Preparing substrates by annealing epitaxial layers in the form o

Details Preparing substrates by annealing epitaxial layers in the form o Preparing substrates by annealing epitaxial layers in the form o

1991-09-12

1994-01-18

Fourson, George

Metal treatment

Barrier layer stock material, p-n type

With recess, void, dislocation, grain boundaries or channel...

148DIG97, 437 82, 437126, 437128, 437939, 437247, H01L 21324, H01L 29161

Patent

active

052796876

DESCRIPTION:

BRIEF SUMMARY
This invention relates to the preparation of substrates used as the starting point for epitaxial growth.
Many varieties of semiconductor devices are produced by depositing one or more single crystal layers onto a substrate. The processes used for the deposition, e.g. molecular beam epitaxy (MBE) and metal organic vapour phase epitaxy (MOVPE), can be adapted to produce a wide range of chemical compositions. However it is necessary to start such process from a primary substrate and the range of primary substrates is very limited because the crystallisation techniques used to prepare them are suitable only for a limited range of pure compounds and elements. It should be appreciated that the size of the unit crystallographic cell of the deposited layer is dependent on the chemical composition and it may not be possible to obtain a primary substrate, e.g. a Czochralski wafer, which has the same cell dimensions.
As an example, the use of alloys of Si/Ge as semiconductors would increase the range of operational properties for high speed circuits but it is not feasible to make Czochralski wafers out of Si/Ge alloys. Thus it is necessary to grow the alloy onto pure Si in which case the unit cell sizes do not match. Another example, selected from a different chemical system, concerns semiconductors based on In, Ga, As, P. The available primary substrates are InP, InAs, GaP and GaAs. In device structures ternary or quaternary layers are usually chosen so that the cell size matches that of the primary substrate. The designer's freedom of choice would be increased if it were possible to reduce the adverse effects of mis-match.
Other reasons for mis-matched cell sizes also exist. For example, it would be convenient to grow ternary or quaternary compounds of In, Ga, P and As onto silicon because the Si is stronger and cheaper than InP. In addition, Si has a higher thermal conductivity than InP. This would be advantageous for integrated devices including, for example, lasers and transistors. In this case the choice of substrate is based upon physical and chemical properties rather than availability but the same epitaxial problem is encountered, namely that the primary substrate does not have the same cell size as the intended epilayers.
The problems caused by growing onto a mis-matched primary substrate will now be explained. If only one or two layers of atoms are deposited, then the epilayer will elastically distort so that its "in-plane" lattice constant matches the lattice constant of the primary substrate. If the epilayer experiences biaxial compression in the substrate plane then its dimension normal to that plane will be increased. As the epilayer is much thinner than the primary substrate, the primary substrate will retain its normal structure and all the strain will be imposed on the epilayer.
The strain in the epilayer represents stored energy which causes instability. More specifically there is an equilibrium critical thickness. Below this thickness the strain energy is insufficient to cause dislocations and the epilayer will remain in a stable strained state; above the equilibrium critical thickness it would be advantageous for the structure to dislocate. However, in most cases the activation energy of dislocation is high and there is a metastable critical thickness. Below this metastable thickness the strain energy is insufficient to provide the activation energy and, therefore, the system is metastable. Above the metastable critical thickness there is enough strain energy to cause the structure to dislocate. Since most useful devices are thicker than the metastable critical thickness it will dislocate.
The misfit dislocations which form when a layer relaxes are generally in the form of dislocation half loops comprising an interfacial component and two arms that thread up through the whole volume of the epitaxial layer to the surface. Clearly the density of interfacial disclocations depends on the degree of mis-match. If the match were perfect there would be no dislocations. If the linear dimensions in the epi

REFERENCES:
patent: 3890170 (1975-06-01), Russ
patent: 4529455 (1985-07-01), Bean et al.
patent: 4696095 (1987-09-01), Thomas
patent: 4952527 (1990-08-01), Calawa et al.
Asai et al: "Heteroepitaxial growth of Ge films on the Si(100)-2X1 surface" p. 257702583, Journal of Applied Physics, vol. 58, No. 97, Oct. 1985.
Applied Physics Letters, vol. 44, No. 12, 15 Jun. 1984, pp. 1149-1151, American Institute of Physics, New York, US; W. I. Wang: "Molecular beam epitaxial growth and material properties of GaAs and AlGaAs on Si(100)".
Journal of Vacuum Science & Technology/B, vol. 3, No. 2, Mar./Apr. 1985, p. 603, American Vacuum Society, Woodbury, N.Y., US; P. N. Uppal et al: "Summary abstract: MBE growth of GaAs and GaP on Si(211)".
Journal of Vacuum Science & Technology/A, vol. 4, No. 3, part 1, May/Jun. 1986, pp. 894-896, American Vacuum Society, Woodbury, N.Y., US; J. M. Seo et al: "Growth and surface structure of Ge-Si alloy films on Si(111)-(7.times.7)".

Affiliated with

Also associated with

Rating

Preparing substrates by annealing epitaxial layers in the form o does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Preparing substrates by annealing epitaxial layers in the form o, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Preparing substrates by annealing epitaxial layers in the form o will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-1133957