Dynamic threshold voltage devices with low gate to substrate resistance

Information

  • Patent Grant
  • 6459106
  • Patent Number
    6,459,106
  • Date Filed
    Wednesday, January 3, 2001
    23 years ago
  • Date Issued
    Tuesday, October 1, 2002
    21 years ago
Abstract
Described is a dynamic threshold field effect transistor (DTFET) that includes a gate-to-body contact structure within the gate. By forming the gate-to-body contact structure that can reduce the gate-to-body contact resistance and increase the device packing density, the DTFET can be used in silicon on insulator (SOI) technologies and take full advantages of the DT-CMOS performance benefit.
Description




FIELD OF THE INVENTION




The present invention relates to a semiconductor apparatus including a dynamic threshold field effect transistor (DTFET), and to a method of producing the same. More particularly, the present invention relates to a method of lowering the resistance of gate interconnections and concurrently making a high device packing density possible on semiconductor substrates.




BACKGROUND OF THE INVENTION




Metal oxide semiconductor field effect transistor (MOSFET) devices have gained wide acceptance in the digital electronics industry. MOSFET devices are characterized by a threshold voltage required at the gate for the transistor to turn on or off and conduct or prevent the flow of current between the source and drain. The state of the MOSFET is therefore changed with the presence or absence of a minimum threshold voltage at the gate.




Recently, manufacturers and developers of integrated circuits have sought to improve performance by lowering the power supply voltages required by the transistors. However, as power supply voltages are scaled down, MOSFET threshold voltages cannot be lowered proportionately. The lowering of the threshold voltage is limited because a minimum threshold voltage has to be maintained, e.g. 200 mV, to ensure the circuits are not sensitive to noise, and the subthreshold leakage currents are not too high. As a result, device performance gain is not as desirable in technologies with low power supply voltages due to low gate-to-source overdrive, wherein the gate-to-source overdrive is proportional to the performance of the device and is defined as the gate-to-source voltage minus the threshold voltage.




One solution that has been proposed to alleviate the low overdrive problem is the dynamic threshold voltage FET (DTFET). In such systems, the threshold voltage dynamically adjusts as the gate input voltage changes. One method in which dynamic characteristics in the DTFET are achieved is by connecting the gate directly to the body of the FET as shown in FIG.


1


. In the case of an N-type DTFET (i.e., a DTNFET), when the gate is low and the DTNFET is off, the threshold voltage is high providing good noise immunity and low leakage current. When the gate switches high to turn the DTNFET on, the threshold voltage is low (near zero) due to the negative body effect, i.e., forward bias voltage from source to body. Since the threshold voltage is low during most of the switching time, the DTNFET can have very good switching speed.




Unfortunately, limitations exist with the above described system. In particular, the use of DTFETs must be limited to silicon on insulator (SOI) technologies where the FET bodies are isolated. One disadvantage of the SOI DTFET is the high substrate resistance in the device body. In general, it is difficult to make good contact (uniformly low resistance) from gate to the body substrate. A typical scheme of an SOI NFET is shown in FIG.


2


. (SOI PFET is similar except for the dopant types). The gate is connected to the body at one end. Since the body substrate resistance is very high, this scheme has the disadvantage that the device width has to be very narrow. In addition, the conventional contact can only reduce body resistance in one direction because the contact can only be made at the DTFET perimeter.




In addition, there is a challenge to develop a method of manufacturing DTFETs on the substrate such that the distance between body contacts is short enough so that the apparatus can take full advantage of DT-CMOS performance benefits. This can be illustrated as follows. The less stringent constraint requires that the body is at the gate voltage at the end of a system cycle such that Tb<Tcycle, where Tb is the body RC time constant and Tcycle is the system cycle time. This less stringent constraint eliminates floating body history effects, but does not take full advantage of the DT-CMOS performance benefit. A more stringent constraint requires that the body voltage must be able to follow the gate voltage while a MOSFET switches to take full advantage of the DT-CMOS performance benefit such that Tb<Tsw, where Tsw is a typical stage delay. The body RC time constant Tb can be expressed as Tb˜(Rb/Leff)×(Csj+Cdj+Cgate×Leff)×(d/2)**2, where Rb is the body sheet resistance ranging between 2˜10 kohm/square, Leff is the MOSFET channel length, Csj and Cdj are the drain and source junction capacitance to the body ranging between 0.5 and 1.5 fF/μm, Cgate is the body capacitance to the MOSFET gate, and d is the distance between body contacts along the width of MOSFET gate. Cgate can be expressed as Cgate˜3.5×(10 nm/Tox) fF/μm, where Tox is the MOSFET effective gate oxide thickness. For a typical 1 GHz microprocessor where Tcyc equals 1 ns and Tsw equals 10˜30 ps, the MOSFETs can have Leff as 0.08 μm, Tox as 2.2 nm, Csj as 1 fF/μm, Cdj as 1fF/μm and Rb as 6 kohm/sq. The less stringent constraint requires that d should be shorter than 4 μm to eliminate history effects in a 1 GHz processor, while the more stringent constraint requires that d should be shorter than 0.4 Mm to take full advantage of DT-CMOS performance benefits.




