Mos transistor for high density integration circuits

Details Mos transistor for high density integration circuits Mos transistor for high density integration circuits

2003-01-06

2004-08-10

Nelms, David (Department: 2818)

Active solid-state devices (e.g., transistors, solid-state diode

Schottky barrier

C257S054000, C257S073000, C257S260000, C257S449000, C257S476000

Reexamination Certificate

active

06774451

ABSTRACT:

TECHNICAL FIELD
This invention concerns a MOS transistor that can be used to make high density integrated circuits.
STATE OF THE PRIOR ART
FIG. 1
is a schematic diagram of the structure of a traditional MOS transistor. It shows an n channel MOS transistor. It is made on a type p solid silicon chip
1
. Two sections
2
and
3
of type n
+
are created, for example by ionic implantation, to form respectively the transistor's source and drain. Reference
4
indicates the gate oxide film created on the surface of the chip
1
under which the sections
2
and
3
extend beyond. A gate electrode
5
is deposited on the gate oxide film
4
. Source
6
and drain
7
electrodes are respectively created on sections
2
and
3
and the transistor is insulated from surrounding devices by an oxide film
8
. The channel
9
is also shown, which is formed under the gate oxide film
4
when the transistor is correctly biased. The following important parameters are also shown in the diagram:
the channel length L,
the thickness of the gate oxide film d,
the depth of the junction r
j
.
The interest sustained in the development of high density integration, high frequency and low consumption silicon technologies is largely due to the new demands for portability, in particular for telecommunications and computer equipment. The exponential growth of technical and economic performances (according to Moore's law) in silicon microelectronics is a strong indicator of the most impressive industrial progress of the twentieth century.
The widespread use of MOSFET transistors in the production of integrated circuits has enabled the density of integration to be increased significantly. However, the sustained growth in density of integration and performances of the electronics runs the risk of finding the physical limits of manufacturing. As an initial estimate, it is accepted that a reduction by a factor &agr; in the length of a MOSFET transistor gate must also be accompanied by a simultaneous reduction by a factor &agr; in the other characteristic dimensions (thickness of the gate oxide, width of channel, depth of junction), by a reduction by a factor &agr; of the input voltage and an increase by a factor &agr; in the doping concentration in the channel.
It is currently possible to obtain a minimum gate dimension of 22 nm. However, such a reduction implies modifications to the other characteristic dimensions of the transistor, which are impossible to achieve either now or in the near future, according to “The International Technology Roadmap for Semiconductors—ITRS”, 1999, SIA Semiconductor Industry Association.
The technological barriers preventing the manufacture of a traditional MOS transistor with a gate length of 22 nm are in particular as follows:
I): very shallow extension from the source and drain to their junction with the channel (from 8 to 13 nm);
II): an extremely abrupt concentration gradient at the source/channel and drain/channel junctions (0.5 nm/dec.);
III): a very thin film of silicon (12 nm);
IV): lower consumption of silicon during silicide coating reaction (from 7 to 17 nm);
V) : low resistance by silicon square due to its reduced thickness (12.5 &OHgr;/for a thickness of silicon of 12 nm);
VI): a very low specific drain and source contact resistance at the silicon—silicide interface (less than 1.5 10
−8
&OHgr;/cm
2
);
VII): a very high level of doping in the channel (3×10
19
cm
3
).
The miniaturisation of MOS devices requires shorter and shorter junctions to be formed, which makes optimising the specific contact resistance extremely difficult. The reduction of this resistance is usually achieved by a silicide deposition process, which permits an alloy to be obtained with low resistance at the silicide—silicon interface, for example the TiS
2
/Si and CoSi
2
/Si interfaces. However, the reaction of the silicide deposition process leads to a consumption of silicon which can lead, on the one hand, to a reduction of the effective contact surface and, on the other hand, to an increase in the specific contact resistance given that the silicide—silicon interface values are too low.
The article “Proposal of a SCHOTTKY barrier SET Aiming at a Future Integrated Device” by FUJISHIMA et al., published in the IEICE Transactions on Electronics, JP Institute of Electronics Information and Comm. Eng. Tokyo, vol. E 80-C, no°7, Jul. 1
st
, 1997, pages 881 to 885, reveals a one electron MOS transistor. This device uses the Coulomb effect, which is to say a load quantification effect, which allows the current to be oscillated according to the gate voltage applied. Several constraints are imposed in this case. The transistor must operate at low temperature (e.g. 10 K) so that the load quantification is not masked by the load temperature fluctuations. It is formed by a silicon channel whose external connection (source and drain contacts) passes via the Schottky barriers, which act as a tunnel barrier. In order to use the Coulomb blocking effect, the tunnel resistance of the Schottky barriers must be greater than the resistance quantum. This constraint justifies the use of relatively high Schottky barriers. Finally, the device described in this article is a quantum wire on non-doped silicon, which means that the width of the device is significantly reduced. This very narrow width is essential for the operation of the SET transistor in order to reduce all of the capacities between the transistor terminal contacts and the silicon channel, as the Coulomb blockage can only be achieved if very low capacities are used. The very narrow width of the device allows a relatively high junction tunnel resistance to be obtained, as the resistance is inversely proportional to the emission surface.
DESCRIPTION OF THE INVENTION
This invention allows the problems of manufacturing MOS transistors for high density integrated circuits to be overcome.
It concerns a MOS transistor made in the thin film of silicon of an SOI chip, said thin film being slightly doped and of less than 30 nm in thickness, the source and drain contacts being of the Schottky type at the lowest level of Schottky barrier possible for majority carriers, with an accumulation type transistor operation.
In preference, the level of doping of the thin film is between 5×10
14
cm
−3
and 10
17
cm
−3
. Advantageously, it is of around 10
15
cm
−3
, for example 1 or 2×10
15
cm
−3
.
If the thin film is p doped, to obtain a p-MOSFET, the source and drain contacts are advantageously composed of a PtGeSi silicide. Annealing, for example at a temperature of around 600° C. for approximately 10 minutes, permits the Schottky barrier of these contacts to be lowered.
If the thin film is n doped to obtain an n-MOSFET, the source and drain contacts can be composed of an erbium based silicide.


REFERENCES:
patent: 4638551 (1987-01-01), Einthoven
patent: 4665414 (1987-05-01), Koeneke et al.
patent: 4696093 (1987-09-01), Welch
patent: 5663584 (1997-09-01), Welch
patent: 6495882 (2002-12-01), Snyder
patent: 0 456 059 (1991-11-01), None
patent: 0 469 611 (1992-02-01), None

Affiliated with

Also associated with

Rating

Mos transistor for high density integration circuits does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Mos transistor for high density integration circuits, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Mos transistor for high density integration circuits will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3286231