U.S. patent application number 10/641522 was filed with the patent office on 2004-02-19 for structure and method of controlling short-channel effect of very short channel mosfet.
Invention is credited to Wann, Hsing-Jen.
Application Number | 20040033665 10/641522 |
Document ID | / |
Family ID | 22355616 |
Filed Date | 2004-02-19 |
United States Patent
Application |
20040033665 |
Kind Code |
A1 |
Wann, Hsing-Jen |
February 19, 2004 |
Structure and method of controlling short-channel effect of very
short channel MOSFET
Abstract
A semiconductor device comprising a gate having an approximately
0.05 .mu.m channel length, an oxide layer below the gate, a
self-aligned compensation implant below the oxide layer, a halo
implant surrounding the self-aligned compensation implant below the
oxide layer; and gate and drain regions on opposite sides of the
halo implant and below the oxide layer.
Inventors: |
Wann, Hsing-Jen; (Briarcliff
Manor, NY) |
Correspondence
Address: |
FREDERICK W. GIBB, III
MCGINN & GIBB, PLLC
2568-A RIVA ROAD
SUITE 304
ANNAPOLIS
MD
21401
US
|
Family ID: |
22355616 |
Appl. No.: |
10/641522 |
Filed: |
August 15, 2003 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10641522 |
Aug 15, 2003 |
|
|
|
09114502 |
Jul 13, 1998 |
|
|
|
6653686 |
|
|
|
|
Current U.S.
Class: |
438/291 ;
257/E21.434; 257/E21.443; 257/E21.618; 257/E21.624; 257/E29.054;
438/289; 438/919 |
Current CPC
Class: |
H01L 21/823456 20130101;
H01L 29/66553 20130101; H01L 21/823412 20130101; H01L 29/1045
20130101; H01L 29/66583 20130101; H01L 29/66537 20130101 |
Class at
Publication: |
438/291 ;
438/289; 438/919 |
International
Class: |
H01L 021/336 |
Claims
I claim:
1. A semiconductor device comprising: a gate; and a compensation
implant below said gate.
2. The semiconductor device as in claim 1, wherein said
semiconductor device has an approximately 0.05 .mu.m channel
length.
3. The semiconductor device as in claim 1, wherein said
compensation implant reduces a threshold voltage of said
semiconductor device.
4. The semiconductor device as in claim 1, wherein said
compensation implant has a doping concentration for controlling a
threshold voltage reduction of said semiconductor device.
5. The semiconductor device as in claim 4, wherein said doping
concentration varies depending upon a channel length of said
semiconductor device.
6. The semiconductor device as in claim 1, wherein said
compensation implant is self-aligned with said gate.
7. The semiconductor device as in claim 1, further comprising: a
halo implant surrounding said compensation implant; and gate and
drain regions on opposite sides of said halo implant.
8. A semiconductor device comprising: a gate; an oxide layer below
said gate; a self-aligned compensation implant below said oxide
layer; a halo implant surrounding said self-aligned compensation
implant below said oxide layer; and gate and drain regions on
opposite sides of said halo implant and below said oxide layer.
9. The semiconductor device as in claim 8, wherein said
self-aligned compensation implant reduces a threshold voltage of
said semiconductor device.
10. The semiconductor device as in claim 8, wherein said
self-aligned compensation implant has a doping concentration for
controlling a threshold voltage reduction of said semiconductor
device that varies depending upon a channel length of said
semiconductor device.
11. A method of manufacturing a semiconductor device comprising
steps of: forming a channel above a substrate; implanting an
impurity in said substrate through said channel to form a
compensation implant in said semiconductor substrate; and forming a
gate above said compensation implant.
12. The method as in claim 11, further comprising steps of: prior
to said step of forming said channel, successively depositing an
oxide layer and a dielectric layer on said substrate, said step of
forming said channel comprising a step of forming said channel in
said oxide layer and said dielectric layer; prior to said step of
forming said gate, depositing an oxide layer in said channel on
said substrate; after said step of forming said gate, removing all
of said dielectric layer and a portion of said oxide layer such
that said oxide layer remains between said gate and said
semiconductor substrate; and doping areas of said semiconductor
substrate adjacent said gate to form a halo implant and source and
drain regions.
