as a p-type semiconductor well dopant that is vertically distributed in a substantially Gaussian manner to reach the maximum subsurface concentration at depth y _W below the upper semiconductor surface. In general, “X” in FIG. 6 represents the location of the maximum subsurface concentration of the p-type well dopant. The concentration of the p-type well dopant at the depth y _W is usually 1 × 10 ¹⁸ -1 × 10 ¹⁹ atoms / cm 3, and usually 5 × 10 ¹⁸ atoms / cm 3. The maximum well concentration depth y _W above the source depth y _S and the drain depth y _D is usually 0.5-1.0 μm, typically 0.7 μm. In addition, the depth y _W is generally 10 times or less, and preferably 5 times or less of the drain depth y _D. That is, the location of the maximum concentration of the p-type well dopant is 10 times or less, preferably 5 times or less, than the drain 104 below the top surface.

The upper boundary and the lower boundary of the heavily doped well portion 116 are slightly inaccurate because the well 116 is located inside the doped semiconductor material of the same conductivity type (p type) as the well 116. The semiconducting material that forms the boundary of the well 116 usually has a relatively uniform low p-type background dopant concentration, as shown below. Typically, the upper and lower boundaries of the well 116 are defined as locations where the concentration of the p-type well dopant is equal to the p-type background dopant concentration. The concentration of the total p-type dopant along the upper and lower boundaries of the well 116 is excluding any location where the well 116 extends to another p-type material more heavily doped than the well 116. It is twice the concentration of the p-type background dopant. Under this boundary definition, the top boundary of the well 116 is usually 0.2-0.5 μm, typically 0.3 μm below the top semiconductor surface. The lower boundary of the well 116 is usually 0.9-1.3 μm, typically 1.1 μm below the top surface.

The depletion region (not shown) extends from the source-body pn junction 110 across the channel zone 106 to the drain-body pn junction 112 along the upper semiconductor surface during the IGFET operation. The average thickness of the surface depletion region is generally less than 0.1 μm, usually around 0.05 μm. Although the upper and lower boundaries of the well portion 116 are slightly inaccurate, the concentration of the p-type well dopant drops to an electrically insignificant level at a depth of less than 0.1 μm below the top surface. Thus, well 116 is located substantially below the surface depletion region.

The p-type upper body-material portion 118 defines a heavily doped pocket portion 120 that extends along the source 102 to the upper semiconductor surface and terminates at a location between the source 102 and the drain 104. Include. 6 shows an example where the p + pocket portion 120 extends deeper below the top surface than the source 102 and drain 104. In particular, FIG. 6 shows an example where pocket portion 120 laterally extends under source 102 and primarily reaches p + well portion 116. As described below with respect to FIGS. 18A-18C, pocket portion 120 may extend below the upper surface to a lower depth than shown in FIG. 6. The remainder of the p-type upper body-material portion 118 (ie, the outer portion of the pocket portion 120) is indicated by item 124 in FIG. 6. Upper body-material remainder 124 is lightly doped and extends along drain 104. Accordingly, between the source 102 and the drain 104, the channel zone 106, which is made of both p-type semiconductor material, has a portion of the source-side p + pocket portion 120 and the drain-side p- upper body-material Formed by a portion of the remainder 124.

Gate dielectric layer 126 is located on the upper semiconductor surface and extends over channel zone 106. Gate electrode 128 is located on gate dielectric layer 126 over channel zone 106. Gate electrode 128 extends partially over source 102 and drain 104. In the example of FIG. 6, gate electrode 128 is made of polycrystalline silicon (“polysilicon”) that is very heavily doped n-type. Gate electrode 128 may be formed of another electrically conductive material (eg, metal or polysilicon) sufficiently doped to p-type to be electrically conductive.

Typically, the top electrode of the source 102, drain 104, and n ++ gate electrode 128 are each provided with a thin layer of electrically conductive metal silicon compound (not shown in FIG. 6), so that regions 102, 104, And 128) facilitate the creation of electrical contact. In this case, the gate electrode 128 and the overlying metal silicon compound layer form a composite gate electrode. The source 102, drain 104, and channel zone 106 are typically laterally surrounded by electrically insulating field regions (likewise not shown in FIG. 6) recessed to the upper semiconductor surface. , And 106, an active semiconductor island. Examples of metal silicon compound layers and field-isolation regions are described below with respect to FIGS. 29.1 and 29.2.

The presence of the p + pocket portion 120 along the source 102 causes the channel zone 106 to be classified in the longitudinal direction (ie, in the direction of the channel length) with respect to the channel dopant concentration. Since the actual mirror image of the source-side pocket portion 120 is not located along the drain 104, the channel zone 106 is asymmetrically doped in the longitudinal direction. The p + well portion 116 is located below the p− upper body-material remainder 124 extending along the drain 104. This configuration of the p + well 116 and the p− upper body-material remainder 124 causes the vertical dopant profile to be hypobrupt in the portion of the body material 108 underlying the drain 102. That is, the concentration of the p-type dopant usually increases at least 10 times as it proceeds from the drain-body junction 112 down through the p- upper body-material rest 124 to the p + well 116. The combination of the longitudinally asymmetric dopant fraction in the channel zone 106 and the hypobolite vertical dopant profile through the drain 104 in the portion of the body material 108 below the drain 104 causes the IGFET 100 to punch through. It is possible to have very good analog characteristics while avoiding

7A-7C (collectively "FIG. 7"), 8A-8C (collectively "FIG. 8"), 9A-9C (collectively "FIG. 9"), and 10A-FIG. 10c (collectively "FIG. 10"), understanding the longitudinally asymmetric dopant classification in the channel zone 106 and the hypobolite vertical dopant profile in the portion of the body material 108 under the drain 104. Becomes easy. 7 shows exemplary dopant concentrations along the upper semiconductor surface as a function of longitudinal distance x. An example dopant concentration as a function of depth y along vertical line 130 through source 102 is shown in FIG. 8. 9 shows an example dopant concentration as a function of depth y along a pair of vertical lines 132 and 134 through channel zone 106. Vertical line 132 passes through the source-side pocket portion 120. Vertical line 134 passes through a vertical position between pocket portion 120 and drain 104. An exemplary dopant concentration as a function of depth y along vertical line 136 through drain 104 is shown in FIG. 10.

FIG. 7A specifically illustrates the concentrations N _I of individual semiconductor dopants along the upper semiconductor surface that largely define regions 102, 104, 120, and 124, thus establishing the longitudinal dopant classification of channel zone 106. Illustrated. 8A, 9A, and 10A define regions 102, 104, 114, 116, 120, and 124 vertically along vertical lines 130, 132, 134, and 136 and thus below drain 104. The concentration N _I of the individual semiconductor dopant, which establishes a hypobolite vertical dopant profile in the portion of the body material 108, is specifically shown. Curves 102 'and 104' represent the concentrations N _I (surface and vertical) of the n-type dopant used to form the source 102 and the drain 104, respectively. Curves 114 ′, 116 ′, 120 ′, and 124 ′ represent concentrations N _I (surface and / or vertical) of p-type dopants used to form regions 114, 116, 120, and 124, respectively. . Items 110 ^# and 112 ^# indicate where the net dopant concentration N _N proceeds to zero and thus marks the positions of the pn junctions 110 and 112, respectively.

The concentrations N _T of the total p-type and total n-type dopants in regions 102, 104, 120, and 124 along the upper semiconductor surface are shown in FIG. 7B. 8B, 9B, and 10B show all p-types and all n-types in regions 102, 104, 114, 116, 120, and 124 along vertical lines 130, 132, 134, and 136. The concentration N _T of the dopant is shown. Curve segments 114 ", 116", 120 ", and 124" corresponding to regions 114, 116, 120, and 124, respectively, represent the total concentration N _T of the p-type dopant. In FIG. 7B, item 106 "corresponds to channel zone 106 and represents the channel-zone portion of curve segments 120" and 124 ". The total concentration N _T of the n-type dopant is represented by source 102 and drain ( Represented by curve 102 " and curve 104 " respectively corresponding to 104. Curves 102 " and 104 " in Fig. 7B are the same as curves 102 'and 104', respectively, in Fig. 7A. Curves 102 "and 104" in FIGS. 8B and 10B are the same as curves 102 'and 104' in FIGS. 8A and 10A, respectively.

7C shows the net dopant concentration N _N along the upper semiconductor surface. Net dopant concentrations N _N along vertical lines 130, 132, 134, and 136 are shown in FIGS. 8C, 9C, and 10C. Curve segments 114 *, 116 *, 120 *, and 124 * represent the net concentrations N _N of p-type dopants in respective regions 114, 116, 120, and 124. Item 106 ^* in FIG. 7C represents the combination of channel zone curve segments 120 ^* and 124 ^* , thus representing the concentration N _N of the net p-type dopant in channel zone 106. The concentrations N _N of the net n-type dopants in the source 102 and drain 104 are represented by curves 102 ^* and 104 ^* , respectively.

Keeping in mind the foregoing general opinion with respect to FIGS. 7-10, FIG. 7A shows that along the upper semiconductor surface, the p-type dopant in the source-side pocket portion 120 has two main components (ie, two separate doping operations). Component provided). One of the main components of the p-type dopant in pocket portion 120 along the top surface is the p-type background dopant, represented by curve 124 'in FIG. 7A. Usually, the p-type background dopant is present at low, mainly uniform concentrations throughout the monosilicon material including regions 102, 104, 114, 116, and 120. Under the pocket 120 and the upper body-material remainder 124, the p-type background dopant is represented by a curve segment 114 ′ as shown in FIGS. 8A, 9A, and 10A. The concentration of the p-type background dopant is usually 1 × 10 ¹⁵ -1 × 10 ¹⁶ atoms / cm 3, and usually 5 × 10 ¹⁵ atoms / cm 3.

Another major component of the p-type dopant in the source-side pocket portion 120 is the p-type pocket (or channel-classification) dopant, indicated by curve 120 'in FIG. 7A. p- type pocket dopant is high the upper portion to define a pocket (120) is provided with ^{- (2 × 10 18 atoms /} ㎤, typically 1 × 10 ¹⁸ atoms / ㎤ usually 5 × 10 ¹⁷⁾ the surface concentration. The specific value of the top-surface concentration of the p-type pocket dopant is typically precisely adjusted to within 10% accuracy to set the threshold voltage of the IGFET 100.

The boundary of the source-side pocket portion 120 is (a) the section of the upper semiconductor surface, (b) the pn junction section formed by the source-body junction 110, and (c) the p- of the body material 108. Consists of a mold section. Although the p-type section of the boundary of the pocket 120 is slightly inaccurate, the p-type pocket section is typically defined as the location where the concentration of the p-type pocket dopant is the same as that of the p-type background dopant. As long as the pocket 120 does not impinge on the well portion 116, the p-type dopant concentration along the p-type section of the boundary of the pocket 120 is such that the p-type pocket-part boundary section is in contact with the upper semiconductor surface. Twice the background dopant concentration, including where it meets.

In addition, the p-type pocket dopant is present at the source 102 as represented by curve 120 'in FIG. 7A. The concentration N _I of the p-type pocket dopant in source 102 is substantially constant along its upper surface. The concentration N _I of the p-type pocket dopant is at a substantially constant upper surface source level to some extent as it moves into the channel zone 106 as it moves from the source 102 along the upper semiconductor surface along the upper semiconductor surface in the longitudinal direction. And then essentially falls to zero from that level at the location between the source 102 and the drain 104.

Through the entire p-type dopant in the channel zone 106 along the top semiconductor surface, which is the sum of the pocket dopant and the p-type background along the top surface, the entire p-type channel-zone dopant along the top surface is shown in FIG. 7B. Curve segment 106 ". The change in curve segment 106" is entirely in the zone 106 along the top surface as it moves across the channel zone 106 longitudinally from the source 102 to the drain 104. The concentration N _T of the p-type dopant is at a substantially constant high level to some extent as it moves into the zone 106, and from its high level at the position between the source 102 and the drain 104 to a low p-type background level After falling down, it remains at a low background level for the remainder of the distance to the drain 104.

In some embodiments, the concentration N _I of the p-type pocket dopant in source 102 may be at a substantially constant source level only along a portion of the top surface of source 102, and inside the top surface of source 102. May decrease as it moves longitudinally along the upper semiconductor surface from the position of to the source-body junction 110. In this case, the concentration N _I of the p-type pocket dopant in the channel zone 106 immediately after crossing the source-body junction 110 when moving across the zone 106 longitudinally towards the drain 104. Begins to fall on. Thus, similarly, the concentration N _T of the total p-type dopant in zone 106 along the top surface is also drained from source 102 instead of at a substantially constant source level to some extent moving into zone 106. As it moves longitudinally across channel zone 106 with 104 it begins to fall just after crossing junction 110.

The concentration N _I of the p-type pocket dopant in the channel zone 106 along the upper semiconductor surface is substantially at source level for a non-zero distance in the longitudinal direction from the source-body junction 110 to the zone 106. Regardless of whether or not the concentration N _T of the total p-type dopant in the zone 106 along the top surface is greater than that where the zone 106 meets the source 102, the drain 106 has a drain 104. Is lower from where it meets. In particular, the concentration N _T of the total p-type dopant in zone 106 is generally at least 10 times more at drain-body junction 112 along the top surface than at source-body junction 110 along the top surface. Low, preferably at least 20 times lower, more preferably at least 50 times lower, and typically at least 100 times lower.

FIG. 7C shows that the net dopant concentration N _N of the net p-type dopant in zone 106 along the top surface drops to zero at pn junctions 110 and 112, as represented by curve 106 ^* . It changes along the top surface in a manner similar to the concentration N _T of the total p-type dopant in the zone 106. Thus, the source side of channel zone 106 has a higher amount of net p-type dopant than the drain side. The source side amount of the high p-type dopant in the channel zone 106 causes the thickness of the channel side portion of the depletion region along the source-body junction 110 to be reduced.

In addition, the high p-type dopant concentration along the source side of the channel zone 106 protects the source 102 from the relatively high electric field at the drain 104. This means that the field line from drain 104 ionized p-type dopant in pocket portion 120 instead of removing ionized dopant atoms in the depletion region along source 102 and adversely lowering the potential barrier to electrons. Occurs because the atom is removed. Thus, the depletion region along the source-body junction 110 suppresses punchthrough to the depletion region along the drain-body junction 112. By appropriately selecting the amount of high source-side p-type dopant in channel zone 106, punchthrough in IGFET 100 is avoided.

The p-type dopant in the portion of the body material 108 under the source 102, the channel zone 106, and the drain 104 contains three main components, as shown in FIGS. 8A, 9A, and 10A. Have One of the main components of the p-type dopant in the body-material portion below the regions 102, 104, and 106 is represented by the curve segment 124 ′ or 114 ′ in FIGS. 8A, 9A, and 10A. P-type background dopant. The second major component is a p-type well dopant that defines well portion 116 as indicated by curve 116 'in FIGS. 8A, 9A and 10A.

The last major component of the p-type dopant in the body-material portion below the regions 102, 104, and 106 is the p-type pocket dopant, indicated by curve 120 'in FIGS. 8A and 9A. The p-type pocket dopant is only present in the portion of the body material 108 between the source 102 and the adjacent portion of the channel zone 106. The amount of p-type pocket dopant present in the portion of the body material 108 below the drain 104 is essentially zero or is substantially unimportantly low. Thus, the entire p-type pocket dopant in the portion of the body material 108 below the drain 104 is represented by curves 116 'and 124' or 114 'taken along the vertical line 136 in FIG. 10A, respectively. It consists essentially of only the background dopants and the p-type well.

The entire p-type dopant in the portion of the body material 108 below the drain 104 is represented by curve segment 116 ″ and its extensions 124 ″ (upward) and 114 ″ (downward) in FIG. 10B. Curve segment 116 ″ represents the sum of the p-type well and the background dopants in well portion 116. Each of the curve segments 114 ″ and 124 ″ corresponds to the p- lower body-material portion 114 and the p- upper body-material remainder 124. Since the concentration N _I of the p-type background dopant is relatively uniform, the concentration N _T of the total p-type dopant in the portion of the body material 108 under the drain 104 is at a subsurface location substantially equal to the depth y _W ( That is, substantially at the point where the p-type well dopant reaches its maximum concentration.

As shown by the change in the combination curve segment 116 ″ / 124 ″ in FIG. 10B, the concentration N _T of the total p-type dopant in the portion of the body material 108 under the drain 104 is determined in the well portion 116. It decreases in a hypobolite manner by at least 1/10 as it moves up along the vertical line 136 from the subsurface location of the maximum concentration of the p-type dopant to the drain 104. The concentration N _T of the total p-type dopant in the portion of the body material 108 below the drain 104 is preferably at least 20 times, more preferably when moving from the location of the maximum p-type well concentration to the drain 104. Preferably at least 40 times, even more preferably at least 80 times, typically approximately 100 times or more. Also, in general, the concentration N _T of the total p-type dopant in the portion of the body material 108 below the drain 104 is the maximum p-type well, as indicated by the combination curve segment 116 "/ 124". It rapidly decreases when moving to the drain 104 at the location of the concentration.

The net dopant in the portion of the body material 108 below the drain 104 is a p-type dopant. FIG. 10C shows the concentration N _N of the net dopant in the portion of body material 108 under drain 104 as represented by the combination of curve segments 116 ^* and 124 ^* , the body under drain 104. Same as the concentration N _T of the total p-type dopant in the portion of the body material 108 under the drain 104, except that the concentration N _N in the portion of the material 108 drops to zero at the drain-body junction. In a vertical fashion. The hypobolite vertical dopant in the body material 108 portion below the drain 104 causes the vapor phase capacitance associated with the drain-body junction 112 to decrease for the reasons described below. This allows the IGFET 100 to have an increased analog speed.

Moving through the source 102 along the vertical line 130 to the dopant distribution, the overall p-type dopant in the portion of the body material 108 below the source 102 is curved 116 ′, 124 ′ in FIG. 8A. And p'-type wells, background and pocket dopants, respectively. The example of FIG. 6 in which the source side pocket portion 120 faces the well 116 below the source 102 such that the concentration N _I of both the p-type well and the pocket dopant exceeds the p-type background concentration at the location where they meet. In, the entire p-type dopant in the portion of the body material 108 below the source 102 is represented by the combination of curve segments 116 ″ and 120 ″ in FIG. 8B. In embodiments where pocket 120 extends below source 102 but does not face well portion 116, the entire p-type dopant of the portion of body material 108 under source 102 is p-top. It is represented by a combination of curve segments 116 ″ and 120 ″ corresponding to the body-material remainder 124.

As shown by the combination curve segment 116 ″ / 120 ″ in FIG. 8B, the concentration N _T of the total p-type dopant of the portion of body material 108 under the source 102 is initially at the well portion 116. It is reduced to 1/10, typically approximately 1/30 when moving upward through the body material 108 from the subsurface location of the maximum concentration of the p-type dopant. Once the local minimum is reached under the source 102, the concentration N _T of the total p-type dopant in the portion of the body material 108 under the source 102 rises before reaching the source 102.

The net dopant in the portion of body material 108 below drain 102 is a p-type dopant. FIG. 8C illustrates the body material under source 102 in the example of FIG. 6 where source side pocket portion 120 meets well portion 116, as represented by a combination of curve segments 116 ^* and 120 ^* . The concentration N _N of the net dopant in the portion of 108 is different except that the net dopant concentration N _N in the portion of the body material 108 below the source 102 drops to zero at the source-body junction 110. The portion of body material 108 below source 102 changes vertically in a manner similar to the concentration N _T of the total p-type dopant. Based on various factors, such as the amount by which the depth of the pocket 120 exceeds the depth of the source 102, the vertical dopant profile underlying the source 102 is typically less than a decrease in parasitic capacitance along the drain-body junction 112. Despite this, the parasitic capacitance along the source-body junction 110 is induced to be reduced.

Returning to FIG. 9, which discusses the dopant concentrations N _I , N _T , and N _N along the vertical lines 132 and 134 through the channel zone 106, the insertion opening behind the reference symbols 120 ′, 120 ″, and 120 ^* . "132" represents the dopant concentration along the vertical line 132. The insertion hole "134" after the reference symbols 124 ', 124 ", and 124 ^* indicates the dopant concentration along the vertical line 134.

11 illustrates an asymmetric stage n-channel IGFET 140 constructed in accordance with the present invention to be particularly suited for high speed analog applications. The short channel IGFET 140 is a variation of the long channel IGFET 100 in which the channel length is shortened to a range in which IGFET operation occurs in the short channel system. The channel length is shortened by appropriately reducing the length of the gate electrode 128. In the example of FIG. 11, the p-type pocket portion 120 extends far enough until it meets the drain 104 across the channel zone 106.

12A-12C (collectively “FIG. 12”) illustrate the IGFET 140 as a function of the longitudinal distance x to facilitate understanding of the asymmetric longitudinal dopant classification in the channel zone 106 of the IGFET 140. Exemplary dopant concentrations N _I , N _T and N _N are respectively shown along the upper semiconductor surface. All of the analyzes described above with respect to IGFET 100 in conjunction with FIG. 7 indicate that the shortened length of channel zone 106 in IGFET 140 is such that the dopant distribution in channel zone 106 of IGFET 140 is equal to FIG. 12. It is applied to IGFET 140, except to induce it to be different from the dopant distribution in channel zone 106 of IGFET 100 as described below.

As in FIG. 7A, the concentration N _I of the p-type pocket dopant along the upper semiconductor surface is represented by curve 120 ′ in FIG. 12A. Unlike what occurs in the channel zone 106 of the IGFET 100, the concentration N _I of the p-type pocket dopant along the upper surface of the channel zone 106 of the IGFET 140 is the source 102 and the drain 104. It doesn't fundamentally drop to zero in the position between. Instead, as shown by curve 120 ′ in FIG. 12A, the concentration N _I of the p-type pocket dopant along the top surface of channel zone 106 of IGFET 140 is the entire top surface of the channel zone. And thus substantially greater than zero, so that the channel zone 106 / pocket portion 120 is at a small finite value that meets the drain 104. The pocket dopant concentration N _I in IGFET 140 is at a much larger value where zone 106 / pocket 120 meets source 102 along the top surface of the channel zone.

The above-described change of the concentration N _I of the p-type pocket dopant along the upper surface of the channel zone 106 in the IGFET 140 is the absolute dopant concentration N _T and the net dopant concentration N along the upper semiconductor surface of the IGFET 140. Reflected in _N As in FIG. 7B, the curve segment 106 ″ in FIG. 12B is a top semiconductor in which the total p-type dopant in the zone 106 along the top surface is the sum of the pocket dopant and the p-type background along the top semiconductor surface. Along the surface is shown the concentration N _T of the total p-type dopant in the channel zone 106. The change in curve segment 106 " in FIG. 12B shows the upper semiconductor surface from the source 102 through the channel zone 106. Along the upper semiconductor surface of IGFET 140 along the top semiconductor surface of the IGFET 140 to the extent that the concentration N _T of the total p-type dopant in the zone 106 moves to the zone 106 in the source 102. It is substantially constant at the high level, and then decreases to a low level slightly above the background concentration when reaching drain 104 at that high level. In particular, the concentration N _T of the total p-type dopant in the channel zone 106 of the IGFET 140 is such that the zone 106 is the source 102 where the zone 106 meets the drain 104 along the upper semiconductor surface. ) Increases rapidly when moving to meet

Similar to that described above for the IGFET 100, the concentration N _I of the p-type pocket dopant along the upper semiconductor surface in the IGFET 140 depends on the channel surface 106 at the source 102 along the upper surface in the longitudinal direction. May begin to decrease immediately after passing through the source-body pn junction 110 as it moves across the drain 104. As observed while moving from drain 104 to source 102 rather than from source 102 to drain 104, the concentration N _T of the total p-type dopant in channel zone 106 in IGFET 140 is , Rapidly increases along the upper semiconductor surface as it moves from where the zone 106 meets the drain 104 to where the zone 106 meets the source 102. However, the concentration N _T of the total p-type dopant in the zone 106 along the top surface of the IGFET 140 is such that the zone 106 has a drain 104 rather than where the zone 106 meets the source 102. Where it meets the above specifications for the lower IGFET 100.

12C shows the concentration N _N of the net p-type dopant along the upper surface of the channel zone 106 of the IGFET 140, as represented by the curve 106 ^* , along the upper semiconductor surface. In the zone 106 of the IGFET 140 along the upper semiconductor surface, except that the concentration N _N of the net p-type dopant in the zone 106 proceeds to zero at the pn junctions 110 and 112. It changes in a similar manner to the concentration N _T of the total p-type dopant. As in the channel zone 106 of the IGFET 100, the source side of the channel zone 106 in the IGFET 140 has a larger amount of net p-type dopant compared to the drain side of the IGFET 140. High source side p-type doping in the channel zone 106 of the IGFET 140 leads to a reduction in the thickness of the channel side portion of the depletion region extending along the source-body junction 110.

Source 102 and drain 104 are located closer to each other in IGFET 140 than in IGFET 100. Thus, there is more probability in IGFET 140 than in IGFET 100 that a depletion region extending along source 102 may punch through for a depletion region extending along drain 104. However, much of the source-side p-type dopant in the channel zone 106 of the IGFET 140 is more likely than the punchthrough that occurs in the IGFET 140 compared to other identical single n-channel IGFETs lacking the pocket portion 120. Reduce the possibility

Dopant concentration along vertical lines 130, 132, and 136 through source 102, channel zone 106, and drain 104, respectively, in IGFET 140 is substantially the same as in IGFET 100. Thus, the concentrations N _I , N _T , and N _{N shown} in FIGS. 8-10 along vertical lines 130, 132, and 136 are applied to IGFET 140. Thus, IGFET 140 has a hypobolite vertical dopant profile in the portion of body material 108 below drain 104. This leads to a reduction in parasitic capacitance associated with drain-body pn junction 112 such that IGFET 140 has increased analog speed, as described below.

Source 102 may be dopant classified in the longitudinal direction to reduce its source (serial) resistance R _S. As described below, reducing the source resistance R _S has particular advantages in analog IGFET applications. The longitudinal dopant fractionation at this source 102 typically involves configuring it as a lightly doped lateral extension and main portion that terminates the channel zone 106 along the upper semiconductor surface. Drain 104 may be provided with a similar longitudinal dopant fraction to reduce hot-carrier injection. Thus, providing longitudinal dopant classification in both S / D zones 102 and 104 depends on whether zones 102 and 104 operate as a source and a drain in the general case, or as a drain and as a source, respectively. Regardless, it has advantages.

FIG. 13 shows an asymmetric long n-channel IGFET configured in accordance with the present invention particularly suitable for high speed analog applications, in particular having a longitudinal source / drain dopant classification for reducing source resistance R _S and drain side hot-carrier injection. 150). IGFET 150 consists of (a) an n-type source 102 consisting of a very heavily doped main portion 102M and a lightly doped lateral extension 102E, and (b) an n-type drain It is arranged identically to IGFET 100, except that 104 consists of a very heavily doped main portion 104M and a lighter doped lateral extension 104E. Although doped lighter than the n ++ main S / D portions 102M and 104M, the lateral extensions 102E and 104E are still heavily doped in sub-μm complementary IGFET applications as present. N + lateral extensions 102E and 104E terminate channel zone 106 along the upper semiconductor surface. Gate electrode 128 extends partially for each of the n + side extensions 102E and 104E, but typically does not extend over n ++ main source portion 102M or n ++ main drain portion 104M.

Main S / D portions 102M and 104M extend deeper below the upper semiconductor surface than respective lateral extensions 102E and 104E. As a result, each of source depth y _S and drain depth y _D in IGFET 150 is the depth of main source portion 102M and main drain portion 104M. Pocket portion 120 extends partially laterally below main source portion 102M such that drain depth y _D again exceeds source depth y _S. In addition, pocket 120 extends laterally under source extension 102E. As a result, the drain extension 104E extends deeper below the top surface than the source extension 102E.

The maximum net dopant concentration N _N in the n ++ main source portion 102M along the upper semiconductor surface is usually at least 1 × 10 ²⁰ atoms / cm 3, typically 4 × 10 ²⁰ atoms / cm 3. The maximum net dopant concentration N _N in the n ++ main drain portion 104M along the upper semiconductor surface is usually at least 1 × 10 ²⁰ atoms / cm 3 to slightly exceed the maximum top-surface N _N concentration in the main source portion 102M, It is usually slightly larger than 4 × 10 ²⁰ atoms / cm 3. The maximum net dopant concentration N _N in the n + source extension 102E along the upper semiconductor surface is usually 1 × 10 ¹⁸ -1 × 10 ¹⁹ atoms / cm 3, typically 3 × 10 ¹⁸ atoms / cm 3. The maximum net dopant concentration N _N at n + drain extension 104E along the upper semiconductor surface is typically slightly greater than 3 × 10 ¹⁸ atoms / cm 3 to slightly exceed the maximum top-surface N _N at source extension 102E.

Subject to longitudinal dopant classification at source 102 and drain 104, channel zone 106 is asymmetrically longitudinal dopant classified substantially the same as at IGFET 150, such as in IGFET 100. 14A-14C (collectively “FIG. 14”) illustrate the upper semiconductor surface of IGFET 150 as a function of longitudinal distance x used to examine longitudinal dopant classification in source 102 and drain 104. Thus, exemplary dopant concentrations are shown. 14A shows the concentrations N _I of individual semiconductor dopants that largely define regions 102M, 102E, 104M, 104E, 120, and 124 along the upper semiconductor surface. Concentrations N _T of the total p-type and total n-type dopants of regions 102M, 102E, 104M, 104E, 120, and 124 along the upper semiconductor surface are shown in FIG. 14B. 14C shows the net dopant concentration N _N along the upper semiconductor surface.

FIG. 14A shows the concentration N _I of the n-type dopant, along which the curves 102M ', 102E', 104M 'and 104E' are used to form regions 102M, 102E, 104M and 104E, respectively, along the upper semiconductor surface. Indicates. 14B shows curves 102 " and (b) main drain portion 104M consisting of (a) segments 102M "and 102E" corresponding to main source portion 102M and source extension 102E, respectively; Similar to FIG. 7B subject to the total concentration N _T of the n-type dopant along the upper semiconductor surface represented by the curve 104 "consisting of segments 104M" and 104E "corresponding to the drain extension 104E, respectively. Do. FIG. 14C shows that (a) a segment 102 ^* showing a net dopant concentration N _N at the source 102 along the upper semiconductor surface, corresponding to the main source portion 102M and the source extension 102E, respectively; 102M ^* and 102E ^* ), and (b) a curve 104 ^* representing the net dopant concentration N _N in the drain 104 along the upper semiconductor surface is the main drain portion 104M and the drain extension 104E. Are similar to FIG. 7C with the conditions formed by the segments 104M ^* and 104E ^* , respectively.

Longitudinal dopant classification at source 102 and drain 104 of IGFET 150 reduces source resistance R _S and reduces drain side hot-carrier injection, but asymmetric longitudinal dopant in channel zone 106. There is no significant effect on classification. Thus, asymmetric channel-zone dopant classification in IGFET 150 avoids punchthrough in much the same way as in IGFET 100.

The construction of the well portion 116 and the upper body-material remainder 124 in the IGFET 150 has a substantially vertical dopant profile through the drain 104 and into the underlying body material 108 as in the IGFET 100. Induced to be the same as the hypobolite. The vertical dopant concentration graphs of FIGS. 8-10 are substantially applied to IGFET 150 via vertical lines 130 and 136 running through n ++ main source portion 102M and n ++ main drain portion 104M, respectively. The resulting reduced parasitic capacitance associated with drain-body pn junction 112 enables IGFET 150 to have increased analog speed. Reducing the source resistance R _S further enhances analog performance in the manner described below.

Drain 104 may be vertical dopant sorted to further reduce parasitic capacitance with respect to drain-body junction 112. Similarly, source 102 may be vertical dopant sorted to reduce parasitic capacitance associated with source-body junction 110. Vertical dopant classification involves configuring S / D zone 102 or 104 as the main portion and the lightly doped lower portion, respectively. The vertical source / drain dopant fraction may be combined with the aforementioned longitudinal dopant fraction of the source 102 and drain 104.

With respect to the foregoing, FIG. 15 illustrates an asymmetric long n-channel IGFET 160 constructed in accordance with the present invention to be particularly suitable for high speed analog applications. IGFET 160 is provided with source / drain longitudinal dopant fractionation to reduce source resistance R _S and drain side hot-carrier injection and source / drain vertical dopant to reduce source / drain parasitic capacitance. IGFET 160 further includes (a) a lower portion 102L in which source 102 is lightly doped in main source portion 102M, and (b) drain 104 in low concentration in main drain portion 104M. Same as IGFET 150 except that it includes a doped bottom portion 104L. Lower source portion 102L and lower drain portion 104L are heavily doped n-type.

Source depth y _S and drain depth y _D in IGFET 160 are equal to n + lower source portion 102L and n + lower drain portion 104L, respectively, below n ++ main source portion 102M and n ++ main drain portion 104M. Since it lies at each, it is the depth of n + lower source portion 102L and n + lower drain portion 104L. Pocket portion 120 extends below lower source portion 102L. As a result, the drain depth y _D again exceeds the source depth y _S.

