US20040004251A1

Movatterモバイル変換

Info

Publication number: US20040004251A1
Application number: US10/413,808
Authority: US
Inventors: Raminda Madurawe
Original assignee: Individual
Current assignee: Liberty Patents LLC; Callahan Cellular LLC
Priority date: 2002-07-08
Filing date: 2003-04-14
Publication date: 2004-01-08
Also published as: US20050121721A1; US20080067594A1; US20040152245A1; US20090004788A1; US7265421B2; US7018875B2

Abstract

A new Insulated-Gate Field-Effect Thin Film Transistor (Gated-FET) is disclosed. A semiconductor Gated-FET device comprises a lightly doped resistive channel region formed on a first semiconductor thin film layer; and an insulator layer deposited on said channel surface with a gate region formed on a gate material deposited on said insulator layer; said gate region receiving a gate voltage having a first level modulating said channel resistance to a substantially non-conductive state and a second level modulating said channel resistance to a substantially conductive state.

Description

This application claims priority from Provisional Application Serial No. 60/393,763 entitled “Wire Replaceable TFT SRAM Cell and Cell Array Technology”, filed on Jul. 8, 2002, Provisional Application Serial No. 60/397,070 entitled “Wire Replaceable Thin Film Fuse and Anti-fuse Technology”, filed on Jul. 22, 2002, Provisional Application Serial No. 60/400,007 entitled “Re-programmable ASIC”, filed on Aug. 1, 2002, Provisional Application Serial No. 60/402,573 entitled “Thin Film SRAM Cell”, filed on Aug. 12, 2002, and Provisional Application Serial No. 60/449,011 entitled “SRAM cell and cell arrays”, filed on Feb. 24, 2003, all of which list as inventor Mr. R. U. Madurawe and the contents of which are incorporated-by-reference.[0001]
This application is also related to application Ser. No. 10/267,484 entitled “Methods for Fabricating Three-Dimensional Integrated Circuits”, application Ser. No. 10/267,483 entitled “Three Dimensional Integrated Circuits”, and application Ser. No. 10/267,511 entitled “Field Programmable Gate Array With Convertibility to Application Specific Integrated Circuit”, all of which were filed on Oct. 8, 2002 and list as inventor Mr. R. U. Madurawe, the contents of which are incorporated-by-reference.[0002]

BACKGROUND

The present invention relates to semiconductor IGFET devices.[0003]

An Insulated-Gate Field-Effect Transistor, or IGFET, is a device of very major importance in the semiconductor IC industry. A Metal-Oxide-Silicon Field-Effect Transistor, or MOSFET, is a sub-class of IGFET devices. An IGFET is a four-terminal device comprising of a source, drain, gate and body nodes; though the body node only allows very limited access to the device. MOSFETs are widely used in the sub micron semiconductor processing technologies to manufacture Ultra Large Scale Integrated Circuits. Ability to form Silicon-oxide interfaces with very low interface states, quality gate oxides with low thickness, reductions in system voltage and reductions in lateral geometries by lithography improvements have all contributed to the popularity of these transistors. Today MOSFETs are used to build ASICs, Memory, FPGA, Gate Array, Graphics, Micro Processors, and a wide variety of semiconductor IC products.[0004]

IGFET differ from a Bipolar Transistor in the power level and power amplification available in the device. Bipolar transistor is a three terminal device with a base, an emitter and a collector node. Compared to the base control terminal of a Bipolar transistor, the gate control terminal of IGFET consumes essentially no power. While the Bipolar can deliver more output power, the gain (defined by the ratio of output current to control current) is infinite for IGFET compared to about 500 for a good Bipolar transistor. This high gain coupled with complementary MOSFET design methodology facilitates low stand-by power in ICs that have over 10 Million transistors. Bipolar is used to build many Analog and Linear ICs such as voltage regulators, power amplifiers, rectifiers, battery regulators, D to A Converters and A to D Converters due to the high output power available. Sub-micron geometry MOSFETs with high current drives are now increasingly used for similar applications.[0005]

IGFET differs from a JFET, also a three terminal device, in the construction of the transistor. In the IGFET the gate is insulated above the transistor body, while in the JFET the gate is formed as a reverse biased junction above the conducting channel. The reverse bias control gate junction consumes a low level of power due to carrier recombination inside the depleted region. The JFET power amplification is better than a Bipolar, but lower than an IGFET. A significant difference between IGFET and JFET occurs in the method of channel conduction. This will be discussed in detail next.[0006]

The MOSFET operates by conducting current between its drain and source through a conducting inversion layer created by the presence of a gate voltage. FIG. 1 shows a cross section of a MOSFET device, and is described herein with reference to an NMOSFET device. In FIG. 1, an[0007]

NMOS transistor body

100 is P-doped, isolating an N+ dopedsource region113 and an N+ dopeddrain region114. The source is connected to afirst voltage103, which may be the ground supply V_S. The drain is connected to asecond voltage104, which may be a switching voltage node for the device. The body region between source and drain under thegate112 is also doped P type same as substrate. The result is the formation of two N+/P− back-to-back reverse-biased diodes. When thevoltage102 atgate112 is zero, or below a threshold voltage V_TN, the N+/P− back-to-back reverse-biased diodes do not conduct and the transistor is off. The surface undergate112 consists of hole carriers. In the embodiment of FIG. 1, thegate112 includes a salicided region shown shaded, and the source and drain salicidation is not shown. When the gate voltage is greater than a threshold voltage (V_TN), aninversion110 occurs undergate112. This inversion layer, called a conducting channel, completes an electron carrier path between thesource103 and drain104 regions. For the MOSFET in FIG. 1, the terminology inversion layer and conducting channel is used interchangeably, and is shown by110. This conducting channel facilitates current flow betweensource113 and drain114 regions. Hole carrier depletion occurs adjacent to thebody region100 under theinversion layer110 and adjacent tosource113 and drain114 regions. This is shown shaded in FIG. 1. This charge is due to the reversed bias electric fields from the gate, source and drain voltages. The component of this depleted charge from the gate voltage determines the magnitude of the V_TN. Trapped oxide charge and Silicon defects affect the V_TNtransistor parameter. The more positive the voltage is at the gate, the stronger is the inversion layer charge and hence the channel conduction. At all levels, thesubstrate100 potential is kept at the lowest voltage level. In most applications, the substrate and source are held at V_S. For special applications, the NMOS body can be pumped to a negative voltage.

A PMOS device is analogous to an NMOS device, with the device operational polarity and doping types reversed. A PMOS is on when the gate is in the voltage range from system ground V[0008]_Sto a threshold difference (V_D−V_TP), and off when the gate is in the voltage range (V_D−V_TP) to system power voltage V_D. Channel conduction is between P+ doped source and P+ doped drain, via a surface inversion P− layer. The body region originally doped N− gets depleted by the gate potential. The body region for a PMOS is termed Nwell and is constructed on a P type substrate wafer as an isolated island. The Nwell is biased to the highest PMOS device potential, and in most applications the source and Nwell are held at V_D. For special applications, the PMOS body can be pumped to a voltage higher than the power supply voltage.

In a MOSFET device, there is a[0009]

body region

100 under the gate. In fact, a conducting channel is not formed until the surface is in inversion with a build up of minority carriers. The gate depletes the body region near the surface to create this inversion layer at the surface. The depletion width reaches a maximum depth at the onset of inversion, and stays constant at higher gate biases. As the body extends well into the bottom surface of the substrate, the gate modulation has little impact on the resistance of the body region between the source and drain regions. A special case of a MOSFET is a depletion device. In the NMOS depletion device, an N− implanted channel is formed under the gate on the device body surface between N+ source and N+ drain regions. This depletion device has a negative threshold voltage, and a negative gate voltage is needed to turn the device off. The channel is modulated by two terminals: the gate above the oxide, and the body below the channel. The body below has a significant impact on the channel resistance, and in some depletion devices a negative body bias is needed to turn the depletion device off completely.