13. The method as in claim 11, further comprising a step, prior to
said implanting step, of forming spacers in said channel.
14. The method as in claim 11, wherein said implanting step aligns
said compensation implant and said gate.
15. The method as in claim 11, wherein said implanting step further
comprises a step of varying a doping concentration of said
compensation implant for controlling a threshold voltage reduction
of said semiconductor device.
16. The method as in claim 15, wherein said step of varying a
doping concentration of said compensation implant is dependent upon
a channel length of said semiconductor device.
17. A method of manufacturing a semiconductor device comprising
steps of: successively depositing an oxide layer and a dielectric
layer on a semiconductor substrate; forming a channel in said oxide
layer and said dielectric layer; forming spacers in said channel;
implanting an impurity in said semiconductor substrate through said
channel to form a self-aligned compensation implant in said
semiconductor substrate; depositing an oxide layer in said channel
on said semiconductor substrate; filling said channel with a
conductive material to form a gate over said oxide layer; removing
all of said dielectric layer; removing a portion of said oxide
layer such that said oxide layer remains between said gate and said
semiconductor substrate; and doping areas of said semiconductor
substrate adjacent said gate to form a halo implant and source and
drain regions in said semiconductor substrate.
18. The method as in claim 17, wherein said implanting step further
comprises a step of varying a doping concentration of said
self-aligned compensation implant for controlling a threshold
voltage reduction of said semiconductor device.
19. The method as in claim 17, wherein said step of varying a
doping concentration of said self-aligned compensation implant is
dependent upon a channel length of said semiconductor device.
20. A method of manufacturing a semiconductor device comprising
steps of: successively depositing an oxide layer and a dielectric
layer on a semiconductor substrate; forming a channel in said oxide
layer and said dielectric layer; depositing an oxide layer in said
channel on said semiconductor substrate; partially filling said
channel with a conductive material; implanting an impurity in said
semiconductor substrate through said conductive material and said
oxide layer in said channel to form a self-aligned compensation
implant in said semiconductor substrate; completely filling said
channel with said conductive material to form a gate over said
oxide layer; removing all of said dielectric layer; removing a
portion of said oxide layer such that said oxide layer remains
between said gate and said semiconductor substrate; and doping
areas of said semiconductor substrate adjacent said gate to form a
halo implant and source and drain regions in said semiconductor
substrate.
21. The method as in claim 20, wherein said implanting step further
comprises a step of varying a doping concentration of said
self-aligned compensation implant for controlling a threshold
voltage reduction of said semiconductor device.
22. The method as in claim 20, wherein said step of varying a
doping concentration of said self-aligned compensation implant is
dependent upon a channel length of said semiconductor device.
23. A method of manufacturing a semiconductor device comprising
steps of: forming a sacrificial mask over a gate area of a
substrate; forming spacers adjacent said sacrificial mask; forming
an opening in said sacrificial mask; implanting an impurity in said
semiconductor substrate through said opening in said sacrificial
mask to form a self-aligned compensation implant in said
semiconductor substrate; removing said sacrificial mask; depositing
an oxide layer between said spacers on said semiconductor
substrate; forming a gate between said spacers and over said oxide
layer; and doping areas of said semiconductor substrate adjacent
said gate to form a halo implant and source and drain regions in
said semiconductor substrate.
24. The method as in claim 23, wherein said implanting step further
comprises a step of varying a doping concentration of said
self-aligned compensation implant for controlling a threshold
voltage reduction of said semiconductor device.
25. The method as in claim 23, wherein said step of varying a
doping concentration of said self-aligned compensation implant is
dependent upon a channel length of said semiconductor device.