Unfortunately, a conventional DT-CMOS gate-to-body contact adds about two lithographic minimum images to a MOSFET width to allow for alignment tolerances and adequate metal-strap contact areas. Therefore the distance between body contacts can be so large in a conventional DTFET apparatus that the apparatus cannot take full advantage of DT-CMOS performance benefits.




Therefore, without a method that allows high performance DTFETs to be implemented in SOI technologies with low gate interconnection resistance and high device packing density, the use of DTFETs will be greatly limited.




SUMMARY OF THE INVENTION




It is an advantage of the present invention to provide a method for reducing gate-to-body resistance to allow high performance of DTFET in SOI technologies.




It is a further advantage of this invention to provide a method for providing a high device packing density on a semiconductor substrate by forming a contact that uses less area than a conventional contact.




It is a further advantage of this invention to provide a method for making body contacts such that the distance between body contacts is short enough to allow a semiconductor device to take full advantage of DT-CMOS performance benefits.




It is still another advantage of this invention to reduce gate-to-body resistance in two directions of MOSFET width and to produce a contact within the gate.




The present invention generally provides a method of fabricating a FET comprising the steps of:




providing a substrate having a device area;




forming a gate on said substrate;




forming an electrically conductive contact structure from said gate to said device area, within said gate; and




forming diffusion regions in said device area having a channel therebetween.




The present invention also provides a semiconductor device comprising:




a substrate;




diffusion regions in said substrate having a channel therebetween; and




a gate on said substrate, wherein said gate includes a contact structure from said gate to said channel, within said gate.




The present invention further provides a semiconductor device comprising:




a substrate;




a source region and a drain region in said substrate having a channel therebetween;




a gate structure on a surface of the channel; and




a contact layer within said gate electrically connected to said channel.




The present invention also provides a method of fabricating a Field Effect Transistor (FET) comprising the steps of:




providing a substrate;




forming an insulator layer over the substrate;




forming a gate on the insulator layer over the substrate;




forming openings through the gate and the insulator layer down to the substrate;




disposing conductive material in the openings in electrical contact with the substrate;




forming a conductor layer over the gate and the conductive material in the openings to electrically connect the gate to the substrate at each of the openings; and




forming diffusion regions in the substrate having a channel therebetween.











BRIEF DESCRIPTION OF THE DRAWINGS




The features of the present invention will become more clearly appreciated as the disclosure of the invention is made with reference to the accompanying drawings. In the drawings:





FIG. 1

depicts a transistor circuit scheme of a DTFET.





FIG. 2

illustrates a top view of a conventional body contact from a gate to one perimeter of a FET.





FIG. 3

illustrates a top view of a body contact from a gate to a substrate, within the gate, in accordance with the present invention.





FIGS. 4



a


and


4




b


are top views illustrating the method for forming a DTFET having a gate-to-body contact within the gate.





FIGS. 5



a


-


9




b


are cross sectional views illustrating the method for forming a DTFET having a gate-to-body contact within the gate.











DETAILED DESCRIPTION OF THE INVENTION




The features and advantages of the present invention are illustrated in detail in the accompanying drawings, wherein like reference numerals refer to like elements throughout the drawings. Although the drawings are intended to illustrate the present invention, the drawings are not necessarily drawn to scale.




All FETs in

FIGS. 1-9



b


are shown to be NFETs, however, it is understood that PFETs or combination NFET/PFET structures can easily be substituted therefore.





FIG. 1

depicts a transistor circuit scheme of a dynamic threshold field effect transistor


10


(DTFET). The DTFET


10


includes a gate


12


that is tied to an input


14


, a drain


16


that is tied to a voltage supply VDD


18


, a source


20


that is tied to an output


22


, and a body


24


.