Source 102 and drain 104 of IGFET 160 include n + side source extension 102E and n + side drain extension 104E, respectively, to achieve longitudinal source-drain dopant classification. See FIG. 15. The longitudinal source / drain dopant classification in IGFET 160 has substantially the same characteristics as in IGFET 150. Thus, the longitudinal top-surface dopant concentration graph of FIG. 14 and the associated description of FIG. 14 associated with IGFET 150 apply to IGFET 160. In IGFET 160, because the longitudinal source / drain dopant classification in IGFET 150 (and therefore in IGFET 160) does not have any significant effect on the asymmetric longitudinal dopant classification in channel zone 106. Asymmetric channel-zone dopant classification of s avoids punchthrough in IGFET 160 in substantially the same manner as in IGFET 100.

Understanding vertical dopant classification in IGFET 160 is based on depth along vertical lines 130 and 136 through source 102 and drain 104, respectively, including lower source portion 102L and lower drain portion 104L. 16A-16C (collectively "FIG. 16") and 17A-17C (collectively "FIG. 17") showing exemplary dopant concentrations as a function of y are facilitated. Specifically, FIGS. 16A and 17A show the concentrations N of individual semiconductor dopants that vertically define regions 102M, 102L, 104M, 104L, 114, 116, 120, and 124 along lines 130 and 136, respectively. _{Show I.} Concentrations N _T of total p-type and total n-type dopants in regions 102M, 102L, 104M, 104L, 114, 116, 120, and 124 along lines 130 and 136 are shown in FIGS. 16B and 17B, respectively. Shown. 16C and 17C respectively show net dopant concentration N _N along lines 130 and 136.

16A and 17A show (a) the concentration along the vertical line 130 of the n-type dopant in which curves 102M 'and 102L' were used to form main source portion 102M and lower source portion 104L, respectively. N _i , and (b) concentrations N _I along the vertical line 136 of the n-type dopant in which curves 104M 'and 104L' are used to form main drain portion 104M and bottom drain portion 104L, respectively. Each is similar to FIGS. 8A and 10A except that is indicated. Similarly, FIGS. 16B and 17B show (a) a segment 102M ″ in which the concentration N _T of the total n-type dopant along line 130 corresponds to the main source portion 102M and the lower source portion 102L, respectively. And 102L ″, and (b) the concentration N _T of the entire n-type dopant along line 136 corresponds to the main drain portion 104M and the lower drain portion 102L, respectively. 8B and 10B, respectively, provided that they are represented by a curve 104 "consisting of segments 104M" and 104L ".

16C and 17C show that (a) a curve 102 ^* representing net dopant concentration N _N at source 102 along line 130 corresponds to main source portion 102M and lower source portion 102L, respectively. segment (102M ^* and 102L ^*) is formed by, (b) line 136, the thus drain 104, net dopant concentration N _N indicating curve (104 *) and the main drain portion (104M) and a lower drain portion in which 8C and 10C, respectively, provided that they are formed of segments 104M ^* and 104L ^* corresponding to 102L, respectively. In addition, the dopant concentration along vertical lines 132 and 134 through channel zone 106 of IGFET 160 is substantially the same as in IGFET 100. Thus, the concentrations N _I , N _T , and N _N along the vertical lines 132 and 134 shown in FIG. 9 are applied to the IGFET 160.

Subject to the above-described interpretation of vertical dopant fractionation at the source 102 and drain 104 of IGFET 160, the configuration of well portion 116 and pocket portion 120 in IGFET 160 is IGFET ( Substantially the same as in 100). Thus, the vertical dopant profile under drain 104 is substantially the same as IGFET 160 as in IGFET 100. For this reason, the parasitic parameters associated with the drain-body junction 112 are reduced in the IGFET 160, thus making it possible to have an increased analog speed. Vertical dopant classification at source 102 and drain 104 may be achieved by IGFET 160 reducing (and further reducing) parasitic capacitance along source-body junction 110 and along drain-body junction 112. Further reduction in parasitic capacitance makes it possible to have a higher analog speed. The longitudinal dopant fractionation at source 102 and drain 104 of IGFET 160 reduces source resistance R _S while simultaneously reducing drain side hot-carrier injection.

18A-18C illustrate each of the asymmetrical long n-channel IGFETs 100, 150, and 160 where pocket portion 120 extends below the upper semiconductor surface to a lower depth than source 102 and drain 104. Versions 170, 180 and 190 are shown. For long n-channel IGFET 180 or 190 where source 102 and drain 104 each include source extension 102E and drain extension 104E, pocket portion 120 extends 102E. And 104E) below.

The p-type section of the boundary of the pocket portion 120 in each IGFET 170, 180 or 190 has a concentration of p-type pocket dopant as described above in connection with the IGFET 100. It is defined as the same position as the concentration of. Next, the total p-type dopant concentration along the p-type section of the boundary of pocket portion 120 is twice the background dopant concentration of IGFET 170, 180 or 190. Thus, some p-type pocket dopant is present at source 102 of IGFET 170, 180, or 190 at a depth below that shown for pocket portion 120 in FIGS. 18A-18C. The additional p-type pocket dopant in source 102 offsets (compensates) some n-type dopants that define source 102 along its lower surface. Thus, drain depth y _D in IGFET 170, 180 or 190 exceeds the source depth y _S , although in a smaller amount than in IGFET 100.

Channel zones 106 of each IGFET 170, 180, and 190 are substantially asymmetrically longitudinal dopant classified as described above for each of IGFETs 100, 150, and 160. In this regard, the dopant concentrations N _I , N _T and N _N along the upper semiconductor surface for IGFET 170 are substantially represented in FIG. 7, respectively. Substantially, FIG. 14 shows the dopant concentrations (N _I , N _T and N _N ) along the top surface for IGFETs 180 and 190. Thus, punchthrough is avoided in IGFETs 170, 180, and 190 in the manner described above for IGFET 100.

As described above for IGFETs 100, 150, and 160, each of IGFETs 170, 180, and 190 has a substantially hypobolite vertical dopant profile below drain 104. FIG. 10 shows concentrations N _I , N _T and N _N along vertical line 136 through drain 104 for IGFETs 170 and 180, respectively. Concentrations N _I , N _T and N _N along vertical line 136 through drain 104 for IGFET 190 are substantially shown in FIG. 17. Accordingly, as previously described for IGFET 100, the drain-in each IGFET 170, 180 or 190 to enable each IGFET 170, 180 or 190 to have an increased analog speed. Parasitic capacitance along the body junction 112 is reduced.

19A-19C (collectively "FIG. 19") and 20A-20C (collectively "FIG. 20") show vertical line 130 through source 102 for IGFETs 170, 180 and 190. FIG. Exemplary dopant concentration as a function of depth y along 19 applies to IGFETs 170 and 180. 20 applies to IGFET 190. Specifically, FIGS. 19A and 20A show the concentrations N _I of individual semiconductor dopants that vertically define regions 102, 114, 116 and 120 along line 130. Concentrations N _T of total p-type and total n-type dopants in regions 102, 114, 116, and 120 along line 130 are shown in FIGS. 19B and 20B. 19C and 20C show net dopant concentration N _N , respectively, along line 130.

As indicated by the change in curve segments 116 ″ and 124 ″ in FIGS. 19B and 20B, the concentration N _T of the total p-type dopant in the portion of the body material 108 under the source 102 is the well portion. At 116 decreases in a hypobolite manner by at least 1/10 as it moves up along the vertical line 130 from the subsurface location of the maximum concentration of the p-type dopant to the source 102. Thus, configuring pocket portion 120 lower than source 102 and drain 104 results in a hypobolite vertical profile for the entire p-type dopant in the portion of body material 108 below source 102. To derive. This occurs because much less amount of p-type dopant is located below IGFETs 170, 180 and 190 than in IGFETs 100, 150 and 160. Curve segments 120 "representing the concentration N _T of the total p-type pocket dopant along line 130 through source 102 in FIGS. 8B, 16B, 19B and 20B are shown in FIGS. 19B and 20B. It is mainly located above the source depth y _S but extends significantly below the depth y _S in FIGS. 8B and 16B.

The hypobubble vertical dopant profile in the portion of the body material 108 below the source 102 of the IGFET 170, 180, or 190 is below the drain 104 of the IGFET 170, 180, or 190, and thus the IGFET ( It is quite similar to the hypobolite vertical dopant profile in the portion of the body material 108 below the drain 104 of 100, 150 or 160. Combination curve segment 116 ″ / 120 ″ along vertical line 130 through source 102 in FIGS. 19B and 20B. Vertical line 136 combination curve segment along drain 104 in FIGS. 10B and 17B. 116 "/ 120"). Similar to what occurs in the portion of body material 108 under drain 104, the overall p-type in the portion of body material 108 under source 102 in each IGFET 170, 180 or 190. The concentration N _T of the dopant is preferably at least 1/20, more preferably at least 1/40 when moving from the subsurface location of the maximum p-type well concentration to the very heavily doped material of the source 102. And even more preferably at least 1/80, typically about 1/100. In IGFET 170, 180, or 190, the hypobolite vertical dopant profile of the body material 108 portion below the very heavily doped material of source 102 reduces parasitic capacitance associated with source-body junction 110. It is induced to. The analog speed of the IGFET 170, 180 or 190 is further increased.

Dopant concentrations N _I , N _T and N _N along vertical line 134 through channel zone 106 of IGFET 170, 180 or 190 appear as shown in FIG. 9. The dopant concentrations N _I , N _T and N _N along the vertical line 132 through the channel zone 106 (including the pocket portion 120) of the IGFET 170, 180 or 190 are IGFETs 170, 180. Or curve 120 ″ for concentration N _I of the p-type pocket dopant along line 132 for 190 is similar to curve 120 ′ along line 130 in FIG. 19A or 20A. Is similar to that shown in FIG.

To illustrate the IGFETs 100, 140, 150, 160, 170, 180 and 190, for simplicity, the concentration of the p-type background dopant is any of the IGFETs 100, 140, 150, 160, 170, 180 and 190. It is substantially constant throughout the semiconductor material, including one. However, the concentration of the p-type background dopant can vary as long as the peak value of the p-type background dopant is relatively low compared to the concentration of other p-type dopants.

The well portion 116 of the body material 108 in each of the IGFETs 100, 140, 150, 160, 170, 180 and 190 is the same as the slightly doped semiconductor material (lower body-material portion 114) directly underneath. It is a challenge type. As described below in connection with the fabrication process of FIGS. 31A-31O and 31P.1-31R.2, the start of the p-type semiconductor material slightly doped with the p + well portion 116 and the p + pocket portion 120. This situation usually occurs when created in a region. Thus, the dopant concentration of the bulk of the upper body-material remainder 124 is approximately equal to the low background dopant concentration of the p- starting region.

Alternatively, the semiconductor material lying directly below the well portion 116 may be of the opposite conductivity type as the well 116. Thus, because the well portion 116 is p-type, the semiconductor material directly below the well 116 is n-type. This alternative typically occurs when p + wells 116 and p + pockets 120 are produced in the starting region of a generally doped n-type semiconductor material at relatively uniform net background concentrations. In one embodiment, the portion of the starting n-type region intended to be the upper body-material portion 118 (ie, the portion of the starting n-type region located above the well portion 116) (or the well 116 The position intended for relative to the p-type compensation dopant has an absolute concentration greater than the n-type background dopant concentration of the starting n-type semiconductor region to induce the entire upper body-material portion 118 to be p-type. Doped with. In another embodiment, the portion of the starting n-type region intended to be the upper body-material portion 118 is p-type conductive by updiffusion of the portion of the p-type well dopant in the well portion 116. Converted to type

Adjacent to the n-type background dopant concentration may be the minimum value of the net concentration N _N of the p-type compensation or well dopant in the upper body-material portion 118. However, to ensure that the entire body-material portion 118 is p-type, the minimum value of the concentration N _N of the p-type compensation or well dopant in the portion 118 is generally significantly greater than the n-type background dopant concentration. Amount (eg, at least twice). Thus, the minimum value of the concentration N _N of the p-type compensation or well dopant of the bulk of the body-material portion 118 outside the pocket portion 120 is substantially greater than the n-type background dopant concentration.

FIG. 21 is a lightly doped n-type lower portion in which p- lower body-material portion 114 forms a lower pn junction 194 with p + well portion 116 in accordance with the present invention, similar to FIG. 6. The variation 100V of the asymmetrical field n-channel IGFET 100 replaced by region 192 is shown. Since the lower region 192 is not p-type conductive, the p-type body material 108 in the IGFET 100V replaces the upper portion 196 that replaces the upper body-material portion 118 of the IGFET 100. ) And well portion 116. A portion of the upper body-material portion 196 of IGFET 100V is formed by p + pocket portion 120. The remainder of the upper body-material portion 196, ie the outer portion of the pocket portion 120, is indicated as item 198 of FIG. 21. The upper body-material remainder 198 is p-type lightly doped at a net concentration slightly higher than the n- lower portion 192. Weak p-type doping of the body-material remainder 198 is achieved through the p-type compensation dopant described above. Except for the aforementioned differences and resulting dopant concentration differences, IGFET 100V is constructed and made substantially the same as IGFET 100.

IGFET 100V has the following characteristics similar to that of IGFET 100: (a) asymmetric longitudinal dopant classification in channel 106 and (b) in the portion of body material 108 under drain 104. Has a hypobolite vertical dopant profile. Understanding these features of IGFET 100V, including how these features may differ slightly from the features of IGFET 100, respectively, is illustrated in FIGS. 22A-22C (collectively "FIG. 22"), FIGS. 23A-23. Reference is made to FIGS. 23C (collectively "FIG. 23") and FIGS. 24A-24C (collectively "FIG. 24"). 22 shows exemplary dopant concentrations along the upper semiconductor surface of IGFET 100V as a function of longitudinal distance x. An exemplary dopant concentration along vertical line 130 through source 102 of IGFET 100V is shown in FIG. 23. 24 shows exemplary dopant concentrations along vertical line 136 through drain 104 of IGFET 100V.

22A, 23A, and 24A show concentrations N _I of individual semiconductor dopants that define regions 102, 104, 116, 120, 192, 196, and 198. Each of the curves 192 'and 198' specifically defines the concentration N _I of the n-type background dopant and the p-type compensation dopant, respectively defining the n-bottom region 192 and the p- upper body-material remainder 198, respectively. Indicates. Item 194 ^# indicates where net dopant concentration N _N proceeds below well portion 116 and thus indicates the location of lower pn junction 194.

Concentrations N _T of total p-type and total n-type dopants in regions 102, 104, 116, 120, 192, 196, and 198 are shown in FIGS. 22B, 23B, and 24B. Curve segments 192 "and 198" in FIGS. 22B, 23B, and 24B correspond to n-bottom regions 192 and p- upper body-material remainder 198, respectively. FIG. 22c, FIG. 23c, and 24c in a region of the net _N concentration N type dopant p- and n- type dopants in the net (102, 104, 116, 120, 192, 196 and 198) is shown variously. Curve segments 192 ^* and 198 ^* in FIGS. 22C, 23C, and 24C correspond to n-lower region 192 and p- upper body-material remainder 198, respectively.

22-24 show that (a) the concentration N _I of the n-type background dopant in IGFET 100V is approximately equal to the concentration N _I of the p-type background dopant in IGFET 100, and (b) the IGFET The concentration N _I of the p-type compensation dopant along the upper semiconductor surface of (100 V) is 2-3 times the concentration N _I of the n-type background dopant in IGFET (100 V), and (c) of the p-type compensation dopant the maximum value of the density N _I N _I 2 at a concentration of p- type compensating dopant along the upper semiconductor surface, three times, and therefore, four of the n- type background dopant concentration N _I-9 shows the times given. Except for this difference, the concentrations N _I of the other dopants in IGFET 100V are mainly the same as in IGFET 100 respectively.

More specifically, the concentration N _I of the p-type pocket dopant in IGFET 100V varies in the longitudinal direction in substantially the same manner as in IGFET 100. The change in curve segment 120 ′ in FIG. 22A results in a concentration N _I of the p-type pocket dopant as it moves longitudinally from the source 102 along the upper semiconductor surface of the IGFET 100V to the channel zone 106. To some extent, moving to zone 106 is at an approximately constant top-surface level, which then essentially falls to zero at the location between source 102 and drain 104 from that level.

The total p-type dopant in channel zone 106 along the upper semiconductor surface of IGFET 100V is the sum of the pocket dopant and the compensation dopant. This is different from IGFET 100, where the total p-type dopant in channel zone 106 along the upper semiconductor surface is the sum of the pocket and background dopants. Concentrations of p- type dopant compensation in the illustrated example N _I 2 of the n- type background dopant concentration of N _I-be triple-2 at a concentration of p- N _I-type background dopant in IGFET thus 3 times (100) Therefore, in the illustrated example, the minimum value of the concentration N _T of the entire p-type dopant along the upper surface of the IGFET 100V is equal to 2 − of the minimum value of the concentration N _T of the entire p-type dopant along the upper surface of the IGFET 100. Three times.

Item 106 " in Fig. 22B represents the channel zone portion of curve segments 120 " and 198 ". Similar to what occurs in IGFET 100, the change in curve 106 " ) zone 106, than where the concentration N _T of the entire p- type dopant at the upper semiconductor surface thus channel zone 106, zone 106 is met and the source 102 in the area of intersection with the drain 104 Shows lower. The concentration N _T of the total p-type dopant in the channel zone 106 of IGFET 100V is generally at drain-body junction 112 along the top surface along the top surface than at source-body junction 110 along the top surface. At least 10 times lower, preferably at least 20 times lower, more preferably at least 50 times lower, and typically about 100 times lower. The reason why the typical value of this concentration difference is not more than 100 times as can occur in the IGFET 100 and is around 100 times in the IGFET 100V is that the minimum value of the concentration N _T of the entire p-type dopant along the upper surface This is because this is 2-3 times higher in the described example of IGFET 100V than in IGFET 100.

Referring to FIG. 22C, item 106 ^* here represents a combination of channel-zone curve segments 120 ^* and 198 ^* . 7A and 7 provided that the segment 198 ^* of the curve 106 * for the IGFET 100V is slightly higher than the segment 124 ^* of the curve 106 * for the IGFET 100 in FIG. 7A. The curve 106 ^* in FIG. 22A is quite similar. Thus, the source side of channel zone 106 in IGFET 100V has a high net amount of p-type dopant compared to the drain side. Accordingly, the thickness of the channel side portion of the depletion region along the source-body junction 110 is reduced in IGFET 100V. Thus, the high p-type dopant concentration along the source side of channel zone 106 in IGFET 100V is relatively high in drain 104 for the reasons described above (field-line termination) with respect to IGFET 100. Protect the source 102 from the electric field. Punchthrough in IGFET 100V is avoided.

The p-type dopants of the body material 108 portion below the source 102 and drain 104 of IGFET 100V are well dopant and compensation dopants, represented by curves 116 'and 198' in FIGS. 23A and 24A, respectively. And for the source 102, as indicated by the curve 120 'in FIG. 23A. The change in curve 198 ′ in FIGS. 23A and 24A indicates that the concentration N _I of the p-type compensation dopant reaches a maximum near the bottom of the source 102 and drain 104. This maximum is 4-9 times the concentration N _I of the n-type background dopant in the specific examples of FIGS. 23A and 24A. In the example of FIGS. 23A and 24A the concentration N _I of the p-type compensation dopant drops essentially to zero at a depth less than the depth y _W of the maximum p-type dopant concentration in the well portion 116.

The total p-type dopant in the portion of the body material 108 below the drain 104 of the IGFET 100V is represented by curve segment 116 ″ and its (upward) extension 198 ″ in FIG. 24B. Since the concentration N _I of the p-type compensation dopant drops essentially to zero at a depth less than the depth y _W , the concentration N _T of the total p-type dopant in the body-material portion below the drain 104 is equal to y _W. A maximum is reached at substantially the same subsurface location. As occurs in the IGFET 100, the change in the combination curve segment 116 ″ / 198 ″ in FIG. 24B results in an overall p− in the portion of the body material 108 under the drain 104 of the IGFET 100V. The concentration N _T of the type dopant decreases in a hypobolite manner by at least 1/10 as it moves upward from the subsurface location of the maximum concentration of the total p-type dopant in the well portion 116 to the drain 104.

Typically, in the particular example of FIG. 24B the concentration N _T of the total p-type dopant in the portion of the body material 108 below the drain 104 of the IGFET 100V is equal to the drain (at the location of the maximum p-type well concentration). It decreases to about 1/15 when moving to 104). The reason why such a typical value for hypobolite concentration reduction is not about 1/100 as occurs in IGFET 100 and 1/15 in IGFET 100V is that in IGFET 100V in the example of FIG. 24A. This is because the concentration N _I of the p-type compensation dopant at the bottom of the drain 104 is 4-9 times the concentration N _I of the p-type background dopant at the bottom of the drain 104 in the IGFET 100. However, the vertical dopant profile for the p-type compensation dopant reduces the value of the concentration N _I of the p-type compensation dopant at the bottom of the drain 104 while the entirety of the upper body-material portion 196 is still p-type. Can be degraded to The concentration N _T of the total p-type dopant in the body material portion below drain 104 in IGFET 100V rises from the location of the maximum concentration of the total p-type dopant in well portion 116 to drain 104. When moving, it can immediately decrease to at least 1/20, typically at least 1/40.

As shown by the combination of curve segments 116 ^* and 198 ^* , FIG. 24C shows that the concentration N _N of the net p-type dopant in the body-material portion below drain 104 in IGFET 100V is determined by the drain ( 104 Concentration N _T of the total p-type dopant in the body-material portion below drain 104, except that concentration N _N in the body-material portion below drops to zero at pn junctions 112 and 194. It changes vertically in a similar way. Due to the hypobolite vertical dopant profile in the body-material portion below the drain 104 of the IGFET 100V, the parasitic capacitance associated with the drain-body junction 112 is reduced for the reasons described below. Although the reduction in parasitic capacitance along junction 104 may be smaller in IGFET 100V than in IGFET 100, IGFET 100V still has increased analog speed.

The presence of the p-type compensation dopant at a concentration greater than that of the p-type background dopant in the IGFET 100 is because the p-type pocket dopant is present below the source 102 in the IGFET 100V. Has a significantly less influence on the vertical dopant profile through the source 102 of the IGFET 100V than along the vertical dopant profile through the drain 104. As is apparent by comparing FIGS. 23 and 24, the foregoing comments regarding the vertical dopant profile through the source 102 of the IGFET 100 generally apply to the vertical dopant profile through the source 102 of the IGFET 100V. do.

FIG. 25 illustrates a variation 150V of the asymmetrical field n-channel IGFET 150 in which the n-lower portion 192 replaces the p-lower body-material portion 114, similar to FIG. 13 in accordance with the present invention. Illustrated. Similarly, IGFET 150V includes p-type upper body-material portion 196 that replaces upper body-material portion 118 of IGFET 150. The p- remainder 198 of the upper body-material portion 196 is at a slightly higher net dopant concentration than the n- lower portion 192. Like IGFET 100V, weak p-type doping of the upper body-material remainder 198 in IGFET 150V is achieved through p-type compensation doping. Subject to the presence of n- lower portion 192 and p- upper body-material remainder 198, IGFET 150V allows longitudinal source / drain to reduce source resistance R _S and drain side hot-carrier injection. It is configured substantially the same as in IGFET 150 to have a dopant classification.

26A-26C (collectively “FIG. 26”) show exemplary dopant concentrations along the upper semiconductor surface of IGFET 150V for purposes of examining longitudinal dopant classification in source 102 and drain 104. Indicates. Concentrations N _I are shown in FIG. 26A along the top surface of the individual semiconductor dopants that largely define regions 102M, 102E, 104M, 104E, 120, 192, and 198. FIG. 26B shows the concentrations N _T of total p-type and n-type dopants of regions 102M, 102E, 104M, 104E, 120, 192, and 198 along the upper surface. Net concentration N _N along the top surface is shown in FIG. 26C.

FIG. 26 repeats FIG. 14 performing each modification represented in FIG. 22 to illustrate the n-lower portion 192 and the p-upper body-material remainder 198. The longitudinal dopant classification in the source 102 and drain 104 of IGFET 150V has no significant effect on the asymmetric longitudinal dopant classification in channel zone 106. Asymmetric channel-zone dopant classification in IGFET 150V avoids punchthrough in the same way as in IGFET 150 and thus in much the same way as in IGFET 100.

The configuration of the p + well portion 116 and the p− upper body-material remainder 198 in the IGFET 150V passes through the drain 104 such that the vertical dopant profile to the lower body material 108 is in IGFET 150. Causing it to be in a substantially hypobolite manner. The vertical dopant concentration graphs of FIGS. 23 and 24 along vertical lines 130 and 136 apply substantially to IGFET 150V. Drain-body junction 112 in IGFET 150V has a reduced parasitic capacitance that is not typically reduced as in IGFET 150 but allows IGFET 150V to have an analog switching speed.

27A and 27B show each asymmetric long field n-channel in which the n-lower portion 192 replaces the p-lower body-material portion 114, similarly as in FIGS. 18A and 18B in accordance with the present invention, respectively. Changes in IGFETs 170 and 180 (170V and 180V) are shown. Likewise, each IGFET 170V or 180V includes a p-type upper body-material portion 196 that replaces the upper body-material portion 118 of the IGFET 170 or 180. The p-rest 198 of the upper body-material portion 196 is also at a slightly higher net dopant concentration than the n- lower portion 192. As in IGFETs 100V and 150V, weak p-type doping of the upper body-material remainder 198 in each IGFET 170V or 180V is achieved through a p-type compensation dopant. Subject to the presence of the n-bottom portion 192 and the p-top body-material remainder 198, the pocket portion 120 of each IGFET 170V or 180V may be a source 102 or below a top semiconductor surface. Extends to a lower depth than the drain 104. IGFET 180V also has a longitudinal source / drain dopant classification of IGFET 150V to reduce source resistance R _S and drain side hot-carrier injection.

Channel zones 106 of each of IGFETs 170V and 180V are asymmetrically longitudinal dopant classified as described above for each of IGFETs 100V and 150V. 22 shows concentrations N _I , N _T , and N _N along the upper semiconductor surface for IGFET 170V. Punchthrough in IGFET 170V is avoided for the reasons described above in connection with IGFET 100V. Along the upper semiconductor surface of IGFET 180V, concentrations N _I , N _T , and N _N appear substantially in FIG. 26, respectively. IGFET 180V avoids punchthrough, as described above for IGFET 170V, and as described above for IGFET 100V.

Each of IGFETs 170V and 180V has a hypobolite vertical dopant profile through drain 104 as described above for IGFET 100V. FIG. 24 shows the concentrations N _I , N _T , and N _N along the vertical line 136 through the drain 104 for each IGFET 170V or 180V. As a result, the parasitic capacitance along the drain-body junction 112 at each IGFET 170V or 180V is reduced for the reasons described above with respect to IGFET 100V. Thus, IGFETs 170V and 180V have increased analog speeds.

28A-28C (collectively “FIG. 28”) show the concentrations N _I , N _T , and N _N along the vertical line 130 through the source 102 for each IGFET 170V or 180V. As shown by the change in curves 116 ″ and 198 ″ in FIG. 28B, the concentration N _T of the total p-type dopant in the portion of the body material 108 under the source 102 is the well portion 116. It decreases in a hypobolite manner by at least 1/10 as it moves up from the subsurface location of the maximum concentration of the p-type dopant to source 102 along line 130. As occurs in IGFETs 170 and 180, constructing pocket portion 120 lower than source 102 and drain 104 in IGFETs 170V and 180V in the body-material portion below source 102. This results in a hypobolite vertical profile for the p-type dopant.

The hypobolite vertical dopant profile in the portion of the body material 108 under the source 102 for each IGFET 170V or 180V may be a hypoabrate in the portion of the body material 108 under the drain 104. It is quite similar to the rupture vertical dopant profile. In the particular example of FIG. 28, the concentration N _T of the total p-type dopant in the body-material portion below the source 102 of the IGFET 170V or 180V is the source 102 at the location of the maximum p-type well concentration. It typically decreases to approximately 1/15 when moving to. Typically 15 times is significantly lower than the typically corresponding 100 times that occurs in IGFET 170 or 180, but the vertical dopant profile for the p-type compensation dopant may be degraded. Similar to that described above for the vertical dopant profile through the drain 104 of the IGFET 100V, the concentration N _T of the total p-type dopant in the body-material portion below the source 102 of the IGFET 170V or 180V is Can be immediately reduced to at least 1/20, typically at least 1/40 when moving up to the source 102 at the location of the maximum concentration of the total p-type dopant in the well portion 116.

The hypobolite vertical dopant profile in the portion of the body material 108 below the source 102 of the IGFET 170V or 180V, although typically less than in the IGFET 170 or 180, is the source-body junction 110. Resulting in a decrease in parasitic capacitance associated with the As a result, the analog speed of each IGFET 170V or 180V is also increased.

In the same way that regions 192 and 198 (or 196) are provided in IGFETs 100V, 150V, 170V, and 180V, changes in IGFETs 140, 160, and 190 have an n- lower region 192 and a p- upper portion. Body-material remainder (or p-type upper body-material portion 196) may be provided. These asymmetric field n-channel changes of IGFETs 140, 160 and 190 are referred to below as IGFETs 140V, 160V and 190V, respectively.

Complementary IGFET Architecture for Mixed Signal Applications

Short-channel versions of long channel IGFETs 150, 160, 170, 180, 190, 100V, 150V, 160V, 170V, 180V and 190V can be manufactured according to the present invention by appropriately reducing the channel length. Similarly, p-channel IGFETs include IGFETs 100, 140, 150, 160, 170, which include short channel variations of IGFETs 150, 160, 170, 180, 190, 150V, 160V, 170V, 180V and 190V. 180, 190, 100V, 140V, 150V, 160V, 170V, 180V and 190V) can be produced according to the present invention by inverting the conductivity type of the semiconductor region.

N-channel IGFETs (100, 140, 150, 160, 170, 180, 190, 100V, including short channel variations of IGFETs (150, 160, 170, 180, 190, 150V, 160V, 170V, 180V and 190V) 140V, 150V, 160V, 170V, 180V and 190V) and p-channel IGFETs can be variously provided with the same semiconductor structure to make complementary-IGFET semiconductor architectures that are particularly suitable for high speed analog applications. For example, one or more n-channel IGFETs (100, 140, 150, 160, 170, 180, and 190) may be combined with one or more p-type variations of IGFETs (100V, 140V, 150V, 160V, 170V, 180V, and 190V). Can be combined. Then, the p- lower body-material portion 114 as the p-type equivalent of the n- lower portion 192 for each p-channel change of the IGFET (100V, 140V, 150V, 160V, 170V, 180V or 190V). The complementary-IGFET structure is created from a lightly doped p-type semiconductor material using < RTI ID = 0.0 > Alternatively, one or more n-channel IGFETs (100V, 140V, 150V, 160V, 170V, 180V and 190V) may be applied to each p-channel change of IGFETs 100, 140, 150, 160, 170, 180 or 190. One or more p-channel change IGFETs 100 fabricated from lightly doped n-type semiconductor material using the n- lower portion 192 as an n-type equivalent of the p- lower body-material portion 114 for the , 140, 150, 160, 170, 180, and 190).

In addition, IGFETs of both n-channel and p-channel that are particularly suitable for digital circuits can be provided in the semiconductor structure. Bipolar transistors npn and / or pnp may be provided in a variety of semiconductor structures. Thus, the resulting semiconductor architecture is suitable for mixed signal applications.

29.1 and 29.2 (collectively "FIG. 29") show two parts of a complementary-IGFET semiconductor structure constructed in accordance with the present invention that is particularly suited for mixed signal applications. The complementary-IGFET structure of FIG. 29 is produced from a doped monosilicon semiconductor body having a lower p-body-material portion 114. The patterned field region 200 of electrically insulating material, which is typically made of silicon oxide, is recessed to the upper surface of the semiconductor body to define a group of side separated active semiconductor islands. Four such islands 202, 204, 206 and 208 are shown in FIG. 29.

Four long channel IGFETs 210, 220, 230 and 240 are formed along the upper semiconductor surface at respective positions of islands 202, 204, 206 and 208. IGFETs 210 and 220 in FIG. 29.1 are asymmetric devices intended primarily for high speed analog applications. IGFETs 230 and 240 in FIG. 29.2 are symmetrical devices intended primarily for digital use. IGFETs 210 and 230 are n-channel devices. IGFETs 220 and 240 are p-channel devices.

Asymmetric n-channel IGFET 210 is an implementation of long n-channel IGFET 180 of FIG. 18B and includes all regions of IGFET 180. Thus, IGFET 210 has a hypobolite vertical dopant under drain 104 as substantially described above with respect to IGFET 180 and as described above with IGFET 100. Similarly, channel zone 106 of IGFET 210 is asymmetrically longitudinal dopant classified as described above for IGFET 180 and thus substantially as described above for IGFET 150.