Description
BACKGROUND OF THE INVENTION
[0001] 1. Field of the Invention
[0002] The present invention generally relates to doped
semiconductor transistors and more particularly to doped
short-channel devices which have channel lengths less than 0.05
.mu.m and which maintain acceptable threshold voltages, and methods
of making such devices.
[0003] 2. Description of the Related Art
[0004] Conventional metal oxide semiconductor field effect
transistor (MOSFET) device structures are continuously being
reduced in size to increase processing speed and decrease
manufacturing cost. Conventional methods of reducing the size of
MOSFET devices shrink all the dimensions of the device
proportionally.
[0005] However, as the channel length in the MOSFET is reduced to
increase speed, a "short-channel" effect often occurs which
severely degrades the device characteristics. More specifically,
the short channel effect is an undesirable increase in the
threshold voltage of the gate as the channel length is reduced.
[0006] Therefore, there is a conventional need for a method and
structure which overcomes the short channel effect and allows
MOSFET structures having channel lengths of approximately 0.051
.mu.m to produce consistent threshold voltages.
SUMMARY OF THE INVENTION
[0007] It is, therefore, an object of the present invention to
provide a structure and method for a doping profile for controlling
short-channel devices having channel lengths down below 0.05 .mu.m,
and the methods of making such profiles.
[0008] More specifically, the invention is a semiconductor device
comprising a gate, an oxide layer below the gate, a self-aligned
compensation implant below the oxide layer, a halo implant
surrounding the self-aligned compensation implant below the oxide
layer, and gate and drain regions on opposite sides of the halo
implant and below the oxide layer. The self-aligned compensation
implant reduces a threshold voltage of the semiconductor device and
has a doping concentration for controlling a threshold voltage
reduction of the semiconductor device that varies depending upon a
channel length of the semiconductor device.
[0009] The invention also includes method of manufacturing a
semiconductor device comprising steps of successively depositing an
oxide layer and a dielectric layer on a semiconductor substrate,
forming a channel in the oxide layer and the dielectric layer,
forming spacers in the channel, implanting an impurity in the
semiconductor substrate through the channel to form a self-aligned
compensation implant in the semiconductor substrate, depositing an
oxide layer in the channel on the semiconductor substrate, filling
the channel with a conductive material to form a gate over the
oxide layer, removing all of the dielectric layer, removing a
portion of the oxide layer such that the oxide layer remains
between the gate and the semiconductor substrate, and doping areas
of the semiconductor substrate adjacent the gate to form a halo
implant and source and drain regions in the semiconductor
substrate.
[0010] The implanting step further comprises a step of varying a
doping concentration of the self-aligned compensation implant for
controlling a threshold voltage reduction of the semiconductor
device. The step of varying a doping concentration of the
self-aligned compensation implant is dependent upon a channel
length of the semiconductor device.
[0011] The invention also includes a method of manufacturing a
semiconductor device comprising steps of successively depositing an
oxide layer and a dielectric layer on a semiconductor substrate,
forming a channel in the oxide layer and the dielectric layer,
depositing an oxide layer in the channel on the semiconductor
substrate, partially filling the channel with a conductive
material, implanting an impurity in the semiconductor substrate
through the conductive material and the oxide layer in the channel
to form a self-aligned compensation implant in the semiconductor
substrate, completely filling the channel with the conductive
material to form a gate over the oxide layer, removing all of the
dielectric layer, removing a portion of the oxide layer such that
the oxide layer remains between the gate and the semiconductor
substrate, and doping areas of the semiconductor substrate adjacent
the gate to form a halo implant and source and drain regions in the
semiconductor substrate.
[0012] Once again, the implanting step further comprises a step of
varying a doping concentration of the self-aligned compensation
implant for controlling a threshold voltage reduction of the
semiconductor device and the step of varying a doping concentration
of the self-aligned compensation implant is dependent upon a
channel length of the semiconductor device.