By connecting the gate


12


the body


24


, the threshold voltage can be dynamically adjusted as the input


14


changes during a switching time. In particular, when the input


14


goes high, the output


22


becomes equal to the drain voltage VDD such that the FET is turned on. Because the body


24


is connected to the input


14


and gate


12


, the threshold voltage is dynamically lowered so that the FET can be turned on with a lower input voltage. Alternatively, when the input


14


is low, i.e., below the threshold voltage, the output


22


remains unchanged such that the FET is turned off. The threshold voltage remains high providing good noise immunity and low leakage current.





FIG. 2

illustrates a conventional way of forming a gate-to-body contact


40


in a conventional DTFET


30


. The DTFET


30


includes a field isolation area


32


, an N+ drain


34


, an N+ source


36


, a P− body


38


, an N+ doped polysilicon gate


42


and a P+ body contact


40


that has a gate to body metal strap


44


. The gate to body metal strap


44


is electronically connected to the P− body


38


so that the P+ body contact


40


is connected to the P− body


38


. However, because the gate to body metal strap


44


can only be connected to the P− body


38


from the perimeter of the P− body


38


, a body resistance


46


can be very high and the use of the DTFET in SOI technologies is greatly limited. Since the body resistance


46


is dependent on the width of the device, the device has to be made very narrow to lower the body resistance


46


.




This invention proposes a new DT-ET structure that eliminates the device width constraint, as shown in FIG.


3


. The DTFET structure


50


includes a field isolation area


52


, an N+ drain


54


, an N+ source


56


, a P− body


58


, an N+ doped polysilicon gate


62


and P+ body contacts


60


A and


60


B. The P+ body contacts


60


A and


60


B connect the N+ doped gate


62


to the P− body


58


within the N+ doped gate


62


. The P+ body contacts


60


A and


60


B can extend into the P− body


58


for about several hundred As. Thus, the gate-to-body resistance is dependent on the polysilicon gate resistance, not the body resistance.





FIGS. 4



a


and


4




b


illustrates top views of fabricating a DTFET having a gate-to-body contact within the gate. A gate conductor


76


is formed along a direction B-B′ on a device area


72


and a field isolation area


74


. The device area


72


is surrounded by the field isolation area


74


. The gate conductor


76


is typically doped polysilicon. Body contacts are formed in body contact regions


78


A and


78


B, and are separated by a distance


79


. Additional details are illustrated in

FIGS. 5



a


-


9




b.






In the following description of the present invention,

FIGS. 5



a,




6




a,




7




a,




8




a,


and


9




a,


are sequential cross-sectional views of the method for producing a DTFET according to the present invention, taken along lines A-A′ of

FIG. 4



b.


Correspondingly,

FIGS. 5



b,




6




b,




7




b,




8




b,


and


9




b,


are sequential cross-sectional views taken along lines B-B′ of

FIG. 4



b.







FIGS. 5



a


and


5




b


are cross sectional views illustrating the method for forming a silicon on insulator dynamic threshold field effect transistor (SOI DTFET) in accordance with the present invention. The method begins by providing an SOI substrate, including the device area


72


, a buried oxide


80


and a Si-wafer


82


. Field isolation areas, such as field oxide layers


74


, are formed surrounding the device area


72


to electronically isolate the device area


72


. On the device area


72


is grown a gate insulator


100


. The gate conductor


76


is then deposited on the gate insulator


100


. The gate insulator


100


can be a silicon dioxide (S


i


O


2


) layer. The gate conductor


76


is typically a polysilicon layer that is doped with an N+ type dopant, such as arsenic (As) either by an ion implantation or by in-situ doping during the polysilicon deposition. The thickness of the gate conductor is preferably about 100-200 nm.




As illustrated in

FIGS. 6



a


and


6




b,


a blanket insulating layer


84


is deposited on the gate conductor


76


and elsewhere on the substrate. The preferred deposition is by LPCVD using, for example, tetraethosiloxane (TEOS). The preferred thickness of the blanket insulating layer


84


is about 5-20 nm. Then, a nitride layer


86


is formed on the blanket insulating layer


84


and elsewhere on the substrate. The nitride layer


86


is thicker than the gate conductor


76


, and the thickness of the nitride layer


86


is preferably about 250-300 nm. The surface of the nitride layer


86


is then polished or etched flat by a well known chemical mechanical polishing (CMP) process. During the CMP process, the thickness of the nitride layer


86


is decreased until the gate conductor


76


is reached. It is shown that the gate conductor


76


is surrounded by the nitride layer


86


and the blanket insulating layer


84


.