Source 102, drain 104, and channel zone 106 of n-channel IGFET 210 are located in island 202. IGFET 210 includes a pair of electrically isolated sidewall spacers 250 and 252 located along the opposite transverse side of gate electrode 128 as well as the region shown in FIG. 18B. Metal silicon compound layers 254, 256, and 258 are located along the top of source 102, drain 104, and gate electrode 128, respectively.

Asymmetric p-channel IGFET 220 is an implementation of a p-channel version of long n-channel IGFET 180V in which n-lower portion 192 has been replaced with p-lower body-material portion 114. IGFET 220 is a p-type source 262 separated by n-type channel zone 266 of n-type body material 268 composed of heavily doped well portion 276 and upper portion 278. ) And p-type drain 264. As an implementation of the p-channel version of the n-channel IGFET 180V, the p-channel IGFET 220 is substantially the same as described above for the n-channel IGFET 180V and thus substantially the same as that described above for the n-channel IGFET 100V. And a hypobolite vertical dopant profile under drain 264, subject to the same conductivity type inversion. Similarly, the channel zone 106 of the p-channel IGFET 210 also has the same conductivity type inversion as described above for the n-channel IGFET 180V and thus the above described for the n-channel IGFET 150V. Subject to the condition, asymmetric longitudinal dopants are classified.

Source 262, drain 264, and channel zone 266 of p-channel IGFET 220 are located within island 204. Each p-type S / D zone 262 or 264 is very heavily doped main portion 262M or 264M to reduce source resistance R _S and drain side hot-carrier injection, and more lightly but still doped And heavily doped lateral extensions 262E or 264E. p + lateral extensions 262E and 264E terminate channel zone 266 along the upper semiconductor surface.

Highly doped pocket portion 280 of n-type upper body-material 278 extends along source 262, mainly source extension 262E. As with the pocket portion 120 of the IGFET 210, the n + pocket portion 280 extends deeper below the upper semiconductor surface than the p + source extension 262E but not as deep as the p ++ main source portion 262M. The remainder 284 of n-type upper body-part material 278 is lightly doped and extends along drain 264. N + well portion 276, n + pocket portion 280, and n− upper body-material remainder 284 in IGFET 220 are p + well portion 116 in IGFET 180V in an inverted conductivity type. , p + pocket portion 120, and p− upper body-material remainder 198, each having the same longitudinal and vertical doping characteristics. Thus, IGFET 220 avoids punchthrough and has reduced parasitic capacitance along the source- and drain-body pn junctions.

In the variation of IGFET 220 described above with respect to FIGS. 32A-32C and 33A-33F, n-top body-material remainder 284 is essentially an extension of n + well portion 276. Weak n-type doping of the n- upper body-material remainder 284 in this variation results in n + wells 276 to avoid individual dopant-introduction steps that cause the remainder 284 to be lightly doped n-type. It is prepared by upward diffusion of a portion of the n-type dopant used to form.

Gate dielectric layer 286 overlies channel zone 266 of IGFET 220. Gate electrode 288 is located on gate dielectric layer 286 over channel zone 266. Gate electrode 288 partially extends over each side S / D extension 262E or 264E. In the example of FIG. 29, gate electrode 288 is composed of very heavily doped p-type polysilicon. A pair of electrically isolated sidewall spacers 290 and 292 are positioned along opposite transverse sidewalls of p ++ gate electrode 288. Each of the metal silicon compound layers 294, 296, and 298 is positioned along the top of the source 262, the drain 264, and the gate electrode 288.

Symmetric n-channel IGFET 230 is a p-type channel zone of p-type body material 308 consisting of a lower p- portion 114, a heavily doped middle well portion 316, and an upper portion 318. It has a pair of n-type S / D zones 302 and 304 separated by 306. S / D zones 302 and 304 and channel zone 306 are located in island 206. Each n-type S / D zone 302 or 304 is heavily doped main portion 302M or 304M and highly doped but lightly doped to reduce drain side hot-carrier injection. Side extensions 302E or 304E. N + lateral extensions 302E and 304E terminate channel zone 306 along the upper semiconductor surface.

The pair of heavily doped halo pocket portions 320 and 322 of the p-type body-material portion 318 extend the S / D zones 302 and 304, respectively, in a symmetrical manner. P + halo pocket portions 320 and 322 extend primarily along n + S / D extensions 302E and 304E. In the example of FIG. 29, p + pocket portions 320 and 322 extend below the upper semiconductor surface deeper than n + extensions 302E and 304E but not deeper than n ++ main drain portions 302M and 304M. Item 324 is a suitably doped p-type remainder of upper body-material portion 318.

Gate dielectric layer 326 overlies channel zone 306. Gate electrode 328 is positioned over gate dielectric layer 326 over channel zone 306. Gate electrode 328 extends partially over each side S / D extension 302E or 304E. Gate electrode 328 is composed of very heavily doped n-type polysilicon in the example of FIG. 29. The pair of electrically isolated sidewall spacers 330 and 332 are located along opposite transverse sidewalls of the n ++ gate electrode 328. Metal silicon compound layers 334, 336, and 338 are located along the top of the S / D zones 302 and 304 and the gate electrode 328, respectively.

Given that it is formed over the p-bottom body-material portion 214, the symmetric p-channel IGFET 240 is a long channel device configured substantially the same as the IGFET 230 in which the conductivity type is inverted. Thus, IGFET 240 is a pair of p separated by n-type channel zone 346 of n-type body-material 348 consisting of heavily doped well portion 356 and top portion 358. -Type S / D zones 342 and 344. S / D zones 342 and 344 and channel zone 346 are located in island 208. Each p-type S / D zone 342 or 344 has a very heavily doped main portion 342M or 344M and side lightly doped but still heavily doped to reduce drain side hot-carrier injection. Section 342E or 344E. p + lateral extensions 342E and 344E terminate channel zone 346 along the upper semiconductor surface.

A pair of heavily doped halo pocket portions 360 and 362 of n-type upper body-material portion 358 extend along S / D zones 342 and 344, respectively, in a symmetrical manner. The n + halo pocket portions 360 and 362 extend mainly along the S / D extensions 342E and 344E, respectively. In the example of FIG. 29, n + pocket portions 360 and 362 extend below the upper semiconductor surface deeper than n + extensions 342E and 344E but not as deep as n ++ main S / D portions 342M and 344M. Item 364 is a suitably doped n-type remainder of upper body-material portion 358.

Gate dielectric layer 366 overlies channel zone 346. Gate electrode 368 is located on gate dielectric layer 366 over channel zone 266. Gate electrode 368 extends partially over each S / D extension 342E or 344E. In the example of FIG. 29, gate electrode 368 is composed of very heavily doped p-type polysilicon. A pair of electrically isolated sidewall spacers 370 and 372 are located along opposite transverse sidewalls of the p ++ gate electrode 368. Metal silicon compound layers 374, 376 and 378 are located along the top of S / D zones 342 and 344 and gate electrode 368, respectively.

Typically, gate dielectric layers 126, 286, 326, and 366 of IGFETs 210, 220, 230, and 240 are primarily composed of silicon oxide, but may be composed of silicon oxynitride or / and other high dielectric constant materials. . The thickness of the gate dielectric layers 126, 286, 326 and 346 is usually 2-8 nm, preferably 3-5 nm and typically 3.5 nm, for operation over the 1.8-V range. The dielectric layer thickness is appropriately increased to operate over the higher voltage range, or appropriately reduced to operate over the lower voltage range. Sidewall spacers 250, 252, 290, 292, 330, 332, 370, and 372 are shown in FIG. 29 in the form of right triangles with approximately convex hypotenuses, but may have other shapes. Silicon compound layers 254, 256, 258, 294, 296, 298, 334, 336, 338, 374, 376 and 378 are typically composed of cobalt silicon compounds.

Channel zones 306 and 346 of IGFETs 230 and 240 have a symmetric longitudinal dopant profile similar to the profile shown in FIG. 2 for symmetric IGFET 20 of FIG. The presence of p + halo pocket portions 320 and 322 in channel zone 306 mitigates threshold voltage roll-off and helps to avoid punchthrough in IGFET 230. Similarly, the presence of n + halo pocket portions 360 and 362 in channel zone 346 mitigates threshold voltage roll-off and helps to avoid punchthrough in IGFET 240.

The vertical dopant profile to the bottom p-type body material 308 through each of the n ++ main S / D portions 302M or 304M of IGFET 230 is similar to that shown in FIG. 3A for IGFET 20, It is also similar to the computer-simulated reference dopant profile shown in FIGS. 40, 44A, and 44B described below. This is based on the body material 348 forming a pn junction with the p-type portion 114 instead of being incorporated into the bottom lightly doped n-type monosilicon ( 348 and to the vertical dopant profile through each p ++ main S / D portion 342M or 344M of IGFET 240. The appropriate but elevated concentration of the p-type dopant in the upper body-material portion 324 of the IGFET 230 is provided by the halo pocket portions 320 and 322 to suppress punchthrough occurring in the IGFET 230. Cooperate with severe p-type dopant concentrations. Similarly, corresponding doping in upper body-material portion 364 and halo pocket portions 360 and 362 of IGFET 240 enables the avoidance of punchthrough.

Typically, the p-type well dopant defining well portion 316 of IGFET 230 has a maximum concentration of p-type well dopant defining well portion 116 of IGFET 210 below the upper semiconductor surface. The maximum concentration is reached at a depth approximately equal to the depth in which it is located. Typically, because the concentration N _I of the p-type background dopant is relatively uniform, the maximum concentration of the entire p-type dopant in the well portion 316 of IGFET 230 is the well portion of IGFET 210 below the top surface. Occurs at 116 approximately the same depth as the maximum concentration of the total p-type dopant. Upper body-material portion 318 of IGFET 230 is provided with a p-type anti-punchthrough (“APT”) dopant to raise the upper portion 318 to an appropriate p-type doping level. The p-type APT dopant in the upper body-material portion 318 reaches a maximum concentration at a depth below the depth at which the maximum concentration of the p-type well dopant of the well portion 316 is below the upper semiconductor surface.

The combination of full p-type dopant (i.e. p-type well), APT, and background dopant in the portion of body material 308 below the n ++ main S / D portion 302M or 304M, in its body-material portion Results in a concentration N _T of the total p-type dopant of to be relatively flat along a vertical line extending from the subsurface location of the maximum p-type dopant concentration in the well 316 to the main S / D portion 302M or 304M. . In particular, the concentration N _T of the total p-type dopant in the portion of the body material 308 underneath the main S / D portion 302M or 302M is determined from the position (from the position of the maximum p-type dopant concentration in the well 316). When moving up to 302M or 304M), it generally changes (decreases) to less than 1/10, typically less than 1/5.

The same happens with the IGFET 240. Typically, an n-type well dopant that defines well portion 356 of IGFET 240 is at a maximum concentration of n-type well dopant that defines well portion 276 of IGFET 220 below the upper semiconductor surface. The maximum concentration is reached at approximately the same depth. Thus, the maximum concentration of the total n-type dopant in the well portion 356 of the IGFET 240 is at a depth below the upper surface with the maximum concentration of the total n-type dopant in the well portion 276 of the IGFET 220. Occurs at approximately the same depth as. Upper body-material portion 358 of IGFET 240 is provided with an n-type APT dopant to raise the upper portion 358 to an appropriate n-type doping level. The n-type APT dopant in the upper body-material portion 358 reaches a maximum concentration at a depth below the depth at which the maximum concentration of the n-type well dopant of the well portion 356 is below the upper semiconductor surface.

The combination of the full n-type dopant (i.e. predominantly n-type well) and the APT dopant in the portion of the body material 348 under the S / D zone 342M or 344M is the total n- in the body-material portion. The concentration NT of the type dopant causes it to be relatively flat along a vertical line extending from the subsurface location of the maximum n-type dopant concentration in the well portion 356 to the main S / D portion 342M or 344M. Specifically, the concentration N _T of the total n-type dopant in the portion of the body material 348 below the main S / D portion 342M or 344M is determined from the position (from the location of the maximum n-type dopant concentration in the well 356). 342M or 344) generally vary by less than 10 times, typically less than 5 times.

Of course, the halo pocket portions 120 and 280 of each IGFET 210 and 220 may alternatively extend deeper than the respective sources 102 and 262 below the upper semiconductor surface. IGFET 210 implements IGFET 150 of FIG. 13, and IGFET 220 implements a p-channel version of IGFET 150V. Halo pocket portions 320 and 322 of n-channel IGFET 230 have an S / D zone 300, as occurs in computer-simulated reference short channel structure B of FIG. 40 below the upper semiconductor surface. And 302). Similarly, halo pocket portions 360 and 362 of p-channel IGFET 240 may extend deeper than S / D zones 340 and 342 below the upper semiconductor surface.

30.1 and 30.2 (collectively "FIG. 30") show two parts of another complementary IGFET semiconductor structure constructed in accordance with the present invention that is particularly suited for mixed signal applications. The complementary-IGFET structure of FIG. 30 has a p- lower portion 114 and a p + well portion for isolating the p + well portion 116 and the p-type upper body-material portion 118 from the p- portion 114. Asymmetric p-channel IGFET 220 and asymmetrical field n− configured identically to asymmetric n-channel IGFET 210 except that a heavily doped n-type subsurface layer 382 lies between 116. Channel IGFET 380. As a result, the p-type body material 108 for the IGFET 380 does not include the p- lower portion 114 but instead into the p + well portion 116 and the p-type upper body-material portion 118. It is composed only.

The complementary-IGFET structure of FIG. 30 has the p + well portion 316 and p- lower portion from the p- portion 114 to separate the p + well portion 316 and the p-type upper body-material portion 318. Symmetric p-channel IGFET 240 and symmetric field n− configured identically to symmetric n-channel IGFET 230 except that a heavily doped n-type subsurface layer 392 lies between 114. Channel IGFET 390 is further included. Thus, p-type body material 308 for IGFET 390 does not include p- lower portion 114, but instead p + well portion 316 and p-type upper body-material portion 318. It consists only of. Except for n + subsurface layers 382 and 392, n-channel IGFETs 380 and 390 operate the same as in n-channel IGFETs 210 and 230, respectively.

Circuit elements other than IGFETs 210, 220, 230, 240, 380 and 390 may be provided in other portions (not shown) of the complementary-IGFET structure of FIG. 29 or 30. For example, short channel versions of IGFETs 210, 220, 230, 240, 380 and 390 may be present in either complementary-IGFET structure. Bipolar transistors along with various types of resistors, capacitors, and / or inductors may be provided in the complementary-IGFET structure of FIG. 29 or 30. Based on the features of the additional circuit elements, appropriate electrical isolation is provided in either complementary-IGFET structure for the additional elements. Of course, in some pure analog complementary-IGFET applications IGFETs 240 and 230 or 390 can be eliminated.

Fabrication of Complementary-IGFET Structures for Mixed Signal Applications

31A to 31O, 31P.1 to 31R.1, and 31P.2 to 31R.2 (collectively "FIG. 31") show the long channel IGFET 210, as generally shown in FIG. A semiconductor process in accordance with the present invention for producing a complementary-IGFET semiconductor structure including 220, 230, and 240 is shown. The steps involved in the manufacture of IGFETs 210, 220, 230 and 240 through stages just prior to the creation of gate sidewall spacers 250, 252, 290, 292, 330, 332, 370 and 372 are described in FIGS. 31A-31O. Shown. 31P.1-31R.1 illustrate the fabrication of spacers 250, 252, 290 and 292 and subsequent steps leading to IGFETs 210 and 220 as shown in FIG. 29.1. 31P.2 through 31R.2 illustrate the fabrication of spacers 330, 332, 370 and 372 and subsequent steps leading to IGFETs 230 and 240 shown in FIG. 29.2.

Short-channel versions of IGFETs 210, 220, 230, and 240 may be manufactured simultaneously, depending on the manufacturing steps used to manufacture long-channel IGFETs 210, 220, 230, and 240. The short channel IGFET is shorter in channel length than the long channel IGFETs 210, 220, 230 and 240, but otherwise has the same intermediate IGFET appearance as shown in FIG. Simultaneous fabrication of the long channel IGFETs 210, 220, 230 and 240 and their short channel versions is realized by masking plates (reticles) with patterns for both long channel and short channel IGFETs.

Except for the pocket ion implantation step (including the halo pocket) and the source / drain extension ion implantation step, all ion implantation steps in the present fabrication process are performed approximately perpendicular to the lower semiconductor surface and thus approximately perpendicular to the upper semiconductor surface. More specifically, all implantation steps except the pocket and source / drain extension ion implantation steps are performed at a small angle to the vertical, typically at 7 °. This slight deviation from normal is used to avoid undesirable ion channeling effects. For simplicity, some deviation from vertical is not shown in FIG. 31.

Unless otherwise indicated, the species of n-type dopant utilized in each of the n-type ion implantation methods in the fabrication process of FIG. 31 consist of a specific n-type dopant in basic form. That is, each n-type ion implantation is performed through the ions of a particular n-type dopant element rather than through the ions of a chemical compound containing the n-type dopant. The species of p-type dopant employed in each of the p-type ion implantation methods consist of the p-type dopant, usually boron, in basic or compound form. Thus, each p-type ion implantation is usually performed through ions of boron-containing compounds such as boron difluoride or through boron ions.

In some of the fabrication steps in FIG. 31, the openings extend (substantially) through the photoresist mask on top of the active semiconductor region for the two IGFETs. When two IGFETs are formed laterally adjacent to each other in the exemplary cross-sectional view of FIG. 31, the two photoresist openings are described as a single opening in FIG. 31 although they may be described below as separate openings.

The letter “P” at the end of the reference symbol appearing in the figure of FIG. 31 represents the precursor to the region shown in FIG. 29 and identified by the portion of the reference symbol prefix “P”. The letter “P” is derived from the reference symbol in the drawing of FIG. 31 when the precursor is sufficiently derived to mainly constitute the corresponding region in FIG. 29.

The starting point for the fabrication process of FIG. 31 is a monosilicon semiconductor body typically composed of a heavily doped p-type substrate 400 and a lightly doped p-type epitaxial layer 114P overlying it. See FIG. 31A. The p + substrate 400 is a semiconductor wafer formed of <100> monosilicon doped with boron to a concentration of approximately 5 x 10 ¹⁸ atoms / cm 3 to achieve a typical resistance of 0.015 Ω-cm. For simplicity, the substrate 400 is not shown in the remainder of FIG. 31. Alternatively, the starting point may simply be a lightly doped p-type subsubstrate substantially the same as the p- epitaxial layer 114P.

The epitaxial layer 114P is a p-type epitaxially grown monosilicon doped with low concentration of boron to a concentration of approximately 5x10 ¹⁵ atoms / cm3 to achieve a typical resistance of 5Ω-cm. It is composed. The thickness of epitaxial layer 114P is typically 5.5 μm. If the starting point for the fabrication process of FIG. 31 is a lightly doped p-type substrate, item 114P is a p-substrate.

Field-isolation region 200 proceeds from left to right in FIG. 31B and is shown in FIG. 31B to define active semiconductor islands 202, 204, 206 and 208 for each of IGFETs 210, 220, 230 and 240. As provided along the top surface of the p-epitaxial layer (or p-substrate). Field isolation 200 is preferably produced in accordance with trench-oxide techniques, but may be produced in accordance with local-oxidation techniques. In providing the field isolation region 200, a thin screen isolation layer 402 of silicon oxide is thermally grown along the top surface of the epitaxial layer 114P.

A photoresist mask 404 having openings over islands 202 and 206 is formed on screen oxide layer 402 as shown in FIG. 31C. A p-type well dopant composed of boron species may be used to define a screen oxide 402 to define (a) p + well portion 116 for IGFET 210 and (b) p + precursor well portion 316P for IGFET 230. Ion is implanted with heavy energy and high energy into the bottom monosilicon through an uncovered section. The portion of epitaxial layer 114P over well portion 116 constitutes p- precursor upper portion-material portion 118P for IGFET 210. Photoresist 404 is removed.

A photoresist mask 406 having an opening over island 206 is formed on screen oxide 402. See FIG. 31D. A p-type APT dopant composed of boron species is routed to the appropriate monosilicon as the bottom monosilicon through an uncovered section of screen oxide 402 to define p precursor upper body-material portion 324P for IGFET 230. Ion implantation. Photoresist 406 is removed.

A photoresist mask 408 with openings over islands 204 and 208 is formed on screen oxide 402 as shown in FIG. 31E. An n-type well dopant composed of phosphorus or arsenic is screen oxide to define (a) n + well portion 276 for IGFET 220 and (b) n + precursor well portion 356P for IGFET 240. The uncovered section of 402 is ion implanted into the bottom monosilicon with high energy and a high concentration.

Through an in place photoresist mask 408, an n-type compensation dopant, likewise composed of phosphorus or arsenic, may be used to define an n-precursor upper body-material portion 278P for IGFET 220. 204 is implanted with low energy and appropriate energy into the bottom monosilicon through the uncovered section of oxide 402 on top. As is present at the stage of FIG. 31E, n-precursor body-material portion 278P overlies n + well portion 276. Usually, the illustration and implant energy of n-type compensation dopant implantation is sufficient to make both precursor body-material portion 278P n-conductive.

The n-type compensation dopant also passes through the uncovered section of oxide 402 on top of island 208 to IGFET 240 to the bottom monosilicon. One of two n-type doping operations using photoresist 408 may be performed first. Photoresist 408 is removed. If it is desired that monosilicon for IGFET 240 not accept any n-type compensation dopant, then the n-type compensation doping operation is not on top of island 208 after the additional photoresist is removed (also islands 202 and 206). It can be done through an additional photoresist mask having an opening on top of island 204 (not on top).

During subsequent fabrication steps, several n-type well dopants used to define the precursor n + well portion 276 for the IGFET 220 may be n + well portion 276 as present at this point in the fabrication process of FIG. 29. Diffusion upwards to the upper semiconductor material. That is, a portion of the n-type well dopant then diffuses upward into the material overlying the initially doped weak p-type island 204. Upward diffusion of the portion of the n-type well dopant that defines n + well portion 276 occurs mainly during subsequent manufacturing steps performed at elevated temperature (ie, significantly above room temperature).

The n + well portion 276 depends on various factors, mainly depending on (a) the time at the elevated temperature of subsequent manufacturing steps and (b) the temperature parameters indicative of increased diffusion of dopant through this elevated temperature. The up-diffused portion of the n-type well dopant defining s may be distributed through island 204 to IGFET 220 in this manner to dedope all of the current p-type dopant in island 204. Thus, ignoring any other dopants subsequently introduced into island 204, the up-diffused portion of the n-type well dopant may cause all of islands 204 to convert to n-conducting. In this case, the step of injecting the n-type compensation dopant can often be omitted to simplify the manufacturing process and reduce the manufacturing cost. 32A-32C and 33A-33C described below show two variations of the manufacturing process of FIG. 31 in which the implantation step of the n-type compensation dopant is omitted.

During subsequent fabrication, some n-type well dopants used to define n + well portion 356P for IGFET 240 may be n + well portion 356P, as is present in the fabrication process of FIG. 29 at this point. Diffusion upwards to the upper semiconductor material. However, as described below, all islands 208 for IGFET 240 have an n-type conductivity at the end of ion implantation of the n-type APT dopant introduced into island 208, and associated activation. As a result, the decision to retain or omit the n-type compensation implant is determined by the conditions of subsequent fabrication steps applied to island 204 for IGFET 220.

A photoresist mask 410 having an opening over island 208 is formed on screen oxide 402. See FIG. 31F. An n-type APT dopant composed of phosphorus or arsenic is passed to the bottom monosilicon through an uncovered section of screen oxide 402 to define the n precursor upper body-material portion 358P for IGFET 240. Ion implanted into a suitable city. Photoresist 410 is removed.

Thermal annealing such as rapid thermal anneal ("RTA") may be performed on the resulting semiconductor structure to effectively locate the implanted p- and n-type dopants in a more stable state and to repair lattice damage. have. See FIG. 31G. The upper semiconductor surface is cleaned. Gate-dielectric-containing dielectric layer 412 is provided along the upper semiconductor surface as shown in FIG. 31H. Dielectric layer 412 is produced by thermal growth techniques.

Precursor gate electrodes 128P, 288P, 328P, and 368P are formed on the gate-dielectric-containing-dielectric layer 412 on top of each of the segments of the upper body-material portions 118P, 278P, 318P, and 358P. . See FIG. 31I. Precursor gate electrodes 128P, 288P, 328P, and 368P are created by depositing a layer of generally undoped (original) polysilicon on dielectric layer 412 and then patterning the polysilicon layer. Portions of dielectric layer 412 at the bottom of precursor gate electrodes 128P, 288P, 328P, and 368P make up gate dielectric layers 126, 286, 326, and 366, respectively. In general, the gate dielectric material formed from the gate dielectric layers 126, 286, 326, and 366 generally includes gate electrode 128P, from body-material segments intended to be channel zones 106, 266, 306, and 346, respectively. 288P, 328P and 368P), respectively.

Dielectric sealing layer 414 grows thermally along the exposed surfaces of precursor gate electrodes 128P, 288P, 328P, and 368P. Reference is again made to FIG. 31I. Between forming the dielectric seal layer 414, the portion of the dielectric layer 412 located on the side of the gate dielectric layers 126, 286, 326, and 366 is slightly thickened to become the composite surface dielectric layer 416. .

In general, photoresist mask 418 having an opening above the intended location for p + pocket portion 120 of IGFET 210 is formed on dielectric layers 414 and 416. See FIG. 31J. Photoresist mask 418 is precisely arranged relative to precursor gate electrode 128P. The p-type pocket dopant composed of boron species is shown obliquely into the bottom monosilicon through the uncovered portion of the surface oxide layer 416 to define the p + precursor pocket portion 120P for the IGFET 210. It is ion implanted. Usually, p-type pocket implantation is performed at two opposite angles of inclination with respect to the vertical. Alternatively, p-type pocket implantation can be performed at a single tilt angle. Photoresist 418 is removed.

Photoresist mask 420 having an opening generally above the intended location for n + pocket portion 280 of IGFET 220 is formed on dielectric layers 414 and 416. See FIG. 31K. Photoresist mask 420 is precisely arranged relative to precursor gate electrode 288P. An n-type pocket dopant composed of phosphorus or arsenic is shown obliquely in the bottom monosilicon through the uncovered portion of the surface dielectric 416 to define the n + precursor pocket portion 280P for the IGFET 220. It is ion implanted. N-type pocket implantation is usually performed at two opposite tilt angles but can be done at a single tilt angle. Photoresist 420 is removed.

Photoresist mask 422 having openings over islands 202 and 206 is formed on dielectric layers 414 and 416 as shown in FIG. 31L. The n-type source / drain extension dopant, consisting of arsenic or phosphorus, includes (a) n + precursor source extension 102EP for IGFET 210, (b) a separate n + precursor drain extension for IGFET 210 ( 104EP), and (c) the lower mono through the uncovered section of the surface dielectric 416 to define a pair of lateral separated n + precursor S / D extensions 302EP and 304EP for IGFET 230. The silicon is ion implanted in heavy dosage form. Photoresist 422 is removed.

Photoresist mask 424 with openings over island 206 is formed on dielectric layers 414 and 416. See FIG. 31M. The p-type halo dopant composed of boron species is an uncovered section of the surface dielectric 416 to define a pair of laterally separated p-type precursor halo pocket portions 320P and 322P for IGFET 230. It is implanted into the bottom monosilicon into the obliquely dark city. Photoresist 424 is removed.

Photoresist mask 426 with openings over islands 204 and 208 is formed on dielectric layers 414 and 416 as shown in FIG. 31N. The p-type source / drain extension dopant, consisting of boron species, comprises (a) p + precursor source extension 262EP for IGFET 220, (b) individual p + precursor drain extension 264EP for IGFET 220. ), And (c) bottom monosilicon through an uncovered section of surface oxide 416 to define a pair of lateral separated p + S / D extensions 342EP and 344EP for IGFET 240. Ion implanted into the dark city. Photoresist 426 is removed.

Photoresist mask 428 having an opening over island 208 is formed on dielectric layers 414 and 416. See FIG. 31O. The s n + halo dopant, composed of phosphorus or arsenic, passes through the uncovered section of the surface dielectric 416 to define a pair of laterally separated n + precursor halo pocket portions 360P and 362P for IGFET 240. Into the bottom monosilicon is implanted into the obliquely dark city. Photoresist 428 is removed.

At this point, a low-temperature furnace anneal may be performed to eliminate the defects caused by the dark illustration of source / drain extension implantation.

In the remainder of the process of FIG. 31, the complementary-IGFET structure at each processing stage is a pair of FIGS. 31z.1 and 31z.2, where “z” is “p” to “r”. It is shown through the letters). Each of FIG. 31z.1 shows processing performed to generate asymmetric IGFETs 210 and 220, and each of FIG. 31z.2 shows processing performed simultaneously to produce symmetric IGFETs 230 and 240. . Each pair of FIGS. 31z.1 and 31z.2 are collectively referred to as "FIG. 31z" (where "z" varies from "p" to "r") for convenience. For example, FIGS. 31P.1 and 31P.2 are collectively referred to as "FIG. 31P".

Gate sidewall spacers 250, 252, 290, 292, 330, 332, 370, and 372 are formed along the transverse sidewalls of precursor gate electrodes 128P, 288P, 328P, and 368P, as shown in Figure 31P. The formation of (250, 252, 290, 292, 330, 332, 370, and 372) is performed primarily by anisotropic etching performed generally perpendicular to the upper semiconductor surface by depositing and then removing the dielectric material on top of the structure. Wherein the dielectric material is not intended to constitute spacers 250, 252, 290, 292, 330, 332, 370 and 372. In addition, portions of the dielectric layers 414 and 416 are partially removed rather than entirely. Items 430 and 432 in FIG. 31P represent the remainder of dielectric layers 414 and 416, respectively, which are not covered by spacers 250, 252, 290, 292, 330, 332, 370, and 372.

Photoresist mask 434 having openings over islands 202 and 206 is formed on dielectric layers 430 and 432 and spacers 290, 292, 370 and 372. See FIG. 31Q. The n-type main source / drain dopant, consisting of arsenic or antimony, comprises (a) n ++ main source portion 102M for IGFET 210 and n ++ main drain portion 104M and (b) n ++ for IGFET 230. In order to define the main S / D portions 302M and 304M, ions are implanted into the lower monosilicon in a very dark manner through the uncovered section of the surface dielectric layer 432. In addition, an n-type main source / drain dopant flows into the precursor electrodes 128P and 328P and converts them to n ++ gate electrodes 128 and 328, respectively. Photoresist 434 is removed.

Portions of regions 102EP, 104EP, and 120P outside of main S / D portions 102M and 104M are n + source extensions 102E, n + drain extensions 104E, and p + pocket portions for IGFET 210, respectively. Configure 120. The p- upper body-material remainder 124 is the precursor upper body-material portion 118P, the remaining lightly doped material of the current p-type upper body-material portion 118. The portions of the regions 302EP, 304EP, 320P, and 322P outside the main S / D portions 302M and 304M are currently the n + S / D extensions 302E and 304E and p + halo pocket portions for the IGFET 230, respectively. And 320 and 322. The p- upper body-material remainder 324 is the precursor upper body-material portion 318P, the remaining lightly doped p-type material of the current p-type upper body-material portion 318.

When the main source / drain dopant is composed of arsenic, thermal annealing may be performed to restore lattice damage, activate the main n-type source / drain dopant, and diffuse it out. This annealing (usually RTA) activates the pocket and / source / drain extension dopant.

Photoresist mask 436 having openings over islands 204 and 208 is formed on dielectric layers 430 and 432 and spacers 250, 252, 330 and 332 as shown in FIG. 31R. The p-type main source / drain dopant, consisting of boron species, includes (a) p ++ main source portion 262M for IGFET 220 and p ++ main drain portion 264M and (b) p ++ main for IGFET 240 To define the S / D portions 342M and 344M, ions are implanted into the dark monosilicon into the bottom monosilicon through an uncovered section of the surface dielectric 432. In addition, the p-type main source / drain dopant enters the precursor electrodes 288P and 368P which are then converted to p ++ gate electrodes 288 and 368, respectively. Photoresist 436 is removed.

The portions of the regions 262EP, 264EP, and 280P outside the main S / D portions 262M and 264M are currently p + source extensions 262E, p + drain extensions 264E, and n + for the IGFET 220, respectively. Configure pocket portion 280. The n- upper body-material remainder 284 is an n- upper body-material portion 278P, the remaining lightly doped n-type material of the current n-type upper body-material portion 278. The portions of the regions 342EP, 344EP, 360P and 362P outside the main S / D portions 342M and 344M are currently present in p + S / D extensions 342E and 344E and n + halo pocket portions for IGFET 240, respectively. 360 and 362). The n- upper body-material remainder 364 is an n-precursor upper body-material portion 358P, the remaining lightly doped n-type material of the current n-type upper body-material portion 358.

A capping layer (not shown) of dielectric material (usually silicon oxide) is formed on top of the structure. The semiconductor structure is then thermally annealed to repair the lattice damage and to activate the implanted main p-type source / drain dopant. If no previous annealing has been performed to activate the main n-type source / drain dopant, this final annealing activates the pocket dopant and all source / drain dopants. Typically, the final annealing is RTA.

