Graded LDD implant process for sub-half-micron MOS devices

Details Graded LDD implant process for sub-half-micron MOS devices Graded LDD implant process for sub-half-micron MOS devices

2000-08-28

2002-09-10

Chaudhari, Chandra (Department: 2813)

Semiconductor device manufacturing: process

Making field effect device having pair of active regions...

Having insulated gate

C438S305000, C438S307000, C438S595000

Reexamination Certificate

active

06448141

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to semiconductor integrated circuit fabrication and, more particularly, to improved processes for fabricating MOS field effect transistors having graded lightly doped drain and source regions.
2. State of the Art
Semiconductor integrated circuits are comprised of a plurality of devices which invariably include transistors. Transistors are of two general types, namely bipolar and field effect transistors (FETs).
FIG. 1
shows the common type of FET structure generally used to form metal oxide substrate (MOS) type circuits. In this case, an N-channel MOS FET (NMOSFET) structure comprises a substrate
1
of semiconducting material such as silicon having a region which has been doped to form a “P-well”
2
. A detailed description of this process may be found in “Silicon Processing For The VLSI Era”—Volume 2, Process Integration, Lattice Press 1990, pp 428-441. An active area
3
is defined between isolating field oxide regions
4
and
5
. A gate region
6
of conductive material such as polysilicon (poly) is separated from the surface by a layer
7
of dielectric material such as silicon dioxide (SiO
2
). Conductive interconnect material
8
such as tungsten silicide is formed above and in contact with the gate region which interconnects the gate to other circuit devices. Implanted into the surface of the P-well
2
astride the gate region
6
are source
9
and drain regions
10
of N type semiconductor material with the FET channel
11
formed in between. A cap layer
12
and sidewall spacer structures
13
of insulating material such as nitride protect the gate structures during subsequent processing such as the self aligned implants of the source and drain regions and the formation of conductive structures which interconnect these regions.
MOSFETs in combination with other devices commonly form dynamic random access memory circuits (DRAM) used in memory systems such as computers. Because of the continuous demand for the further miniaturization and speed increase of DRAMs, MOSFET devices have been scaled to the point where the channel length from source to drain falls below 0.5 micron (sub-half micron). As the channel shrinks, the maximum electric field (E-field) in the channel region increases, thereby resulting in higher substrate current and short/long term hot electron reliability problems. Electrons traveling through the channel become more energized by the E-field and have a greater tendency to cross into the gate dielectric region
7
and become trapped. These problems are discussed in detail in “Silicon Processing For The VLSI Era—Volume 2”, Lattice Press, 1990, pp 428-441.
The reference cited above also discusses various methods employed to partially overcome these problems and maximize performance and reliability. One common method involves adding a first lightly doped region between the drain and channel regions and a second lightly doped region between the source and channel regions.
FIGS. 2 and 3
show the typical fabrication sequence for this structure. In
FIG. 2
, after formation of the insulating sidewall spacer structures
13
, a low dosage phosphorous implant and drive create N− regions
14
,
15
. Due to the relatively high diffusivity of phosphorous, the N− regions extend underneath the spacers toward the FET channel area
11
. In
FIG. 3
, a high dosage arsenic implant and drive creates N+ source
16
and drain
17
regions which supersede most of the lightly doped N− regions. What remains are lightly doped regions
18
,
19
separating the source and drain from the channel. This structure has come to be known as a lightly doped drain (LDD) structure. The use of LDD structures to relax the E-field is well known.
However, as the devices get smaller, and FET channels become shorter than 0.4 microns, limitations on fabrication precision result in structures that are far from the ideal one shown in FIG.
3
. Due to its high diffusivity, the phosphorous in the N− regions further diffuses into the channel during the high heat drive processes required to create the N+ source and drain regions. This causes severe short channel problems resulting in increased sub-threshold leakage which adversely affects refresh time in DRAMs.
An alternative to the phosphorous LDD (phos-LDD) approach is to use arsenic to create the LDD structures as proposed by H. R. Grinolds, et al. in “Reliability and Performance of Submicron LDD NMOSFET's with Buried-As n-Impurity Profiles,” IEDM Tech. Dig., 1985, pp. 246-249 and by C.-Y. Wei, et al. in “Buried and Graded/Buried LDD Structures for Improved Hot-Electron Reliability,” IEEE Electron Device Lett., vol. EDL-7, np. Jun. 6, 1986.
The fabrication processes required to create an arsenic LDD (As-LDD) proceed similarly to the phosphorous LDD processes.
FIGS. 4 and 5
show that an LDD structure can be created by first implanting a low dosage, self-aligned arsenic implant prior to sidewall spacer formation. This implant is then diffused into the substrate through a heating drive process, resulting in the lightly doped N− regions
20
and
21
.
In
FIG. 5
, after the formation of insulating sidewall spacers
13
, the N+ source
22
and drain
23
regions are created using a high dosage, self-aligned arsenic implant and drive. The N+ regions supplant portions of the N− regions. There remains, however, a first lightly doped N− LDD region
24
existing between the N+ source region
22
and the channel area
11
, and a second lightly doped N− LDD region
25
existing between the N+ drain region
23
and the channel area
11
.
Since low diffusivity arsenic was used to create the N− LDD regions, the resulting LDD structures are much more predictable and do not suffer from the short channel problems plaguing phosphorous LDD structures. However, arsenic's low diffusivity also causes the N− LDD regions
24
,
25
to have an abrupt end
26
,
27
below the edges of the gate region
6
. This abruptness creates an E-field which is still unsuitable in sub half-micron devices due to the resulting hot electron reliability problem.
To alleviate this problem, a combination phos/As LDD structure has been developed where a phos-LDD implant occurs immediately after an As-LDD implant to grade the channel to LDD junction. Again, due to the diffusivity of phosphorous, during subsequent processing, the short channel characteristics are degraded.
Another method to reduce the E-field involves burying the drain/channel and source/channel junctions.
FIG. 6
shows an NMOSFET having buried drain/channel and source/channel junctions. In this FET, the N+ source and drain regions
28
,
29
each have a projection
30
,
31
which terminates at a junction
32
,
33
with the channel
11
. The projections exist a distance below the channel/gate dielectric interface
7
. This moves the highest concentration of hot-electrons deeper into the channel area and away from the gate dielectric region. Buried structures incorporating LDD regions and graded combination structures have also been created, but at the expense of device speed. Although these structures offer promise, they are quite costly to construct and, therefore, not currently viable solutions, economically.
It would be desirable, therefore, to have a process which produces a sub half-micron MOSFET with a low E-field and improved short channel characteristics and reliability in an efficient and economical manner.
SUMMARY OF THE INVENTION
The primary and secondary objects of this invention are to provide a process for creating reliable and inexpensive sub-half-micron NMOSFETs.
These and other objects are achieved by a process wherein a low dosage N− phosphorous implant occurs after the high dosage N+ arsenic implant and drive which creates the source and drain.


REFERENCES:
patent: 4939386 (1990-07-01), Shibata et al.
patent: 4949136 (1990-08-01), Jain
patent: 5021851 (1991-06-01), Haken et al.
patent: 5217910 (1993-06-01), Shi

Affiliated with

Also associated with

Rating

Graded LDD implant process for sub-half-micron MOS devices does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Graded LDD implant process for sub-half-micron MOS devices, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Graded LDD implant process for sub-half-micron MOS devices will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2823580