Now referring to

FIGS. 7



a


and


7




b,


a damascene process is undertaken to form a contact structure within the gate conductor


76


. First, a resist layer


88


is deposited to protect the area other than the body contact regions


78


. The exposed body contact regions


78


are etched in a reactive ion etcher (RIE) using a suitable etch gas mixture. As shown in

FIG. 7



b,


portions of the gate conductor


76


are etched in the body contact regions


78


.




As illustrated in

FIGS. 8



a


and


8




b,


the resist layer


88


is removed in a known manner. Oxide spacers


90


are formed on the sidewalls of the nitride layer


86


to prevent diffusion between body contact and source/drain junctions. The oxide spacers


90


are formed by depositing a conformal sidewall material layer preferably having a thickness of about 5-20 nm, which is then etched back. The etch back can be carried out by RIE.




Further referring to

FIGS. 8



a


and


8




b,


several methods can be chosen to form a body contact structure through the gate conductor


76


. One method is to deposit a contact layer


92


into the contact regions


78


that has the same doping as the device area


72


. The preferred contact layer


92


is typically doped with a P+ type dopant. The thickness of the contact layer


92


is about 10-20 nm. The out diffusion layer


94


is formed by an out diffusion process. An alternate way is to form a low resistance contact by implanting, and then to deposit contact material (e.g. polysilicon, tungsten, etc.) into the contact regions


78


. Here, a low resistance L


2


gate-to-body contact has been produced within the gate.




Still referring to

FIGS. 8



a


and


8




b,


the surface of the contact layer


92


is polished or etched using a CMP process. The nitride layer


86


is removed and the blanket insulating layer


84


is exposed.




The last step of the invention is a standard MOSFET processing, as shown in

FIGS. 9



a


and


9




b.


Gate sidewall spacers


96


are formed on the sidewalls of the gate conductor


76


. These sidewall spacers


96


are formed by a similar method as that used to form the oxide spacers


90


. After the sidewall spacers


96


are formed, an ion implantation is used to form a source area


98


and a drain area


102


. Typically the heavily doped source area


98


and drain area


102


are formed by implanting a dopant such as arsenic or phosphorous. The N-channel DTFET is formed on SOI, and an extra step can be taken to improve the conductivity of the gate, source and drain by depositing a silicide layer


104


.




The foregoing description of the present invention has been presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed, and many modifications and variations are possible in light of the above teaching. Such modifications and variations that may be apparent to a person skilled in the art are intended to be included within the scope of this invention as defined by the accompanying claims.



Claims
  • 1. A semiconductor device, comprising:a substrate; diffusion regions in said substrate, wherein said diffusion regions extend in a first direction and are separated by at least one channel therebetween; and a gate on said substrate, wherein said gate includes a plurality of electrically conductive contact structures from said gate to said at least one channel, said electrically conductive contact structures located within said gate and extending in a direction perpendicular to said first direction, and wherein said at least one channel is located between said electrically conductive contact structures.
  • 2. The semiconductor device of claim 1, wherein said substrate is a semiconductor-on-insulator (SOI) type semiconductor substrate.
  • 3. A semiconductor device, comprising:a substrate; a source region and a drain region in said substrate extending in a first direction and separated by at least one channel therebetween; a gate structure on a surface of the channel; and a contact layer formed within said gate, said contact layer forming a series of contact structures which extend in a direction perpendicular to said channel, and which are electrically connected to said channel, and wherein said at least one channel is located between said contact structures.
  • 4. The semiconductor device of claim 3, wherein said gate structure further includes:a gate insulator layer on said substrate; and a gate body on the gate insulator layer.
  • 5. The semiconductor device of claim 4, wherein said gate body comprises doped polysilicon.
  • 6. The semiconductor device of claim 4, wherein said gate insulator layer is a silicon dioxide layer.
  • 7. The semiconductor device of claim 3, wherein said substrate is a semiconductor-on-insulator (SOI) type semiconductor substrate.
Parent Case Info

This application is a divisional of Ser. No. 09/447,122, filed on Nov. 22, 1999.

US Referenced Citations (11)
Number Name Date Kind
5320975 Cederbaum et al. Jun 1994 A
5571738 Krivokapic Nov 1996 A
5668051 Chen et al. Sep 1997 A
5670426 Kuo et al. Sep 1997 A
5677557 Wuu et al. Oct 1997 A
5710073 Jeng et al. Jan 1998 A
5733812 Ueda et al. Mar 1998 A
5780899 Hu Jul 1998 A
5834846 Shinriki Nov 1998 A
5869867 Takeuchi Feb 1999 A
5998847 Assaderaghi Dec 1999 A