A thin layer of dielectric material, including dielectric layers 430 and 432, is removed along the top surface of gate electrodes 128, 288, 328, and 368 and along the top semiconductor surface. The metal silicon compound layers 254, 256, 258, 294, 296, 298, 334, 336, 338, 374, 376 and 378 have regions 102M, 104M, 128, 262M, 264M, 288, 302M, 304M, 328, 342M, 344M and 368 respectively along the top surface. This typically involves depositing a thin layer of a suitable metal (usually cobalt) on the top surface of the structure and performing a low temperature step to react the metal with the bottom silicon. Unreacted metal is removed. The second low temperature step completes the reaction of the metal with the bottom silicon to thereby form the metal compound layers 254, 256, 258, 294, 296, 298, 334, 336, 338, 374, 376 and 378. To be performed. Metal silicon compound formation completes the basic fabrication of IGFETs 210, 220, 230 and 240. The resulting complementary-IGFET structure appears as shown in FIG.

In general, the p-type wells, p-type APT, n-type wells, n-type compensation, and n-type APT implantation of FIGS. 31C-31F may be performed in any order. In general, the p-type pocket, n-type pocket, n-type source / drain extension, p-type halo, p-type source / drain extension, and n-type halo implant of FIGS. 31J-31O are in any order. It can be carried out as. The n-type main source / drain implant of FIG. 31Q is usually performed before the p-type main source / drain implant of FIG. 31R, especially when the main source / drain dopant is composed of arsenic. However, p-type main source / drain implantation can often be performed prior to n-type main source / drain implantation.

The inclination angles for the p-type pockets, n-type pockets, p-type halo, and n-type halo implants of FIGS. 31J, 31K, and 31M and 31O are usually at least 15 °. Although typically changing from one of the inclined injection methods to another, the inclination angle for each inclined injection method is typically 25 to 45 °.

Typically, the complementary-IGFET structure of FIG. 30 is complementary-IGFET of FIG. 29 except for the n + isolation layers 382 and 392 that convert IGFETs 210 and 230 into IGFETs 380 and 390, respectively. Manufactured according to substantially the same steps as the structure. Usually, isolation layers 382 and 392 are formed between the stages of FIGS. 31B and 31C using additional photoresist masks having openings over islands 202 and 206. In addition, additional photoresist masks have openings used to create heavily doped n-type regions that connect isolation layers 382 and 392 to the upper semiconductor surface to accommodate the appropriate isolation voltage. An insulating dopant consisting of arsenic or phosphorus includes (a) n + isolation layers 382 and 392 and (b) n + isolation layer interconnection regions for IGFETs 380 and 390, respectively. It is ion implanted into the bottom of the monosilicon through the uncovered section into dark city.

The fabrication process of FIG. 31 shows that from the implementation of IGFET 180, n-type S / D zones 102 and 104 each have an n + lower S / D underneath each of n ++ main S / D portions 102M and 104M. The implementation of the asymmetric n-channel IGFET 190 of FIG. 18C, which further includes portions 102L and 104L, respectively, can be modified as described below to change the asymmetric n-channel IGFET 210. Since the lower S / D portions 102L and 104L are n-type doped at a lower concentration than the main S / D portions 102M and 104M, the lower S / D portions 102L and 104L are the IGFET 160 of FIG. 15. A source / drain vertical dopant classification is provided to further reduce source / drain parasitic capacitance as described above.

This process variant starts at the stage of FIG. 31Q through photoresist mask 434 throughout which is used to ion implant the n ++ main S / D portions 102M and 104M. The n-type bottom source / drain dopant, consisting of phosphorus or arsenic, is routed to the bottom monosilicon through the uncovered section of the surface dielectric layer 432 to define the n + bottom S / D portions 102L and 104L. Ion implanted into the dark city. Injection of the n + lower S / D portions 102L and 104L may be performed before or after injection of the n ++ main S / D portions 102M and 104M.

The implantation energy for the n-type main and bottom source / drain dopants is selected such that the n-type bottom source / drain dopant is in a larger implant range than the n-type main source / drain dopant. Since both the n-type main source / drain implant and the n-type bottom source / drain implant are performed solely through the surface dielectric layer 432, the n-type bottom source / drain dopant is n-type below the upper semiconductor surface. It is injected to an average depth deeper than the main source / drain dopant. Through the n-type main source / drain dopant, which is injected into a larger city than the n-type bottom source / drain dopant, so that the n + lower S / D portions 102L and 104L are n ++ main S / D portions ( Doped lighter than 102M and 104M, and extend deeper below the upper semiconductor surface than the n ++ main S / D portions 102M and 104M.

Symmetric n-channel IGFET 230 further includes a pair of lower S / D portions each of n-type S / D zones 302 and 304 doped at a lower concentration than main S / D portions 302M and 304M. Is simultaneously converted into a change. Compared to the n + lower S / D portions 102L and 104L for the aforementioned changes in IGFET 210, the lower S / D portions for changes in IGFET 230 are heavily doped n-type. If IGFET 230 is not converted to change with n + lower S / D portion, implantation of n-type lower source / drain dopant is not island 206 (or after additional photoresist mask is removed) (or This may be done through an additional photoresist mask having an opening on top of island 202 (rather than islands 204 and 208).

Similarly, the fabrication process of FIG. 31 is based on the implementation of n-channel IGFET 180 as long as p-type S / D zones 262 and 264 are underneath each of p ++ main S / D portions 262M and 264M. Implementation of the p-channel version of the n-channel IGFET 190V further comprising a pair of heavily doped p-type lower S / D portions may be modified to change the asymmetric p-channel IGFET 220. The p + lower S / D portion provides source / drain vertical dopant classification to further reduce source / drain parasitic capacitance similar to that described above for IGFET 160 in FIG.

This additional process variant begins at the stage of FIG. 31R through photoresist mask 436 throughout which was used to ion implant p ++ main S / D portions 262M and 264M. The p-type bottom source / drain dopant, consisting of boron species, is bottomed through an uncovered section of surface dielectric layer 432 to define two p + bottom S / D portions for changes in IGFET 220. Monosilicon is implanted into the dark city. The n + bottom source / drain injection may be performed before or after the n ++ main source / drain injection. In one example, the p-type bottom source / drain dopant is composed of elemental boron and the p-type main source / drain dopant is composed of boron difluoride.

The implantation energy for the p-type main and bottom source / drain dopants is selected such that the p-type bottom source / drain dopant has a larger implant range than the p-type main source / drain dopant. Both the p-type main source / drain implant and the p-type lower source / drain implant are performed through the surface dielectric layer 432 only, and the p-type lower source / drain dopant is transferred to the p-type main source / drain below the upper semiconductor surface. It is injected to a greater average depth than the drain dopant. Since the p-type main source / drain dopant is implanted into a very dark town and therefore larger than the p-type bottom source / drain dopant, the p + lower S / D portion of the change in IGFET 220 is defined as p ++ main S /. More lightly doped than the D portions 262M and 264M and extending deeper below the upper semiconductor surface than the p ++ main S / D portions 262M and 264M.

Symmetric p-channel IGFET 240 further includes a pair of lower S / D portions each of which are p-type S / D zones 362 and 364 doped lightly than the main S / D portions 362M and 3604M. Is simultaneously converted into a change. Compared to the p + lower S / D portion for the change of IGFET 220, the lower S / D portion for the change of IGFET 240 is heavily doped p-type. If IGFET 240 does not change with a change with a lower S / D portion, then the p-type lower source / drain implant is not island 208 (also, islands 202 and 202) after additional photoresist is removed. May be performed through an additional photoresist mask having an opening over island 204 (rather than 206).

Manufacturing process variations to avoid n-type compensation injection

32A-32C (collectively “FIG. 32”) show an alternative to the step of FIG. 31E for fabricating changes in the complementary-IGFET semiconductor structure of FIG. 29, in accordance with the present invention. The use of n-type compensation injection in island 204 (and island 208) is avoided in the manufacturing process of FIG. 31 as modified to incorporate the alternative of FIG. 32. As a result, the complementary-IGFET structure produced by utilizing the alternative of FIG. 32 includes a variation 220V of the asymmetric p-channel IGFET 220.

The process alternative of FIG. 32 begins with the structure of FIG. 31D repeated here as FIG. 32A. The n-type well doping step described above in connection with FIG. 31E is performed in the structure of FIG. 32A. In particular, photoresist mask 408 is formed on screen oxide 402 as shown in FIG. 32B. Photoresist mask 408 has openings over islands 204 and 208. The n-type well dopant consisting of phosphorus or arsenic is (a) precursor n + well portion 276P for asymmetric p-channel IGFET 220V and (b) n + precursor well portion for symmetric p-channel IGFET 240 In order to define 356P, ions are implanted at high concentrations and with high energy into the bottom monosilicon through an uncovered section of screen oxide 402.

No n-type compensation implantation into island 204 (and island 208) with photoresist 408 everywhere is not performed at this point. Instead, the photoresist 408 is simply removed. After removal of photoresist 408, lightly doped p-type portion 278Q of island 204 for IGFET 220V is present above precursor n + well portion 276P. Similarly, lightly doped p-type portion 358Q of island 208 for IGFET 240 resides above precursor n + well portion 356P.

The alternative of FIG. 32 continues through the formation of photoresist mask 410 on screen oxide 402. See FIG. 32C. Again, photoresist 410 has an opening over island 208. An n-type APT dopant composed of phosphorus or arsenic is passed to the bottom monosilicon through an uncovered section of screen oxide 402 to define the n precursor upper body-material portion 358P for IGFET 240. Ion implanted into a suitable city. Photoresist 410 is removed.

The n-type APT dopant implanted into island 208 converts all of the p- portion 358Q of island 208 to n-conducting. As a result, the p- portion 358Q disappears. Since trace amounts of n-type APT dopant do not enter island 204, the p- portion 278Q of island 204 is still present at this stage of manufacture.

The structure of FIG. 32C is further processed in accordance with the fabrication steps described above in connection with FIGS. 31G-31O, 31P.1-31R.1 and 31P.2-31R.2, including associated annealing operations. Some of these additional steps are performed at elevated temperatures (significantly higher than room temperature). During the elevated temperature step, the portion of the n-type well dopant used to define the precursor n + well portion 276P for IGFET 220V diffuses upward into p- portion 278Q. The up-diffused portion of the n-type well dopant is such that the p- material that has not undergone p-type and / or n-type doping following the n-type well doping step is n-conductive at the end of manufacture. Distributed through island 204 to be transformed. The p-part 278Q employs the alternative of FIG. 32, subject to the n-type doped slightly below the n-type upper body-material portion 278 fabricated in accordance with the basic manufacturing process of FIG. N-type upper body-material portion 278 of a complementary-IGFET structure fabricated following the fabrication process of FIG. The n-top body-material remainder 284 is the remaining lightly doped n-type material of the n-type body-material portion 278.

Complementary-IGFET structures fabricated according to the fabrication process of FIG. 31, such as modified to incorporate the alternative of FIG. 32, appear roughly as shown in FIG. A generalized version of the asymmetric p-channel IGFET 220V fabricated by utilizing the alternative of FIG. 32 is shown in FIG. 34 described below.

33A-33F (collectively “FIG. 33”) illustrate an alternative to the steps of FIGS. 31C-31F for fabricating changes in the complementary-IGFET semiconductor structure of FIG. 29, in accordance with the present invention. . As in the alternative of FIG. 32, the alternative of FIG. 33 avoids the use of n-type compensation injection into island 204 (and island 208). As a result, the complementary-IGFET structure fabricated according to the process of FIG. 31 as modified to incorporate FIG. 33 includes an asymmetric p-channel IGFET 220V instead of IGFET 220.

The process alternative of FIG. 33 begins through the structure of FIG. 31B repeated as FIG. 33A. Screen oxide layer 402 is formed along the top surface of epitaxial layer 114P at the stage of FIG. 33A. However, no ion implantation has yet been made to any of the islands 204, 204, 206 and 208.

As shown in FIG. 33B, a photoresist mask 408 is formed on screen oxide 402 with openings over islands 204 and 208. An n-type well dopant composed of phosphorus or arsenic is screened to define (a) precursor n + well portion 276P for IGFET 220V and (b) n + precursor well portion 356P for IGFET 240. The uncovered section of the oxide 402 is ion implanted at high energy and in dark illustration to the bottom monosilicon. Photoresist 408 is removed. After removal of the photoresist 408, the p− portion 278Q of the island 204 for the IGFET 220V is above the precursor n + well portion 276P. Similarly, p- portion 358Q of island 208 for IGFET 240 resides above precursor n + well portion 356P.

At this point, thermal annealing (preferably RTA) is usually performed on the resulting semiconductor structure in order to restore lattice damage and effectively bring the atoms of the implanted n-type well dopant into a more stable state. See FIG. 33C. The portion of the n-type well dopant used to define precursor n + well portion 276P for IGFET 220V diffuses upward into p- portion 278Q. Similarly, the portion of the n-type well dopant used to define precursor n + well portion 356P for IGFET 240 diffuses upward to p- portion 358Q. Typically, upward diffusion of these portions of the n-type well dopant is sufficient to convert each lower portion of the p- portions 278Q and 358Q to n-conductive. These converted lower portions of p- portions 278Q and 358Q are labeled as precursor n- upper body-material portions 278P and 358P, respectively, in FIG. 33C. Due to the formation of the precursor n-body-material portions 278P and 358P, the p- portions 278Q and 358Q shrink vertically in size as generally indicated in FIG. 33C.

As described below, the larger amount of n-type well dopant used to define precursor n + well portion 276P may be modified during subsequent steps of the manufacturing process of FIG. 31 as modified to utilize the alternative of FIG. 33. -Spread upward to portion 278Q. Similar to what occurs in the alternative of FIG. 32, the entire up-diffused portion of the n-type well dopant in the alternative of FIG. 33 is either p-type and / or n-type doping subsequent to the n-type well doping step. All of this p-material, which was not actually performed, is distributed throughout the island 204 in such a way as to be converted to n-conductive at the end of manufacture. Importantly, since the steps for introducing such dopants into the semiconductor structure have not yet been performed, the n-type well dopant for defining the precursor n- upper body-material portion 278P as present in the stage of FIG. 33C Partial upward diffusion occurs without affecting the p-type well dopant, n-type APT dopant, p-type APT dopant, or any source / drain dopant. Thus, at this point, performing n-type well implantation in the fabrication process is such that all of the p- material of island 204 where no other (subsequent) p-type and / or n-type doping has been performed is p- It does not result in undesirable diffusion of type well dopants, n-type APT dopants, p-type APT dopants, or any source / drain dopants and eventually converts to n-conducting.

Photoresist mask 410 is formed on screen oxide 402. See FIG. 33D. Photoresist 410 also has an opening over island 208. An n-type APT dopant composed of phosphorus or arsenic is suitable as the bottom monosilicon through the uncovered section of screen oxide 402 to define precursor upper body-material portion 358P for IGFET 240. Ion implanted into the city. The n-type APT dopant implanted into island 208 converts all of the p- portion 358Q of island 208 to n-conductive, thus causing the p- portion 358Q to disappear. Photoresist 410 is removed.

Phonoresist mask 404 is formed on screen oxide layer 402 as shown in FIG. 33E. Photoresist 404 has openings over islands 202 and 206. The p-type well dopant composed of boron species is screened to define (a) p + type well portion 116 for IGFET 210 and (b) p + precursor well portion 316P for IGFET 230. The uncovered section of the oxide 402 is ion implanted at high energy and in dark illustration to the bottom monosilicon. A portion of epitaxial layer 114P over well portion 116 constitutes p- precursor upper body-material portion 118P for IGFET 210. Photoresist 404 is removed.

Photoresist mask 406 is formed on screen oxide 402. See FIG. 33F. Photoresist 406 has an opening over island 206. The p-type APT dopant composed of boron species is suitable as the bottom monosilicon through the uncovered section of screen oxide 402 to define the p- precursor upper body-material portion 324P for IGFET 230. Ion implanted into the city. Photoresist 406 is removed.

The structure of FIG. 33F is further processed in accordance with the fabrication steps described above with respect to FIGS. 31G-31O, 31P.1-31R.1, and 31P.2-31R.2 including associated annealing operations. do. During the subsequent elevated-temperature step, all of the p-materials in the islands where no other p- and / or n-type doping was performed during the manufacturing process were converted to n-conducting, until IGFET 220V. Using more of the n-type well dopant used to define precursor n + well portion 276P diffuses upward into p- portion 278Q of island 204. Subject to the n-type doped slightly lighter than the n-type upper body-material portion 278 fabricated generally in accordance with the basic manufacturing process of FIG. 31, the p- portion 278Q is primarily an alternative to FIG. 33. N-type upper body-material portion 278 of a complementary-IGFET structure fabricated according to the fabrication process of FIG. 31 as modified to utilize. The n-top body-material remainder 284 is the remaining lightly doped n-type material of the n-type body-material portion 278.

The complementary-IGFET structure fabricated in accordance with the fabrication process of FIG. 31 utilizing the alternative of FIG. 33 is shown roughly as shown in FIG. The p-channel IGFET shown in FIG. 34 described below is a generalized version of IGFET 220V fabricated by utilizing the alternative of FIG. 33.

As mentioned above, the semiconductor portion of p-channel IGFET 220V is produced from the lightly doped p-type material of island 204. To ensure that all island p-materials that do not have p-type and / or n-type doping other than n-type well doping during the manufacturing process are converted to n-conductive at the end of the preparation, At the end of, the concentration of n-type well dopant along the top surface of island 204 must exceed the initial concentration of p-type dopant in island 204. Since island 204 is formed from a portion of p- epitaxial layer 114P (or a p-type substrate doped substantially lightly with epitaxial layer 114P), the island at the end of manufacturing The top-surface concentration of the n-type well dopant in 204 must exceed the p-type background dopant concentration in epitaxial layer 114P.

Doping and thermal processing conditions in one implementation of the fabrication process of FIG. 31, such as modified to incorporate the alternatives of FIG. 32 or 33, may be applied to an n-type well along the top surface of island 204 at the end of fabrication. The concentration of dopant is selected to be at least twice the p-type background dopant concentration in epitaxial layer 114P. Selecting the doping and thermal processing conditions in this manner indicates that the top-surface concentration of the n-type well dopant in island 204 at the end of the fabrication is the p of epitaxial layer 114P in terms of typical manufacturing process variations. Increases the likelihood that the type background dopant concentration will be substantially exceeded. Thus, the selection of the doping and thermal processing conditions involves changing the p-type background dopant concentration in the epitaxial layer 114P from the foregoing and / or changing the n-type well doping conditions from the foregoing. You may. Such changes are described below with respect to FIGS. 35A-35C, 36A-36C, and 37A-37C, and 38A-38C.

Usually, channel zone 266 of p-channel IGFET 220V is asymmetrically longitudinal dopant classified (typically slightly larger) similar to that described above for p-channel IGFET 220. Thus, subject to conductivity type inversion, the asymmetric longitudinal classification in channel zone 266 of IGEFT 220V is similar (typically slightly larger) as described above for IGFETs 180V and 150V.

Typically, p-channel IGFET 220V has a hypobolite vertical dopant profile similar to that described above for p-channel IGFET 220 but slightly below drain 264. In particular, the concentration of the total n-type dopant in the portion of the body material 268 below the drain 264 of the IGFET 220V is determined from the subsurface location of the maximum concentration of the n-type dopant in the well portion 276. At least when moving vertically from the well portion 276 to a drain 264 having a location of the maximum concentration of the n-type dopant at a depth of 10 times or less, usually 5 times or less, below the upper semiconductor surface than 264. To 1/10, typically to approximately 1/15. Thus, subject to the conductivity type inversion, the hypobolite vertical dopant profile under the drain 262 of IGFET 220V is similar to that described above for n-channel IGFETs 180V and 100V but is typically slightly weaker. For the reasons described above in connection with FIGS. 38A-38C, the hypobolite vertical dopant profile below the drain 264 of the IGFET 220V makes it possible to have an increased analog speed.

In an alternative process embodiment, the p-type background dopant concentration, n-type well doping conditions, and subsequent thermal processing conditions in the epitaxial layer 114P may be applied to the body material 268 under the drain 264 of the IGFET 220V. The concentration of the total n-type dopant in the portion of N) is reduced to less than 1/10 as it moves from the subsurface location of the maximum concentration of n-type dopant in the well portion 276 to the drain 264. . As a result, the analog speed may not be as large as when the vertical dopant profile underlying the drain 264 of the IGFET 220V is a hypobolite, but manufacturing an IGFET 220V according to this process embodiment still simplifies the manufacturing process. And reduce the manufacturing cost.

Asymmetric p-channel IGFET 220V may replace IGFET 220 in the complementary-IGFET semiconductor structure of FIG. Typically, as described above-an alternative version of the complementary-IGFET structure of FIG. 30, except for the n + isolation layers 382 and 392, which convert each of IGFETs 210 and 230 into IGFETs 380 and 390. Is manufactured substantially in accordance with the process of FIG. 31 as modified to incorporate the alternative of FIG. 32 or 33. Usually, when utilizing the alternative of FIG. 32, isolation layers 382 and 392 are formed between the stages of FIGS. 31B and 31C using the additional photoresist mask described above with openings over islands 202 and 206. . If the alternative of FIG. 33 is utilized, isolation layers 382 and 392 are formed between the stages of FIGS. 33A and 33B in the same manner as in the alternative of FIG. 32. Normally, the insulating dopant diffusion that occurs during the thermal annealing performed just after n-type well implantation utilizing the alternative of FIG. 33 has substantially no detrimental effect on the resulting complementary-IGFET structure.

Further, in utilizing the alternatives of FIG. 32 or 33, additional photoresist masks are used to create highly doped n-type regions that connect isolation layers 382 and 392 to the upper semiconductor surface to accommodate the isolation voltage. Has an opening. An insulating dopant composed of arsenic or phosphorus may be used to define (a) the n + isolation layers 382 and 392 and (b) the n + isolation layer interconnection region for the IGFETs 380 and 390. It is ion implanted into the bottom of the monosilicon through the uncovered section into dark city.

P-channel IGFET with hypobolite vertical body-material dopant profile below drain due to subsurface maximum well dopant concentration but avoiding n-type compensation implantation

FIG. 34 shows a generalized version 220U of an asymmetric p-channel IGFET 220V that is opposite conductivity to the p- semiconductor material-portion 114 where the n + well portion 276 lies directly below, in accordance with the present invention. . Asymmetric p-channel IGFET 220U is complementary to the p-type portion of island 204 directly over n + well portion 276 as initially defined by ion implantation of the n-type well dopant. IGFET features produced and produced in accordance with the present invention without the use of dopant implants. In essence, an IGFET 220V fabricated in accordance with the process of FIG. 31 as modified to incorporate the alternative of FIG. 32 or 33 is one implementation of the IGFET 220U.

P-channel IGFET 220U includes two portions of p-type source 262, two portions of p-type drain 264, n-type body material 268, gate dielectric layer 286, and a gate electrode. 288. N-type body material 268 is n-type upper body-material portion 278 and n + well portion 276 composed of n + source side pocket portion 280 and n-type body-material remainder 394. Is formed. Similarly, n-type channel zone 266 of n-type upper body-material portion 278 laterally separates n-type S / D zones 262 and 264. Components 262, 264, 266, 268, 276, 278, 280, 286, and 288 of IGFET 220U are constructed and doped to be substantially the same as IGFET 220V.

Item 396 of FIG. 34 shows the pn junction between the source 262 and the body material 268. Item 398 represents a pn junction between drain 264 and body material 268. Similar to the n-channel IGFETs, items y _S and y _D represent the depths at which the source 262 and drain 264 of the p-channel IGFET 220U extend below the upper semiconductor surface, respectively.

For simplicity, the n-type upper body-material remainder 394 of the IGFET 220U is referred to herein as how the n-type upper body-material remainder 284 of the IGFET 220 is identified by the label "n-". Similarly it is labeled "n-" in FIG. In fabricating the IGFET 220U according to the process of FIG. 31, as modified to incorporate the alternative of FIG. 32 or 33, the upper body-material remainder 394 usually diffuses upward from the underlying n + well portion 276. Only by substantially accepting the n-type dopant. As further described below with respect to FIGS. 37A-37C, the n-type dopant concentration in the body-material remainder 394 usually decreases rapidly from the well portion 276 to the upper semiconductor surface. Since well portion 276 is heavily doped n-type, body-material remainder 394 is lightly doped n-type surface adjacent portion, and n + well portion 276 and lightly doped n-type It can be vividly observed as consisting of a suitably doped n-type intermediate portion located between the surface-adjacent portions.

Understanding the doping characteristics of IGFET 220U is shown in FIGS. 35A-35C (collectively "FIG. 35"), FIGS. 36A-36C (collectively "FIG. 36"), and FIGS. 37A-37C (collectively). "FIG. 37", and FIGS. 38A to 38C (collectively "FIG. 38"). 35 is an exemplary dopant concentration along the upper semiconductor surface as a function of longitudinal distance x. Exemplary dopant concentration as a function of depth y along vertical line 130U through source 262 is shown in FIG. 36. FIG. 37 shows exemplary dopant concentrations as a function of depth y along a pair of vertical lines 132U and 134U along channel zone 266. Vertical line 132U passes through source side pocket portion 280. Vertical line 134U passes through a vertical position between pocket portion 280 and drain 264. An exemplary dopant concentration as a function of depth y along vertical line 136U through drain 264 is shown in FIG. 38. Vertical lines 130U, 132U, 134U and 136U for p-channel IGFET 220U each correspond to vertical lines 130, 132, 134 and 136 for n-channel IGFET of the present invention.

35A shows the concentrations N _I of individual semiconductor dopants along the upper semiconductor surface that largely define regions 262, 264, 280 and 394 and thus establish the longitudinal dopant classification of channel zone 266. 36A, 37A and 38A define regions 114, 262, 264, 276, 280 and 394 vertically along vertical lines 130U, 132U, 134U and 136U and thus below drain 264. The concentration N _I of the individual semiconductor dopant that establishes the hypobolite vertical dopant profile in the portion of the body material 268 is shown. Curves 262 'and 264' show the concentrations N _I (surface and vertical) of the p-type dopant used to form the source 262 and drain 264, respectively. Curves 276 ', 280' and 394 'represent the concentrations N _I (surface and / or vertical) of the n-type dopants used to form regions 276, 280 and 394, respectively. The items 396 ^# and 398 ^# indicate where the net dopant concentration N _N proceeds to 0 and thus indicate the positions of the pn junctions 396 and 398, respectively.

Concentrations N _T of total p-type and total n-type dopants in regions 262, 264, 280, and 394 along the upper semiconductor surface are shown in FIG. 35B. 36B, 37B, and 38B show the concentration of total p-type and total n-type dopants in regions 114, 262, 264, 276, 280, and 394 along vertical lines 130U, 132U, 134U, and 136U. N _T is shown. Curve segments 276 ", 280", and 394 "corresponding to regions 276, 280, and 394, respectively, represent the total concentration N _T of the n-type dopant. Item 226" in FIG. 35B shows channel zone 266. Corresponding to the channel zone portions of the curve segments 280 " and 394 &quot;. The total concentration N _T of the p-type dopant is represented by curves 262 ″ and 264 ″ corresponding to the source 262 and drain 264, respectively.

35C shows the net dopant concentration N _N along the upper semiconductor surface. Net dopant concentrations N _N along vertical lines 130U, 132U, 134U, and 136U are shown in FIGS. 36C, 37C, and 38C. Curve segments 276 ^* , 280 ^* and 394 ^* represent the net concentration N _N of the n-type dopant in each region 276, 280 and 394. Item 266 ^* in FIG. 35C shows the combination of the channel zone curve segments 280 ^* and 394 ^* , and thus the concentration N _N of the net n-type dopant in the channel zone 266. The concentrations N _N of the net p-type dopant in the source 262 and the drain 264 are represented by curves 262 ^* and 264 ^* , respectively.

The top-surface dopant distribution shown in FIG. 35 for the p-channel IGFET 220U is (a) the background p-type dopant for the p- lower portion 114 as indicated by curve 114 'in FIG. 35A. The concentration is less than the background n-type dopant concentration for the n-lower portion 192 as indicated by item 192 'in FIG. 26A, and (b) items 394', 394 "and 394 ^{* in} FIG. The top-surface dopant concentration for the n-top body-material remainder 394 as indicated by p-top body-material remainder 198 as indicated by items 198 ', 198 "and 198 ^{* in} FIG. It is substantially the same as the top-surface dopant distribution shown in FIG. 26 for the n-channel IGFET 180V, except that it is below the top-surface dopant concentration for. Nevertheless, the p-channel IGFET 220U is configured similarly to the n-channel IGFET 180V. Thus, the opinion regarding the top-surface dopant distribution shown in FIG. 26 for the n-channel IGFET 180V is expressed in the regions 102, 102M, 102E, 104, 104M, 104E, 106, 120, for IGFET 180V. 192 and 198 are replaced with regions 262, 262M, 262E, 264, 264M, 264E, 266, 280, 114, and 394 for IGFET 220U, respectively, in FIG. 35 for p-channel IGFET 220U. It applies greatly to the top-surface dopant distribution shown.

Consider FIG. 37 again. The n-type dopant in the n + pocket portion 280 is (a) a major portion consisting of the n-type pocket dopant indicated by the dopant concentration curve 280 'in FIG. 37A, and (b) the dopant concentration curve 276 in FIG. 37A. It consists of a minor portion consisting of an upwardly diffused portion of the n-type well dopant indicated by '). Practically, the n-type dopant in the n-top body-material remainder 394 consists only of the updiffused portion of the n-type well dopant. The updiffused concentration N _I of the n-type well dopant in the n-body-material remainder 394 is represented as part 394 'of the n-type well dopant concentration curve 276' in FIG. 37A. Similarly, portions 394 " and 394 ^* of the respective dopant concentration curves 276 " and 276 ^* in FIGS. 37B and 37C show the total dopant concentration N _T and net dopant concentration in the n-body material remainder 394. N _N is represented, respectively.

Source-side pocket portion 280 of IGFET 220U is heavily doped n-type, and n-type doping of n- upper body-material remainder 394 is only caused by upward diffusion from n + well portion 276. Through substantially provided, the lowest n-type dopant concentration in the channel zone 266 determines whether any significant amount of the updiffused portion of the n-type well dopant is deposited along the top surface of the remainder 394. Depending on whether it occurs in the n-body-material remainder 394 along or close to its upper surface. 37 shows an illustration without any n-type well dopant deposition along the top surface of n-body material remainder 394.

p-channel IGFET 220V in which a semiconductor portion is formed in an island 204 formed from a portion of p- epitaxial layer 114P (or a p-type substrate lightly doped substantially equal to epitaxial layer 114P). As described above in connection with the fabrication of, the concentration of the n-type well dopant along the top surface of island 204 at the end of the preparation should exceed the p-type background dopant concentration in epitaxial layer 114P. When applied to IGFET 220U, the manufacturing doping requirement is the concentration of _I -type well dopant along the upper surface of channel zone 266 (in particular, the upper surface of n- upper body-material remainder 394). Exceeds the concentration N _I of the p-type background dopant in the p- lower portion 114. A preferred fabrication objective is that the top-surface concentration of the n-type well dopant in island 204 at the end of the fabrication of IGFET 220V is at least twice the concentration of p-type background dopant in epitaxial layer 114P. Similarly, the anticipated doping and thermal processing conditions are expected to be such that the concentration N _I of the n-type well dopant along the top surface of the n-body-material remainder 394 is at least 2 of the concentration N _I of the p-type background dopant. The transition is to multiplying structural doping requirements.

35-38 show an n-type well dopant concentration N _I as shown by curve segment 394 'in FIGS. 35A and 37A along the upper surface of n-body-material remainder 394. As shown by curve 114 ′ in FIG. 37A, an illustration is shown, which is approximately twice the p-type background dopant concentration of p-bottom region 114. Thus, the examples of FIGS. 35-38 illustrate specific structural doping in which the concentration NI of the n-type well dopant along the upper surface of the n-body-material remainder 394 is at least twice the p-type background dopant concentration N _I. Satisfy the requirements.

The concentration N _I of the p-type background dopant is approximately 1 × 10 ¹⁵ atoms / cm 3 in the example of FIGS. 35 to 38. This is the lower limit of 1 × 10 ¹⁵ -1 × 10 ¹⁶ atoms / cm 3 given above for the p-type background dopant concentration. However, using the alternative of FIG. 32 or 33 to implement IGFET 220U as IGFET 220V, the lower limit of the range for the p-type background dopant concentration is downconverted to 5 × 10 ¹⁴ atoms / cm 3 or less. . For island 204 used to create IGFET 220V, this means that all island p-materials that do not perform p-type and / or n-type doping other than n-type well doping during the manufacturing process are More flexibility is provided while ensuring that the conversion is n-conductive at the end.