As discussed in U.S. Pat. No. 5,537,078, conventional JFET transistors are of two main types: P-channel (PJFET) and N-channel (NJFET). FIG. 2 shows a cross section of a JFET device. The description that follows is for an NJFET device. In FIG. 2, a[0010]

semiconductor channel

206 which has been doped N− is positioned between two

N+ diffusions

213 and214. These diffusion nodes are connected to two

terminals

203 and204 respectively. The terminal supplying the majority carrier to the channel (which is the lowest potential) is designated the source (S) while the other terminal is designated the drain (D). Across the N−channel206 there are two gates which are referred to as thetop gate212 and thebottom gate222. Top gate is connected toterminal202, and the bottom gate is connected toterminal232. In some embodiments, the two

gate terminals

202 and232 may be common. Each gate is doped with P+ type dopant to create two back to back P+/N− diodes perpendicular to the channel. When drain and source voltages are different, the drain to source current passes entirely through the conducting N−channel206. This current increases with higher voltage drop between the terminals, reaching a saturation value at high biases. At saturation, the depletion regions meet at a pinch-off point230 near the drain edge as shown in FIG. 2. The gates are biased to keep the gate to channel P+/N− junctions reversed biased. The reversed biased voltage createsdepletion regions210 and220 that penetrate into the channel reducing the channel height available for current flow. The gate voltages also control the flow of current between the source and drain by modulating the channel height. When the gate reverse bias is sufficiently large, the entire channel is pinched-off causing no current flow between drain and source. In on and off states, there is no current flow through the gate terminal of a JFET due to reverse bias junction voltages, except for junction leakage current. For the device in FIG. 2 a negative gate voltage (lower than V_S) creates the channel off condition. Such a negative gate voltage increases the operating voltage of this process, a draw back for JFET scheme.

A PJFET device is analogous to an NJFET device, with the device operational polarity and doping types reversed. A PJFET is on when the gate is at V[0011]_D, and off when the gate is more positive than V_Dfurther increasing the voltage level of the process. Channel conduction is between P+ doped source and drain regions via a P− doped channel sandwiched between two N+ doped gate regions. For source and drain terminals at voltages in the range from V_Sto V_D, operating range of NJFET gate is less than V_Sto V_S, while the operating range for PJFET gate is V_Dto more than V_D.

Compared to the[0012]

non-conducting body

100 of MOSFET on FIG. 1, the JFET has a conductingchannel206 between source and drain. Due to non-overlapping gate voltages and the high voltage range thus needed, a complementary JFET process is impractical to realize. Hence there is no low cost process that provides CJFET devices analogous to CMOS devices. Compared to the MOSFET in FIG. 1, a JFET conducting channel is formed inside the body of the switching device. This channel current is not affected by trapped oxide charges near the gate, a draw back with MOSFETs. Compared to MOSFETs, JFETs also have poorer switching characteristics due to higher depleted charge stored in the channel and the transient times required to store and remove this depletion charge. Reverse biased junctions hurt JFET device ease of use and popularity in modern day ICs.

A special MOSFET device constructed in Silicon-on-Insulator (SOI) is shown in FIG. 3. This three terminal device is constructed as either an NMOSFET or a PMOSFET. The difference in FIG. 1 and FIG. 3 is in the thickness of the[0013]

body region

306 of the device, and in its body isolation. In the SOI device, the

regions

313,306 and314 are constructed on a thin film semiconductor material. Thesubstrate300 is isolated from the device region byinsulator307, hence there is no fourth terminal to this device. This isolation helps with lower junction capacitance and no body effect for SOI MOSFET.Source313, drain314,gate312,spacer320, and

salicided regions

312 and325 are all similar to the standard MOSFET in FIG. 1. Two conditions differ in SOI MOSFET when the device in on. In PD SOI, thebody306 is only partially depleted (PD) when the body is thicker than a maximum depletion width. Then a neutral floating body exists insideregion306 causing deleterious effects on device performance. For thinner FD SOI devices the body is fully depleted (FD) and a neutral body region does not occur. These tend to show short channel effects from the drain and source reverse biased depletions into body region.

Analogous to standard MOSFET, SOI MOSFET also has a non-conducting body under the[0014]

gate

312. Thechannel310 is only formed by inverting the surface. Thebody306 is fully isolated with no access points. The gate modulation of the body has no influence to access ports. Unlike the body, the conducting channel can be accessed via source and drain nodes. There is no analogous device to depletion MOSFET in SOI. This is due to the floating body in an SOI and the inability to control body voltage. Depletion device behavior strongly depends on the body voltage control.

SUMMARY

In one aspect, a semiconductor Gated-FET device comprises of a lightly doped resistive channel region formed on a first semiconductor thin film layer; and an insulator layer deposited on said channel surface with a gate region formed on a gate material deposited on said insulator layer; said gate region receiving a gate voltage having a first level modulating said channel resistance to a substantially non-conductive state and a second level modulating said channel resistance to a substantially conductive state.[0015]

In a second aspect said channel region is formed between a source region and a drain region in the said first semiconductor thin film; and said source region coupled to a source voltage; and said drain region coupled to a drain voltage; and said source and drain regions having a higher level of the same dopant type as said channel region.[0016]

In a third aspect, the Gated-FET device further comprises of an off state with said gate voltage below a first threshold voltage level, and said thin film channel substantially not conducting a current between said drain and source regions for a differential bias voltage ranging from zero to a system power supply voltage; and an on state with said gate voltage above a first threshold voltage level, and said thin film channel substantially conducting a current between said drain and source regions for a differential bias voltage ranging from zero to a system power supply voltage.[0017]

The Gated-FET device is a subset of IGFET devices where the gate is insulated from the channel. This terminology is used to distinguish the new device from MOSFET and JFET devices. A Gated-FET device is a hybrid device between an SOI MOSFET device and a conventional JFET device. The Gated-FET device has a channel region like that of the JFET device: entirely comprising of a thin film resistive channel between its source and drain regions. There is no inversion layer like in an SOI MOSFET to conduct current with no floating body. The gate node of the Gated-FET device is like that of a MOSFET device: the gate constructed above a dielectric material insulating gate from the channel. There is no reverse biased gate junction like in a JFET. The gate voltage thus modulates the channel through the oxide similar to the gate modulation of the SOI MOSFET body region. Unlike in the SOI MOSFET, this modulation occurs in the channel region, which connects the source and drain regions.[0018]

Advantages of the invention may include one or more of the following. A Gated-FET device is used with no increase in voltage range compared to JFET. A Gated-FET device has a threshold voltage not degraded by fixed charge and surface states. A Gated-FET has a channel conductance not degraded by lower surface mobility. A Gated-FET channel current is better controlled with thin film physical properties such as thickness, doping and work function. A Gated-FET has lower charge storage in the channel and faster switching speeds. A Gated-FET has only one gate. A Gated-FET has very low junction capacitance and no body effect. A Gated-FET has no isolated body and no charge trapping effects. A Gated-FET is constructed in a second semiconductor plane, different from a first plane used for logic transistor construction. A Gated-NFET and a Gated-PFET is built on the same process. Gated-FETs are used to build 3D dense integrated circuits. Complementary Gated-FET devices are fabricated in conjunction with regular CMOS devices in a single process. A Gated-NFET and Gated-PFET share a common drain node on a single geometry. The CGated-FETs share a common gate voltage. A switching device is built with as a CGated-FET inverter. The CGated-FET inverter has identical gate voltage range as power supply voltage. A latch is constructed with two CGated-FET inverters connected back to back.[0019]

An off-state Gated-FET thin film transistor body is fully depleted. The depleted channel resistance is non-conductive with no current flow between source and drain. An on-state Gated-FET has a surface accumulation. The accumulation enhances the channel conduction beyond the original dopant level. The enhanced Gated-FET channel conduction is 2 to 100 times more than the intrinsic channel conductance. Thin film Gated-FET has superior device on and off characteristics.[0020]

The method of fabricating the Gated-FET may have one or more of following advantages. A Gated-FET is constructed with III-V semiconductor material. A Gated-FET is constructed with poly-crystalline semiconductor thin film transistors. A Gated-FET is constructed with amorphous poly-Silicon semiconductor thin film transistors. A Gated-FET is constructed with laser re-crystallized poly-Silicon semiconductor thin film transistors. A Gated-FET is constructed on SOI material, or thinned down region of SOI material. A thinned down crystalline SOI Gated-FET has very high performance. The Gated-FET is fabricated in poly-crystalline Silicon layers with good on and off device characteristics. A circuit may be constructed with a conventional MOSFET device, and a new Gated-FET vertically integrated. A TFT module layer may be inserted to a logic process module. The TFT module layer may be inserted to SOI process module. The module insertion may be at a first contact layer. The module insertion may be at a later via layer.[0021]

Implementation of the new device may have one or more following advantages. Gated-FETs are used to build circuits and latches. Inexpensive latches are built with 3D integrated Gated-FET devices. Latches are vertically integrated to a logic process for FPGA applications. A split latch is constructed with regular MOSFET in a first layer, and vertically integrated Gated-FET in a second layer connected back to back. A split latch is used to construct high density SRAM memory. A split SRAM memory is used for high memory content FPGA applications. A split SRAM is used for high density stand alone memory. The split level latch cells have very high performance similar to full CMOS latches. The split level latches have very low power consumption similar to full CMOS SRAM memory. New latches can be used for very fast access embedded memory applications. Thinned down split latch SOI memory allows very high memory densities. The complete TFT latch can be stacked above logic transistors, further reducing Silicon area and cost. Full TFT latches have longer access times, but useful for slow memory applications. Slow TFT latches can be used in PLDs (Programmable Logic Devices) and subsequently mapped to ASICs (Application Specific Integrated Circuit). The PLDs are used for prototyping and low volume production, while the ASICs are used for high volume production. Programmable TFT latches are used in PLD's. Programmable elements are replaced with hard wires in ASICs.[0022]

