Process for producing heteropitaxial diamond layers on...

Details Process for producing heteropitaxial diamond layers on... Process for producing heteropitaxial diamond layers on...

1995-07-17

2001-08-14

Tran, Minh Loan (Department: 2826)

Semiconductor device manufacturing: process

Having diamond semiconductor component

C257S077000, C257S627000

Reexamination Certificate

active

06274403

ABSTRACT:

BACKGROUND OF THE INVENTION
The invention relates to a method for producing heteroepitaxial diamond layers from the vapor phase on a silicon substrate, as well as to layers produced by this method and the use thereof.
As is known from M. W. Geis et al, IEEE Electron Device Letters EDL-8, 1987, 341 and M. W. Geis et al, J. Vac. Sci. Technol., A 6, 1988, 1953, as a result of its excellent physical properties such as
band gap
5.5
eV
hole mobility
1800
cm
2
/V
s
electron mobility
2000
cm
2
/V
s
electron saturation drift rate
2 × 10
7
cm/s
disruptive field
1 × 10
7
V/cm
thermal conductivity
20
W/cmK
relative permativity
5.7
diamond is an excellent material, which is suitable for high temperature, high frequency and high power semiconductor components.
After Japanese scientists described at the beginning of the 1980's that diamond can be synthesized as a layer or coating from gaseous substances, numerous efforts were made throughout the world to find new applications for this promising material with its great technological potential.
The efforts to use CVD-diamond as an active or passive electronic material have failed up to now because it has not been possible to heteroepitaxially deposit diamond from the most standard substrate, namely silicon. However, epitaxy is the prerequisite for obtaining electrical characteristics as are expected of such layers. In order to develop and test new solutions for this problem, numerous developments have been made in the past. Only localized epitactic growth has been obtained on silicon substrates (D. G. Jeng, H. S. Tuan, R. F. Salat and G. J. Fricano, Appl. Phys. Lett. 56, 1968 (1990); J. Narayan, A. R. Srivatra, M. Peters, S. Yokota and K. V. Ravi, Appl. Phys. Lett. 53, 1823, (1988)). Although heteroepitaxy is successful on C-BN, this is only so on particles, because at present it is not possible to produce a crystalline, large-surface C-BN (S. Koizumi, T. Murakami and T. Inuzuka and K. Suzuki, Appl. Phys. Lett. 57, 563 (1990)). On silicon carbide (1-wafer), a carrier material which is very difficult to produce, only a partially oriented growth of diamond (50% of the grown crystals) has been performed with bias nucleation (B. R. Stoner and J. T. Glass, Appl. Phys. Lett. 60, 698 (1992)).
Of late further publications have appeared which deal with this problem.
In physical review B V. 45, no.19, p.11067ff, Stoner et al reported on the deposition of diamond on Si. The procedure was such that a two-stage process is proposed, involving a pretreatment (nucleation) accompanied by the application of a negative bias voltage in the case of a 2% methane/hydrogen plasma. This was followed by a growth phase while maintaining the bias voltage. The crystallites in the layers obtained have a random orientation.
The working group of Yugo et al (Appl. Phys. Lett. 58 (10), p.1036ff) have also proposed a two-stage process. The first stage is performed accompanied by the application of a negative bias voltage with a CH
4
/H
2
plasma, the methane proportion being 40%. The diamond deposition is carried out without bias under “normal” conditions (approx. 0.5% methane in H
2
). However, this process leads to randomly oriented, polycrystalline layers and not to heteroepitaxial layers.
Thus, no method is known which makes it possible to heteroepitaxially deposit diamond on the most widespread substrate, namely Si.
SUMMARY OF THE INVENTION
Therefore the problem of the present invention is to provide a method for the production of heteroepitaxial diamond layers on silicon substrates.
With regards to the method, this problem is solved by the characterizing features of claim
1
and with regards to the layers by the features of claim
11
. The subclaims give advantageous further developments.
As compared with diamond, silicon has widely differing lattice parameters and a low surface energy, it is not possible under normal conditions to form a high nucleus density on not pretreated silicon and thereby achieve diamond deposition.
Surprisingly and unexpectedly on the basis of the hitherto known results, it has been found that through a planned pretreatment of the Si substrate, an excellently oriented diamond nucleation is achieved without any mechanical diamond prenucleation.
The invention proposes a three-stage process. In the first stage the Si-surface is pretreated. Preferably the procedure is such that there is either a cleaning or purification in ultra-high vacuum at <10
−9
mbar, which requires a substrate temperature of >950° C., or a plasma cleaning takes place with H
2
with or without a bias voltage. This is followed by nucleation and diamond growth.
It has proved important that prior to the nucleation phase said pretreatment is performed in order to remove from the Si-surface impurities and in particular oxides and/or oxygen compounds. This is achieved according to the first variant (high vacuum) by a desorption of the impurities and in the second variant by H
2
plasma treatment.
It has been found that very brief pretreatments (a few seconds) can be adequate
It is advantageous if the pretreatment (Si-surface cleaning) in the case of plasma treatment takes place at 300 to 1100° C. with an almost 100% H
2
plasma. When applying a bias voltage it is preferable to work at +50 to −300 V and particular preference is given to −50 V.
In the case of the plasma pretreatment it is sufficient if first a pure H
2
plasma is ignited and immediately following this (i.e. after a few seconds) the nucleation phase is performed.
In the method according to the invention, it is particularly advantageous if during plasma cleaning a negative bias voltage is also applied to the substrate. This development of the invention also favors the subsequent nucleation.
The method according to the invention can be performed with all standard process gases. The following process gases can be used: CO+CH
4
, C
2
H
5
OH, acetylene and acetone or e.g. CF
4
and methane. According to a particularly preferred embodiment of the invention methane and hydrogen are used as process gases.
In the case of CH
4
/H
2
it has also proved advantageous if the negative bias during nucleation is in the range 60 to 300 V, preferably approximately 150 V and a 0.1 to 10% process gas flow of methane in hydrogen is supplied. Particular preference is given to a 0.2 to 2% process gas flow.
The growth phase is then performed without applying a bias voltage under known conditions. It is advantageous to have substrate temperatures of 400 to 1100° C. and a 0.1 to 10% process gas flow of CH
4
in H
2
. As for nucleation, it is preferable to work with a 0.2 to 2% process gas flow.
According to the invention the desired nucleus density can be chosen over the nucleation phase duration. The desired thickness of the diamond layer can be varied as a function of the deposition period.
Thus, compared with the prior art, the method according to the invention leads to important advantages.
1. As a result of the pretreatment according to the invention Si-surfaces are effectively cleaned, so that during the second stage oriented nuclei can be produced on the silicon.
2. Compared with most other materials silicon is the basic material for modern microelectronics. Thus, as a result of the method according to the invention completely new application possibilities are provided and layers can be produced, whose electrical properties are close to those of perfect diamond monocrystals.
3. The method according to the invention makes it possible to perform the surface cleaning of substrates, nucleation and the growth of the diamond layers in a single process.
The layers produced with the method according to the invention are characterized in that the individual crystallites have edge dimensions of 0.1 to 100 &mgr;m and are oriented to 20 to 100% on the Si-substrate lattice. The method is preferably used for (100), (111) or (110)-oriented Si-substrates.
The orientation is in particular such that within the framework of a possible misorientation of less than 10°, corresponding crystal orientations <

Affiliated with

Also associated with

Rating

Process for producing heteropitaxial diamond layers on... does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Process for producing heteropitaxial diamond layers on..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Process for producing heteropitaxial diamond layers on... will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2491133