[0013] The invention also includes a method of manufacturing a
semiconductor device comprising steps of forming a sacrificial mask
over a gate area of a substrate, forming spacers adjacent the
sacrificial mask, forming an opening in the sacrificial mask,
implanting an impurity in the semiconductor substrate through the
opening in the sacrificial mask to form a self-aligned compensation
implant in the semiconductor substrate, removing the sacrificial
mask, depositing an oxide layer between the spacers on the
semiconductor substrate, forming a gate between the spacers and
over the oxide layer, and doping areas of the semiconductor
substrate adjacent the gate to form a halo implant and source and
drain regions in the semiconductor substrate.
[0014] Again, the implanting step further comprises a step of
varying a doping concentration of the self-aligned compensation
implant the step of varying a doping concentration of the
self-aligned compensation implant is dependent upon a channel
length of the semiconductor device.
BRIEF DESCRIPTION OF THE DRAWINGS
[0015] FIG. 1 is a schematic diagram of a cross-sectional view of a
semiconductor device;
[0016] FIG. 2 is a schematic diagram of a cross-sectional view of a
semiconductor device;
[0017] FIG. 3 is a chart showing the relationship between a target
threshold voltage and a channel length of a semiconductor
device;
[0018] FIG. 4 is a schematic diagram of a cross-sectional view of a
semiconductor device;
[0019] FIG. 5 is a chart showing the relationship between a target
threshold voltage and a channel length of a semiconductor
device;
[0020] FIG. 6 is a schematic diagram showing a cross-sectional view
of a stage of development of a semiconductor device;
[0021] FIG. 7 is a schematic diagram showing a cross-sectional view
of a stage of development of a semiconductor device;
[0022] FIG. 8 is a schematic diagram showing a cross-sectional view
of a stage of development of a semiconductor device;
[0023] FIG. 9 is a schematic diagram showing a cross-sectional view
of a stage of development of a semiconductor device;
[0024] FIG. 10 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0025] FIG. 11 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0026] FIG. 12 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0027] FIG. 13 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0028] FIG. 14 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0029] FIG. 15 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0030] FIG. 16 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0031] FIG. 17 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0032] FIG. 18 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0033] FIG. 19 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0034] FIG. 20 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0035] FIG. 21 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0036] FIG. 22 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device;
[0037] FIG. 23 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device; and
[0038] FIG. 24 is a schematic diagram showing a cross-sectional
view of a stage of development of a semiconductor device.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS OF THE INVENTION
[0039] Referring now to the drawings, and more particularly to FIG.
1, a MOSFET device having a gate 10, gate oxide layer 11, source
region 12 and drain region 13 is illustrated. When reducing the
size of the MOSFET, the dimension of the oxide thickness 11 and the
junction depth 15 can preferably be decreased and the depletion
region 16 can preferably be effectively reduced by increasing the
substrate doping.
[0040] When the channel length 14 is reduced to 0.1 .mu.m, a "halo
implant" 20, as shown in FIG. 2, can preferably be used to control
the short channel effect. As mentioned above, the short channel
effect is an undesirable increase in the threshold voltage of the
gate 10 as the channel length 14 is reduced. The halo implant 20 is
a ring-shaped structure which is positioned between the gate 12 and
drain 13 and below the oxide layer 11.
[0041] The halo implant 20 reduces the short channel effect by
surrounding the source-drain extensions to prevent electrical field
line penetration. The halo implant 20 introduces a lateral
nonuniformity, which is essential for controlling the short-channel
effect and is useful down to channel lengths around 0.08 .mu.m.
[0042] However, at approximately 0.05 .mu.m, the halo implant 20
begins to reach a limitation in the ability to control the
short-channel effect. More specifically, when channel lengths are
reduced below 0.05 .mu.m, the distribution tails 21 of the halo
implant 20 tend to be too long, which increases the threshold
voltage unacceptably, producing a situation known as "Vt roll-off".
Also, the concentration of halo implant 20 can be too high, which
causes tunneling leakage current to increase.