The invention thus provides an attractive solution for two separate industries: (i) very high density stand alone or embedded low power, fast access SRAM memory and (ii) high-density, low-cost SRAM for PLD and FPGA with convertibility to ASIC. It also provides an alternative set of complementary devices to a traditional SOI MOSFET process for very high density integrated circuit fabrication.[0023]

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 shows a conventional MOSFET device conduction channel.[0024]

FIG. 2 shows a conventional JFET device conduction channel.[0025]

FIG. 3 shows a conventional SOI MOSFET device.[0026]

FIG. 4 shows a Gated-FET device.[0027]

FIGS. 5A and 5B shows a cross sectional view and top view of a Gated-NFET device.[0028]

FIGS. 6A and 6B shows a band diagram for off state Gated-NFET device.[0029]

FIGS. 7A and 7B shows a band diagrams for a first on state Gated-NFET device.[0030]

FIGS. 8A and 8B shows a band diagrams for a second on state Gated-NFET device.[0031]

FIGS. 9A and 9B shows a cross sectional view and top view of a Gated-PFET device.[0032]

FIGS. 10A and 10B shows a band diagram for off state Gated-PFET device.[0033]

FIGS. 11A and 11B shows a band diagrams for a first on state Gated-PFET device.[0034]

FIGS. 12A and 12B shows a band diagrams for a second on state Gated-PFET device.[0035]

FIGS. 13A and 13B shows a top view and cross sectional view of a fabricated Gated-FET device.[0036]

FIG. 14 show constructional cross sections of processing steps showing fabrication of complementary Gated-FET devices.[0037]

DESCRIPTION

The terms wafer and substrate used in the following description include any structure having an exposed surface with which to form the Gated-FET structure of the invention. The term substrate is understood to include semiconductor wafers. The term substrate is also used to refer to semiconductor structures during processing, and may include other layers that have been fabricated thereupon. The term layer is used for processing steps used in the manufacturing process. The term layer also includes each of the masking layers of the process. Both wafer and substrate include doped and undoped semiconductors, epitaxial semiconductor layers supported by a base semiconductor or insulator, SOI material as well as other semiconductor structures well known to one skilled in the art. The term conductor is understood to include semiconductors, and the term insulator is defined to include any material that is less electrically conductive than the materials referred to as conductors. The term geometry is used to define an isolated pattern of a masking layer. One mask layer is a collection of geometries in the mask pattern. The term module includes a structure that is fabricated using a series of predetermined process steps. The boundary of the structure is defined by a first step, one or more intermediate steps, and a final step. The term channel is used to identify a region that connects two other regions. The term body identifies a region common to a plurality of devices. The term body is also used to identify a substrate or a well region. The term body is also used to identify a region other than a conducting region. The following detailed description is, therefore, not to be taken in a limiting sense.[0038]

One embodiment of the Gated-FET is shown in FIG. 4. This device differs from the MOSFET shown in FIG. 1 in three aspects: absence of an inversion layer at the surface, absence of a body region and the thinness of the channel region. This device differs from the JFET device shown in FIG. 2 in three aspects: absence of diffused gate junction, absence of dual gates, and presence of an insulated gate. This device differs from the SOI MOSFET shown in FIG. 3 in two aspects: absence of an inversion layer at the surface, absence of a floating body region.[0039]

Gated-FET in FIG. 4 is comprised of a lightly doped[0040]

resistive channel region

406 formed on a first semiconductorthin film layer480; and aninsulator layer405 deposited on saidchannel406 surface; and agate region412 formed on a gate material deposited on saidinsulator layer405; and saidgate region412 coupled to agate voltage402; and saidgate voltage402 at a first level modulating saidchannel406 resistance to a substantially non-conductive state; and said gate voltage at a second level modulating saidchannel406 resistance to a substantially conductive state.

Gated-FET in FIG. 4 further comprises of said[0041]

channel region

406 formed between asource region413 and adrain region414 in the said first semiconductorthin film480; and said source region coupled to asource voltage403; and said drain region coupled to adrain voltage404; and said source and drain

regions

413 and414 having a higher level of the same dopant type as saidchannel region406.

In the shown embodiment in FIG. 4, the Gated-FET device is constructed on an[0042]

isolation layer

407 that is deposited on asubstrate layer400. In a preferred embodiment theisolation layer407 is an insulator. In another embodiment,isolation layer407 is a semiconductor material, a body region with opposite type dopant to reverse bias and isolate

regions

413,406 and414. Thesubstrate400 can be doped P type or N type, and contain N-well, P-well and any other diffused region. In another embodiment, thesubstrate400 also contains other transistors constructed upon the substrate surface and isolated by theregion407. In FIG. 4 thegate412, drain414 andsource413 regions are shown salicided. These salicided regions for the gate and drain are shown as422 and425. In another embodiment these are not salicided. A vertical side wall of thegate412 is covered by aspacer420, and thesalicided region425 is separated from the gate edge by the width of that spacer. Thespacer425 is self aligned togate412.Salicided region425 is self aligned tospacer420. The device in FIG. 4 can be either a Gated-NFET or a Gated-PFET depending on the dopant types chosen in theregion480.

Regions

413,406 and414 have substantially the same type of dopant.Channel406 is doped lighter than the source and drain

regions

413 and414. A gate material is chosen to satisfy a work function requirement for proper functionality of the device.Insulator405 thickness andsemiconductor480 thickness and dopant level are optimized for device performance. Device design criteria will be discussed in detail next.

FIG. 5 shows a Gated-PFET device. Top view in FIG. 5B shows a[0043]

first semiconductor geometry

5080 orthogonal to agate geometry5022. These two layers are separated by an insulator deposited in between. Thegate geometry5022 is surrounded by aspacer ring5020. Asource voltage5003 is connected via a contact to sourceregion5013 formed onsemiconductor geometry5080. A drain voltage5004 is connected via a contact to drainregion5014 formed onsemiconductor geometry5080.Geometry5080 is subdivided by implant type into different regions. The separation occurs by overlapping nature ofgate geometry5022 andspacer ring5020 abovegeometry5080. The middle region undergate geometry5022 is defined as the channel region. The two regions under thespacer ring5020 are defined as the lightly doped drain (LDP) regions. These regions are better seen in the cross sectional view in FIG. 5A aschannel region506 and

LDP regions

584 and586. In this embodiment, the source and drain

regions

513 and514 are shown completely salicided with no P+ implanted semiconductor regions. The source and drains are entirely determined by the

LDP regions

584 and586. FIG. 5A also shows a poly-Silicon gate material512 partially salicided.Source513, drain514 andgate512 salicidations occur simultaneously and a thicker poly-Silicon film forgate512 ensures partial consumption.

Terminals

502,503 and504 are coupled to the

regions

512,513 and514 respectively completing a three terminal device. There is no fourth terminal in this device as thechannel506 is electrically coupled betweensource513 and drain514. There is no body region for this device. Thechannel506 height is same as the depositedsemiconductor film5080 thickness.

The operation of Gated-PFET is described next. The device has an on threshold voltage V[0044]_TP. Gated-PFET source503 is connected to the higher voltage compared to drain504. Device on-off is determined bygate502 over voltage with respect tosource503. For this discussion, power supply voltage V_Dis taken as the higher voltage. The other voltage is taken as a ground supply voltage V_S. Furthermore, thesource terminal503 is assumed connected to the system power voltage V_D. When thegate voltage502 is between V_Dand (V_D−V_TP), the device is off with no significant current flow between drain and source. When thegate voltage502 is between (V_D−V_TP) and V_Sthe device is on. The drain to source current flow depends on the voltage difference between the two terminals V_DS.

FIG. 6A shows an energy band diagrams for a Gated-PFET comprised of an N+ poly[0045]

Silicon gate material

512,oxide insulator505, and P− dopedchannel506 in FIG. 5A when the gate and channel are both biased tolevel600.Reference level600 is at V_Dvolts. This bias condition occurs near the source edge of an off Gated-PFET device with gate and source at V_Dvoltage.

Levels

601,602 and603 represent the conduction band, mid gap band and the valence band energy levels for N+ doped poly Silicon. For Si semiconductor, the band gap is about 1.12 eV at 300 Kelvin temperature as shown by the difference between

levels

601 and603. For N+ dopant, the Fermi level is at theconduction band level601.