The concentration N _I of the n-type well dopant in IGFET 220U reaches a maximum at depth y _W below the upper semiconductor surface. The total n-type dopant in the portion of body material 268 below source 262 and drain 264 of IGFET 220U is an n-type well as indicated by curve 276 ′ in FIGS. 36A and 38A. It consists only of dopants. As a result, the concentration N _T of the total n-type dopant in the portion of p-body material 268 underlying source 262 and drain 264 of IGFET 220U is curve 276 ″ in FIGS. 36B and 38B. The maximum value is reached at a depth y _W as indicated by.

As shown by the change in curve 276 ″ in FIG. 38B, the concentration N _T of the total n-type dopant in the portion of the body material 268 below the drain 264 of the IGFET 220U is determined by the vertical line ( 136U decreases by at least 1/10, typically approximately 1/15 when moving from the subsurface location of the maximum concentration of n-type dopant of well portion 276 to drain 264. The location of the maximum concentration of the entire n-type dopant in 276 is at a depth below the upper semiconductor surface of 1/10 or less, usually 1/5, below the drain 264. Thus, the drain of the IGFET 220U ( 264) The vertical dopant profile underneath is hypobolite, and the concentration N _T of the total n-type dopant in the portion of the body material 268 below the drain 264 of the IGFET 220U is below the drain 264. Dre from the depth y _W of the maximum concentration of the total n-type dopant in the portion of the body material 268 of During movement from the location of the maximum n-type well concentration to the drain 264, as indicated by the portion of the curve 276 ′ extending to the item 398 ^# representing the in-body junction 398, it gradually decreases.

The net dopant in the portion of body material 268 below drain 264 of IGFET 220U is an n-type dopant. 38C shows that the concentration N _N of the net dopant in the portion of the body material 268 below the drain 264 of the IGFET 220U, as indicated by the curve 276 ^* , represents the body material (below the drain 264). 268 a similar manner to the concentration N _T of the entire n-type dopant in the portion of the body material 268 below the drain 264, except that the concentration N _N of the portion drops to zero at the drain-body junction 398. Changes vertically. In the portion of the body material 268 below the drain 264 of the IGFET 220U, the hypobolite vertical dopant profile is characterized in that the parasitic capacitance associated with the drain-body junction 398 is determined by the p-channel IGFET 220 and the present invention. Cause to be reduced for the reasons described above in connection with the asymmetric n-channel IGFET. As in IGFET 220U using n-type compensation injection, where the hypobolite vertical dopant profile in the portion of body material 268 below drain 264 avoids n-type compensation injection, The hypobolite vertical dopant profile in the portion of the body material 268 below the drain 264 causes the drain-body junction 398 to have reduced parasitic capacitance. Thus, IGFET 220U has increased analog speed.

The concentrations N _I and N _T of the total n-type dopant in the portion of the body material 268 under the source 262 of the IGFET 220U are all n-type in the portion of the body material 268 under the drain 264. It changes significantly perpendicular to the concentrations of dopants N _I and N _T. Curve 276 of FIGS. 36A and 36B taken along vertical line 130U through source 262 and curve 276 of FIGS. 38A and 38B taken along vertical line 136U along drain 264. 'And 276 "). Note that the net dopant in the portion of body material 268 under source 262 of IGFET 220U is an n-type dopant, and the concentration of the net dopant in the portion of body material 268 under source 262. N _N changes the concentration N _N of the net dopant in a portion of the body material 268 below the drain 264 to be substantially vertically equal. Thus, as represented by curve 276 ^* in FIG. 36C, the concentration N _N in the portion of the body material 268 below the source 262 drops to zero at the source-body junction 396. , In a portion of the body material 268 below the source 262, changes vertically in a manner similar to the concentration N _T of the total n-type dopant. Normally, the vertical dopant profile underlying the source 262 of the IGFET 220U causes the parasitic capacitance along the source-body junction 396 to be reduced.

General purpose computer simulation

Computer simulations were performed to check the device characteristics and performance benefits of IGFETs constructed in accordance with the present invention, particularly for analog applications. This simulation includes (a) a Micro Tec two-dimensional device simulator supplied by Siborg System and (b) Avant! Corp. It was performed by a Medici two-dimensional device simulator supplied by. The Micro Tec simulator was mainly used for large signal (DC) simulation. Medici simulator is mainly used for small signal simulation.

Two types of n-channel IGFETs generally correspond to device level: (a) asymmetric n-channel IGFETs constructed in accordance with the present invention and (b) novel computer-simulated new n-channel IGFETs (but computer-simulating). Computer simulated in a symmetric reference n-channel IGFET (different from the new n-channel IGFET). In general, new computer-simulated asymmetric IGFETs are represented below by the structure “A”. The computer-simulated novel asymmetric IGFET of structure A generally corresponds to the long n-channel IGFET 150 or short-channel version of IGFET 150 of FIG. 13. In general, the reference computer-simulated symmetric IGFET is represented below as structure “B”. The computer-simulated reference IGFET of structure B is generally the chapter n of FIG. 29, with the halo pocket portions 320 and 322 of IGFET 230 extending below the S / D zones 302 and 304, respectively. Correspond to the channel IGFET 230 or the short-channel version of the IGFET 230.

Structures (A and B) are based on an analytical-profile model that generally uses Gaussian dopant profiles. Each set of structures A and B was compared and considered to be manufactured according to the same process flow as in FIG. 31. In addition to the variations that produce the body-material doping features of the present invention, except for the cases expressed differently below, the comparison of structure A and structure B in each set had substantially the same dopant distribution. In addition, the comparison of structure A and structure B in each set had essentially the same geometric dimensions, eg, gate length, gate stack height, and source / drain length. Computer simulations for structures A and B represent devices manufactured at 0.18-μm technology nodes (ie, made with design rules that had a minimum printable feature size of 0.18 μm).

In general, computer simulations have been ordered to enhance analog performance. Thus, structure B is computer simulated at the expected parameter values to yield enhanced analog performance for structure A. Given that the basic architecture of structure B is for digital use, the values for certain parameters used in computer simulation structure B to achieve enhanced analog performance are different from those that yield enhanced digital performance. It is different.

Structures A and B are arranged in a multi-IGFET structure, and trenches filled with insulator to implement field-isolation regions, such as field-isolation region 200 in FIG. 29, to laterally insulate the IGFET in the multiple IGFET structures. Was considered to be used. Trench having the same dimensions for structures A and B is up to 50% deeper and up to 20% wider than the dimensions needed to optimize digital performance for reference structure B in many IGFET structures. The IGFET wells corresponding to each of the p + well portions 116 and 316 for asymmetric n-channel IGFET 150 in FIG. 13 and for symmetric n-channel IGFET 230 in FIG. 29 optimize the digital performance of structure B. Up to 20% deeper than the dimensions needed to The novel structure A may require wider / deeper trenches to maintain good inter-IGFET isolation in the presence of wells where the average dopant concentration is lower than that of structure B optimized for digital performance.

Both structures A and B have the same threshold voltage V _T and the same background p-type dopant concentration, ie 0.4V and 5 × 10 ¹⁵ atoms / cm 3. For a given value of the p-type background dopant concentration, the p-type implants that define pocket portions along the upper semiconductor surfaces of structures A and B control their threshold voltages V _T. In view of this, the peak top-surface concentration of the p-type pocket dopant in the single pocket portion of the novel structure A arises from the peak top-surface concentration of the p-type halo dopant in the two halo pocket portions of the reference structure B. Is adjusted appropriately to achieve the same threshold voltage. More specifically, the new asymmetric structure A accommodates the darker depictions of the pocket dopant than the reference symmetric structure B to accommodate the fact that the reference structure B has twice the pocket portions of the new structure A. The higher doping in the single pocket portion of the novel asymmetric structure A is the same asymmetric IGFET structure as the structure A, except that the pocket portion of the comparative asymmetric structure has the same doping as the pocket portion of either of the reference symmetric structures B. Increase the punch-through resistance in comparison with

39 shows a three-dimensional dopant profile for the implementation of a short channel version of the novel asymmetric structure A. FIG. The three-dimensional dopant profile for the implementation of the corresponding short channel version of the reference symmetric structure B is shown in FIG. 40. 39 and 40 specifically illustrate net dopant concentration N _N as a function of longitudinal distance x and depth y. Only the monosilicon portions of the implementations of structures A and B are shown in FIGS. 39 and 40. Although both Structure A and Structure B are considered to be fabricated with 0.18 μm technology nodes, implementations of both Structure A and Structure B in FIGS. 39 and 40 in computer simulations have a channel length L of 0.12 to 0.14 μm due to lithographic alignment. Induced gate length L _G of 0.2 μm.

The implementation of the novel structure A in FIG. 40 basically corresponds to the short channel version of IGFET 150. Basically, the implementation of reference structure B in FIG. 40 is generally simulated at parameter values where the implementation of structure B of FIG. 40 enhances the analog performance of the implementation of structure A, and the halo pocket portion of IGFET 230 ( Corresponding to the short channel version of IGFET 230 provided that 320 and 322 extend below S / D zones 302 and 304, respectively. For convenience, the implementation of the short channel structures A and B in FIGS. 39 and 40 are labeled with reference symbols employed to identify the monosilicon regions of IGFETs 150 and 230. Since structure B is simulated for analog use, S / D zones 302 and 304 are labeled as source 302 and drain 304 in FIG. 40, respectively. Except as otherwise indicated below, all references to short channel versions of structures A and B or to short channel structures A and B refer to the implementations shown in FIGS. 39 and 40, respectively.

Each S / D zone of each structure A or structure B consists of a very heavily doped main portion and a lightly doped but still heavily doped lateral extension. Thus, the drain 104 of the short channel version of structure A consists of the main portion 104M and the lateral extensions 104E. However, regions 104M and 104E of drain 104 are difficult to distinguish in FIG. 39 and are not separately labeled in FIG. 39.

41 and 42 show two-dimensional dopant contours for each short channel version of structure A and structure B. FIGS. Dopant contours are taken along the vertical plane through the short channel structures A and B. The regions 104M and 104E of the short channel structure A can be clearly distinguished in FIG. 41 and are labeled in FIG. 41. As shown by the positions of the pn junctions 110 and 112 in FIG. 41, the main drain portion 104M of the short channel structure A extends deeper below the upper semiconductor surface than the main source portion 102M. In other words, the drain depth y _D for the novel asymmetric structure A is deeper than the source depth y _S. In contrast, the irradiation of FIG. 42 shows that the main source portion 302M and the main drain portion 304M in the reference symmetry structure B extend to substantially the same depth below the upper semiconductor surface.

43 shows the net dopant concentration N _N as a function of the longitudinal distance x from the measurement-reference S / D zone position for the short channel versions of structures A and B below the upper semiconductor surface. The measurement-reference S / D zone position is approximately 3.5 μm from the channel-zone center. In the following graph showing computer-simulated data for both Structure A and Structure B, and in FIG. 43, the curve representing data for the new structure A is plotted against the reference structure B where the data curve is not further marked. It is marked with a small empty circle to distinguish it from the data. At locations where Structure A and Structure B have substantially the same data, the curve segments for Structure A and Structure B are visually indistinguishable from each other.

Similar to the curve segment of FIG. 14C, the curve segments 102M ^* , 104M ^* , 102E ^*, and 104E of FIG. 43 are in regions 102M, 104M, 102E, and 104E along the upper semiconductor surface of the new short channel structure A. FIG. Each shows the concentration N _N of the net n-type dopant. Curve segments 106 ^* and 120 ^{* in} FIG. 43 represent the concentration N _N of the net p-type dopants in each of regions 106 and 120 along the upper semiconductor surface of short channel structure A. In FIG. Curve segments 302M ^* , 304M ^* , 302E ^*, and 304E ^* determine the concentration N _N of p-type dopant in each region 302M, 304M, 302E, and 304E along the upper semiconductor surface of reference short channel structure B. Indicates. Curve segments 306 ^* , 320 ^*, and 322 ^* represent the concentration of net p-type dopants in regions 306, 320, and 322 along the upper semiconductor surface of short channel structure B.

Curve segment 106 ^{* in} FIG. 43 illustrates asymmetric dopant classification in channel zone 106 of novel short channel structure A. FIG. In particular, the curve segment 106 ^* has a high value after the net dopant concentration N _N reaches a high value of approximately 1 × 10 ¹⁸ atoms / cm 3 near the source 102 along the upper semiconductor surface of the short channel structure A. Decrease rapidly as it moves across channel zone 106 toward drain 104 from the position of. Although not shown in the computer simulation, the concentration N _T of the total p-type dopant in the channel zone 106 of the short channel structure A is greater than the zone 106 where the zone 106 meets the drain 104 along the upper surface. ) Is at least 10 times or less, typically greater than 100 times or less, where it meets the source 102 along the top surface. In contrast, the symmetric dopant classification in the channel zone 316 of the short channel structure B has a concentration N _N along the upper surface of the short channel structure B at approximately the same peak value close to both the source 102 and the drain 104. It is shown by the curve segment 306 ^* indicating that it is at a slightly lower value in the center of the channel zone 306.

FIG. 44A shows the absolute (full p-type and full n-type) vertical dopant profiles through the S / D positions for the short channel versions of structure A and structure B. FIG. Using the same measurement-referenced S / D zone position where the longitudinal distance x was measured in FIG. 43, the absolute dopant concentration N _T for the new short channel structure A is the main source portion 102M at a distance x equal to 0.0 μm. 44a along a vertical line (or vertical plane) through and along a vertical line through main drain portion 104M at a distance x equal to 0.7 μm. Similarly, FIG. 44A shows a reference short channel along a vertical line extending through main source portion 302M at a distance x equal to 0.0 μm and along a vertical line extending through main drain portion 304M at a distance x equal to 0.7 μm. The concentration N _T for structure B is shown.

Similar to the curve segments in FIGS. 8B and 10B, the curve segments 102 ″ and 104 ″ in FIG. 44A correspond to the source 102 and drain 104 of the new short channel structure A, respectively, and thus The concentration N _T of the total n-type dopant along the vertical line at a distance x of 0.0 and 0.7 μm, respectively, through the source 102 and drain 104, specifically the main source portion 102M and the main drain portion 104M, respectively. Indicates. In general, each of the curve segments 114 ", 116", 118 ", 120", and 124 "of FIG. 44A corresponds to the regions 114, 116, 118, 120, and 124 of the short channel structure A, and accordingly, source 102 and drain 104, different concentrations of N _T the entire p- type dopant along the vertical lines at distances x of 0.0 and 0.7㎛ shows through. Item ^(# 110 and ^# 112), each stage channel structure a Pn junctions 110 and 112 for.

The maximum concentration of the p-type well dopant in the well portion 116 of the novel short channel structure A occurs at a depth y _W equal to 0.7 μm as indicated by curve segment 116 ″ in FIG. 44A. The combination of 116 " and 124 &quot;) exhibits the hypobrupt property of the vertical dopant profile underlying the drain 104 along the vertical line at a distance x equal to 0.7 [mu] m in the short channel structure A. In particular, the combined curve segment 116 " / 124 &quot;) indicates that the concentration N _T of the total p-type dopant in the portion of the body material just below the drain 104 is at the location of the maximum p-type well dopant concentration to the pn junction 112 at the bottom of the drain. It decreases by approximately 100 times when moving, and therefore by a factor that significantly exceeds 10.

The depth y _D of the drain-body junction 112 is approximately 0.2 μm in the novel short channel structure A in FIG. 44A. Since the depth y _W of the maximum concentration N _T of the total p-type dopant in the body material portion below the drain 104 is 0.7 μm, the total p-type dopant in the body material portion below the drain 104 of the short channel structure A is The location of the maximum concentration N _T is approximately 3.5 times deeper than the drain 104 below the upper semiconductor surface. As a result, the concentration N _T of the total p-type dopant in the body material 108 is drained from the maximum p-type well dopant concentration position, which is 5 times or less deeper than the drain 104 below the upper semiconductor surface. Decrease to approximately 1/100 when moving upward.

Each of curve segments 302 ″ and 304 ″ in FIG. 44A corresponds to source 302 and drain 304 of reference short channel structure B, and through 0.0 and 0.7 μm through source 302 and drain 304. The concentration N _T of the total n-type dopant along the vertical line at distance x is shown. Each of the curve segments 114 ", 316", 318 ", 320", and 322 "corresponds to the regions 114, 316, 318, 320, and 322 of the short channel structure B, and thus, the source 302 and the drain ( Concentration N _T along the vertical line at a distance x of 0.0 and 0.7 [mu] m through 304. In this respect, curve segments 114 "are utilized for both short channel structures A and B. .

As in the situation in the well portion 116 of the novel short channel structure A, the maximum concentration of the p-type well dopant in the well portion 316 of the reference short channel structure B is the curve segment 316 ″ in FIG. 44A. ) Occurs at a depth y _W equal to 0.7 μm, but the combination of curve segments 316 ″, 318 ″, and 322 ″ (or 320 ″) may result in drain 304 (or Relatively flat in the p-type body-material portion directly underneath the source 302. The combined curve segment 316 "/ 318" / 322 "is the entire p in the body-material portion directly underneath the drain 304. The concentration N _T of the -type dopant varies considerably less than 5 times and further significantly less than 10 times as it moves from the location of the maximum p-type well dopant concentration to the pn junction at the bottom of the drain 304. . That is, the vertical dopant profile in the body-material portion just below the drain 304 in the short channel structure B is not hypobolite.

The net vertical dopant profile corresponding to the absolute vertical dopant profile of FIG. 44A is shown in FIG. 44B for structures A and B of the short channel version. Curve segments 102 ^* and 104 ^* in FIG. 44B are 0.0 and 0.7 [mu] m respectively through source 102 and drain 104, specifically through main source portion 102M and main drain portion 104M, respectively. The concentration N _N of the net n-type dopant in each of the source 102 and the drain 104 of the new short channel structure A along the vertical line at a distance x is shown. Curve segments 114 ^* , 116 ^* , 120 ^*, and 124 ^* are regions of short channel structure A along a vertical line at a distance x of 0.0 and 0.7 μm through source 102 and drain 104. 120 and 124), respectively, show the concentration N _N of the net p-type dopant.

Curves 302 ^* and 304 ^* in FIG. 44B are respectively 0.0 and 0.7 μm distances through source 302 and drain 304, specifically through main source portion 302M and main drain portion 304M. The concentration N _N of the net n-type dopant of each of the source 302 and the drain 304 of the reference short channel structure B along the vertical line at x is represented. Curve segments 114 ^* , 316 ^*, and 318 ^* are regions of short-channel structure B, 114, 316, and 318, respectively, along a vertical line at a distance x of 0.0 and 0.7 μm through source 302 and drain 304, respectively. Denotes the concentration N _N of the net p-type dopant. Curve segments 114 ^* are used for both short channel structures A and B.

45A and 45B respectively show the linear linear-range transconductance g _mw and the linear saturation transconductance as a function of the gate-source voltage V _GS for the short channel versions of structures A and B with a gate length L _G of 0.2 μm. g _msatw . 46A and 46B each show a function of the gate-source voltage V _GS for the long channel version structures A and B configured essentially the same as the short channel structures A and B, except that the gate length L _G was 0.5 μm. Linear linear-range transconductance g _mw and linear saturation transconductance g _msatw . The drain-source voltage V _DS for the linear-range g _mw graphs of FIGS. 45A and 46A was 0.1V. The voltage V _DS for the saturation g _msatw graph of FIGS. 45B and 46B was 2.0V. 45A, 45B, 46A and 46B also show the change in the drain current I _Dw of the direct _{series in} which the linear-range transconductance g _mw and the saturated transconductance g _msatw are determined.

46A and 46B, the long channel version of the novel structure A exhibited significantly higher transconductance (both linear-range transconductance g _mw and saturated transconductance g _msatw ) than the long channel version of the reference structure B. The short channel version of structure A exhibited approximately 10% higher linear-range transconductance g _mw , slightly higher than the short channel version of structure B, as shown in FIG. 45A. 45B shows that the short channel versions of structures A and B have nearly the same g _msatw characteristics. In general, higher values of transconductances g _mw and g _msatw for the novel structure A make it possible to have higher voltage gains and thus improve their analog performance.

FIG. 47 shows the current-voltage transfer characteristics (i.e., the change in the drain current I _Dw of the direct series with the gate-source voltage V _GS) for the short channel versions of structures A and B at a drain-source voltage V _DS of 2.0 V. FIG. do. As shown in Fig. 47, the short channel structures A and B have almost the same current-voltage characteristics. The long channel versions of structures A and B are expected to have largely identical current-voltage characteristics.

Note that the upper body-material portion 318 of the reference short channel structure B is provided with a high concentration of p-type APT dopant to help avoid punchthrough in the short channel structure B, and the p in the reference structure B The presence of the p-type APT dopant in the novel structure A, which is generally at a position generally similar to that of the -type APT dopant, indicates that the nearly identical current-voltage characteristics for the short channel structures A and B are punchthrough in the new structure A. It does not lead to A qualitative physical explanation for this result is that the p-type pocket dopant performs the anti-punchthrough function in structure A. More specifically, the p-type pocket implant in the pocket portion 120 of the novel structure A is one of the two pocket portions 320 and 322 of the reference structure B so that the structure A has the same threshold voltage V _T as that of the structure B. Higher doping is provided than a single p-type halo implant. This difference can be observed by comparing the curve segments 320 ^* and 322 ^* with the curve segments 120 ^* in FIG. 43. Higher concentrations of p-type pocket implants in the novel structure A make it possible to avoid punchthrough while simultaneously operating at low current leakage similar to the concentration of the reference structure B.

In addition, the foregoing conclusions are supported by computer-simulated data obtained in another reference short channel IGFET structure C, which lacks APT injection of the reference short channel structure B than the other as the short channel structure B. The current-voltage transfer characteristic for the reference structure C at the V _DS value 2.0V is indicated by curve tag C in FIG. 47. The leakage current I _D0w is the value of the drain current I _Dw at the zero value of the gate-source voltage V _GS . As shown in FIG. 47, the drain leakage current I _D0w is approximately 50 times higher than the reference structure C than in the new short channel structure A. FIG. This indicates that punchthrough occurs in the reference structure C.

48 shows the direct current drain current I _Dw as a function of the drain-source voltage V _DS for the short-channel versions of structures A and B at the value of the gate-source voltage V _GS in the range of 0.5V to 2.0V. As shown in FIG. 48, the new short channel structure A generally achieves a drain current I _Dw of slightly higher value than the reference short channel structure B at each indicated V _GS value. Thus, the novel structure A has a lower channel resistance than the reference structure B. In addition, the drain current I _Dw increases the drain-source voltage V _DS at the high voltage V _DS in the novel structure A less than in the reference structure B. This indicates that avalanche multiplication or / and channel-width modulation occurs less in novel structure A than in reference structure B.

Analytical Analysis and Performance Benefits of New Asymmetric IGFETs

In good analog performance, the source of the IGFET should be as shallow as reasonably possible to avoid roll-off of the threshold voltage V _T at short channel length. In addition, the source must be doped as hard as possible to maximize the effective transconductance g _meff of the IGFET in the presence of the source resistance R _S. The effective transconductance g _meff is determined from the intrinsic transconductance g _m of the IGFET as follows.

As shown in equation 1, decreasing the source resistance R _S causes to increase the effective transconductance g _meff . In addition, the voltage drop across the source resistance R _S is subtracted from the intrinsic gate-source voltage so that the actual gate-source voltage V _GS is at a lower value. This divides the IGFET at its gate electrode. In short, the source resistance R _S should be as low as reasonably possible.

In addition to minimizing the source resistance R _S to maximize the effective transconductance g _meff , _there is also a need to have a lower series resistance at the source and drain of the IGFET to achieve lower values of the on-resistance R _on of the IGFET. . Specifically, the voltage drop across the source resistance R _S is added to the overall source-drain voltage drop. This causes an increase in the _on -resistance R _on .

In order to achieve high voltage performance and reduce hot-carrier injection, the drain of the IGFET should be as reasonably deep and weakly doped as possible. This need must be met without significantly increasing the on resistance R _on and causing a roll-off of the short channel threshold voltage.

Parasitic capacitances of IGFETs play an important role in setting the speed performance of circuits including IGFETs, especially in small signal small frequency operation. 49 shows that C _DB represents drain-body capacitance, C _SB represents source-body capacitance, C _GB represents gate-body capacitance, C _GD represents gate-drain capacitance, and C _GS represents gate-source Shows parasitic capacitances C _DB , C _SB , C _GB , C _GD and C _GS that are variously associated with drain electrode D, source electrode E, gate electrode E, and body-region electrode B of n-channel IGFET Q exhibiting capacitance. . The small signal equivalent model of IGFET Q is shown in FIG. 50 where V _BS is the body-source voltage, g _mb is the transconductance of the body electrode, and items 440 and 442 are current sources.

(대략 0.707) 으로 떨어지는 주파수 값으로서 정의된다. 일반적으로, 증폭기의 대역폭은 가능한 한 큰 것이 바람직하다.The bandwidth of the amplifier, the gain of the amplifier

It is defined as a frequency value falling to (approximately 0.707). In general, the bandwidth of the amplifier is preferably as large as possible.

IGFET Q of FIG. 49 can be arranged in three main amplifier configurations for providing an amplified output voltage V _out as a function of input voltage V _in according to the following relationship.

Where H _A is the complex transition function of the IGFET. These three configurations have the common-source, common-gate, and common-drain configurations shown in Figs. 51A-51C, where C _L is the load capacitance, V _DD is the high supply voltage, and V _SS is the low supply voltage. to be. The amplifier input voltage V _in is supplied from the voltage source 444. Element 446 in FIGS. 51B and 51C is a current source, and signal V _G in FIG. 51B is a gate voltage. Examination of the transfer function H _A for the three configurations of FIGS. 51A-51C shows that reducing the parasitic drain-body capacitance C _DB and / or parasitic source-body capacitance C _SB improves the IGFET performance of each of these configurations. Indicates.

The transfer function H _A for the common-source amplifier configuration of FIG. 51A is an input pole / output pole function.

Where R _D is the drain (serial) resistance, ω _in is the angular frequency at the input pole, ω _out is the angular frequency at the output pole, and s is the complex frequency operator equal to jω where ω is the angular frequency. The parasitic capacitance of the IGFET Q in the common-source configuration is substituted into equation (3) in the manner of the pole frequencies ω _in and ω _out , respectively, given below.

The parasitic drain-body capacitance C _DB is represented by the output pole frequency ω _out of equation (5). In the situation where the source resistance R _S is 0 in the common-source configuration, the input pole frequency ω _in is infinite according to equation (4). Next, the bandwidth of IGFET Q in FIG. 51A is equal to ω _out as given by equation (5). The output pole frequency ω _out increases with increasing drain-body capacitance C _DB for a given value of drain resistance R _D and parasitic gate-drain capacitance C _GD . Thus, reducing the parasitic drain-body capacitance C _DB preferably increases the bandwidth for the common-source configuration of IGFET Q in FIG. 51A.

In addition, the parasitic drain-body capacitance C _DB of the common-source configuration is in parallel with the load capacitance C _L as shown in FIG. 51A. Thus, reducing the drain-body capacitance CDB advantageously reduces its output loading effect.

The transfer function H _A for the common-gate amplifier configuration of FIG. 51B is an input pole / output pole function.

Here, the input pole frequency ω _in for the common-gate configuration is as follows.

Here, decreasing the source-body capacitance C _SB causes increasing the input pole frequency ω _in . This makes it possible to improve the performance of IGFET Q in the common-gate configuration in FIG. 51B.

The output pole frequency ω _out is provided by equation (5) for the common-gate amplifier configuration of FIG. 51B. Thus, the parasitic drain-body capacitance C _DB increases the bandwidth of the common-gate configuration of FIG. 51B.

The transition function H _A for the common-drain amplifier configuration of FIG. 51C is a single-zero / single-pole function.

Where ω _z is the angular frequency at zero and ω _p is the angular frequency at the pole. Parasitic capacitance is substituted into equation (8) in the manner of the zero frequency ω _z and the pole frequency ω _p respectively provided as follows.

Here, the parasitic source-body capacitance C _SB is shown at the pole frequency ω _p in equation (10). By decreasing the capacitance C _SB , the pole frequency ω _p increases. This improves the frequency characteristic of IGFET Q in the common-drain configuration of FIG. 51C.

Similar to what occurs through the drain-body capacitance C _DB in the common-source configuration of FIG. 51A, the parasitic source-body capacitance C _SB is in parallel with the load capacitance C _L in the common-drain configuration, as shown in FIG. 51C. Thus, reducing the source-body capacitance C _SB advantageously reduces its loading effect in the common-drain configuration.

FIG. 52 illustrates a small signal model of the shorted-output version of the common-source amplifier configuration of FIG. 51A. In the small signal model of FIG. 52, the drain electrode D of IGFET Q is electrically shorted to the source electrode S. FIG. FIG. 53 shows a small signal equivalent circuit of the model of IGFET Q in FIG. 52. Element 448 in FIG. 53 is a voltage-controlled current source. Items v _gs , i _i and i _o in FIGS. 52 and 53 are each a small signal gate-source (input) voltage, a small signal input current, and a small signal output current.

The cut-off frequency f _T of the IGFET is defined as the value of the frequency f at which the absolute value of the current gain A _I of the shorted-output common-source configuration of the IGFET falls to one. In other words,

The cut-off frequency f _T comes from the small signal equivalent circuit of FIG.

The capacitance C _GB in equation (12) is the parasitic capacitance between the gate electrode G and the body region of the IGFET outside the active region occupied by IGFET Q.

In general, increasing the transconductance g _m of an amplified IGFET improves its performance capacity because its voltage gain generally increases. Since the cutoff frequency f _T increases at the same time as the transconductance g _m increases according to equation (12), the increase in the cut-off frequency f _T is an indication of improved IGFET performance.

In a typical long channel model of IGFET Q with a source resistor R _S of zero, the transconductance g _m is:

Where W is the IGFET width, L is the channel length, μ _n is the electron mobility, and C _GIa is the gate dielectric capacitance per unit area. In the short channel speed saturation model of IGFET Q, the transconductance g _m is:

Where _nsat is the electron saturation rate since IGFET Q is an n-channel device. The examination of equations (13) and (14) shows that the transconductance g _m is proportional to the local gate dielectric capacitance C _GIa in both the long channel and short channel models.

For a typical long channel model of saturated IGFET Q, the capacitances C _GS , C _GD , and C _GB are:

Where L _GSoverlap and L _GDoverlap are the longitudinal distances at which the gate electrodes overlap the source and drain of IGFET Q, respectively. The term WL _GSoverlap C _GIa is a parasitic capacitance resulting from the gate electrode overlapping the source. The term WL _GDoverlap C _GIa is a parasitic capacitance resulting from the gate electrode overlapping the drain. Introducing equations (15) through (17) into equation (12) yields the cut-off frequency f _T for the ideal long channel IGFET in saturation.

Equations (15) and (16) are not expected to be accurate for the asymmetric IGFET of the present invention due to the asymmetric longitudinal dopant classification in its channel zone. However, equations (15) and (16) can be used as trend indicators to calculate parasitic capacitances C _GS and C _GD to evaluate the cut-off frequency f _T of current asymmetric IGFETs. More accurate values of capacitances C _GS and C _GD can be determined by computer simulation.

The cut-off frequency f _T semantically entails a short-circuit condition at the output in the common-source configuration. As a result, the frequency f _T essentially eliminates the effect of the parasitic drain-body capacitance C _DB . In addition, the frequency f _T does not cause the effect of parasitic source-body capacitance C _SB because it utilizes a common-source configuration.

The cut-off frequency f _T has a peak cut-off value f _Tpeak according to the operating current, ie the drain current I _D. Although the peak cut-off value f _Tpeak is useful for evaluating high frequency IGFET performance, the circuit typically operates at frequencies 10 to 20 times lower than the peak value f _Tpeak . In addition to the preferably high peak value f _{Tpeak for IGFETs} , it is generally desirable to reduce the change in cut-off frequency f _T while reducing the operating current below the operating current level corresponding to the peak value f _Tpeak .

Usually, source-body such as pn junctions 110 and 112 at novel IGFETs (100, 140, 150, 160, 170, 180, 190, 210, 100V, 140V, 150V, 160V, 170V, 180V and 190V) and The drain-body junction is reverse biased. When the pn junction is in the reverse bias state, the depletion region along the junction shows the capacitance C _da of the small signal region as follows.

Where ε ₀ is the absolute permittivity, K _S is the relative permittivity of the semiconductor material, and t _d is the voltage-dependent thickness of the depletion region.

For pn junctions formed along a uniformly doped substrate, the thickness t _d for the depletion region of this ideal pn junction is:

Where V _R is the applied reverse voltage, V _BI is the built-in voltage of the junction, q is the charge, and N _B0 is the uniform background dopant concentration in the substrate. The built-in voltage V _BI changes with the background dopant concentration N _B0 in accordance with the following relationship.

Where k is Boltzmann's constant, T is temperature and n _i is the intrinsic carrier concentration.

54 illustrates a p-type substrate material of a model of a pn junction that may have any one of three basic types of dopant profiles in a p-type substrate where N _D and N _A are absolute donor and acceptor dopant concentrations, respectively. It shows how it changes with distance y. In addition, the bonding model is shown in FIG. 54. As indicated by the bonding model shown, the p-type material is thicker than the n-type material of the bond, and thus is doped at a lower concentration. Each of curves 450, 452, and 454 in FIG. 54 exhibits hypobrupt, flat, and hyperabrupt dopant profiles in the p-type material. The distance y _d represents the thickness of the p-type portion of the depletion region along the junction.