[0043] The halo implant 20, which has a three-dimensional ring or
doughnut-shaped structure, tends to lose the opening or "hole" at
the center of the structure when the size of the channel is
reduced. In other words, when the channel length is reduced below
0.05 .mu.m, the inner surfaces of the halo structure may join, due
to the lateral distribution tails 21 overlapping, thereby severely
altering the threshold voltage (Vt) of the device.
[0044] Because of this problem, it is difficult to produce a device
having a channel length less than 0.08 .mu.m (within the normally
acceptable range of manufactureable tolerances) and a stable
threshold voltage.
[0045] This problem is graphically illustrated in FIG. 3. A typical
range of manufactureable tolerances of channel lengths is shown on
the horizontal axis of FIG. 3. More specifically, FIG. 3
illustrates a mean value for an effective channel length (e.g.,
nominal leff) which is preferably approximately 0.05 .mu.m and a
tolerance between a worse case long channel length and a worse case
short channel length. The tolerance range of channel length results
from variations which would be expected in a normal manufacturing
process. For example, the tolerance range could be +/0.017 .mu.m
the preferable target channel length of 0.05 .mu.m.
[0046] The vertical axis in FIG. 3 represents the threshold voltage
Vt and more specifically shows a target threshold voltage range
(Vt) for the device. For example, the target threshold voltage
range is preferably 0.35 V+/0.15 V.
[0047] Line 30 represents the threshold voltages of more heavily
doped devices and line 31 represents the threshold voltages of
lighter doped devices, where each of the devices includes a
conventional halo implant.
[0048] More specifically, lines 30, 31 illustrate the
"short-channel effect" (e.g., threshold voltage roll-off). Line 30
illustrates that only those heavily doped devices which have
channel lengths substantially shorter than the standard channel
length (e.g., nominal leff) will perform adequately and exhibit the
target threshold voltage and that nominal leff and larger channel
length devices will have a threshold voltage which is above the
target threshold voltage.
[0049] Similarly, line 31 illustrates that even if the doping of
the devices is changed (e.g., made lighter), not all ranges of
channel lengths expected within a given manufacturing tolerance
will produce the desired target threshold voltage. Specifically,
those devices having a channel length which is less than nominal
leff will have a threshold voltage which is below the target
threshold voltage and only those devices having a nominal leff and
larger channel lengths will have a threshold voltage within the
range of target threshold voltages.
[0050] Additionally, FIG. 3 illustrates that the longer channel
lengths tend to have a higher threshold voltage. The shape of lines
30 and 31 is caused by the tails 21 of the halo implant 20
overlapping partially or completely, as discussed above with
respect to FIG. 2. Thus, FIG. 3 illustrates that even if devices
having channel lengths less than 0.05 .mu.m include a halo implant,
such devices will not likely produce target threshold voltages
(throughout the acceptable channel length tolerance range)
regardless of doping concentrations. Thus, the process margin will
be very small and the junction leakage will be too high.
[0051] Thus, as discussed above, when channel lengths are reduced
below 0.05 .mu.m, the distribution tails of the halos tend to
overlap, which increases the threshold voltage unacceptably
producing the situation known as "Vt roll-off". The invention
reduces the size of the lateral distribution tail of the halo
implant so that the distribution tails do not overlap. In order to
accomplish this, the invention uses a self-aligned gate
compensation implant 42, as shown in FIG. 4, to counteract the
short-channel Vt roll-off.
[0052] More specifically, FIG. 4 illustrates a gate 40, a gate
oxide layer 41 below the gate 40, a source region 42 and a drain
region 43 having a junction depth 45, a channel length 44, a halo
implant 46, a compensation implant 47 and lateral distribution
tails 48. The compensation implant 47 is positioned below the gate
oxide 41 and is surrounded by the halo implant 46. The processing
steps used to create the structure illustrated in FIG. 4 are
discussed below with respect to FIGS. 6-24.