Energy levels

631,632 and634 represent the conduction band, mid gap band and the valence band energy levels for P− channel Silicon. Again the band gap is 1.12 eV for Silicon at 300 Kelvin, shown by the energy difference between631 and634. The Fermi level for P− Silicon is shown bylevel633. The P− Silicon has a higher vacuum level electron energy compared to N+ poly Silicon. In the diagram, no oxide charge is assumed, and the difference in energy level is the work function difference between the two materials. In the diagram work function difference is assumed to equal flat band voltage. Presence of fixed oxide charge reduces the flat band voltage. There is a net electron transfer from Silicon side to the Gate side causing band bending as shown in FIG. 6A. This creates voltage drop across the oxide and Silicon according to the laws of Semiconductor Physics. The total voltage drop equals the flat band voltage V_FB. The voltage drop in the oxide is shown by thedifference651 with a uniform electric field in the absence of fixed charge inside the oxide. The voltage drop inside Silicon creates aband bending region640 with adepth641 into Silicon from the oxide interface. This region is depleted of majority carrier holes. There is no supply of minority carriers (an N type diffusion region) to cause a surface inversion layer and pin the band bending at the surface as shown byregion110 in FIG. 1 for a MOSFET device. In FIG. 6A, athickness642 is shown to be significantly depleted of carriers. That region has no conducting majority carrier holes. Choosing an appropriateSilicon channel height642 allows building a Gated-PFET channel that does not conduct when a V_Dvoltage is applied on the gate.

FIG. 6B is a Gated-PFET energy band diagram when the gate is at V[0046]_Dand the channel at zero or V_Svolts. This condition can occur at the drain edge of an off Gated-PFET when the drain has to support V_S. In FIG. 6B, theSilicon Fermi level6033 is at V_Swhile thegate Fermi level6001 is at V_D. The two

Fermi levels

6033 and6001 are separated by thebias voltage6061 which has a value V_D. Thedepletion region6040 is deeper than640, extending to adepth6041 into the Silicon. A thin layer channel that is fully depleted in FIG. 6A remains fully depleted in FIG. 6B. Such a channel has no conduction between a source and a drain that are biased to V_Dand V_Svoltages respectively. Regardless of drain voltage, the Gated-PFET is off when the gate is at V_D. This can be seen by the increased band bending shown in FIG. 6B.

As the gate voltage decrease from V[0047]_Dto a value (V_D−V_TP), the voltage drop across the oxide decreases, and theSilicon depletion width641 also decreases. V_TPis chosen such that there is a clear noise margin on the threshold level of the Gated-PFET against power and ground voltage variations. At that threshold, thedepletion width641 falls to within thefilm thickness642 shown in FIG. 6A at the source edge. That creates an onset of conduction current between the source and drain, even if the drain edge is still fully depleted and pinched off. The drain edge pinch off contributes to creating a saturation current.

At a gate voltage higher than the threshold the bands attain a flat band level as shown in FIG. 7A. This flat band voltage is defined V[0048]_FBand is shown byvoltage level761. In FIG. 7A the gate is biased to V_FBwhile the channel is held at V_D. Hence, Fermi level733 in the Silicon is at V_D, while the Fermi level of N+ poly is at V_FB. The voltage drop across theoxide751 is zero, and the band bending in the Silicon is also zero. When there is no fixed oxide charge in the oxide, V_FBequals the work-function difference between gate and Silicon. There is no meaning to a depletion region under this bias condition, and the entireSilicon film thickness742 has majority carriers at the doping level of the Silicon. This is seen by the flat Fermi level733 in FIG. 7A. FIG. 7B shows the gate at V_FBand the channel at drain edge biased to V_S.Silicon Fermi level7033 is at zero (V_S) volts, whileGate Fermi level7001 is at V_FB. Voltage level7061 shows the applied channel voltage V_Sagainstreference level700 at V_D. The over voltage between V_FBand V_Sis now dropped across the oxide and the Silicon creating adepletion region7040 in Silicon. Such depletion causes a current saturation in the conducting channel. Hence at a bias V_G=V_FB, the Gated-PFET conducts with a current saturation occurring at a voltage when the channel depth is pinched-off near the drain edge.

A Gated-PFET with a gate biased at ground (V[0049]_S) is shown in FIGS. 8A and 8B. In FIG. 8A the channel near source edge is at voltage V_D.Gate Fermi level801 is at V_S, while the channel Fermi level833 is at V_D. The over voltage between V_FBand V_Dis now dropped across the oxide and the Silicon creating anaccumulation region840 near the channel surface. The majority carrier concentration is now far higher than the original doping level of the Silicon layer. This provides enhanced channel conduction beyond the doping level chosen for the film. Thechannel film thickness842 is chosen thicker than the accumulation width to facilitate a strong on current for the Gated-PFET. FIG. 8B shows the gate and channel both at V_S. Both

Fermi levels

8001 and8033 are at V_S, above reference level800 by a value V_D. The band diagram is identical to FIG. 6A when both sides were biased at V_D. Again the drain edge of the channel is pinched-off demonstrating the existence of current saturation in the device.

The diagrams shown in FIGS. 6, 7 and[0050]8 have consistent labels. All diagrams consistently show the existence of a thin film channel region that will allow construction of a Gated-PJFET device by ensuring an on state and an off state. The channel is fully depleted of majority carriers in the off state. The gate and the semiconductor material properties need to be chosen to ensure this condition. The flat band voltage for the system needs to be large enough to fully deplete the chosen channel thickness when the device is off. In the embodiment chosen an N+ doped poly Silicon gate material, an oxide dielectric and a P− doped Silicon channel meets that condition. A thinner dielectric thickness and a lower dielectric constant material for the gate insulator allow a lower voltage loss across the dielectric and a larger channel modulation in the Silicon.

Next we will discuss a Gated-NFET device as shown in FIG. 9. Top view in FIG. 9B shows a[0051]

first semiconductor geometry

9080 orthogonal to agate geometry9022. These two layers are separated by an insulator deposited in between. Thegate geometry9022 is surrounded by aspacer ring9020. Asource voltage9003 is connected via a contact to sourceregion9013 formed onsemiconductor geometry9080. Adrain voltage9004 is connected via a contact to drainregion9014 formed onsemiconductor geometry9080.Geometry9080 is subdivided by implant type into different regions. The separation occurs by overlapping nature ofgate geometry9022 andspacer ring9020 abovegeometry9080. The middle region undergate geometry9022 is defined as the channel region. The two regions under thespacer ring9020 are defined as the lightly doped drain (LDN) regions. These regions are better seen in the cross sectional view in FIG. 9A aschannel region906 and

LDN regions

984 and986. In this embodiment, the source and drain

regions

913 and914 are shown completely salicided with no N+ implanted semiconductor regions. The source and drains are entirely determined by the LDN regions. FIG. 9A also shows a poly-Silicon gate material912 partially salicided.Source913, drain914 andgate912 salicidations occur simultaneously and a thicker poly-Silicon film forgate912 ensures partial consumption.

Terminals

902,903 and904 are coupled to the

regions

912,913 and914 respectively completing a three terminal device. There is no fourth terminal in this device as thechannel906 is electrically coupled betweensource913 and drain914. There is no body region for this device. Thechannel906 height is same as the depositedsemiconductor film9080 thickness.

The operation of Gated-NFET is described next. The device has an on threshold voltage V[0052]_TN. Gated-NFET source903 is connected to the lower voltage compared to drain904. Device on-off is determined bygate902 over voltage with respect tosource903. For this discussion, ground supply voltage V_Sis taken as the lower voltage. The other voltage is taken as a power supply voltage V_D. Furthermore, thesource terminal903 is assumed connected to the system ground voltage V_S. When thegate voltage902 is between V_Sand V_TN, the device is off with no significant current flow between drain and source. When thegate voltage902 is between V_TNand V_Dthe device is on. The drain to source current flow depends on the voltage difference between the two terminals V_DS.

FIG. 10A shows an energy band diagrams for a Gated-NFET comprised of a P+ poly[0053]

Silicon gate material

1010,oxide insulator1020, and N− dopedchannel1040 when the gate and channel are biased at V_Svolts atlevel1000. This bias condition occurs near the source edge of an off device with gate and source at V_Svoltage.

Levels

1001,1002 and1003 represent the conduction band, mid gap band and the valence band energy levels for P+ poly Silicon. For Si semiconductor, the band gap is about 1.12 eV at 300 Kelvin temperature as shown by the difference between

levels

1001 and1003. For P+ dopant, the Fermi level is at thevalence band level1003.