Roughly, the hypobolite profile curve 450 is a novel n-channel IGFET (100, 140, 150, 160, 170, 180, 190, 210, 100V, 140V, 150V, 160V, 170V, 180V and 190V). In each case a vertical dopant profile below the drain-body junction 112 is shown. In IGFETs 150, 160, 180, 190, 210, 150V, 160V, 180V and 190V where drain 104 includes main drain portion 104M and side drain extension 104E, curve 450 is specifically defined. As a vertical dopant profile below the portion of the drain-body junction 112 along the bottom of the main portion 104M. For IGFETs 170, 180, 190, 170V, 180V and 190V where source 102 extends deeper than pocket portion 120 below the upper semiconductor surface, curve 450 is below source-body junction 110. Specifically, the vertical dopant profile, below the junction portion along the bottom of the main source portion 102M for each of the IGFETs 180, 190, 180V and 190V, is shown. Provided that the conductivity type is inverted, curve 450 further shows a vertical dopant profile along the bottom of main portion 264M of drain 264 in each of the new p-channel IGFETs 220, 220U and 220V. . The flat curve 452 represents the p-type material of the ideal pn junction covered by equations (18) through (20).

FIG. 55 shows how the parasitic region capacitor C _da in the depletion region changes with reverse voltage VR across the pn junction modeled in FIG. 54. Curves 460, 462, and 464 of FIG. 55 show the C _da changes for each of curves 450, 452, and 454 of FIG. 54. In particular, curve 462 shows the power of capacitance C _da for an ideal pn junction as determined from equation (18) using t _d data from equation (19) (and V _BI data from equation (20)). Qualitative change of the law.

Parasitic drain along drain-body junction 112 in a novel n-channel IGFET (100, 140, 150, 160, 170, 180, 190, 210, 100V, 140V, 150V, 160V, 170V, 180V or 190V) The body capacitance C _DB is approximately proportional to the depletion capacitance C _da of the region as represented in FIG. 55 by the curve 460 corresponding to the hypobolite profile curve 450 of FIG. 54. As shown in FIG. 55, the depletion capacitance C _da is either (a) in the abrupt-junction-profile (flat) curve 462 in FIG. 54 or (b) in the hyperabrupt-profile curve 454 in FIG. 54. Lower than the capacitance for the corresponding curve 464. Thus, a hypobolite under the drain-body junction 112 at each IGFET 110, 140, 150, 160, 170, 180, 190, 210, 100V, 140V, 150V, 160V, 170V, 180V or 190V. The vertical dopant profile causes the drain-body capacitance C _DB to be reduced. The same applies to the capacitance C _DB along the bottom of the drain 264 in the novel p-channel IGFET 220, 220U or 220V. In addition, the parasitic source-body capacitance C _SB is particularly characterized by the n-channel IGFETs 170, 180, 190, 170V, 180V and 190V in which the source 102 extends deeper than the pocket portion 120 below the upper semiconductor surface. Usually reduced.

Additionally, the local depletion capacitance C _da for curve 460 corresponding to hypobolite profile curve 450 in FIG. 54 is shown by comparing curve 460 in FIG. 55 with curves 462 and 464. Change according to the reverse voltage V _R more slowly than the curve 462 or 464 as shown. Thus, parasitic drain-body at each new IGFET (100, 140, 150, 160, 170, 180, 190, 210, 100V, 140V, 150V, 160V, 170V, 180V, 190V, 220, 220U or 220V) Capacitance C _DB reduces the change with reverse voltage V _R. This is advantageous because it requires a smaller amount of compensation to account for the change in capacitance C _DB . The same discussion applies in particular to parasitic source-body capacitance C _SB in IGFETs 170, 180, 190, 170V, 180V and 190V.

In further reviewing the hypobolite dopant profile under the drains 104 and 204, the selected distance along which the net dopant concentration N _B of the semiconductor material is sufficiently close to the junction from the first dopant-concentration value along the lightly doped side of the pn junction. Consider an extreme example of a hypobrupt junction profile that steps to a higher second dopant-concentration value in to affect parasitic capacitance along the junction. This example is modeled in FIG. 56 showing how the net dopant concentration N _B varies with the distance y from the junction.

The two-step pn-junction model of FIG. 56 is constructed in the manner described below. The lightly doped side of the pn junction is a p-type with a first value N _B0 for the distance extending from the junction to a distance y _d0 , where concentration N _B causes one step to increase to a second value N _B1 . It is formed of a material. Y distance _d0 is at least a distance y _d0 the past, the concentration N _B of the low value in the p- type material to do not significantly impact position to the parasitic capacitance according to the change in the bonding of the concentration N _B of the p- type material If it is in N _B0 , it is a position that constitutes a p-type boundary of the depletion region with a zero value of reverse voltage.

The depletion region along the junction in the model of FIG. 56 extends from distance y _d0 to maximum distance y _dmax as the reverse voltage V _R increases from 0 to some maximum value V _Rmax . Over the distance y _dmax , the net dopant concentration N _B can have any profile in the p-type material as shown in FIG. 56. As shown in FIG. 56, the concentration N _B of the more heavily doped n-type material along the junction is at a much larger uniform value than N _B1 .

Figure 56 is a high density value is N _B1 as a p- type dopant profile of the material from the usual y _d0 (missing a step change in concentration at distance _d0 y) in y value _d0 and of a conventional 0.2㎛ 3 × 10 ¹⁶ atoms / The range is up to 20 N _{B0 in} cm 3. If the modeled pn junction is a drain-body junction at depth y _D below the upper semiconductor surface, a step increase from N _B0 to N _B1 of concentration N _B results in depth y _D + y _d0 below the upper surface.

Depletion capacitance C _da of the region of the two-step pn junction of FIG. 56 is controlled by the differential equation as follows:

Integrating equation (21) provided that the depletion region extends to distance y _d0 when the reverse voltage V _R is 0 yields the following value for distance y _d0 :

Combining equations (18) and (22) produces the following results for the depletion capacitance C _da :

57 shows how the local depletion capacitance C is through the reverse voltage V _R for values of high concentration values N _B1 ranging from N _B0 (again, above the modeled junction) to 20 N _B0 , as determined in equation (23). _Shows whether _da changes. FIG. 57 shows that increasing the concentration ratio N _B1 / N _B0 changes the capacitance C _da more slowly with the voltage V _R. For this reason, it is desirable that the concentration ratio N _B1 / N _B0 be as high as reasonably possible in order to allow the parasitic capacitances C _DB and C _SD to change more slowly in the asymmetric IGFET of the present invention.

The local depletion capacitance is at the initial value C _d0a when the reverse voltage V _R is zero. Setting voltage V _R to 0 in equation (23) yields the following.

The initial depletion capacitance value C _d0a is, as expected, a typical value for an ideal pn junction at zero reverse voltage. According to equation (24), and the capacitor value C is _d0a reduced while reducing the low density value N _B0 according to a low value N _B0 root (square root). Low concentration values N to ensure low capacitance C _DB and C _SB at zero reverse voltage V _R , along with selecting a high value of concentration ratio N _B1 / N _B0 to have slow changes in parasitic capacitances C _DB and C _SB . _B0 must be low.

Computer simulation related to capacitance and frequency parameters

Keeping in mind the above information on capacitance and frequency parameters, small signal simulations were performed through the Medici simulator to characterize the junction capacitance of the novel structure A. 58A and 58B respectively show the short channel and long channel versions of structure A generated by the Medici simulator for junction capacitance characteristics. Items 470 and 472, respectively, represent source and drain contacts (or source and drain electrodes) in FIGS. 58A and 58B. Metal silicon compound layers 254 and 256 are included in contacts 470 and 472, respectively. Doping contours of the p-type body material (or region) 108 in FIGS. 58A and 58B, respectively, illustrate the classified properties of doping in the body material 108. The short channel IGFET of FIG. 58A had a gate length L _G of 0.15 μm. The gate length L _G for the long channel IGFET in FIG. 58B was 1.0 μm.

FIG. 59 illustrates the size-method and doping features of the present invention for the short channel implementation of novel structure A and in the short channel implementation of structure A in FIG. 59, which is substantially similar to the short channel implementation of structure A in FIG. 58A. Except, the parasitic drain-body capacitance C _DBw as a function of the drain-body voltage VDB for the short channel implementation of the reference structure B substantially corresponding to the dopant scheme is shown. The gate length L _G is 0.2 μm in FIG. 59 rather than 0.15 μm in FIG. 58A. Gate-source voltage V _GS was 0.9V for C _DBw of FIG. 59. As shown in FIG. 59, the drain-body capacitance C _DBw is significantly lower for this short channel version of the new structure A than for the corresponding short channel version of the reference structure B. In particular, the reviewed short channel version of the novel structure A was about 50% of the capacitance C _DBw for the reviewed short channel version of the reference structure B in the V _DS range of 0 to 2V.

FIG. 60 shows parasitic linear source-body capacitance C _SBw as a function of source-body voltage V _SB for the short channel implementation of structures A and B discussed in FIG. 59. The gate-source voltage V _GS was again 0.9V. As shown in FIG. 60, the source-body capacitance C _SBw is considerably lower for the examined short channel version of the new structure A than for the corresponding short channel version of the reference structure B. C _SBw reduction is reduced by a greater C _DBw does, the review team channel version of structure A is approximately 35 than the review team channel version of structure B in a 2.0V V _SB - 40% C _SBw low value, and The short-channel version of the reviewed structure B of reference structure B at 0 V _SB had a C _SBw value as low as 25-35%.

Since the overall p-type dopant in the source 102 is increased in understanding p-type pocket implantation, a small minor improvement in the source-body capacitance C _SBw for the short channel version of structure A discussed in FIG. 59 is expected. Also, because source 102 is shorted to body material 108, in many cases source-body capacitance C _SB is less important than drain-body capacitance C _DB . As desired, an additional reduction in source-body capacitance C _SB for the short channel version of structure A can be achieved by deepening well portion 116.

FIG. 61 shows the cut-off frequency f _T as a function of the direct drain current I _DW for the short channel implementation of structures A and B discussed in FIG. 59. 61 also shows the change in cut-off frequency f _T according to the direct drain current I _DW for the change A 'of the new short channel structure A. FIG. In FIG. 61, and in the current computer-simulated data for new structures A and A ', the curve representing the data for structure A' is to distinguish the data from the data marked with an empty circle for structure A. Marked with filled circles. Additional features of the novel novel structure A 'are described below in connection with FIGS. 63 and 64. As shown in FIG. 61, the cut-off frequency f _T was approximately the same for the simulated short channel structures A, A 'and B. FIG.

The change in cut-off frequency f _T of the direct drain current I _Dw is shown in FIG. 62 for the long channel version of the implementation of structures A, A 'and B in FIG. As shown in FIG. 62, the cut-off frequency f _T was approximately the same for the long channel version of the novel structures A and A '. Importantly, the frequency f _T for the long channel versions of the new structures A and A 'was significantly greater than the frequency for the long channel version of the reference structure B. Thus, the long channel versions of the novel structures A and A 'had better performance capacity than the long channel versions of the reference structure B.

An additional IGFET that is hypobrupt due to the vertical body-material dopant profile below the drain being the subsurface maximum of the well dopant concentration.

Figure 63 illustrates a short n-channel implementation 480 of the asymmetric IGFET structure A 'in accordance with the present invention. Doping contours for IGFET 480 generally represent structural details as well as shown in FIG. 63 as a function of depth y and longitudinal distance x from the source location. The source location for measuring the distance is approximately 0.35 μm from the center of the channel zone.

IGFET 480 consists of n ++ main source portion 102M and n + lightly doped lateral extensions 102E as source 102 in IGFET 480 as in long channel IGFET 150 of FIG. Except for the configuration, generally similar to the short channel IGFET 140 of FIG. This enables reduction of source resistance R _S at IGFET 480, thus improving its analog performance. As in IGFETs 140 and 150, p-type pocket portion 120 extends deeper than source 102 below the upper semiconductor surface. Drain depth y _D for IGFET 480 is approximately 50% deeper than source depth y _S.

64 illustrates a net dopant along the upper semiconductor surface as a function of the longitudinal distance x from the previous source location for the novel structure A of FIG. 39 and the novel structure A ′ of FIG. 63 (ie, IGFET 480). The concentration N _N is shown. As in FIGS. 12C and 14C, the curve segments 106 ^* and 120 ^* represent the concentration N _N of the net p-type dopant in the respective regions 106 and 120, and the curve segments 102M ^* , 102E ^* , 104M ^* , 104E ^*, and 104 ^* ) represent the concentrations N _N of the net n-type dopants in the respective regions 102M, 102E, 104M, 104E, and 104. Although only marked as open circles (empty circles), curve segments 102M ^* , 102E ^* , 106 ^* and 120 ^* apply to both structure A and structure A '.

As the curve segments 104 ^* and 102M ^* of FIG. 64 show, the short channel IGFET 480 of the novel structure A 'has a slightly lower maximum at the drain 104 along the upper surface than at the main source portion 102M. Net dopant concentration is reached. More specifically, the maximum value of net dopant concentration N _N at drain 104 along the top surface of IGFET 480 is usually 20-50 of the maximum top-surface value of concentration N _N at main source portion 102M. %, Typically 30-40%. FIG. 64 shows an example where the maximum top-surface value of the concentration N _N in the drain 104 exceeds 1 × 10 ²⁰ atoms / cm 2, but the maximum top in the drain 104 of the short channel IGFET 480. The surface N _N concentration is easily considerably lower (eg, from 5 × 10 ¹⁹ atoms / cm 2 to 1 × 10 ¹⁹ atoms / cm 2 or less) based on the maximum top-surface N _N concentration in the main source portion 102M. Can be. In addition, drain 104 in IGFET 480 extends slightly deeper below the top surface than main source portion 102M. In essence, in a IGFET, such as IGFET 150, whose source consists of the main portion and the lightly doped lateral extension, the two-part drain formed from the main portion and the lightly doped lateral extension is deeper in the IGFET 480. It is replaced by a lightly doped drain. Reduced doping in drain 104 of IGFET 480 results in a lower electric field in drain 104 and enables IGFET 480 to operate from an electric field scale where undesirable drain impact ionization occurs.

FIG. 65 shows a computer simulation of an additional symmetrical reference structure D, a reference structure B and an IGFET of the novel structure A ', which lacks both halo pocket portions of structure B but is substantially the same size- and dopant-type as structure B. FIG. The threshold voltage V _T as a function of the gate length L _G for. The gate dielectric thickness t _GI was 4.0 nm in the simulation of FIG. 65.

As FIG. 65 shows, the threshold voltage roll-off has been converted to a lower threshold voltage V _T in the novel structure A 'than in the reference structure B or D. FIG. In addition, FIG. 65 shows that the novel structure A 'produced less undesirable inverted short channel effects than the reference structure B or D. FIG. That is, the novel structure A experienced less (typically less) change in the threshold voltage V _{T as} the gate length L _G was increased in the longer channel domain than the reference structure B or D. Thus, structure A had better short channel and longer channel characteristics than structure B or D.

Fabrication of Additional IGFETs

The n-channel IGFET 480 that implements the asymmetric IGFET structure A 'of FIG. 63 is typically subject to appropriate modifications in the n-type source / drain extension and main source / drain implantation steps and additional masking steps and Subject to the use of an associated ion implantation operation, it is fabricated in accordance with the present invention according to the steps used to fabricate an asymmetric n-channel IGFET 210 in the process of FIG. This difference is described below with respect to IGFET 480 as appropriate using the same reference signs employed to describe the fabrication of IGFET 210.

In particular, n + precursor source extension 102EP is defined for IGFET 480 in the stage of FIG. 31L without defining the corresponding n + precursor drain extension for IGFET 480. This is because, in other cases, the precursor drain extension has an opening above the location for the precursor source extension 102EP with respect to the IGFET 480 but may be formed for the IGFET 480 above the location of the photoresist mask 422. It involves the step of constructing the extension. In so doing, photoresist 422 is precisely aligned with respect to precursor gate electrode 128P for IGFET 480. N-type source / drain extension implantation is performed as described above with respect to FIG. 31L after photoresist 422 is removed. Because photoresist 422 marked the position for precursor drain extension for IGFET 480, precursor source extension 102EP is formed for IGFET 480 without forming a corresponding precursor drain extension.

After the stage of FIG. 31Q, the photoresist mask 434 is configured to extend above the position for the drain 104 of the IGFET 480 but has an opening above the position for the main source zone 102M of the IGFET 480. do. Photoresist 434 is precisely aligned with respect to precursor gate electrode 128P of IGFET 480. The n-type main source / drain implantation is performed as generally described above with respect to FIG. 31Q after the photoresist 434 is removed. Because photoresist 422 masks the location for drain 104 of IGFET 480, main source zone 102M is defined for IGFET 480 without yet defining drain 104. The portion of the precursor source extension 102EP outside the main source zone 102M constitutes the source extension 102E. Through the precursor gate electrode 128P portion of the IGFET 480 exposed during the implantation, an n-type main source / drain dopant was introduced into the exposed portion of the electrode 128P.

Additional photoresist masks (not shown) with openings above the intended locations in source 102 of IGFET 480 include gate sidewall spacers for dielectric layers 430 and 432 and IGFET 210, and gate sidewall spacers ( 290, 292, 330, 332, 370 and 372. Additional photoresist is precisely aligned with respect to precursor gate electrode 128P of IGFET 480. The n-type drain dopant is ion implanted into the bottom monosilicon to a very high figure through the uncovered portion of the surface dielectric layer to define the n ++ drain 104 of the IGFET 480. The illustration of the n-type drain dopant used to define the drain 104 of the IGFET 480 is very high, while the illustration of the n-type drain dopant is used to define the main source zone 102M of the IGFET 480. It is lower than the very high metropolis of -type main source / drain dopant. As a result, the drain 104 of the IGFET 480 is doped at a lower concentration than the main source zone 102M.

The n-drain implant for IGFET 480 is performed under conditions where the drain 104 extends deeper than both the main source zone 102M and its precursor source extension 102EP below the upper semiconductor surface. For example, n-type main source / drain injection and n-type drain injection for IGFET 480 may be performed through one of the same n-type dopants, arsenic or antimony. In this case, the n-type drain implant for IGFET 480 is performed at a higher implantation energy than the n-type main source / drain implant. Alternatively, two implants may be performed using different n-type dopants in which the n-type drain dopant of IGFET 480 has a lower molecular weight than the n-type main source / drain dopant. In one example, arsenic is the main source / drain dopant and phosphorus is the n-type drain dopant of IGFET 480. The implanted energies are closer to each other in this case than in the case mentioned above. However, the range of n-type drain dopant is larger than the range of n-type main source / drain dopant in both cases. Additional photoresist is removed after n-type drain injection to IGFET 480.

The portion of precursor gate electrode 128P for IGFET 480 covered during main source / drain injection was heavily exposed during n-type drain injection for IGFET 480. This enabled the n-type drain dopant for IGFET 480 to enter the portion of electrode 128P covered during n-type main source / drain injection. As a result, substantially all of the precursor gate electrode 128P of IGFET 480 is heavily doped n-type. Thus, precursor gate electrode 128P of IGFET 480 becomes n ++ gate electrode 128.

The n-type drain implant for IGFET 480 may be performed before rather than after the n-type main source / drain implant. In any case, as described above, the rest of the manufacture of IGFET 480 is performed as described above for IGFET 210.

When the IGFET 210 is present in this semiconductor structure, the configurations of the photoresist masks 422 and 434 on the positions intended for the IGFET 210 are the same as described above with respect to Figs. 31L and 31Q, respectively. The formation of IGFET 480 does not affect the formation of IGFET 210.

Additional Complementary-IGFET Architecture Suitable for Mixed-Signal Applications

66 illustrates a variation of the complementary-IGFET structure of FIG. 29.1, in accordance with the present invention. The complementary-IGFET structure of FIG. 66 is particularly suitable for mixed-signal applications. The main structural difference between the complementary-IGFET structures of FIGS. 29.1 and 66 is that the complementary-IGFET structure of FIG. 66 is created from the starting structure, such as a bonded wafer.

In the complementary-IGFET structure of FIG. 66, the subsurface electrical isolation layer 482, which is typically composed of silocon oxide, is an island 202 and 204 separated by a field-isolation region 200 below the upper semiconductor surface. And separates the lower semiconductor layer 484 from the upper semiconductor layer. Usually, the lower semiconductor layer 484 is composed of monosilicon (either p-type or n-type). 66 shows an example where the lower semiconductor layer 484 is lightly doped p-type. Electrically isolated extensions 486, which are typically made of trench type and typically made of silicon oxide, extend from the field-isolation region 200 to the subsurface isolation layer 482. Field isolation 200 and isolation extension 486 laterally surround islands 202 and 204 such that they are completely insulated from each other with electrical insulators.

Usually, islands 202 and 204 are composed of doped <100> monosilicon. Island 202 has a low and substantially uniform n-type background dopant concentration imposed with a low but slightly high substantially uniform p-type background concentration typically provided by p-type semiconductor dopant aluminum. As a result, the portion of island 202 that does not accept any other dopant (p-type or n-type) is lightly doped p-type. In brief, island 204 has a low substantially uniform n-type background dopant concentration.

Island 202 provides monosilicon for the change 210W of the long n-channel IGFET 210. Source 102 and drain 104 of long n-channel IGFET 210W are p-type body material 108 composed of lightly doped lower portion 488, p + well portion 116 and upper portion 490. Are separated by the channel portion. Each of the p- lower body-material portion 488 and the p-type upper body-material portion 490 is each a p- lower body material portion 114 and a p-type upper body-material portion 118 of the IGFET 210. Corresponds to. Upper body-material portion 490 of IGFET 210W is lightly doped residue 492 corresponding to source-contacting p + pocket portion 120 and p-top body-material remainder 124 of IGFET 210. It consists of. Due to the imposition of a low p-type background dopant concentration on the lower n-type background concentration in island 202 of IGFET 210W, the net dopant concentration N _N in the bulk of each region 488 or 492 is p-type and There is a difference between the n-type background dopant concentration.

Except for the aforementioned structural differences and the presence of two background dopant concentrations in island 202, n-channel IGFET 210W is configured and set substantially the same as n-channel IGFET 210. The p− lower body-material portion 488 may be removed so that the p + well portion 116 extends downward into the subsurface isolation layer 482.

Island 204 provides monosilicon for change 220W of long p-channel IGFET 220. Source 262 and drain 264 of long p-channel IGFET 220W are top portion 496, n + well portion 276 corresponding to n-type upper body-material portion 278 of IGFET 220. And a lightly doped lower portion 494. The upper body-material portion 496 of the IGFET 220W is a lightly doped residue 498 corresponding to the n- upper body-material remainder 284 of the IGFET 220, and the source-contact n + pocket portion 280 ) Unlike IGFET 220, IGFET 220W does not have a low p-type background dopant concentration and does not use an n-type compensation dopant to ensure that the entire body-material portion 496 is n-conducting. . The net dopant concentration NN in each region 494 or 498 is simply the n-type background dopant concentration.

Except for the aforementioned structural differences and the presence of an n-type compensation dopant to ensure that the entire body-material portion 496 is n-type, the p-channel IGFET 220W is equal to the p-channel IGFET 220. It is configured and set up substantially the same. The n− lower body-material portion 494 may be removed so that the n + well portion 276 extends downward into the subsurface isolation layer 482.

Fabrication of Additional Complementary-IGFET Structures

The complementary-IGFET structure of FIG. 66 is provided by the method described below in accordance with the present invention. A structure is first provided in which a subsurface isolation layer 482 is sandwiched between the lower semiconductor layer 484 and the upper semiconductor region consisting of a low uniform dopant concentration of n-type monosilicon. This initial structure can be created by, for example, bonding two semiconductor wafers together through an electrically insulating material forming subsurface isolation layer 482. One of the wafers provides a <100> n-type monosilicon for the upper semiconductor region. The other wafer provides a lower semiconductor layer 484 that is typically constructed of monosilicon (either p-type or n-type as in the illustrated example).

Isolation extensions 486 are formed inside the n-top semiconductor region in accordance with deep trench-insulation techniques. A field-isolation region 200 is then formed along the outer (top) surface of the n-top semiconductor region according to shallow trench-insulation techniques to define islands 202 and 204. Using a photoresist mask with openings over the island 202, the p-type semiconductor dopant, which is usually composed of aluminum, is weakly high enough to convert the entire material of the island 202 into a low net concentration p-conductor. Is introduced into the island of Ireland (202). When aluminum is used to perform p-type doping of island 202, aluminum diffuses relatively quickly throughout island 202 so that it becomes a substantially uniformly doped p-type in a relatively short time.

p + well portion 116 and n + well portion 276 are formed in islands 202 and 204, respectively, in the manner described above in connection with the manufacture of IGFETs 210 and 220. A portion of island 202 lies under well 116 and constitutes p-bottom body-material portion 488. Similarly, a portion of island 204 lies under well 276 and constitutes n-bottom body-material portion 494. Regions 102, 104, 120, 126, 128, 250, 252, 254, 256, and 258 for IGFET 210W and regions 262, 264, 280, 286, 288, 290, 292 for IGFET 220W , 294, 296 and 298 are formed as described above for IGFETs 210 and 220. The p-type bonosilicon on top of well portion 116 may have a p-type upper body-material portion 490 where the portion outside p + halo pocket portion 120 constitutes the p- upper body-material remainder 492. Configure. The n-type monosilicon on top of well portion 276 may form p-type complementary body-material portion 496 where a portion outside n + halo pocket portion 280 constitutes n-top body-material remainder 498. Configure.

IGFET with vertical burst of vertical body-material dopant profile under drain due to step change in body-material dopant concentration

In asymmetric IGFETs constructed in accordance with the present invention, the vertical dopant profile below the drain is sharply at least 1/10 as it proceeds from the location of the maximum well dopant concentration to the drain, rather than having a concentration N _T of the conductivity-defined dopant in the body material. It can be made into a hyperabrupt in a decreasing manner. In particular, the vertical dopant profile below the drain includes: (a) the conductive-definition dopant at a first uniform concentration, (b) the conductive-definition dopant at the drain-adjacent portion- It can be made with hypobolite by placing the body material under the drain, including the drain-remote portion just below it at a substantially uniform second concentration that is significantly greater than the concentration of the positive dopant (usually at least 10 times larger).

Thereafter, the concentration of the conductivity-defined dopant experiences a step decrease, usually at least 1/10, as it proceeds from the drain-remote body-material portion to the drain through the drain-adjacent body-material portion. The n-channel IGFET provided with this second type of hypo-blow dopant dopant profile under drain according to the asymmetric channel-zone doping feature of the novel structure A or A 'is generally referred to herein as the novel structure E.

In general, FIG. 67 shows the vertical dopant profile through the drain to the underlying body region for the n-channel IGFET configured as structures A / A ', B and E. FIG. More specifically, the change in absolute dopant concentration NT as a function of depth y along the vertical line through the drain is shown in FIG. 67 for the n-channel IGFET of structures A / A ', B and E, respectively. The concentration of the Likewise, the entire p- type dopant along the vertical line through the drain of n- channel IGFET of structure E as that shown in FIG. 56 N _T are the same depth and drain depth y _D y + distance _d0 from the upper semiconductor surface _ST y The uniform concentration value is at N _B0 . Thus, the concentration N _T of the p-type dopant in the drain-adjacent body-material portion extending the distance y _d0 from the depth y _D to the depth y _ST is N _B0 . The distance y _d0 is usually 0.05-1.0 μm and typically 0.1 μm.

In depth y _ST, absolute dopant concentration N _T makes a transition to step usually at least 10 times larger than N N _{B1 B0} from N _B0. Drains extending downward from the depth y _ST-body of the remote-concentrations of p- type dopant in the material portion _T is N, the concentration of p- type dopant in the body material of the drain-body junction characterized in, in particular the drain- It is at the value N _B1 beyond some depth beyond the line which does not have any significant effect on the body capacitance C _DB . Therefore, the concentration of the N _T p- type dopant in the body material, the concentration of p- type dopant N _T and N _B1 same drain-remote body-material portion from the concentration of p- type dopant is equal to N _T N _B0 While traversing to the drain-adjacent body-material portion, a step reduction is usually produced by at least about 1/10, and then maintained at N _B0 until the drain-body junction.

68A shows an asymmetric long n-channel IGFET 500 constructed in accordance with the present invention such that structure E is particularly suited to high speed analog applications. IGFET 500 is illustrated in that the p-type body material 108 is comprised of an upper surface-adjacent portion 504 extending to the upper semiconductor surface and a heavily doped lower subsurface portion 502. Arranged substantially the same as in IGFET 170 of 18a. The p + subsurface body-material portion 502 lies under the source 102, the drain 104, and the channel zone 106. The upper boundary (top) of the subsurface body-material portion 502 is at a depth y _ST below the upper semiconductor surface. The depth y _ST is usually 10 times or less, preferably 5 times or less, of the drain depth y _D. In the portion closest to the source 102 and the drain 104, the subsurface portion 502 is usually 10 times or less deeper than the source 102 and the drain 104 below the upper semiconductor surface, and preferably 5 Deeper than pear

The p-type surface-adjacent body-material portion 504 lies below the p + subsurface body-material portion 502 and faces the p + subsurface body-material portion 502. Channel zone 106 is part of surface-adjacent body-material portion 504. Here, the p + pocket portion 120 that is shallower than the source 102 is the portion of the surface-adjacent body-material portion 504. Currently, item 124 in FIG. 68A is a lightly doped material of surface-adjacent body-material portion 504, ie, a segment of portion 504 outside pocket portion 120.

The p-type dopant in the segment of the surface-adjacent body-material portion 504 underlying the drain 104 is present at a substantially uniform concentration equal to N _B0 . Typical values of the concentration N _B0 are 5x10 ¹⁵ atoms / cm 3. The p-type dopant in the surface body-material portion 502, thus underlying the previous segment of the surface-adjacent body-material portion 504 and thus underlying the drain 104, is present at a substantially uniformly higher concentration equal to NB1. . Value NB1 is normally at least 10 times the N _B0, preferably at least 20 times the N _B0, and more preferably at least 40 times the N _B0, typically about 100 times that of N _B0.

68B shows another asymmetric long n-channel IGFET 510 configured in accordance with the present invention such that structure E is particularly suited to high speed analog applications. IGFET 510 includes IGFET 500 except that subsurface electrical isolation layer 512, consisting largely of silicon oxide, typically contacts subsurface body-material portion 502 along its bottom surface. Are arranged identically. In IGFET 510, the p-type dopant in a segment of subsurface body-material portion 502 underlying drain 104 down to depth subsurface isolation layer 512 from depth y _ST is largely uniform with concentration N _B1 . Doped.

Understanding of the hypobolite vertical dopant profile under drain 104 in body material 108 underlying IGFETs 500 and 510 is shown in FIGS. 69A-69C (collectively "FIG. 69"), FIG. 70A. To FIG. 70C (collectively "FIG. 70"), and FIGS. 71A to 71C (collectively "FIG. 70"). FIG. 69, generally similar to FIG. 8, shows an exemplary dopant concentration along vertical line 130 through source 102 of IGFET 500 or 510. Exemplary dopant concentrations along vertical lines 132 and 134 through channel zone 106 of IGFET 500 or 510 are shown in FIG. 70, which is generally similar to FIG. 9. 71, which is generally similar to FIG. 10, shows an exemplary dopant concentration along the vertical line 136 through the drain 104 of the IGFET 500 or 510.

69A, 70A, and 71A show source 102, drain 104, subsurface body-material portion 502, surface-adjacent body-body, along vertical lines 130, 132, 134, and 136. The concentration N _I of the individual semiconductor dopant forming the pocket portion 120 of the material portion 504 and the remainder 124 of the portion 504 is shown. The concentrations N _T of total p-type dopants and total n-type dopants in regions 102, 104, 502, 120, and 124 along vertical lines 130, 132, 134, and 136 are shown in FIGS. 69B, 70B, and 71B. Is shown. 69C, 70C, and 71C show net dopant concentrations N _N along lines 130, 132, 134, and 136.

The curves / curve segments 102 ', 102 ", 102 ^* , 104', 104", 104 ^* , 120 ', 120 ", 120 ^* , 124', 124" and 124 ^* in FIGS. 69-71 are Each has the same meanings described above in connection with FIGS. 8 to 10. Curve 502 'in FIGS. 69A, 70A, and 71A shows the concentration of n-type dopant used to form subsurface body-material portion 502 along vertical lines 130, 132, 134, and 136. N _I is represented. Subsegments 502 "in FIGS. 69B, 70B, and 71B show the concentration N _T of the total n-type dopant in subsurface portion 502 along lines 130, 132, 134, and 136. Curve segment 502 ^* in 69c, 70c and 71c represents the concentration N _N of the net n-type dopant in portion 502 along vertical lines 130, 132, 134 and 136. FIG.

