Active solid-state devices (e.g. – transistors – solid-state diode – Field effect device – Having insulated electrode
Reexamination Certificate
1999-09-07
2001-02-20
Carroll, J. (Department: 2811)
Active solid-state devices (e.g., transistors, solid-state diode
Field effect device
Having insulated electrode
C257S344000, C257S369000, C257S371000, C257S391000, C257S392000, C438S231000, C438S232000, C438S279000
Reexamination Certificate
active
06191460
ABSTRACT:
BACKGROUND
1. Field of Invention
This invention relates to semiconductor integrated circuit memory devices, and more particularly to static random access memory (SRAM) devices in which some insulated gate transistors have N-type conductivity gates and other insulated gate transistors have P-type conductivity gates.
2. Related Art
Static random access memory (SEAM) cells using cross-connected complementary metal oxide semiconductor (CMOS) inverters are widely used. 
FIG. 1
 is a schematic drawing showing a typical prior art six transistor (6T) SRAM cell 
2
. As shown, transistors N
1
, N
2
, N
3
, and N
4
 are N-type MOS (NMOS) transistors and transistors P
1
 and P
2
 are P-type MOS (PMOS) transistors.
As shown, transistors P
1
 and N
1
 form inverter 
10
. Transistor P
1
 is coupled between a voltage source (not shown) supplying voltage V
DD 
(typically 5.0 or 3.5 volts, although other voltages are possible, e.g., 2.5 volts) and node A. Transistor N
1
 is coupled between node A and a second voltage source (not shown) supplying voltage V
SS 
(typically a reference potential, e.g., ground). Similarly, transistor P
2
 is coupled between V
DD 
and node B, and transistor N
2
 is coupled between node B and V
SS
. Inverters 
10
 and 
12
 are cross-connected such that node A is coupled to the gates of transistors P
2
 and N
2
, and node B is coupled to the gates of transistors P
1
 and N
1
. Thus if a logic low state, for example low voltage, exists at node A, transistor P
2
 conducts and transistor N
2
 does not conduct. The voltage at node B is therefore approximately V
DD 
and a logic high state, for example high voltage, exists at node B. Due to the logic high state at node B, transistor P
1
 does not conduct and transistor N
1
 conducts. The voltage at node A is therefore approximately V
SS 
and a logic low state exists at node A. Accordingly, the cross-connected inverter configuration is stable.
FIG. 1
 also shows select transistor N
3
 coupled between node A and bit line 
14
, and select transistor N
4
 coupled between node B and {overscore (bit)} line 
16
. The gates of transistors N
3
 and N
4
 are each coupled to word line 
18
. Of importance is that current flows in transistors P
1
 and P
2
 as nodes A and B, respectively, change logic state. Over time, and in large memory cell arrays, the cumulative current flows may be a significant factor in overall circuit power consumption.
FIG. 1
 includes a conventional sensing circuit 
20
 and conventional control circuit 
22
. Sensing circuit 
20
 is coupled to bit line 
14
 and {overscore (bit)} line 
16
. Control circuit 
22
 is coupled to bit line 
14
, {overscore (bit)} line 
16
, and word line 
18
. Persons skilled in the art will understand that circuits 
20
 and 
22
 provide voltage sources that control writing and reading of information stored as logic states in SRAM cell 
2
. Circuits 
20
 and 
22
 typically contain both NMOS and PMOS transistors in a variety of conventional configurations.
In early 6T-SRAM integrated circuits the NMOS and PMOS transistors both in the memory cells, and in the control and sensing circuits, were formed with conductive polycrystalline silicon (polysilicon) gate electrodes (gates) heavily doped to N-type conductivity. Later, a “dual gate process” fabrication approach was used in which all NMOS transistors were formed with polysilicon gates having N-type conductivity and all PMOS transistors were formed with polysilicon gates having P-type conductivity. The dual gate process is typically used to form 3.3V and 2.5V technology SRAM integrated circuits. For example, PMOS transistors having P-type gates are typically used for sensing and control circuits, e.g., circuits 
20
 and 
22
, in SRAM integrated circuits because their lower threshold voltage (V
T
), compared with PMOS transistors having N-type gates, and higher drain-source current (I
DS
) increases operating speed.