Energy levels

1031,1033 and1034 represent the conduction band, mid gap band and the valence band energy levels for P− channel Silicon. Again the band gap is 1.12 eV for Silicon at 300 Kelvin, shown by the energy difference between1031 and1034. The Fermi level for N− Silicon is shown bylevel1032. The N− Silicon has a lower vacuum level electron energy compared to P+ poly Silicon. In the diagram, no oxide charge is assumed, and the difference in energy level is the work function difference between the two materials. There is a net electron transfer from Gate side to the Silicon side causing band bending as shown in FIG. 10A. This creates voltage drop across the oxide and Silicon according to the laws of Semiconductor Physics. The voltage drop in the oxide is shown by the difference1051 with a uniform electric field in the absence of fixed charge inside the oxide. The voltage drop inside Silicon creates aband bending region1040 with adepth1041 into Silicon from the oxide interface. This region is depleted of majority carrier electrons. There is no supply of minority carriers (a P type diffusion region) to cause a surface inversion layer and pin the band bending at the surface as shown byregion110 in FIG. 1 for a MOSFET device. In FIG. 10A, athickness1042 is shown to be significantly depleted of carriers near the surface. That region has no conducting majority carrier electrons. Choosing an appropriateSilicon channel height1042 allows constructing a Gated-NFET channel that does not conduct when a V_Svoltage is applied on the gate.

FIG. 10B is a Gated-NFET energy band diagram when the gate is at V[0054]_Sand the channel at V_Dvolts. This can occur at the drain edge of an off Gated-NFET when the drain has to support a voltage V_D. In FIG. 10B, theSilicon Fermi level10032 is at V_Dwhile the gate Fermi level10003 is at V_S. The twoFermi levels10032 and10003 are separated by thebias voltage10061 which has a value V_D. Thedepletion region10040 is deeper than1040, extending to adepth10041 into the Silicon. A thin layer channel that is fully depleted in FIG.10A remains fully depleted in FIG. 10B. Such a channel has no conduction between a source and a drain that are biased to V_Sand V_Dvoltages respectively. Regardless of drain voltage, the Gated-NFET is off when the gate is at V_S. This can be seen by the enhanced band bending shown in FIG. 10B.

As the gate voltage increase from V[0055]_Sto a value V_TN, the-voltage drop across the oxide decreases, and the Silicon depletion width also decreases. V_TNis chosen such that there is a clear noise margin on the threshold level of the Gated-NFET against power and ground voltage variations. At that threshold, thedepletion width1041 falls to within thefilm thickness1042 shown in FIG. 10A at the source edge. That creates an onset of conduction current between the source and drain, even if the drain edge is still fully depleted and pinched off. The drain edge pinch off contributes to creating a saturation current.

At a gate voltage higher than the threshold the bands attain a flat band level as shown in FIG. 11A. This flat band voltage is defined V[0056]_FBand is shown byvoltage level1161. Fixed oxide charge affect V_FB. In FIG. 11A the gate is biased to V_FBwhile the channel is held at V_S. Hence, Fermi level1132 in the Silicon is at V_S, while the Fermi level ofP+ poly1103 is at V_FB. The voltage drop across the oxide1151 is zero, and the band bending in the Silicon is also zero. When there is no fixed oxide charge in the oxide, V_FBequals the work-function difference between gate and Silicon. There is no meaning to a depletion region under this bias condition, and the entireSilicon film thickness1142 has majority carriers at the doping level of the Silicon. This is seen by the flat Fermi level1132 in FIG. 11A. FIG. 11B shows the gate at V_FBand the channel at drain edge biased to V_D.Silicon Fermi level11032 is at V_Dvolts, whileGate Fermi level11003 is at V_FB. Voltage level11061 shows the applied channel voltage V_Dagainstreference level1100 at V_S. The over voltage between V_FBand V_Dis now dropped across the oxide and the Silicon creating adepletion region11040 in Silicon. Such depletion causes a current saturation in the conducting channel. Hence at a bias V_G=V_FB, the Gated-NFET conducts with a current saturation occurring at a voltage when the channel depth is pinched-off near the drain edge.

A Gated-NFET with a gate biased at power supply V[0057]_Dis shown in FIGS. 12A and 12B. In FIG. 12A the channel is at source edge with a voltage V_S. Gate Fermi level1203 is at V_D, while thechannel Fermi level1232 is at V_S. The over voltage between V_FBand V_Dis now dropped across the oxide and the Silicon creating anaccumulation region1240 near the channel surface. The majority carrier concentration is now far higher than the original doping level of the Silicon layer. This provides enhanced channel conduction beyond the doping level chosen for the film. Thechannel film thickness1242 is chosen thicker than the accumulation width to facilitate a strong on current for the Gated-NFET. FIG. 12B shows the gate and channel both at V_D. BothFermi levels12003 and12032 are at V_D, abovereference level1200 by a value V_D. The band diagram is identical to FIG. 10A when both sides were biased at V_S. Again the drain edge of the channel is pinched-off demonstrating the existence of current saturation in the device.

The diagrams shown in FIGS. 10, 11 and[0058]12 have consistent labels. All diagrams consistently show the existence of a thin film channel region that will allow construction of a Gated-NJFET device by ensuring an on state and an off state. The channel is fully depleted of majority carriers in the off state. The gate and the semiconductor material properties need to be chosen to ensure this condition. The flat band voltage for the system needs to be large enough to fully deplete the chosen channel thickness when the device is off. In the embodiment chosen a P+ doped poly Silicon gate material, an oxide dielectric and an N− doped Silicon channel meets that condition. A thinner dielectric thickness and a lower dielectric constant material for the gate insulator allow a lower voltage loss across the dielectric and a larger channel modulation in the Silicon.

The lightly doped resistive channel region formed on a first semiconductor[0059]

thin film geometry

480 forming the conducting paths betweensource413 and drain414 in FIG. 4 can be a thinned down SOI single crystal Silicon film, or a deposited thin Poly-crystalline Silicon film, or a post laser annealed as deposited amorphous Poly-crystalline Silicon film. The thickness and doping of thechannel region406 are optimized with theinsulator405 thickness T_Gand gate material work function to get the required threshold voltage Vt, on-current and off-current for these devices. Thechannel406 thickness T_Soptimization to contain the fully depleted channel as discussed earlier is discussed in detail next. Two thickness parameters X and Y for a semiconductor material are defined by:

X=ε_S*T_G/ε_GAngstroms (EQ 1)

Y=[(2*ε_S*V_FB)/(q*D)]^0.5Angstroms (EQ 2)

X_D=(X²+Y²)^0.5−XAngstroms (EQ 3)

T_S<X_DAngstroms (EQ 4)

where, ε[0060]_Sis channel semiconductor permittivity, ε_Gis gate insulator permittivity, T_Gis gate insulator thickness, V_FBis gate to semiconductor absolute flat band voltage, q is electron charge, D is channel doping level, X_Dis the depletion depth and T_Sis channel semiconductor layer thickness. EQ-3 denotes the maximum depletion width for the off Gated-FET shown asdepth641 in FIG. 6A anddepth1041 in FIG. 10A. The inequality in EQ-4 ensures film thicknesses642 and1042 shown FIGS. 6A and 10A respectively are within the

maximum depletion depths

641 and1041 into Silicon channel. Preferably T_Schosen to be in the range 0.2*X_Dto 0.9*X_D, and more preferably T_Sis chosen to be in the range 0.4*X_Dto 0.8*X_Drange.

For most practical doping levels and oxide thicknesses, X is much larger than Y value. EQ-3 can be simplified to:[0061]

X_D=Y−X+X²/(2*Y) Angstroms (EQ 5)

For Poly-Oxide-Silicon Gated-FET devices, when D is 2E17 Atomc/cm[0062]³doping density (i.e. 2E−7 Atoms/Å³), T_G=70 Å, ε_S/ε_OX=3, and assuming no fixed charge in the oxide the following is easily shown: the flat band voltage V_FB=0.987V, X=210 Å, and Y=799 Å. Using EQ-3, X_D=616 Å. EQ-5 also yields X_D=616 Å as Y>X criterion is met. Hence a semiconductor layer preferably 120-550 Å, more preferably 250-490 Å meets the channel thickness requirement. For the Poly-Oxide-Silicon Gated-FET device, a simplified practical criterion can be extracted from EQ-5 as:

X_D˜{square root}D*(0.36/D+12.5*T_OX²)−3*T_OXAngstroms (EQ 6)

Where, D is in Atoms/Å[0063]³. This expression assumes a V_FB=1V. For the example discussed earlier, EQ-6 yields X_D˜622A, in fairly good agreement to the correct 616A.