Returning to FIG. 71A, the p-type dopant in the portion of the body material 108 below the drain 104 of the IGFET 500 or 510 is referred to herein as a "lower" p-type dopant and a "top" p-type dopant. It has two main components referred to. The lower p-type dopant is at a high fixed concentration N _B1 in the subsurface body-material portion 502, represented by the curve segment 502 ′. The upper p-type dopant is at a low defined concentration N _B0 in the remainder 124 of the surface-adjacent body-material portion 504, as indicated by curve 124 ′. In addition, the upper p-type dopant is present inside drain 104 as indicated by the extension of curve 124 'to the area surrounded by curve 104'.

The entire p-type dopant in the portion of the body material 108 below the drain 104 of the IGFET 500 or 510 is represented by the combination of curve segments 502 ″ and 124 ″ in FIG. 71B. As shown by the variation in the combined curve 502 "/ 124", the concentration N _T of the total p-type dopant in the portion of the body material 108 below the drain 104 is substantially the concentration N _B1. the subsurface body in-material portion 502, the upper body in the density N _B0 in-the crossing of a material remainder 124 when experiencing step decreases and then moved further upward to the drain (104) to maintain a concentration N _B0 . Usually, since the high temperature N _B1 is at least 10 times N _B0 , the concentration N _T of the total p-type dopant in the portion of the body material 108 under the drain 104 is from the sub body-material portion 502. As it moves upward through the upper body-material remainder 124 to the drain 104, it decreases to at least 1/10 in a hypobolite manner.

As described above, the high-concentration N value _B1 is more preferably preferably at least 20 times the N _B0 and at least 40 times the N _BO. Therefore, the hypobolite reduction in the concentration N _T of the total p-type dopant in the portion of the body material 108 below the drain 104 is preferably at least 1/20, more preferably at least 1/40.

71C shows the net p-type dopant in the portion of body material 108 under drain 104 in IGFET 500 or 510, as represented by a combination of curve segments 502 ^* and 124 ^* . N _N concentration is, the drain 104 in the portion of body material 108 below the entire p- type dopant concentration of the drain depth y N _{T D} (in other words, the drain-body junction 112) that drops to zero at Except that, similarly fluctuate vertically to the concentration N _T of the total p-type dopant in the portion of the body material 108 below the drain 104. As with the aforementioned IGFETs of the present invention, the hypobolite dopant profile in the portion of the body material 108 below the drain 104 reduces the parasitic capacitance along the drain-body junction 112 of the IGFET 500 or 510. Decrease. Thus, increased analog speeds for IGFETs 500 and 510 are achieved.

Using the vertical dopant distribution below the source 102 of the IGFET 500 or 510, the curve segments 502 ′ and 124 ′ in FIG. 69A have substantially the same shape as in FIG. 71A. Although curve 120 ′ is shown in FIG. 69A, the portion of body material 108 under source 102 because p-type pocket portion 120 is shallower than source 102 in the examples of FIGS. 68A and 68B. The total p-type dopant in represents the upper p-type dopant and the lower p-type dopant at concentrations N _B0 and N _B1 , respectively. The portion of the combination curve segments 502 "and 124" at a depth greater than the source depth y _S in FIG. 69B is the portion of the combination curve segment 502 "/ 124" at a depth greater than the drain depth y _D of FIG. 71B. It is shaped substantially the same. Thus, the source 102, the concentration of p- type dopant in the portion of body material 108, the body material 108 below, the total concentration of the N _T p- type dopant in the portion of the drain 104 of the following N _T In the same way as in the change to hypobolite. As a result, the parasitic capacitance along the source-body junction 110 is reduced to further enhance the analog performance of the IGFET 500 or 510.

Similar to what occurs in IGFET 100 of FIG. 6, p-type pocket portion 120 in IGFET 500 or 510 extends deeper than source 102 and drain 104 below the upper semiconductor surface. It can be modified. In this case, the p-type pocket dopant of the pocket portion 120 causes the concentration N _T of the total p-type dopant in the portion of the body material 108 below the source 102 to be directly under the source-body junction 110. And slightly rise above, causing it to be slightly larger than N _B0 just below the bottom of the source 102. The parasitic capacitance along the source-body junction 110 is larger in the example of FIGS. 68A and 68B, but is still reduced, depending on the depth of the pocket 120 and the appropriate choice for doping. This enhances the analog performance of the IGFETs 500 and 510. Deforming the pocket portion 120 below the upper semiconductor surface to extend deeper than the source 102 and drain 104 is substantially because the p-type pocket dopant is not located inside the drain 104. There is no significant effect on the drain characteristics of 500 or 510.

Channel zone 106 of IGFET 500 or 510 is doped asymmetrically in the lengthwise manner in the same manner as in channel zone 106 of IGFET 170 in FIG. 18A. Since the dopant distribution of FIG. 7 and related information described above with respect to FIG. 7 are applied to IGFET 170, this information is generally applied to IGFETs 500 and 510. Thus, punchthrough in IGFETs 500 and 510 is avoided. The channel length of the IGFET 500 or 510 can be sufficiently reduced to convert it to a short channel device. In this case, the surface dopant distribution of FIG. 12 and the related information described above with respect to FIG. 12 apply generally to IGFETs 500 and 510.

Each S / D zone 102 or 104 of the IGFET 500 or 510 can be modified to make up the main portion 102M or 104M and the lightly doped lateral extensions 102E or 104E. Alternatively or additionally, each S / D zone 102 or 104 of IGFET 500 or 510 may include a lightly doped lower portion 102L or 104L. In this case, the dopant distributions shown in FIGS. 14, 16, and 17 and the relevant information about such dopant distributions are shown in FIGS. 16 and 17, for the curves / curve segments 116 'and 114', 116 "and 114". , And 116 ^* and 114 ^* ) are generally applied to IGFETs 500 and 510 with conditions replacing curves / curve segments 502 ′, 502 ″ and 502 ^* , respectively.

Additional complementary-IGFET structure suitable for mixed-signal applications and with step changes in body-material dopant concentration

72A shows another complementary-IGFET structure constructed in accordance with the present invention, which is particularly suited for mixed-signal applications. The complementary-IGFET structure of FIG. 72A is produced from a doped silicon material, such as a bonded wafer structure. The patterned field region 520 of the electrically insulating material extends along the upper portion of the silicon material to define a group of laterally separated semiconductor islands that include islands 522 and 524. Two asymmetric long channel IGFETs 530 and 540 are formed along the upper semiconductor surface at the locations of islands 522 and 524, respectively.

IGFET 530 is an n-channel device implementing IGFET 510 of FIG. 68B. Source 102, drain 104, and channel zone 106 are located in island 522. Body material 108 of IGFET 530 is comprised of <100> p-type monosilicon. The lower semiconductor layer 550, composed of lightly doped p-type monosilicon, is placed under the isolation layer 512 and is in contact with the isolation layer 512 such that the layer is a subsurface layer. Typically, trench-type field-isolation regions 520 are vertically separated from subsurface isolation layer 512.

IGFET 540 is a p-channel device configured substantially the same as n-channel IGFET 500 of FIG. 68A with an inverted dielectric type. Thus, the IGFET 540 is in the channel zone 566 of the n-type body material 568 consisting of an upper surface-adjacent portion 574 extending to the upper semiconductor surface and a heavily doped lower subsurface portion 572. A heavily doped p-type source 562 and a heavily doped p-type drain 564 separated by a. Source 562, drain 564 and channel zone 566 are located in island 524.

Body material 568 is formed of <100> n-type monosilicon. Subsurface portion 572 of n-type body material 568 extends over p-bottom semiconductor layer 550, thus forming lateral pn junction 576 with semiconductor layer 550. Subsurface body-material portion 572 also forms a vertical pn junction 578 with p + subsurface body-material portion 502 of IGFET 530. Reverse bias is applied across the pn junction 578 to isolate IGFETs 530 and 540 from each other.

Highly doped pocket portion 580 of n-type surface-adjacent body-material portion 574 extends along source 562 of IGFET 540. The n + pocket portion 580 causes the channel zone 566 to be an asymmetric longitudinal dopant classified in a manner similar to the asymmetric longitudinal dopant classification of the channel zone 106 in the IGFET 530. Item 584 is a lightly doped n-type remainder of surface-adjacent body-material portion 574. Gate dielectric layer 586, which is typically composed of silicon oxide, overlies channel zone 566. Gate electrode 588 overlies gate dielectric layer 586 over channel zone 566. Gate electrode 588 extends partially over source 562 and drain 564. In the example of FIG. 72A, gate electrode 588 is composed of very heavily doped p-type polysilicon.

The n-type dopant in the subsurface body-material portion 572 is in large scale uniform concentration N _B0 ′. The n-type dopant in the segment of the surface-adjacent body-material portion 574 underlying the drain 564 is present in a segment of large uniform concentration N _B1 ′ that is greater than N _B0 ′. Concentration N analogy _B1 and N _B0, concentration N _B1 'is usually N _B0' of at least 10 times, preferably N _B0 'of at least 20 times, more preferably N _B0' at least 40 times the, typically N _B0 ′ is approximately 100 times. Thus, in the portion of the body material 568 underlying the drain 564, the IGFET 540 is a hypothesis generally having the same characteristics as the IGFET 530 has in the portion of the body material 108 below the drain 104. Have an abrupt dopant profile. Similarly, the vertical dopant profile in the portion of body material 568 underlying source 562 of IGFET 540 is substantially similar to the vertical dopant profile of the portion of body material 108 under source 102 of IGFET 530. Do. Thus, IGFET 540 has reduced parasitic capacitance along its drain-body and source-body junctions.

FIG. 72B illustrates a change in the complementary-IGFET structure of FIG. 72A. In the variation of FIG. 72B, the field-isolation region 520 is provided with an electrically isolated extension 590 (typically a trench type) that reaches the subsurface isolation layer 512. The combination of the field-isolation region 520 and the isolation extension 590 laterally surrounds the subsurface body-material portion 502 of the IGFET 530. This insulates the IGFETs 530 and 540 from one another with a dielectric.

In the complementary-IGFET structure of Figs. 72A and 72B, the conductivity type can be reversed. The resulting n-type body material and n-bottom semiconductor layer, respectively, along the p-type body material 108 and the p- lower semiconductor layer 550 are both n-type monosilicon. The p-type body material corresponding to n-type body material 568 is <110> p-type monosilicon.

FIG. 72C shows a further variation of the complementary-IGFET structure of FIG. 72A. FIG. 72D shows the corresponding change in the complementary-IGFET structure of FIG. 72B. 72C and 72D, the lower semiconductor layer 592 composed of lightly doped n-type monosilicon replaces the p- lower semiconductor layer 550. The n-type body material 568 of the p-channel IGFET 540 is formed of <110> n-type monosilicon rather than <100> n-type monosilicon in the complementary-IGFET structures of FIGS. 72C and 72D. . The p-type body material 108 for the n-channel IGFET 530 leads to the <100> p-type monosilicon in the complementary-IGFET structures of FIGS. 72C and 72D.

The conductivity type in the complementary-IGFET structure of FIGS. 72C and 72D can be reversed. In this case, the resultant p-type body material and p-bottom semiconductor layer corresponding to n-type body material 568 and n-bottom semiconductor layer 592, respectively, are both <100> p-type monosilicon. The n-type body material corresponding to the p-type body material 108 is <110> n-type monosilicon.

Fabrication of Additional Complementary-IGFET Structures with Step Changes in Body-Material Dopant Concentration

The complementary-IGFET structure of FIG. 72A is fabricated by the method described below in accordance with the present invention. (a) a subsurface semiconductor region composed of <100> p-type monosilicon heavily doped with high uniform concentration N _B1 is adjacent to the subsurface electrical isolation layer, and (b) low concentration with low uniform concentration N _B0 A surface-adjacent semiconductor region composed of doped <100> p-type monosilicon lies adjacent to and over the subsurface semiconductor region, and (c) a lower semiconductor composed of lightly doped <100> p-type monosilicon. The layer is first provided with a structure over and adjacent to the subsurface isolation layer. The lightly doped lower semiconductor layer constitutes the p− lower semiconductor layer 550.

The initial structure can be created by, for example, bonding two semiconductor wafers together through an electrically insulating material forming a subsurface isolation layer. One of the wafers has a lightly doped p-type monosilicon surface that forms the lower semiconductor layer 550. The other wafer is a heavily doped p-type monosilicon substrate and a low concentration overly doped, respectively, substantially uniformly doped at concentrations NB1 and NB0 to form a subsurface semiconductor region and a surface-adjacent semiconductor region, respectively. It has a <100> p-type monosilicon epitaxial layer doped with.

Field-isolation region 520 defines the outer (upper) surface of the p- surface-adjacent semiconductor region to define island 522 for IGFET 520 and island 524 for IGFET 540. Thus formed. Field isolation 520 is such that field isolation 520 is deep (but not completely passed) into p-type surface-adjacent body-material portion 504 of the completed complementary-IGFET structure as shown in FIG. 72A. May extend partially through the p- surface-adjacent semiconductor region. Alternatively, field isolation 520 may extend completely through the p− surface-adjacent semiconductor region and partially extend to the underlying p + subsurface semiconductor region. The portion of the p- surface-adjacent semiconductor region in island 522 constitutes a precursor to the p-type surface-adjacent body-material portion 504. Substantially, the portion underlying the p + subsurface semiconductor region constitutes the p + subsurface body-material portion 502.

At the position relative to island 524, the lower section of the subsurface isolation layer downward through the p− surface-adjacent semiconductor region, through the bottom section of the p + subsurface semiconductor region, and down to the p− lower semiconductor layer 550. Through the cavity is formed. The remaining portion of the subsurface isolation layer constitutes the subsurface isolation layer 512. Highly doped <100> n-type monosilicon is epitaxially at a uniform concentration N _B1 ′ over the exposed section of the lower semiconductor layer 550 to substantially form the n + subsurface body-material portion 572. Is grown. The lightly doped <100> n-type monosilicon is epitaxial with a uniform concentration N _B0 'in the cavity above the subsurface portion 572 to form the precursor into the n-type surface-adjacent body-material portion 574. It grows tactically. Body-material portion 572 and precursors to body-material portion 574 form island 524.

Gate dielectric layers 126 and 586 are connected to p-type surface-adjacent body-material portion 504 for IGFEt 530 and n-type surface-adjacent body-material portion 574 for IGFET 540. Each along the exposed (top) surface of the precursors. Gate electrodes 128 and 588 are formed over gate dielectric layers 126 and 586, respectively. The n ++ source 102, n ++ drain 104, and p + pocket portion 120 are formed inside the precursor for the surface-adjacent body-material portion 504. Thereafter, the remaining portion of the precursor for the body-material portion 504 substantially constitutes the portion 504 for the IGFET 530. The p ++ source 562, the p ++ drain 564, and the n + pocket portion 580 are similarly formed inside the precursor for the n-type surface-adjacent body-material portion 574. The remaining n-type portion of the precursor for the surface-adjacent body material portion 574 then substantially constitutes the portion 574 for the IGFET 540. Involved in forming gate electrodes 128 and 588, n ++ source 102, n ++ drain 104, p + pocket portion 120, p ++ S / D zones 562 and 564, and n + pocket portion 580 The operations may be performed in various orders.

The complementary-IGFET structure of FIG. 72B is except that isolation extension 590 to field-isolation region 520 is formed inside the p + subsurface semiconductor region during formation of field isolation 520. It is made according to the invention in the same manner as the complementary-IGFET structure of 72a.

72C and 72D show the complementary-IGFET structure of FIGS. 72A and 72 except that the lower semiconductor layer composed of lightly doped n-type monosilicon replaces the p- lower semiconductor layer 550. In accordance with the present invention in each manner identical to the complementary-IGFET structure of FIG. 72B. The lightly doped bottom semiconductor layer makes up the n- bottom semiconductor layer 592. The initial structure used to create the complementary-IGFET structure of FIG. 72C or 72D is a <110> n-type mono lightly doped, rather than a lightly doped <100> p-type monosilicon substrate. Except having a silicon substrate, it can be produced in the same manner as the initial structure used to create the complementary-IGFET structure of FIG. 72A or 72B.

The cavity is also formed through the p− surface-adjacent semiconductor region, through the bottom section of the p + subsurface semiconductor region, and through the bottom section of the subsurface isolation layer descending to the n− lower semiconductor layer 592. N + subsurface body-material portion 572 is epitaxially grown on the lower semiconductor layer 592 within the cavity as a heavily doped n-type monosilicon at a concentration N _B1 ′. Next, the precursor for the n- surface-adjacent body-material portion 574 is lightly doped to a concentration N _B0 ′ over the n + subsurface portion 572 inside the cavity as n-type monosilicon. It is grown epitaxially.

change

Although the present invention has been described with reference to specific embodiments, this detailed description is for illustrative purposes only and is not intended to limit the scope of the invention as claimed below. For example, silicon in the semiconductor body and / or gate electrodes 128, 288, 328, 368 and 588 may be replaced with other semiconductor materials. Alternative candidates include group 3a-group 5a alloys such as germanium, silicon-germanium alloys, and germanium arsenide.

Metal silicon along the top electrodes of the sources 102 and 562, the drains 104 and 564, and the gate electrodes 128 and 588 in the IGFETs 530 and 540 of the complementary-IGFET structure of FIGS. 72A-72D. Compound layers may be provided. Composite gate electrodes 128/258, 288/298, 328/338, and 368/378 in IGFETs 210, 220, 230, 240, 380, and 390 of the complementary-IGFET structures of FIGS. 29 and 30 And gate electrodes 128 and 588 in IGFETs 530 and 540 of the complementary-IGFET structure of FIGS. 72A-72D are provided in a substantially complete metal or silicon compound gate electrode to control its working function. It can be replaced with a gate electrode made of a substantially complete metal silicon compound (eg, a cobalt silicon compound or a nickel silicon compound) with a dopant. Accordingly, various modifications may be made by one skilled in the art without departing from the true scope of the invention as defined in the appended claims.

Claims (90)