[0053] The compensation implant 47 formed below the gate 40 will
reduce the undesirable Vt roll-off. However, a uniform "blanket"
compensation implant will equally lower the Vt of MOSFET of all
channel lengths. The inventive compensation implant 47 is different
than a "blanket" compensation implant because it is self-aligned
with the gate 40. By self-aligning the compensation implant 47, the
implant dose of the compensation implant 47 will be a function of
the gate length 44.
[0054] With the self-aligned compensation implant, the compensation
dose is lower for a smaller gate length and the threshold voltage
reduction is correspondingly lower. For a longer gate length the
compensation dose is greater and the threshold voltage is reduced
more. In other words, the inventive compensation implant will
reduce long channel threshold voltage more than short-channel
threshold voltage, and will counteract the Vt roll-off and result
in better threshold voltage stability.
[0055] As mentioned above with respect to FIG. 3, a longer channel
length will have a higher threshold voltage. The inventive
compensation implant matches this characteristic by reducing the
threshold voltage more for longer channel lengths and less for
shorter channel lengths. Further, the self-aligned nature of the
compensation implant permits the variation in doping concentration
which adjusts the reduction of the threshold voltage. Therefore,
the inventive compensation implant tends to maintain the threshold
voltage of semiconductor devices closer to the target threshold
voltage because the compensation implant lowers the threshold
voltage of devices having higher threshold voltages (e.g., longer
channel lengths) more than it lowers the threshold voltages of
devices which have lower threshold voltages (e.g., shorter channel
lengths).
[0056] This is graphically illustrated in FIG. 5 which is
substantially similar to FIG. 3 except for the inclusion of line
50, which represents the threshold voltage of various devices
produced with the inventive self-aligned compensation implant. As
demonstrated in FIG. 5, throughout the acceptable variations of
channel lengths, the devices with the inventive self-aligned
compensation implant exhibit a threshold voltage which is within
the target threshold voltage range.
[0057] A first preferred method for forming the inventive
structure, which, for example, could be a damascene gate process,
is illustrated in FIGS. 6-11. More specifically, FIG. 6 illustrates
the process where a silicon substrate 60 is formed with dielectric
areas 61. An oxide layer 62 is formed over the silicon substrate
60. Subsequently a dielectric 63 is formed over the oxide layer 62.
A channel 64 is formed through the dielectric layer 63 and the
oxide 62, exposing the surface of the substrate 60.
[0058] Referring to FIG. 7, nitride, oxide or other similar spacers
70, are formed in the channel 64. Subsequently, the inventive
self-aligned compensation implant 71 is formed by an impurity
implantation through the channel 64. For n-type transistors, the
donor impurities could include, for example, phosphorus, arsenic
and antimony. For p-type transistors, the donor impurities could
include, for example, boron, gallium and indium.
[0059] In FIG. 8, the self-aligned compensation implant 71 is
covered again with the oxide layer 62 and a conductive material 80
(e.g., gate) over the oxide layer 62. The conductive material 80
could include, for example, polysilicon or metal.
[0060] With respect to FIG. 9, the dielectric layer 63 and the
oxide layer 62 are etched (except for the portion of the oxide
layer 62 beneath the conductive layer 80). For example, the etching
process could include wet or dry chemical etching.
[0061] Referring to FIG. 10 the source/drain extension implant 100,
101 is illustrated. The source/drain extension implant 100, 101
could utilize the same impurity material used for the self-aligned
compensation implant 71.
[0062] FIG. 11 shows the more heavily doped source 110 and drain
regions 111, which are formed in a second heavier doping step.
Additionally, oxide spacers 112 are formed adjacent the gate layer
80.
[0063] FIGS. 6-11 illustrate a process in which the compensation
implant 71 is self-aligned with the gate 80 by the channel 64. The
channel 64 defines the location of the compensation implant 71, the
gate oxide layer 62 and the gate 80. Further, the gate 80 defines
the source and drain regions 110, 111. Therefore, since the
position of all elements of the transistor are defined by the
location and length of the channel 64, the compensation implant 71
will automatically be aligned beneath the gate 80 and will be
equally spaced between the source of 110 and the drain 110
regions.