The insulator thickness and channel doping also needs to satisfy the threshold voltage for the Gated-FET device. This threshold voltage is preferable selected in the range 0.18*V[0064]_Dto 0.4*V_D, and more preferably 0.2*V_Dto 0.3*V_D, where V_Dis the power supply voltage. This puts an added constraint on the

semiconductor film thickness

642 and1042 shown in FIG. 6A and FIG. 10A respectively. When a voltage V_Tis applied at the gate, the depletion width641 (or1041) equals semiconductor thickness642 (or1042) for the Gated-FET device. At that bias, EQ-2 is modified by the additional bias, to a new thickness defined by:

Z=[(2*ε_S*(V_FB−V_T))/(q*D)]^0.5Angstroms (EQ 7)

T_S=(X²+Z²)^0.5−XAngstroms (EQ 8)

EQ-8 shows the relationship between doping level D and semiconductor film thickness T[0065]_Srequired to satisfy the Gated-FET design. EQ-8 satisfies the constraint for maximum semiconductor film thickness in EQ-4 trivially. For the example discussed earlier, for a 1.5V process with V_Tat 0.42V, EQ-7 yields Z=606 Å, and from EQ-8 T_S=431 Å well within the desired thickness range 300-500 Å. The gate dielectric thickness and dielectric fixed charge density impacts this threshold voltage.

Other embodiments may use gate and substrate materials different from Silicon. Gate dielectrics can be oxide, oxy-nitride, nitride, or multi-layered insulators. The semiconductor material may be Silicon, Silicon-germanium, gallium-arsenide, germanium, or any other III-V material. The gate material may be poly-Silicon, aluminum, tungsten, or any other metal. The value of X in equation-1 will change based on the physical properties of the materials chosen to form the Gated-FET device.[0066]

The total resistance of the conducting body region for Gated-FET under conducting mode is determined by the applied voltage difference between drain and source nodes, and gate over voltage above threshold. A typical device top vies and cross section is shown in FIG. 13. In addition, the device channel width[0067]1391 (W_S), device channel length1392 (L_S) and film thickness1393 (T_S) all determine the device on current. The channel resistance is given by:

R=ρ_S*L_S/(W_S*T_S) Ohms (EQ 9)

where, ρ[0068]_Sis the resistivity of as dopedchannel region1340. Gate voltage and channel depletion heavily modulates resistivity ρ_S. Parameters are chosen for R to be preferably in the 1 KOhm to 1 Meg-Ohm range, more preferably 10 KOhm to 100 KOhms, when the channel is on. As an example, for P− doping 2E17 atoms/cm³, under flat-vand conditions in FIG. 7A, the resistivity for single crystal Silicon is 0.12 Ohm-cm. When L_S=0.3μ, W_S=0.3μ, T_S=431 Angstroms, R is 27.8 KOhms. When V_DS=0.2V (drain to source bias), the channel current I_ONis 7 μA. This is the conduction level under flat band bias condition in FIG. 7A for the Gated-PFET. Poly-Silicon mobility is lower than single crystal Silicon degrading the on current for non single-crystal films. The surface accumulations shown in FIG. 8A and FIG. 12A for Gated-PFET and Gated-NFET devices enhance the channel I_ONcurrent for higher gate biases. An effective film thickness increase is used to express the channel resistance as defined by:

R=ρ_S0*L_S/[W_S*((1+γ)T_S)] Ohms (EQ 10)

Where γ absorbs the channel modulation effects, and ρ[0069]_S0remains the resistivity at doping level D of the thin film channel region. The γ value depends on the depth of the accumulation region into thin film channel, and the surface potential at the Semiconductor-Insulator interface in the band diagrams in FIG. 8A and FIG. 12A. If the absolute surface potential is Φ_Sin the semiconductor, a system of equations can be solved by trial and error to converge on the correct Φ_Svalue.

N_S=D*exp(qΦ_S/kT) (EQ 11)

L_D=[ε_S*kT/(q²N_S)]^0.5 (EQ 12)

X_A={square root}2*L_D*[(N_S/D)^0.5−1] (EQ 13)

Q_S=q*N_S*X_A/(1+X_A/({square root}2*L_D)] (EQ 14)

(Φ_S=V_D−V_FB−T_G*Q_S/ε_G (EQ 15)

Referring to FIG. 8A (or equivalently FIG. 12A), EQ-11 denotes the excess surface concentration of majority carriers due to the surface potential Φ[0070]_Sat the surface. EQ-12 denotes the Debye length at the surface concentration N_S. EQ-13 denotes the depth of accumulation layer penetration into semiconductor region. At this depth, the doping level drops off to the channel doping level D, and no accumulation is further observed. EQ-14 denotes the total excess accumulation charge in the semiconductor thin film due to accumulation. EQ-15 is the voltage balance where the over voltage above flat-band is distributed across gate insulator and semiconductor film. A consistent solution to the system of equations 11-15 can be achieved iteratively. To achieve the full benefit from the enhanced conduction due to accumulation, the film thickness is further chosen such that:

T_S>X_A (EQ 16)

Preferably T[0071]_Sis chosen in the range 1*X_Ato 10*X_A, and more preferably T_Sis chosen in the range 2*X_Ato 6*X_A. When EQ-16 is met, the effective thickness increase factor γ due to accumulation can be expressed as:

γ={square root}2*(L_D/T_S)*[(N_S/D)−(N_S/D)^0.5] (EQ 17)

This shows that the accumulation effectively acts so as to increase the thin film thickness beyond the chosen T[0072]_Svalue at the same doping density D. This enhancement can be quite significant. For the example chosen earlier, for 1.5V power supply, the over voltage (V_D−V_FB)=0.513V. Start with a guess surface voltage Φ_S=0.0927V. Substituting in EQ-11 through EQ-15: N_S=7.21E18 Atoms/cm³, L_D=15.2 Å, X_A=108 Å, Q_S=2.07E−7 C. Substituting into EQ-15 the surface voltage is recalculated as Φ_S=0.0926V same as the starting point. Thickness enhancement factor is calculated from EQ-17 as γ=1.50. This γ is very sensitive to V_Dover voltage above V_FB. For a 1.8V power supply, γ=2.47. When L_S=0.3μ, W_S=0.3μ, T_S=431 Angstroms, new R under accumulated surface condition is 11.1 KOhms. When V_DS=0.2V (drain to source bias), the channel current I_ONis 18 μA, a significant increase over the flat-band gate voltage bias condition. For this example, the condition in EQ-17 is met as the film thickness 431 Å is larger than the accumulation depth 108 Å. Under flat-band condition, or when V_D<V_FB, there is no surface accumulation and EQ-16 simply reduces to T_S>0 Å.

The following terms used herein are acronyms associated with certain manufacturing processes. The acronyms and their abbreviations are as follows:[0073]



	V_T	Threshold voltage
	V_TN	Gated-NFET Threshold voltage
	V_TP	Gated-PFET Threshold voltage
	LDN	Lightly doped Gated-NFET drain
	LDP	Lightly doped Gated-PFET drain
	LDD	Lightly doped drain
	RTA	Rapid thermal annealing
	Ni	Nickel
	Ti	Titanium
	Co	Cobalt
	Si	Silicon
	TiN	Titanium-Nitride
	W	Tungsten
	S	Source
	D	Drain
	G	Gate
	ILD	Inter layer dielectric
	C1	Contact-1
	M1	Metal-1
	P1	Poly-1
	P−	Positive light dopant (Boron species, BF₂)
	N−	Negative light dopant (Phosphorous, Arsenic)
	P+	Positive high dopant (Boron species, BF₂)
	N+	Negative high doparit (Phosphorous, Arsenic)
	Gox	Gate oxide
	C2	Contact-2
	CVD	Chemical vapor deposition
	LPCVD	Low pressure chemical vapor deposition
	PECVD	Plasma enhanced CVD
	ONO	Oxide-nitride-oxide
	LTO	Low temperature oxide

The device shown in FIG. 13, and discussed in the example earlier has P+ doped poly-Silicon gate over N− doped channel for the Gated-NFET, and N+ doped poly-Silicon gate over P− doped channel for the Gated-PFET. This is easily achieved in the fully salicided source/drain embodiment shown in FIG. 13. The Gated-NFET and Gated-[0074]

PFET gate regions

1312 are first doped P+ and N+ respectively before the gates are etched. After gates are etched, prior tospacer1320 formation, Gated-NFETs are implanted with N type LDN tip implant and Gated-PFETs are implanted with P type LDP tip implant. The LDD tip-implant dose is much lower than the gate doping to affect gate doping type. The Source & Drain regions are now defined by the self aligned LDDtip implants regions1326 shown under thespacer oxides1320 adjacent to thegate1312 regions in FIG. 13B. As thedrain1314 andsource1313 regions outside the spacer are fully consumed by salicide, those regions do not need heavy doping. Thechannel1306 doping levels N− for Gated-NFET and P− for Gated-PFET are chosen to achieve the desirable V_Tas discussed earlier. The first semiconductorthin film layer1306 forming thesource1313,LDD tips1326,channel1306 and drain1314 can be thinned down SOI single crystal Silicon material, or a first thin-film PolySilicon layer, or a laser crystallized amorphous Silicon layer, or any other thin film semiconductor layer. A thicker first film allows higher current.