A structure comprising a main field-effect transistor ("FET"),
The main field-effect transistor ("FET") is:
A main channel zone of a main body material of a semiconductor body having an upper semiconductor surface, wherein the main body material is sufficiently doped with a semiconductor dopant of a first conductivity type to become the first conductivity type;
A pair of primary sources / drains ("S / D") located in the semiconductor body along the upper semiconductor surface and laterally separated by the primary channel zone and of a second conductivity type opposite to the first conductivity type Zones;
A main gate dielectric layer overlying the main channel zone; And
A main gate electrode overlying the main gate dielectric layer above the main channel zone,
The main body material laterally extends below the main S / D zones,
The first conductivity type semiconductor dopant in the main body material is a major lower subsurface body-material location up to 10 times deeper below the upper semiconductor surface than a particular major S / D zone of the major S / D zones. and a primary FET having a concentration that decreases by at least 1/10 when moving upward from an underlying subsurface body-material location to the particular primary S / D zone. The method of claim 1,
The main lower subsurface body-material location comprises a main FET up to five times deeper below the upper main semiconductor surface than the particular main S / D zone. The method according to claim 1 or 2,
The main lower subsurface body-material location comprises a main FET underlying almost all of the main channel zone and the main S / D zones. The method according to claim 1 or 2,
The concentration of the first conductivity type semiconductor dopant in the main body material decreases to at least 1/20 as it moves upward from the main lower subsurface body-material position toward the particular main S / D zone. The containing structure. The method according to claim 1 or 2,
The concentration of the first conductivity type semiconductor dopant in the main body material is such that where the main channel zone meets the particular main S / D zone along the upper semiconductor surface, the main channel zone is directed to the upper semiconductor surface. And therefore a primary FET, which is lower than where it encounters the remaining major S / D zones of the major S / D zones. The method of claim 5, wherein
The concentration of the first conductivity type semiconductor dopant in the main body material is such that where the main channel zone meets the particular main S / D zone along the upper semiconductor surface, the main channel zone is directed to the upper semiconductor surface. A main FET, thus, at least 10 times lower than where it meets the remaining main S / D zones. The method of claim 5, wherein
The concentration of the first conductivity type semiconductor dopant in the main body material is such that where the main channel zone meets the particular main S / D zone along the upper semiconductor surface, the main channel zone is directed to the upper semiconductor surface. A main FET, thus, at least 20 times lower than where it meets the remaining main S / D zones. The method of claim 5, wherein
The particular major S / D zone includes a major FET extending deeper below the upper semiconductor surface than the remaining major S / D zones. The method according to claim 1 or 2,
The concentration of the first conductivity type semiconductor dopant in the main body material substantially experiences a step decrease as it moves upward from the main lower subsurface body-material position toward the particular main S / D zone. A structure containing the primary FET. The method according to claim 1 or 2,
Each major S / D zone comprises a main S / D portion and a lightly doped lateral extension continuous with the main S / D portion,
Wherein the major channel zone is terminated by the lateral extension along the upper semiconductor surface. The method according to claim 1 or 2,
The remaining major S / D zones of the main S / D zones include a main S / D portion and a lightly doped lateral extension continuous with the main S / D portion,
Wherein the major channel zone is terminated by the lateral extension of the remaining major S / D zone and the particular major S / D zone along the upper semiconductor surface. The method of claim 11,
Each major S / D zone is defined as a semiconductor dope along the upper semiconductor surface defined by a semiconductor dopant that reaches a lower maximum net concentration in the particular major S / D zone than in the remaining main S / D zones Structure comprising a. The method of claim 12,
The maximum net concentration of the semiconductor dopant in the main S / D zone is at least 20% lower in the particular main S / D zone than in the remaining main S / D zone. The method according to claim 1 or 2,
Each major S / D zone comprising a main FET comprising a main S / D portion and a lightly doped lower portion underlying the main S / D portion and continuous with the main S / D portion. . The method according to claim 1 or 2,
Further comprising an additional FET of the same polarity as the main FET,
The additional FET,
An additional channel zone of the main body material;
A pair of additional S / D zones located in the semiconductor body along the upper semiconductor surface and laterally separated by the additional channel zone and of the second conductivity type;
An additional gate dielectric layer overlying the additional channel zone; And
An additional control electrode overlying said additional gate dielectric layer over said additional channel zone,
The concentration of the first conductivity type semiconductor dopant in the main body material is each further S from an additional lower subsurface body-material position that is approximately as deep as the main lower subsurface body-material position below the upper semiconductor surface. A structure comprising a primary FET that changes substantially constant or less than 1/10 when moving up towards the / D zone. The method of claim 15,
The concentration of the first conductivity type semiconductor dopant in the main body material is such that where the main channel zone meets the particular main S / D zone along the upper semiconductor surface, the main channel zone is directed to the upper semiconductor surface. Thus lower than where it encounters the remaining major S / D zones of the major S / D zones;
The concentration of the first conductivity type semiconductor dopant in the main body material is such that the additional channel zone meets the additional S / D zone of one of the additional S / D zones and the additional channel zone meets the additional S / D zone. A structure comprising a primary FET, which is approximately identical where it encounters another additional S / D zone among the D zones. The method according to claim 1 or 2,
Further comprising an additional FET of opposite polarity to said primary FET,
The additional FET is:
An additional channel zone of additional body material of the semiconductor body, wherein the additional body material is sufficiently doped with a semiconductor dopant of a second conductivity type to become the second conductivity type;
A pair of additional S / D zones located in the semiconductor body along the upper semiconductor surface and laterally separated by the additional channel zone and of the first conductivity type;
An additional gate dielectric layer overlying the additional channel zone; And
An additional gate electrode overlying said additional gate dielectric layer over said additional channel zone,
The second conductivity type semiconductor dopant in the additional body material is further formed from the additional lower subsurface body-material position below the upper semiconductor surface, up to ten times deeper than a particular additional S / D zone of the additional S / D zones. A structure comprising a primary FET having a concentration that decreases by at least 1/10 when moving up from a particular additional S / D zone. The method of claim 17,
The concentration of the first conductivity type semiconductor dopant in the main body material is such that where the main channel zone meets the particular main S / D zone along the upper semiconductor surface, the main channel zone is directed to the upper semiconductor surface. Therefore, it is lower than where it meets the remaining major S / D zones of the main S / D zones,
The concentration of the second conductivity type semiconductor dopant in the additional body material is such that where the additional channel zone meets the particular additional S / D zone along the upper semiconductor surface, the additional channel zone is directed to the upper semiconductor surface. And thus a primary FET, which is lower than where it encounters the remaining additional S / D zones of the additional S / D zones. The method of claim 17,
Further comprising a third FET of the same polarity as the main FET,
The third FET is,
A third channel zone of said main body material;
A pair of third S / D zones located in the semiconductor body along the upper semiconductor surface, laterally separated by the third channel zone, and of the second conductivity type;
A third gate dielectric layer overlying said third channel zone; And
A third control electrode overlying the third gate dielectric layer over the third channel zone,
The concentration of the semiconductor dopant of the first conductivity type in the main body material is reduced from each of the third lower subsurface body-material positions approximately as deep as the main lower subsurface body-material position below the upper semiconductor surface. A structure comprising a primary FET that changes substantially constant or less than 1/10 when moving up to 3 S / D zones. The method of claim 19,
Further comprising a fourth FET of the same polarity as said additional FET and thus of a polarity opposite to said primary FET,
The fourth FET is,
A fourth channel zone of said additional body material;
A pair of fourth S / D zones located in the semiconductor body along the upper semiconductor surface and laterally separated by the fourth channel zone and of the first conductivity type;
A fourth gate dielectric layer overlying said fourth channel zone; And
A fourth gate electrode overlying the fourth gate dielectric layer over the fourth channel zone,
The concentration of the second conductivity type semiconductor dopant in the additional body material is each fourth S from a fourth lower body-material position that is approximately as deep as the additional lower subsurface body-material position below the upper semiconductor surface. A structure comprising a primary FET that changes substantially constant or less than 1/10 when moving up to the / D zone. A structure comprising a main field-effect transistor ("FET"),
The main field-effect transistor ("FET"),
A main channel zone of a main body material of a semiconductor body having an upper semiconductor surface, the main body material being of a first conductivity type;
A pair of primary sources / drains ("S / D") located in the semiconductor body along the upper semiconductor surface and laterally separated by the primary channel zone and of a second conductivity type opposite to the first conductivity type Zones;
A main gate dielectric layer overlying the main channel zone; And
A main gate electrode overlying the main gate dielectric layer above the main channel zone,
A main well portion of the main body material extends below the main channel zone and the main S / D zones,
The main well portion is defined by a main semiconductor well dopant of the first conductivity type such that the main well portion is more heavily doped than the upper and lower portions of the main body material,
The primary semiconductor well dopant has a concentration that reaches a major subsurface maximum value along a position up to 10 times deeper below a particular major S / D zone of the major S / D zones, thus providing the first within the main body material. All dopants of the conductivity type have a concentration that decreases by at least 1/10 when moving upward from the position of the main subsurface maximum in the concentration of the main semiconductor well dopant toward the particular main S / D zone, Structure containing the main FET. The method of claim 21,
Wherein the location of the major subsurface maximum in the concentration of the major semiconductor well dopant is up to five times deeper below the upper semiconductor surface than the particular major S / D zone. The method of claim 21 or 22,
A pocket portion of the main body material extends along the remaining main S / D zones of the main S / D zones to surround a portion of the main channel zone to the upper semiconductor surface,
The pocket portion is defined by the semiconductor pocket dopant of the first conductivity type such that the pocket portion is more heavily doped than the adjacent portion of the main body material,
The concentration of all the dopants of the first conductivity type in the main body material is such that where the main channel zone meets the particular main S / D zone along the upper semiconductor surface, the main channel zone is in contact with the upper semiconductor surface. Hence, a structure comprising a primary FET that is lower than where it meets the remaining major S / D zones. The method of claim 23,
The concentration of all dopants of the first conductivity type in the main body material is such that the main channel zone is directed to the upper semiconductor surface than where the main channel zone meets the remaining main S / D zone along the upper semiconductor surface. The primary FET, thus, at least 10 times lower in the encounter with the particular primary S / D zone. The method of claim 23,
The particular major S / D zone includes a major FET extending deeper below the upper semiconductor surface than the remaining major S / D zones. The method of claim 21 or 22,
Further comprising an additional FET of the same polarity as the main FET,
The additional FET,
An additional channel zone of the main body material;
A pair of additional S / D zones located in the semiconductor body along the upper semiconductor surface and laterally separated by the additional channel zone and of the second conductivity type;
An additional gate dielectric layer overlying the additional channel zone; And
An additional gate electrode overlying said additional gate dielectric layer over said additional channel zone,
An additional well portion of the main body material extends below the additional channel zone and the additional S / D zones,
The additional well portion is substantially defined by an additional semiconductor well dopant of the first conductivity type such that the additional well portion is more heavily doped than the lower portion of the main body material,
The additional semiconductor well dopant has a concentration that reaches an additional subsurface maximum value along the location below the additional channel zone and the additional S / D zones, thus reducing the concentration of all dopants of the second conductivity type in the main body material. Wherein the concentration changes by less than 10 times as the concentration moves upwards from each position of the additional subsurface maximum toward each additional S / D zone. A structure comprising a main field-effect transistor ("FET"),
The main FET is,
A main channel zone of a main body material of a semiconductor body having an upper semiconductor surface, wherein the main body material is sufficiently doped with a semiconductor dopant of a first conductivity type to become the first conductivity type;
A pair of primary sources / drains ("S / D") located in the semiconductor body along the upper semiconductor surface and laterally separated by the primary channel zone and of a second conductivity type opposite to the first conductivity type Zones;
A main gate dielectric layer overlying the main channel zone; And
A main gate electrode overlying the main gate dielectric layer above the main channel zone,
The main body material (a) lies below the main channel zone and the main S / D zones, and closest to the main S / D zones, below 10 the main S / D zones below the upper semiconductor surface. A major subsurface body-material portion deeper by a fold and (b) extending into the upper semiconductor surface, including the main channel zone, overlying the main subsurface body-material portion and overlying the main subsurface body-material portion A main surface-adjacent body-material portion which meets the
The first conductivity type semiconductor dopant in the main body material generally experiences step reduction by at least 1/10 as it crosses upward from the main subsurface body-material portion to the main surface-adjacent body-material portion. And remain at least 10 times less than in the main subsurface body-material portion as it moves further upward through the main surface-adjacent body-material portion and upwards to a particular major S / D zone of the main S / D zones. A structure comprising a primary FET having a concentration. The method of claim 27,
The main subsurface body-material portion comprising a main FET, closest to the main S / D zones, up to five times deeper than the main S / D zones below the upper semiconductor surface. The method of claim 27 or 28,
The concentration of the first conductivity type semiconductor dopant in the main body material is such that where the main channel zone meets the particular main S / D zone along the upper semiconductor surface, the main channel zone is directed to the upper semiconductor surface. And therefore a primary FET, which is lower than where it encounters the remaining major S / D zones of the major S / D zones. The method of claim 29,
The concentration of the first conductivity type semiconductor dopant in the main body material is such that where the main channel zone meets the particular main S / D zone along the upper semiconductor surface, the main channel zone is directed to the upper semiconductor surface. A main FET, thus, at least 10 times lower than where it meets the remaining main S / D zones. The method of claim 27 or 28,
And wherein the concentration of semiconductor dopant of the first conductivity type is approximately constant throughout the major subsurface body-material portion. The method of claim 27 or 28,
Further comprising an additional FET of opposite polarity to said primary FET,
The additional FET,
An additional channel zone of additional body material of the semiconductor body, wherein the additional body material is sufficiently doped with the second conductivity type semiconductor dopant to become the second conductivity type;
A pair of additional S / D zones located in the semiconductor body along the upper semiconductor surface and laterally separated by the additional channel zone and of the first conductivity type;
An additional gate dielectric layer overlying the additional channel zone; And
An additional gate electrode overlying said additional gate dielectric layer over said additional channel zone,
The additional body material (a) lies beneath the additional channel zone and the additional S / D zones and closest to the additional S / D zones, 10 less than the additional S / D zones below the upper semiconductor surface. An additional subsurface body-material portion extending deeper up to the belly and (b) extending to the upper semiconductor surface, including the additional channel zone, overlying the additional subsurface body-material portion and over the additional subsurface body- An additional surface-adjacent body-material portion that meets the material portion,
The second conductivity type semiconductor dopant in the additional body material generally experiences step reduction by at least 1/10 when crossing upwards from the additional subsurface body-material portion to the additional surface-adjacent body-material portion. And maintain a concentration that is at least 10 times less than in the additional subsurface body-material portion when further moved upwards to a particular additional S / D zone of the additional S / D zones with the additional surface-adjacent body-material portion. A branch contains a main FET. 33. The method of claim 32,
The concentration of the first conductivity type semiconductor dopant in the main body material is such that where the main channel zone meets the particular main S / D zone along the upper semiconductor surface, the main channel zone is directed to the upper semiconductor surface. Thus lower than where it encounters the remaining major S / D zones of the major S / D zones;
The concentration of the second conductivity type semiconductor dopant in the additional body material is such that where the additional channel zone meets the particular additional S / D zone along the upper semiconductor surface, the additional channel zone causes the upper semiconductor surface. And thus a primary FET lower than where it encounters the remaining additional S / D zones of the additional S / D zones. 33. The method of claim 32,
The concentration of the semiconductor dopant of the first conductivity type is approximately constant throughout the main subsurface body-material portion;
Wherein the concentration of the semiconductor dopant of the second conductivity type comprises a main FET approximately constant throughout the additional subsurface body-material portion. Introducing a first semiconductor dopant of the first conductivity type into a semiconductor body to define a major well portion of the first conductivity type for a major field-effect transistor (“FET”);
Providing a main gate electrode on a segment of the semiconductor body intended to be a main channel zone of the first conductivity type, the main gate electrode being separated from the segment by a main gate dielectric material;
The second body in the semiconductor body to form a pair of primary source / drain ("S / D") zones of a second conductivity type opposite to the first conductivity type laterally separated by the primary channel zone. Introducing a conductive first semiconductor dopant; And
(a) the semiconductor body has an upper semiconductor surface, and (b) the main channel zone and the main well portion are portions of the main body material that are of the first conductivity type and laterally extend below the main S / D zones. and (c) a first semiconductor dopant of the first conductivity type from a major lower subsurface body-material position below the upper semiconductor surface up to ten times deeper than a particular major S / D zone of the major S / D zones. Performing additional processing to complete the fabrication of the primary FET such that it has a concentration that decreases by at least 1/10 when moving up to the particular primary S / D zone. 36. The method of claim 35 wherein
Introducing the first semiconductor dopant of the first conductivity type and performing the further processing may be performed when moving from the major lower subsurface body-material location to the particular major S / D zone. Carried out under conditions that make it possible to reduce the concentration of the first semiconductor dopant to at least 1/10. The method of claim 35 or 36,
And wherein said major lower subsurface body-material location is up to five times deeper below said upper semiconductor surface than said particular major S / D zone. The method of claim 35 or 36,
And the concentration of the first semiconductor dopant of the first conductivity type decreases by at least 1/20 as it moves up from the major lower subsurface body-material location to the particular major S / D zone. The method of claim 35 or 36,
After the operation of performing the further processing to define a pocket portion extending along the remaining major S / D zones of the major S / D zones to the then-existing top surface, (a) the pocket portion is the primary (B) all dopants of the first conductivity type in the main body material, wherein the main channel zone is along the upper semiconductor surface The semiconductor body of the first conductivity type such that where it encounters a particular major S / D zone has a lower concentration than where the major channel zone meets the remaining major S / D zones along the upper semiconductor surface Introducing a second semiconductor dopant. The method of claim 39,
After the operation of performing the further processing, the concentration of all the dopants of the first conductivity type in the main body material is at the point where the main channel zone meets the particular main S / D zone along the upper semiconductor surface. At least 10 times lower than where the major channel zone meets the remaining major S / D zone along the upper semiconductor surface. 36. The method of claim 35 wherein
Introducing a second semiconductor dopant of the first conductivity type into the main body material to define the precursor upper body-material portion such that the precursor upper body-material portion of the first conductivity type overlies the main well portion Further comprising, the method. The method of claim 35 or 36,
A portion of the semiconductor body is (a) overlying the main well portion as it is immediately after the operation of introducing the first semiconductor dopant of the first conductivity type, and (b) then the second conductivity type;
In the portion of the first semiconductor dopant of the first conductivity type, another doping of the first conductivity type or the second conductivity type is not remarkably performed following the operation of introducing the first semiconductor dopant of the first conductivity type. Not diffused upwardly into the portion of the semiconductor body during the operation of performing the further processing to cause substantially all of the portion of the semiconductor body to transition to the first conductivity type. The method of claim 35 or 36,
Introducing the first semiconductor dopant of the second conductivity type,
The side-extending mask, the main gate electrode, and the dopant-blocking through an opening in the side-extension mask and through the then-existing upper surface of the semiconductor body to the side-separated main portion of the pair of semiconductor bodies. Introducing a side-extending semiconductor dopant of the second conductivity type using any material along the main gate electrode as a shield;
Providing a spacer material across the main gate electrode; And
The main-part mask, the main gate electrode, and the dopant-blocking through an opening in the main-part mask to a pair of laterally separated additional parts of the semiconductor body through the then-existing top surface of the semiconductor body. Using the spacer material as a shield, introducing a main-part semiconductor dopant of the second conductivity type. The method of claim 35 or 36,
Introducing the first semiconductor dopant of the second conductivity type,
(a) introducing a main-part semiconductor dopant of the second conductivity type into the semiconductor body at a main-part average depth below the then-existing upper surface of the semiconductor body and in a main part dosing; and (b) introducing a lower-part semiconductor dopant of the second conductivity type into the semiconductor body at a lower-part average depth below the then-existing upper surface of the semiconductor body and in a lower-part illustration. Introducing the main-part semiconductor dopant, wherein the main-part shown is larger than the bottom-part shown, and wherein the bottom-part average depth is greater than the main-part average depth. Performing one of the steps of introducing a semiconductor; And
The bottom-part semiconductor dopant is separated by a pair of side pairs of each of the main S / D zones such that the main-part semiconductor dopant defines a pair of laterally separated main S / D portions of each of the main S / D zones. To define a lower S / D portion, and wherein the lower S / D portion is lightly doped than the main S / D portion, respectively under the main S / D portion, and perpendicular to the main S / D portion, respectively. Introducing the main-part semiconductor dopant and introducing the bottom-part semiconductor dopant so as to be continuous. The method of claim 35 or 36,
The operation of introducing the first semiconductor dopant of the first conductivity type may be performed on the semiconductor body to define an additional well portion of the first conductivity type for an additional FET of the same polarity as the primary FET. Introducing an additional first semiconductor dopant;
The main gate electrode providing operation includes providing an additional gate electrode vertically separated by an additional gate dielectric material from the segment over the segment of the semiconductor body intended to be an additional channel zone of the first conductivity type. Including;
Introducing the first semiconductor dopant of the second conductivity type comprises: forming the pair of additional S / D zones of the second conductivity type separated by the additional channel zone in the semiconductor body; Introducing an additional first semiconductor dopant of a conductivity type;
The performing of further processing may be performed such that the main body material extends below the additional S / D zones, such that the additional channel zone and the additional well portion become part of the main body material, and The additional first semiconductor dopant is substantially constant as it moves up from the additional lower subsurface body-material position to each additional S / D zone approximately as deep as the main lower subsurface body-material position below the upper semiconductor surface; or Completing the fabrication of the additional FET to have a concentration that varies by less than 10 times. The method of claim 35 or 36,
The method includes introducing a second semiconductor dopant of the second conductivity type into the semiconductor body to define an additional well portion of the second conductivity type for an additional FET of opposite polarity to the main FET;
The main gate electrode providing operation includes providing an additional gate electrode vertically separated from the segment by an additional gate dielectric material over the segment of the semiconductor body intended to be an additional channel zone of the second conductivity type. Including;
The method includes introducing a second semiconductor dopant of the first conductivity type into the semiconductor body to form a pair of additional S / D zones of the second conductivity type that are laterally separated by the additional channel zone. More;
The performing of further processing is performed such that the additional channel zone and the additional well portion are part of an additional body material of the second conductivity type extending below the additional S / D zones, and the second of the second conductivity type. At least when a semiconductor dopant moves upwards below the upper semiconductor surface from the additional lower subsurface body-material position up to ten times deeper than a particular additional S / D zone of the additional S / D zones. Completing the fabrication of the additional FET to have a concentration decreasing to one tenth. The method of claim 46,
Introducing the first semiconductor dopant of the first conductivity type comprises: first defining a third well portion of the first conductivity type with respect to a third FET having the same polarity as the main FET; Introducing an additional first semiconductor dopant of a conductivity type;
The main gate electrode providing operation provides a third gate electrode vertically separated from the segment by a third gate dielectric material over a segment of the semiconductor body intended to be a third channel zone of the first conductivity type. Including a step;
Introducing the first semiconductor dopant of the second conductivity type into the semiconductor body to form a pair of third S / D zones of the second conductivity type that are laterally separated by the third channel zone. Introducing an additional first semiconductor dopant of the second conductivity type;
The further processing operation may be such that the third channel zone and the third well portion are part of the main body material extending below the third S / D zones, and the additional first of the first conductivity type. Is substantially constant as the semiconductor dopant moves upward from the third lower subsurface body-material position to each of the third S / D zones approximately below the upper semiconductor surface approximately as deep as the main lower subsurface body-material position; or Completing the fabrication of the third FET to have a concentration that varies by less than 10 times. The method of claim 47,
The operation of introducing the second semiconductor dopant of the second conductivity type is to define a fourth well portion of the second conductivity type for the fourth FET that is of the same polarity as the additional FET and thus of the opposite polarity to the main FET. Introducing an additional second semiconductor dopant of the second conductivity type into the semiconductor body;
The main gate electrode providing operation provides a fourth gate electrode vertically separated from the segment by a fourth gate dielectric material over a segment of the semiconductor body intended to be a fourth channel zone of the second conductivity type. Including a step;
Introducing the second semiconductor dopant of the first conductivity type into the semiconductor body to form a pair of fourth S / D zones of the first conductivity type that are laterally separated by the fourth channel zone. Introducing an additional second semiconductor dopant of the first conductivity type;
The further processing operation may be such that the fourth channel zone and the fourth well portion are part of the additional body material extending below the fourth S / D zones, and the additional second of the second conductivity type. The semiconductor dopant is substantially constant as it moves upward from the fourth lower subsurface body-material position to each of the fourth S / D zones approximately below the upper semiconductor surface approximately as deep as the additional lower subsurface body-material position. Completing the fabrication of the fourth FET to have a concentration that varies by less than 10 times. So that the portion of the semiconductor body is (a) on the well portion as if it is present immediately after the operation of introducing the first semiconductor dopant of the first conductivity type, and (b) thereafter is a second conductivity type opposite to the first conductivity type. Introducing a first semiconductor dopant of the first conductivity type into the semiconductor body to define a well portion of the first conductivity type for a field-effect transistor (“FET”);
Providing a gate electrode vertically separated from a portion of the semiconductor body by a gate dielectric material over a portion of the semiconductor body intended to be a channel zone of the first conductivity type;
Introducing a first semiconductor dopant of the second conductivity type into the semiconductor body to form a pair of source / drain ("S / D") zones of the second conductivity type that are laterally separated by the channel zone ; And
Subsequently, the operation of introducing the first semiconductor dopant of the first conductivity type causes the portion of the first body to be substantially all of the semiconductor body in which no other doping of the first conductivity type or the second conductivity type is performed significantly. Performing the further processing to complete the fabrication of the FET such that a portion of the first semiconductor dopant of the first conductivity type upwardly diffuses into the portion of the semiconductor body during an additional processing perform operation to cause conversion to Comprising a step. The method of claim 49,
Introducing the first semiconductor dopant of the first conductivity type comprises ion implanting the first semiconductor dopant of the first conductivity type. 51. The method of claim 49 or 50,
The portion of the first semiconductor dopant of the first conductivity type diffuses upward during at least some of the additional processing performed at elevated temperatures. 51. The method of claim 49 or 50,
Defining a pocket portion extending along the remaining S / D zones of the S / D zones to the then-present upper surface, following the further processing operation, (a) the semiconductor body has an upper semiconductor surface, (b) the channel zone and the well portion are portions of the first conductivity type body material extending laterally below the S / D zones, and (c) the pocket portion constitutes a portion of the body material and the body Doped at a higher concentration than adjacent portions of material, and (d) all dopants of the first conductivity type in the body material are such that the channel zone meets the particular S / D zone along the upper semiconductor surface. Introducing a second semiconductor dopant of the first conductivity type into the semiconductor body such that it has a lower concentration than where it meets the remaining S / D zones along the upper semiconductor surface; Including, method. The method of claim 52, wherein
Subsequent to the further processing operation, the concentration of all dopants of the first conductivity type in the body material is determined by the channel zone where the channel zone meets the specific S / D zone along the upper semiconductor surface. At least 10 times lower than where it meets the remaining S / D zones along the upper semiconductor surface. 51. The method of claim 49 or 50,
The operation of introducing the first semiconductor dopant of the second conductivity type is as follows:
The side-extending mask, the gate electrode and the dopant-blocking shield, through the opening in the side-extension mask, through the then-existing upper surface of the semiconductor body, to a pair of laterally separated main portions of the semiconductor body. Introducing a side-extending semiconductor dopant of the second conductivity type using any material along the gate electrode as a;
Providing a spacer material across the gate electrode; And
The main-part mask, the gate electrode, and the dopant-blocking, through the opening in the main-part mask, through the then-existing upper surface of the semiconductor body, to a pair of laterally separated additional parts of the semiconductor body. Introducing the second conductivity type main-part semiconductor dopant using the spacer material as a shield. (a) a subsurface semiconductor region of a first conductivity type over and adjacent to a subsurface electrical isolation layer, and (b) a lightly doped surface-adjacent semiconductor region of the first conductivity type is the subsurface semiconductor region. A concentration which substantially experiences step reduction by at least 1/10 as the dopant of the first conductivity type crosses upward from the subsurface semiconductor region to the surface-adjacent semiconductor region. Providing a primitive structure, wherein the semiconductor regions are doped with a dopant of the first conductivity type, and (d) a lower semiconductor layer lies adjacent to and below the subsurface electrical isolation layer;
(a) the remaining portion of the sub-surface semiconductor region constitutes the main sub-surface body-material portion of the first conductivity type, and (b) the remaining portion of the surface-adjacent semiconductor region is of low concentration of the first conductivity type. Forming a cavity downwardly through the semiconductor regions and the subsurface electrical isolation layer into the lower semiconductor layer, constituting a doped precursor major surface-adjacent body-material portion;
(a) an additional subsurface body-material portion of the second conductivity type opposite to the first conductivity type extending downward to the lower semiconductor layer to form a pn junction, and (b) at a low concentration of the second conductivity type. Introducing the second conductivity type semiconductor material into the cavity to create a doped precursor additional surface-adjacent body-material portion, the dopant of the second conductivity type upwardly from the additional sub-surface body-material portion. The second conductivity type, wherein the additional semiconductor portions are doped with a dopant of the second conductivity type so as to have a concentration that substantially experiences a step reduction by at least 1/10 when traversed to the additional surface-adjacent body-material portion. Introducing a semiconductor material; And
(a) the precursor major along its upper surface such that the major channel zone of the major surface-adjacent body-material portion laterally separates a pair of major source / drain ("S / D") zones of the second conductivity type; Provide the primary S / D zones within a surface-adjacent body-material portion, and (b) the additional channel zone of the additional body-material portion to laterally separate the pair of additional S / D zones of the first conductivity type. Providing the additional S / D zones within the precursor additional surface-adjacent body material portion along its upper surface, and (c) providing a primary gate dielectric layer and an additional gate dielectric layer over the primary channel zone and the additional channel zone, respectively. And (d) transfer the main gate over each of the main gate dielectric layer and the additional gate dielectric layer over each of the main channel zone and the additional channel zone. Providing a pole and an additional gate electrode. The method of claim 55,
The concentration of the semiconductor dopant of the first conductivity type is approximately constant throughout the main subsurface body-material portion;
And the concentration of the dopant of the second conductivity type is approximately constant throughout the additional subsurface body-material portion. The method of claim 56, wherein
Selectively introducing a major semiconductor pocket dopant of the first conductivity type into the precursor major surface-adjacent body-material portion to define a major pocket portion extending along the remaining S / D zones of the major S / D zones Selectively introducing the primary semiconductor pocket dopant, wherein the primary pocket portion is more heavily doped with the first conductivity type than the adjacent material of the primary body-material portion;
Selectively introducing an additional semiconductor pocket dopant of the second conductivity type into the precursor additional surface-adjacent body-material portion to define an additional pocket portion extending along the remaining S / D zones of the additional S / D zones Wherein the additional pocket portion optionally includes introducing the additional semiconductor pocket dopant more heavily doped to the second conductivity type than the adjacent material of the additional surface-adjacent body-material portion. The method of any one of claims 55-57,
The concentration of semiconductor dopant of the first conductivity type in the main surface-adjacent body-material portion along a position extending from the main subsurface body-material portion to a particular major S / D zone of the main S / D zones is: Also 10 times lower than in the main subsurface body-material portion;
The concentration of the dopant of the second conductivity type in the additional surface-adjacent body-material portion along the position extending from the additional subsurface body-material portion to a particular additional S / D zone of the additional S / D zones is: At least 10 times lower than in the additional subsurface body-material portion. Introducing a main dopant of a first conductivity type opposite to the second conductivity type into a portion of the starting semiconductor material of the second conductivity type, placing the portion of the starting semiconductor material into the surface-adjacent major of the first conductivity type. Switching to body material such that the main dopant has a relatively uniform concentration throughout the main body material, and the remaining portion of the starting semiconductor material constitutes the surface-adjacent additional body material of the second conductivity type, Introducing a primary dopant of the first conductivity type;
Introducing the primary well dopant of the first conductivity type into the primary body material to form a primary well portion that reaches a maximum concentration of primary well dopant at a subsurface location within the primary body material;
Introducing the additional well dopant of the second conductivity type into the additional body material to form an additional well portion that reaches a maximum concentration of additional well dopant at a subsurface location within the additional body material;
(a) the main channel zone of the main body material laterally separates a pair of surface-adjacent main source / drain ("S / D") zones and from the location of the maximum concentration of the main well dopant The second conductivity type in the main body material portion such that all dopants of the first conductivity type in the main body material have a concentration that decreases by at least 1/10 when moving to a particular major S / D zone of the D zones. Provide a pair of surface-adjacent major S / D zones, (b) the additional channel zone of the additional body material laterally separates the pair of additional S / D zones, and the additional well dopant in the additional body material All dopants of the second conductivity type in the additional body material are at least 1/10 when moving from the position of the maximum concentration of to a particular additional S / D zone of the additional S / D zones. Provide the pair of additional S / D zones of the first conductivity type in the additional body material to have a decreasing concentration, and (c) a primary gate dielectric layer and an additional over the primary channel zone and the additional channel zone, respectively; Providing a gate dielectric layer, and (d) providing a primary gate electrode and an additional gate electrode over the primary gate dielectric layer and the additional gate dielectric layer, respectively, on each of the primary channel zone and the additional channel zone. How to. The method of claim 59,
The first conductivity type and the second conductivity type are p-type and n-type, respectively;
And the main dopant of the first conductivity type comprises aluminum. A first body-material region and a second body-material region of a semiconductor body having an upper surface, wherein the first body-material region and the second body-material region are doped with a semiconductor dopant of a first conductivity type to form the first body-material region; The first body-material region and the second body-material region, which are of a first conductivity type; And
A first zone and a second zone of a second conductivity type opposite the first conductivity type located in the semiconductor body along the upper surface,
The first and second body-material regions extend below the first and second zones, respectively, and the first and second body-material regions respectively correspond to the first and second zones and the first and second pn. Meet the first and second zones to form a junction, respectively, so that (a) each pn junction reaches a maximum depth below the upper surface of the semiconductor body, and (b) a semiconductor dopant of the first conductivity type Are in both the first and second zones, and the first conductivity type semiconductor dopant extends laterally below the first and second zones, respectively, and is located within the first and second body-material regions, respectively. Having a concentration locally reaching a first and second subsurface maximum concentration at each of said first and second subsurface body-material locations, and (c) said first subsurface body-material location and said second subsurface Body-material position, the upper table of the semiconductor body Down to 10 times deeper than the maximum depth of the first and second pn junctions, respectively, and (d) the concentration of the semiconductor dopant of the first conductivity type is (i) along the selected first vertical line. Decrease at least 1/10 when moving upward from the first subsurface body-material position to the upper surface of the semiconductor body through the first zone, and (ii) from the first subsurface body-material position to the first surface; Decrease substantially monotonously when moving to the first pn junction along a vertical line, and (iii) upward from the second subsurface body-material position through the second zone to the upper surface of the semiconductor body along a selected second vertical line At least one additional subsurface maximum concentration when moving. 62. The method of claim 61,
The concentration of the semiconductor dopant of the first conductivity type is reduced to at least 1/20 when moving from the first sub-surface body-material position to the upper surface of the semiconductor body through the first zone along the first vertical line. , Semiconductor structure. 62. The method of claim 61,
The concentration of the semiconductor dopant of the first conductivity type is reduced to at least 1/40 when moving from the first sub surface body-material position to the upper surface of the semiconductor body through the first zone along the first vertical line. , Semiconductor structure. 64. The method of any of claims 61-63,
The concentration of the semiconductor dopant of the first conductivity type is reduced to less than 1/10 when moving upward from the second sub-surface body-material position through the second zone along the second vertical line to the upper surface of the semiconductor body. Semiconductor structure. The method of claim 64, wherein
Wherein the concentration of the semiconductor dopant of the first conductivity type varies substantially monotonically along the first vertical line through the first zone at the depth of each additional subsurface maximum concentration in the second zone. 64. The method of any of claims 61-63,
Wherein the first subsurface maximum concentration moves from the first subsurface body-material position downward along the first vertical line to a depth ten times the maximum depth of the first pn junction. A semiconductor structure, wherein the substantially unique local subsurface maximum of the concentration of dopant. Co-polar first field-effect transistor ("FET") and second field-effect provided along an upper surface of a semiconductor body having a body material doped with a first conductivity type semiconductor dopant to become the first conductivity type A structure comprising a transistor,
Each FET is
A channel zone of the body material;
A first conductivity type located within the semiconductor body along the upper surface of the semiconductor body and laterally separated by the channel zone and of a second conductivity type opposite to the first conductivity type to form respective pn junctions with the body material And a second source / drain ("S / D") zone, wherein (a) each pn junction reaches a maximum depth below the upper surface of the semiconductor body, and (b) the body material is first S / D Extending laterally under both the zone and the second S / D zone, and (c) the first conductivity type semiconductor dopant is present in both the first and second S / D zones, and generally in all of the respective channel zones and Having a concentration locally reaching the main subsurface maximum concentration at a main subsurface body-material location extending laterally below the S / D zone, and (d) the main subsurface body-material location is below an upper surface of the semiconductor body. For each S / D zone said first and second S / D zones being up to ten times deeper than the maximum depth of a pn junction;
A gate dielectric layer overlying said channel zone; And
A gate electrode overlying the gate dielectric layer above the channel zone, wherein the concentration of the semiconductor dopant of the first conductivity type is (i) from the main subsurface body-material position for the first FET along the selected first vertical line; Decreases by at least 1/10 when moving upwards to a top surface of the semiconductor body through a particular S / D zone of the S / D zones of the first FET, and (ii) the main subsurface body for the first FET Substantially monotonically decrease when moving from a material location along the first vertical line to the pn junction for the particular S / D zone of the first FET, and (iii) the main subsurface body for the second FET. At least one additional subsurface maximum concentration is reached between a material location and an upper surface of the semiconductor body and each additional subsurface maximum for the second FET is reached. At an additional subsurface body-material location where the concentration extends laterally under substantially all material of at least a portion of each of the S / D zones of the second FET and the gate electrode of the second FET overlying the channel zone of the second FET. 12. A structure comprising first and second FETs of the same polarity, comprising a gate electrode that occurs. The method of claim 67 wherein
The concentration of the semiconductor dopant of the first conductivity type is determined in the semiconductor body through the specific S / D zone of the first FET along the first vertical line from the main sub-surface body-material location for the first FET. 1. A structure comprising first and second FETs of the same polarity that decrease to at least 1/20 when moving to a top surface. The method of claim 67 wherein
The concentration of the semiconductor dopant of the first conductivity type is determined in the semiconductor body through the specific S / D zone of the first FET along the first vertical line from the main subsurface body-material location for the first FET. 1. A structure comprising first and second FETs of the same polarity that decrease to at least 1/40 when moving to a top surface. 70. The method of any of claims 67-69,
The concentration of the semiconductor dopant of the first conductivity type is determined from the main subsurface body-material location for the second FET along the selected second vertical line through the S / D zone of any one of the second FETs. A first and second FETs of the same polarity that decrease by less than one tenth as they move up to the top surface of the second polarity. 71. The method of claim 70,
The concentration of the semiconductor dopant of the first conductivity type includes first and second FETs of the same polarity that vary substantially monotonically along the first vertical line at a depth of each additional subsurface maximum concentration for the second FET. Structure. 70. The method of any of claims 67-69,
The main subsurface maximum concentration for the first FET is equal to the pn for the particular S / D zone of the first FET along the first vertical line from the main subsurface body-material location for the first FET. And first and second FETs of the same polarity, which are substantially unique local subsurface maximums of the concentration of the first conductivity type semiconductor dopant when moving downward to a depth ten times the maximum depth of the junction. 70. The method of any of claims 67-69,
Each S / D zone of each FET includes a main portion and a lightly doped side extension extending laterally with the main portion, wherein the lateral extension includes a channel zone of each of the FETs of the semiconductor body. And first and second FETs of the same polarity extending laterally below the gate electrode of the FET so as to be terminated by side extensions of each FET along an upper surface. 70. The method of any of claims 67-69,
The pocket portion of the body material of each FET extends along the first S / D zone of each FET to the channel zone of each FET and is more heavily doped than the side adjacent material of the body material, A structure comprising first and second FETs of the same polarity. The method of claim 74, wherein
The pocket portion of the first FET includes first and second FETs of the same polarity, causing the channel zone of the first FET to be asymmetrical with respect to the S / D zones of the first FET. The method of claim 74, wherein
Another pocket portion of the body material extends along the second S / D zone of the second FET to the channel zone of the second FET and is doped at a higher concentration than the side adjacent material of the body material. A structure comprising first and second FETs. As a method of manufacturing a semiconductor structure,
A semiconductor dopant of the first conductivity type in the semiconductor body to define the first body-material region and the second body material region such that each of the first body-material region and the second body material region is of a first conductivity type; Introducing a; And
Introducing a semiconductor dopant of the second conductivity type into the semiconductor body to define a first zone and a second zone of a second conductivity type opposite to the first conductivity type, respectively,
Upon completion of fabrication of the semiconductor structure, (a) each of the first body-material region and the second body-material region forms a 1 pn junction and a second pn junction with the first zone and the second zone and And (b) each pn junction extends to a maximum depth below an upper surface of the semiconductor body, and (c) a semiconductor of the first conductivity type, respectively, laterally extending below the first zone and the second zone. Dopants are present in both the first zone and the second zone, and (d) all semiconductor dopants of the first conductivity type in the semiconductor body are in the first body-material region and the second body-material region. A first subsurface maximum concentration and a second subsurface maximum at each of the first subsurface body-material position and the second subsurface body-material position respectively positioned and laterally extending below the first zone and the second zone, respectively. Locally at concentration And (e) the first subsurface body-material position and the second subsurface body-material position are the maximum depths of the first pn junction and the second pn junction below the upper surface of the semiconductor body. Each of which is up to ten times deeper, and (f) the concentrations of all dopants of the first conductivity type are (i) from the first subsurface body-material location along the first zone along the selected first vertical line through the first zone. Decreases by at least 1/10 when moving upward to the upper surface of the semiconductor body, and (ii) decreases substantially monotonously when moving from the first subsurface body-material position to the first pn junction along the first vertical line, (iii) at least one weight when moving upwardly from the second subsurface body-material position through the second zone to the upper surface of the semiconductor body along a selected second vertical line; A method for manufacturing a semiconductor structure that reaches the surface of the sub-maximum concentration. 78. The method of claim 77 wherein
At the completion of manufacture of the semiconductor structure,
The concentration of all dopants of the first conductivity type decreases to less than 1/10 when moving upward from the second subsurface body-material position to the upper surface of the semiconductor body along the second vertical line. Way. 79. The method of claim 77 or 78,
At the completion of manufacture of the semiconductor structure,
And wherein the concentration of all dopants of the first conductivity type varies substantially monotonically along the first vertical line through the first zone at the depth of each additional subsurface maximum concentration in the second zone. 79. The method of claim 77 or 78,
At the completion of manufacture of the semiconductor structure,
The first subsurface maximum concentration is a semiconductor dopant of the first conductivity type as it moves downward from the first subsurface body-material position to a depth ten times the maximum depth of the first pn junction along the first vertical line. Wherein the structure is a substantially unique local subsurface maximum of concentration. A method of manufacturing a structure comprising a first field-effect transistor ("FET") and a second field-effect transistor, the same being polar,
Upon completion of fabrication of the structure, defining a first body-material region and a second body-material region for the first FET and the second FET, respectively, such that each body-material region is of a first conductivity type. Introducing a semiconductor dopant of a first conductivity type into the semiconductor body;
The gate electrode of each numbered FET is located above a portion of the same-numbered body-material region intended to be a channel zone for that FET and corresponds from a portion of the same-numbered body-material region. Defining a pair of gate electrodes for each of the FETs so as to be vertically separated by a dielectric layer; And
For each FET, the semiconductor body to form a pair of source / drain ("S / D") zones of a second conductivity type opposite to the first conductivity type laterally separated by the channel zone of the FET. Introducing a second dopant semiconductor dopant into the
Upon completion of fabrication of the structure, (a) each of the numbered body-material regions each form a pair of pn junctions with the S / D zones of the same numbered FET and laterally extend below the S / D zone. (B) each pn junction extends to a maximum depth below the upper surface of the semiconductor body, (c) the first conductivity type semiconductor dopant is present in each of the S / D zones, and (d ) All semiconductor dopants of the first conductivity type in the semiconductor body reach a first main subsurface maximum concentration at a first main subsurface body-material position below the upper surface of the semiconductor body relative to the first FET. And for the second FET has a concentration reaching a second main subsurface maximum concentration at a second main subsurface body-material location below the upper surface of the semiconductor body, and (e) each numbered main Sub surface body An end position extends laterally under almost all of the channel zones and the S / D zones of the same-numbered FET and below the top surface of the semiconductor body for the respective S / D zone of the FET. Wherein the concentration of all dopants of the first conductivity type is (i) from the first main subsurface body-material location along the selected first vertical line; Decreases by at least 1/10 when moving upward through the particular S / D zone of the first FET to the upper surface of the semiconductor body, and (ii) from the first main subsurface body-material position Decrease substantially monotonously when moving to the pn junction for the particular S / D zone of the first FET along a first vertical line, and (iii) the channel zone of the second FET and the S / D of at least the second FET Zone each division The upper surface of the semiconductor body and the second main so that each additional subsurface maximum concentration occurs at a corresponding additional subsurface body-material location laterally extending below almost all material of the gate electrode of the second FET overlying. 16. A method of fabricating a structure comprising first and second FETs of the same polarity reaching at least one additional subsurface maximum concentration between subsurface body-material locations. 82. The method of claim 81 wherein
Upon completion of fabrication of the structure, the concentration of all dopants of the first conductivity type is determined by removing any one S / D zone of the second FET from the second main subsurface body-material location along the selected second vertical line. A first and second FETs of the same polarity that decrease by less than one tenth as they move upward to the upper surface of the semiconductor body. 83. The method of claim 81 or 82,
Upon completion of fabrication of the structure, the concentrations of all dopants of the first conductivity type vary substantially monotonically along the first vertical line at a depth of each additional subsurface maximum concentration for the second FET of the same polarity. A method of making a structure comprising first and second FETs. 83. The method of claim 81 or 82,
The first main subsurface maximum concentration is ten times the maximum depth of the pn junction for the particular S / D zone of the first FET along the first vertical line from the first main subsurface body-material location. A method of fabricating a structure comprising first and second FETs of the same polarity that are substantially unique local subsurface maximums of the concentration of the first conductivity type semiconductor dopant when moved to depth. 83. The method of claim 81 or 82,
Defining the gate electrode includes first and second FETs of the same polarity, which are primarily performed subsequent to introducing the first conductive semiconductor dopant. 83. The method of claim 81 or 82,
The operation of introducing the second conductivity type semiconductor dopant comprises a first and a second FET of the same polarity, which is mainly performed following the operation of defining the gate electrode. 83. The method of claim 81 or 82,
Introducing the second conductivity type semiconductor dopant may comprise forming each S / D zone of each FET to include a main portion and a lightly doped side extension continuous to the main portion. and,
The lateral extension is below the gate electrode of the FET such that upon completion of fabrication of the structure, the channel zone of each FET is terminated by the lateral extension of each FET along the upper surface of the semiconductor body. A method of manufacturing a structure comprising laterally extending first and second FETs of the same polarity. 83. The method of claim 81 or 82,
The pocket portion of each of the numbered body-material regions extends along the first S / D zone of the same-numbered FET into the channel zone of the same-numbered FET and on the side of the body-material region Further comprising introducing additional dopants of the first conductivity type into the semiconductor body to be more heavily doped than adjacent materials. 1. A method of fabricating a structure comprising first and second FETs of the same polarity. 90. The method of claim 88,
Wherein the pocket portion of the first FET comprises first and second FETs of the same polarity such that the channel zone of the first FET is asymmetrical with respect to the S / D zone of the first FET. 90. The method of claim 88,
Introducing the additional dopant of the first conductivity type may include extending another pocket portion of the second body-material region along the second S / D zone of the second FET to the channel zone of the second FET. And first and second FETs of the same polarity to cause a higher concentration of doping than the side adjacent material of the second body-material region.

Images