[0064] As discussed above, by having a self-aligned compensation
implant 71, the dose of impurity within the compensation implant 71
is controlled by the length of the channel 64. Therefore, a longer
channel 64 will allow a higher impurity dosage to be implanted in
the compensation implant 71. Similarly, a shorter channel length
will reduce the doping of the compensation implant. As mentioned
above, reducing the amount of impurity doping limits the reduction
of the threshold voltage. Further, the thickness of the spacer 70
can also be used in a similar way to optimize the dose and energy
of the implant.
[0065] In FIGS. 12-14 another preferred method of the invention is
illustrated. More specifically, the method illustrated in FIGS.
12-14 is similar to the method illustrated in FIGS. 6-11. However,
FIGS. 12-14 illustrate a more efficient method of forming the
compensation implant.
[0066] More specifically, in a similar manner as discussed above
with respect to FIG. 6, a silicon substrate 120 includes dielectric
areas 121 and has an oxide layer 122 formed above the silicon
substrate 120 and a dielectric layer 123 formed over the gate oxide
layer 122. A channel 124 is formed through the dielectric layer 123
and the oxide layer 122, exposing the substrate 120.
[0067] As shown in FIG. 13, the channel 124 is partially filled
with a conducting material 130, such as metal or polysilicon.
Unlike the process shown in FIG. 7, no spacers are formed in FIG.
12. Further, the compensation implant 131 is formed by making the
conductive layer 130 thin enough to allow an ion implantation to
pass. Therefore, in this preferred embodiment, the impurity is
implanted directly through the conductive layer 130. In this way,
this embodiment is more efficient than the previous embodiment
because the steps of forming the nitride spacers 70 (shown in FIG.
7) is eliminated. Further, such a structure minimizes oxidation and
enhances diffusion.
[0068] As shown in FIG. 14, the remainder of the channel 124 which
was not completely filled by the first layer of conductive material
130 is filled completely with more of the conductive material 140.
The structure is then substantially similar to that shown in FIG. 8
and the processing shown in FIGS. 9-11 is completed on the
structure illustrated in FIG. 14, as discussed above.
[0069] FIGS. 15-20 illustrate the advantages produced by the
self-aligned compensation implant. FIGS. 15-20 follow essentially
the same method discussed above with respect to FIGS. 6-11. Thus,
many of the details have been omitted from FIGS. 15-20 to clearly
illustrate how the doping concentration of the self-aligned
compensation implant is controlled by the channel length.
[0070] More specifically, FIG. 15 illustrates a structure similar
to that shown in FIG. 6, having a dielectric/oxide layer 151 above
a substrate 150. The dielectric layer has two channels, 152, 153
formed therein. However, because of manufacturing variations, the
lengths of the channels 152, 153 are different. In this example,
both channels are considered to have been produced within normal
manufacturing tolerances.
[0071] FIG. 16 illustrates spacers 160, 161 which respectively line
the channels 152, 153, in a similar manner as discussed above with
respect to FIG. 7. However, while spacers 160 only line the channel
152, spacers 161 fill the channel 161.
[0072] FIG. 17 illustrates an ion implantation 170, which is
similar to the implantation 71, shown in FIG. 7. However, FIG. 17
illustrates that the compensation implant 171 is formed only below
channel 152 and is not formed in channel 153 because the spacers
161 block the implant.
[0073] FIGS. 18 and 19 illustrates the formation of the gates 180,
181, as discussed above with respect to FIGS. 8 and 9 and FIG. 20
illustrates the extension/drain implant 200 which is similar to the
implant discussed above with respect to FIG. 10.
[0074] Therefore, as shown in FIGS. 15-20, the length of the
channel controls the doping concentration of the compensation
implant 171 by limiting the opening (e.g., 152, 153) through which
the impurity 170 may travel. The impurity concentration is adjusted
in this manner to completely prevent the ion implantation, to
severely limit ion implantation or to increase ion
implantation.