The gate dielectric[0075]1305 is grown either thermally or deposited by PECVD. The first thin film layer1306 (P1) forms the body of the transistor. The P1 layer is deposited above the insulator layer1307. The insulator is oxide, or nitride, or a reversed bias doped semiconductor region (in the case when source and drain regions are not fully salicided) that can isolateP1 geometry1380. A P1 mask is used to define and etch these P1 islands. Gated-PFET regions are mask selected and implanted with P− doping, and gated N-FET devices are implanted with N− doping, the channel doping V_Tlevels required for Gated-FET devices. Thegate1312 is deposited after the gate insulator1305 is deposited as a second thin film semiconductor layer (P2). In the embodiment shown, the second thin film layer is a ploySilicon layer. The Gated-PFET gate poly1312 is mask selected and implanted N+ and Gated-NFET is implanted with P+ prior to gate definition and etch. The gate regions are then defined and etched. A P tip (LDP) implant is used over all Gated-PFET devices, and an N tip (LDN) implant is used for Gated-NFET devices. This can be done by open selecting Gated-PFET devices, and not selecting Gated-NFET devices and visa-versa. The N+ and P+ doped gates are not affected by the lower N and P tip implant level.Gate1312 blocks P and N tip implants getting intochannel region1340, and only P1 regions outside P2 gets this tip implant.Spacer oxide regions1320 are formed on either side of gate by conventional oxide deposition and etch back techniques. In FIG. 13A, theP2 gate1312 is perpendicular toP1 geometry1380. TheP2 gate1312 andspacers1320 sub-divide the P1 geometry into five regions: (1)source region1313, (2)source spacer region1326 doped with LDD tip implant, (3)channel region1306 doped with V_Timplant, (4)drain spacer region1326 also doped with LDD tip implant and (5)drain region1314. The source and drain regions are fully salicided and need no implant. After the spacer etch, exposed P2 and P1 regions are reacted with deposited Nickel (or Cobalt) and salicided using Rapid Thermal Annealing. The LDD tip implant after P2 etch forms self-aligned Source/Drain LDD tip regions and salicidation after spacer etch forms self aligned Source/Drain salicide regions adjacent to spacer regions. After excess Ni etch, a dielectric film1366 is deposited and C21367 is defined and etched. These are W-filled and polished. A M11368 is deposited, defined and etched, and a dielectric1369 deposited to add multiple layers of metal in the process.

For the device in FIG. 13 a high quality P1 film is beneficial. The terms P1 refers to the first semiconductor layer in FIG. 13 and P2 refers to the second semiconductor layer in FIG. 13 forming the gate. An ideal film is a single crystal Silicon with a precise thickness control over an insulator. In SOI technology, the single crystal Silicon layer above an insulator meets this criterion. Inside a Gated-FET array, P1 is mask selected and thinned down to the required thickness as defined by EQ-8. This allows formation of two sets of transistors adjacent to each other: regular SOI MOSFET and thinned down SOI Gated-FET.[0076]

In one embodiment, a logic process is used to fabricate CMOS devices on a substrate layer. These CMOS devices may be used to build AND gates, OR gates, inverters, adders, multipliers, memory and other logic functions in an integrated circuit. A Complementary Gated-FET TFT module layer is inserted to a logic process at a first contact mask to build a second set of TFT Gated-FET devices. An exemplary logic process may include one or more following steps:[0077]

P-type substrate starting wafer[0078]

Shallow Trench isolation: Trench Etch, Trench Fill and CMP[0079]

Sacrificial oxide[0080]

PMOS V[0081]_Tmask & implant

NMOS V[0082]_Tmask & implant

Pwell implant mask and implant through field[0083]

Nwell implant mask and implant through field[0084]

Dopant activation and anneal[0085]

Sacrificial oxide etch[0086]

Gate oxidation/Dual gate oxide option[0087]

Gate poly (GP) deposition[0088]

GP mask & etch[0089]

LDN mask & implant[0090]

LDP mask & implant[0091]

Spacer oxide deposition & spacer etch[0092]

N+ mask and NMOS N+ G, S, D implant[0093]

P+ mask and PMOS P+ G, S, D implant[0094]

Ni deposition[0095]

RTA anneal—Ni salicidation (S/D/G regions & interconnect)[0096]

Unreacted Ni etch[0097]

ILD oxide deposition & CMP[0098]

FIG. 14 shows an exemplary process for fabricating a thin film Gated-FET device in a module layer. In one embodiment the process in FIG. 14 forms a Gated-FET device in a layer substantially above the substrate layer. The processing sequence in FIGS. 14.[0099]1 through14.8 describes the physical construction of a Gated-FET device shown in FIG. 13. The process shown in FIG. 14 includes adding one or more following steps to the logic process after ILD oxide CMP step.

C1 mask & etch[0100]

W-Silicide plug fill & CMP[0101]

˜400A poly P1 (crystalline poly-1) deposition[0102]

P1 mask & etch[0103]

Blanket V[0104]_TNN− implant (Gated-NFET V_T)

V[0105]_TPmask & P− implant (Gated-PFET V_T)

TFT Gox (70A PECVD) deposition[0106]

600A P2 (crystalline poly-2) deposition[0107]

Blanket P+ implant (Gated-NFET gate & interconnect)[0108]

N+ mask & implant (Gated-PFET gate & interconnect)[0109]

P2 mask & etch[0110]

Blanket LDN Gated-NFET N tip implant[0111]

LDP mask and Gated-PFET P tip implant[0112]

Spacer LTO deposition[0113]

Spacer LTO etch to form spacers & expose P1[0114]

Ni deposition[0115]

RTA salicidation and poly re-crystallization (exposed P1 and P2)[0116]

Fully salicidation of exposed P1 S/D regions[0117]

Dopant activation anneal[0118]

Excess Ni etch[0119]

ILD oxide deposition & CMP[0120]

C2 mask & etch[0121]

W plug formation & CMP[0122]

M1 deposition and back end metallization[0123]

The TFT process technology consists of creating Gated-PFET and Gated-NFET poly-Silicon transistors. In the embodiment in FIG. 14, the module insertion is after the substrate device gate poly etch and the ILD film deposition. In other embodiments the insertion point may be after M1 and the subsequent ILD deposition, prior to V1 mask, or between two other metal definition steps.[0124]

In the logic process, after gate poly of regular transistors are patterned and etched, the poly is salicided using Cobalt or Nickel & RTA sequences. Then an ILD is deposited, and polished by CMP techniques to a desired thickness. In the shown embodiment, the contact mask is split into two levels. The first C1 mask contains all contacts that connect Gated-FET outputs to substrate transistor gates or diffusion nodes. Then the C1 mask is used to open and etch contacts in the ILD film. Ti/TiN glue layer followed by W-Six plugs, W plugs or Si plugs may be used to fill the plugs, then CMP polished to leave the fill material only in the contact holes. The choice of fill material is based on the thermal requirements of the TFT module in subsequent steps. Si plugs allow RTA thermal oxidation of P1 at a subsequent step.[0125]

Then, a first P1 poly layer, amorphous or crystalline, is deposited by LPCVD to a desired thickness as shown in FIG. 14.[0126]1. The P1 thickness is between 50A and 1000A, and preferably between 200-600A. This poly layer P1 is used for the channel, source, and drain regions for both Gated-FETs. P1 is patterned and etched to form the transistor geometries. In other embodiments, P1 is used as contact pedestals to stack a C2 contact above C1. Gated-NFET transistors are blanket implanted with N− doping, while the Gated-PFET transistor regions are mask selected and implanted with P− doping. This is shown in FIG. 14.2. The implant doses and P1 thickness are optimized to get the required threshold voltages for Gated-PFET & Gated-NFET devices under fully depleted transistor operation, and maximize on/off device current ratio. The pedestals implant type is irrelevant at this point. In another embodiment, the V_Timplantation is done with a mask P− implant followed by masked N− implant. First doping can also be done in-situ during poly deposition or by blanket implant after poly is deposited.