FIG. 2
 is a simplified cross-sectional view showing a typical CMOS structure. (Note that the overlying passivation layers and source/drain/gate electrode contacts are not shown.) The NMOS and PMOS transistors of the type shown are typically used in, for example, memory cell 
2
, sensing circuit 
20
, and control circuit 
22
 (FIG. 
1
). As shown, a conventional P-Well semiconductor area 
30
 (of P-type conductivity) and a conventional N-Well semiconductor area 
32
 (of N-type conductivity) are each formed in crystalline substrate 
34
. Conventional field oxide layer 
36
 and conventional gate oxide layer 
40
 are each formed on substrate 
34
's surface 
38
 as shown. Conductive gate 
42
A is formed on oxide layer 
40
 above region 
30
 and conductive gate 
42
B is formed on oxide layer 
40
 above region 
32
. In single gate technology, gates 
42
A and 
42
B are typically heavily N-type (N+) doped polysilicon. For dual gate technology, gate 
42
A is typically N+ polysilicon, and gate 
42
B is typically heavily P-type (P+) doped polysilicon.
As shown, NMOS transistor 
44
 is formed with N+ regions 
46
, and with lightly doped N-type (N−) regions 
48
 underlying conventional sidewall spacers 
50
 as shown. The N+ regions 
46
 are approximately aligned (self-aligned) with spacers 
50
 as shown. The N− regions 
48
 are approximately aligned (self-aligned) with gate 
42
A. Also shown is PMOS transistor 
52
 formed with P+ regions 
53
, and with lightly doped P-type (P−) regions 
54
 underlying spacers 
50
 as shown. The P+ regions 
53
 are approximately aligned (self-aligned) with spacers 
50
 as shown. The P− regions 
54
 are approximately aligned (self-aligned) with gate 
42
B as shown. Persons skilled in the art will be familiar with various conventional fabrication methods used to form transistors 
44
 and 
52
.
Stability is the ability of a memory cell to retain its programmed state; SRAM cell stability is desirable. In 6T SRAMs, e.g., SRAM cell 
2
, memory cell stability is enhanced by reducing I
DS 
in the pull-up PMOS load transistors when these transistors conduct. One method of reducing I
DS 
is to make the PMOS transistor “weaker” or less conductive. One way of making the PMOS weaker is to lightly dope source and/or drain regions.
U.S. Pat. No. 5,804,477 ('477 Patent), under common assignment with the present invention and incorporated herein by reference, discloses “a 6-T SRAM cell which occupies less chip area, [and] has improved write speed and improved latch-up immunity.” See U.S. Pat. No. 5,804,477, col. 3 line 66. In the '477 patent a MOS transistor is made weaker by omitting heavily doped portions of source/drain regions. As described in detail in the '477 patent, lightly doped source and/or drain regions provide increased resistance with a consequent I
DS 
decrease and enhanced SRAM stability.
The '477 patent further discloses a range of dopant concentration ratios between source/drain regions and the substrate region in which the source/drain regions are formed. For example, referring to 
FIG. 2
, if N-Well 
32
 has a particular dopant concentration, and if the P− source/drain regions 
54
 comprised the entire source/drain regions, the P− regions 
54
 would have a dopant concentration of 1 to 2.5 times the particular dopant concentration of N-Well 
32
. Such a range of ratios also applies to NMOS transistors formed in P-wells.
Although the '477 patent demonstrates lowered I
DS 
and improved stability over previously known 6T SRAMs, further I
DS 
reductions and enhanced memory cell stability are desirable. Improved stability is especially important in low voltage memory cells, such as those operating at approximately 2.5 volts.
SUMMARY
In accordance with the invention, the dual gate process is modified so that both NMOS and PMOS memory cell transistors have gates of the same conductivity type. As in conventional SRAM integrated circuits, the associated NMOS sensing and control circuit transistors have N-type gates and the associated PMOS sensing and control circuit transistors have P-type gates. However,
Choi Jeong Y.
Lien Chuen-Der
Carroll J.
Integrated Device Technology Inc.
Skjerven Morrill MacPherson Franklin & Friel
LandOfFree
Identical gate conductivity type static random access memory... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Identical gate conductivity type static random access memory..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Identical gate conductivity type static random access memory... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2577431