[0075] Further, the self-aligned nature of the compensation implant
171 automatically compensates for manufacturing variations by
basing the impurity concentration based on the length of the
channel, as discussed above.
[0076] FIGS. 21-24 similarly demonstrate the advantages of the
self-aligned compensation implant and illustrate another embodiment
of the invention. More specifically, FIGS. 21-24 illustrate a
method of utilizing the channel lengths to control the impurity
concentration of the compensation implant.
[0077] FIG. 21 illustrates a substrate 210, source and drain
regions 211, and sacrificial masks 212. The sacrificial masks 212
are formed to occupy space that will eventually become the
conductive gates 240 (FIG. 24). Once again, acceptable variations
in manufacturing tolerances are illustrated in the differences in
the lengths of the sacrificial masks 212. As with the previous
embodiments, these variations will be automatically compensated for
by the self-aligned nature of the compensation implant.
[0078] In FIG. 22, the spacers 220 are formed adjacent the masks
212 in a similar manner as the spacers 112, shown in FIG. 11, were
formed adjacent the conductive gate 80.
[0079] In FIG. 23, the sacrificial masks 212 have been partially
etched to create openings 230, 231 to allow the compensation
implant 232 to be formed by impurity implant 233. More
specifically, the etching of the sacrificial masks 212 is halted
before all the sacrificial ask material has been removed.
[0080] Again, depending upon the variation within the manufacturing
process, different sized sacrificial masks will be formed which
causes different sized openings 230, 231 in the sacrificial masks
to form during the partial etch of the sacrificial material.
Therefore, different sized openings 230, 231 allow differing
concentrations of impurity to be implanted in the compensation
implant 232. In the example shown in FIG. 23, only one of the
openings (e.g, 230) will receive the compensation implant 233
because the opening 231 in the sacrificial mask did not fully form.
However, as discussed above, the impurity concentrations of the
compensation implants can vary from a heavy implant, light implant
to no implant.
[0081] After the doping process and the formation of the
compensation implants, the remainder of the sacrificial masks 212
are removed and the gates 240 are formed, as shown in FIG. 24. More
specifically, the gates 240 are selectively deposited, using
processes well known to those ordinarily skilled in the art, in the
openings between the spacers 220. As with the gates discussed in
the previous embodiments, the gates preferably comprise a
conductive material, such as metal or polysilicon.
[0082] FIGS. 21-24 again illustrate that the impurity
concentrations of the compensation implants are automatically
varied depending upon the channel length because the compensation
implants are self-aligned. Therefore, the compensation implants
adjust the amount by which the threshold voltage is reduced
depending upon the channel length.
[0083] By matching the reduction in threshold voltage to the
channel length, the invention substantially eliminates the short
channel effect for devices having channel lengths as low as 0.05
.mu.m.
[0084] Thus, with the invention, a channel length which is in the
upper range of manufacturing tolerances, which would normally
exhibit an increase in threshold voltage will be matched with a
more heavily doped compensation implant. As discussed above, the
heavier doped compensation implant will cause a larger reduction in
threshold voltage and, therefore, counterbalance the increased
threshold voltage associated with the larger channel length.
[0085] Similarly, a channel length which is in the lower range of
manufacturing tolerances, which would normally exhibit less of an
increase in threshold voltage will be matched with a less heavily
doped compensation implant. As discussed above, the lighter doped
compensation implant will cause a smaller reduction in threshold
voltage and, therefore, counterbalance the smaller increase in
threshold voltage associated with the smaller channel length.
[0086] Therefore, the self-aligned nature of the inventive
compensation implant permits devices having channel lengths as
small as 0.05 .mu.m to be produced which do not suffer the adverse
effects of threshold voltage roll-off.
[0087] While the invention has been described in terms of preferred
embodiments, those skilled in the art will recognize that the
invention can be practiced with modification within the spirit and
scope of the appended claims.
* * * * *