Patterned and implanted P1 may be subjected to dopant activation and crystallization. In one embodiment, RTA cycle is used to activate & crystallize the poly after it is patterned to near single crystal form. In a second embodiment, the gate dielectric is deposited, and buried contact mask is used to etch areas where P1 contacts P2 layer. Then, Ni is deposited and salicided with RTA cycle. All of the P1 in contact with Ni is salicided, while the rest poly is crystallized to near single crystal form. Then the unreacted Ni is etched away. In a third embodiment, amorphous poly is crystallized prior to P1 patterning with an oxide cap, metal seed mask, Ni deposition and MILC (Metal-Induced-Lateral-Crystallization).[0127]

Then the TFT gate dielectric layer is deposited followed by P2 layer deposition. The dielectric is deposited by PECVD techniques to a desired thickness in the 30-200A range, desirably 30-100A thick. The gate may be grown thermally by using RTA when C1 plug fill is doped Silicon. This gate material could be an oxide, nitride, oxynitride, ONO structure, or any other dielectric material combination used as gate dielectric. The dielectric thickness is determined by the voltage level of the process. At this point an optional buried contact mask BC may be used to open selected P1 contact regions, etch the dielectric and expose P1 layer. BC could be used on P1 pedestals to form P1/P2 stacks over C1. In the P1 salicided embodiment using Ni, the dielectric deposition and buried contact etch occur before the crystallization. In the preferred embodiment, no BC is used.[0128]

Then second poly P2 layer, 200A to 1000A thick, preferably 300-800A is deposited as amorphous or crystalline poly-Silicon by LPCVD as shown in FIG. 14.[0129]3. Then Gated-NFET devices & P+ poly interconnects are blanket implanted with P+ implant. The implant energy ensures full dopant penetration into the P2 layer. This doping gets to only P2 regions as no P1 regions are exposed. An N+ mask is used to select Gated-PFET devices and N+ interconnect, and implanted with N+ dopant as shown in FIG. 14.4. Transistor gate regions of Gated-PFET and Gated-NFET are doped to the correct dopant type. Source and drain regions are blocked by P1 and not implanted. This N+/P+ implants can be done with N+ mask followed by P+ mask. The V_Timplanted P1 regions are completely covered by P2 layer and form channel regions of Gated-NFET & Gated-PFET transistors.

P2 layer is defined into Gated-NFET & Gated-PFET gate regions intersecting the P1 layer channel regions, C1 pedestals if needed, and local interconnect lines and then etched as shown in FIG. 14.[0130]5. The P2 layer etching is continued until the dielectric oxide is exposed over P1 areas uncovered by P2 (source, drain, P1 resistors). As shown in FIG. 13A, the source & drain P1 regions orthogonal to P2 gate regions are now self aligned to P2 gate edges. The S/D P2 regions may contact P1 via buried contacts. Gated-NFET devices are blanket implanted with LDN N dopant. Then Gated-PFET devices are mask selected and implanted with LDP P dopant as shown in FIG. 14.5. The implant energy ensures full dopant penetration through the residual oxide into the S/D regions adjacent to P2 layers. The N and P type dopant level is much lower than the N+ and P+ dopant levels used to dope P2 regions. Hence P2 is unaffected by these LDD implants.

A spacer oxide is deposited over the LDD implanted P2 using LTO or PECVD techniques as shown in FIG. 14.[0131]6. The oxide is etched to formspacers1320 shown in FIG. 13. The spacer etch leaves a residual oxide over P1 in a first embodiment, and completely removes oxide over exposed P1 in a second embodiment. The latter allows for P1 salicidation at a subsequent step. After the spacer etch Nickel is deposited over P2 and salicided to form a low resistive refractory metal on exposed poly by RTA. Un-reacted Ni is etched as shown in FIG. 14.7. This 100A-500A thick Ni-salicide connects the opposite doped poly-2 and poly-1 regions together providing low resistive poly wires for data transfer. In one embodiment, the residual gate dielectric left after the spacer prevents P1 layer salicidation. In a second embodiment, as the residual oxide is removed over exposed P1 after spacer etch, P1 is salicided. The thickness of Ni deposition may be used to control full or partial salicidation of P1 regions in FIG. 13. Fully salicided S/D regions up to spacer edge facilitate high drive current due to lower source and drain resistances.

An LTO film is deposited over P2 layer, and polished flat with CMP. A second contact mask C2 is used to open contacts into the TFT P2 and P1 regions in addition to all other contacts to substrate transistors. In the shown embodiment, C1 contacts connecting Gated-FET outputs to substrate transistors require no C2 contacts. Contact plugs are covered with a glue layer, filled with tungsten, CMP polished, and connected by metal as done in standard contact metallization of IC's as shown in FIG. 14.[0132]8.

In another embodiment, thinned down SOI is used to construct the Gated-FET shown in FIG. 13. The SOI starting wafer is chosen to have the correct P1 thickness as given by EQ-8. This can be achieved by thinning down an existing SOI wafer to Silicon thickness ˜400A as discussed in the example. The process sequence is similar to the Gated-FET TFT device fabrication, except for the starting point. There is no preceding logic process, and the P1 definition is the starting point for the Gated device fabrication as detailed in the TFT process sequence. In another embodiment, an SOI logic process is used to fabricate CMOS devices on a substrate layer, and a second thinned down Gated-FET device for special applications. Standard SOI devices may be used to build AND gates, OR gates, inverters, adders, multipliers, memory and other logic functions in an integrated circuit. Thinned down SOI Gated-FET devices may be constructed to integrate a high density of latches or memory into the first fabrication module. A thinned down SOI module is inserted to an exemplary SOI logic process which includes one or more of following steps:[0133]

SOI substrate wafer[0134]

Shallow Trench isolation: Trench Etch, Trench Fill and CMP[0135]

Sacrificial oxide[0136]

Periphery PMOS V[0137]_Tmask & implant

Periphery NMOS V[0138]_Tmask & implant

Gated-FET mask and Silicon etch[0139]

Gated-FET blanket V[0140]_TN implant

Gated-FET V[0141]_TP mask and P implant

Dopant activation and anneal[0142]

Sacrificial oxide etch[0143]

Gate oxidation/Dual gate oxide option[0144]

Gate poly (GP) deposition[0145]

Gated-FET N+ mask and N+ implant[0146]

Gated-FET P+ mask and P+ implant[0147]

GP mask & etch[0148]

LDN mask & N− implant[0149]

LDP mask & P− implant[0150]

Spacer oxide deposition & spacer etch[0151]

Periphery N+ mask and N+ implant[0152]

Periphery P+ mask and P+ implant[0153]

Ni deposition[0154]

RTA anneal—Ni salicidation (S/D/G regions & interconnect)[0155]

Dopant activation[0156]

Unreacted Ni etch[0157]

ILD oxide deposition & CMP[0158]

C mask and etch[0159]

In this embodiment, the Gated-FET body doping is independently optimized for performance, but shares the same LDN, LDP implants. The Gated-FET gates are separately doped N+ & P+ prior to gate etch and blocked during N+/P+ implants of peripheral SOI devices as the dopant types differ. In other embodiments, Gated-FET devices and periphery MOSFET devices may share one or more V[0160]_Timplants. One P2 is used for latch and peripheral device gates. In another embodiment, SOI substrate devices may be integrated with a TFT latch module. This allows for a SOI inverter and TFT inverter to be vertically integrated to build high density, fast access memory devices.

Processes described in the incorporated-by-reference Provisional Application Serial Nos. 60/393,763 and 60/397,070 support poly-film TFT-SRAM cell and anti-fuse construction. This new usage differs from the process of FIG. 15 in doping levels and film thicknesses optimized for Gated-FET applications. The thin-film transistor construction and the Thin-Film Anti-Fuse construction may exist side by side with this Thin-Film Gated-FET device if the design parameters overlap.[0161]

These processes can be used to fabricate a generic field programmable gate array (FPGA) with the inverters connecting to form latches and SRAM memory. Such memory in a TFT module may be replaced with hard wired connections to form an application specific integrated circuit (ASIC). Multiple ASICs can be fabricated with different variations of conductive patterns from the same FPGA. The memory circuit and the conductive pattern contain one or more substantially matching circuit characteristics. The process can be used to fabricate a high density generic static random access memory (SRAM) with inverters connecting to form latches and SRAM memory. A TFT module may be used to build a vertically integrated SRAM cell with one inverter on a substrate layer, and a second inverter in a TFT layer.[0162]

Although an illustrative embodiment of the present invention, and various modifications thereof, have been described in detail herein with reference to the accompanying drawings, it is to be understood that the invention is not limited to this precise embodiment and the described modifications, and that various changes and further modifications may be effected therein by one skilled in the art without departing from the scope or spirit of the invention as defined in the appended claims.[0163]