RU2489533C1 - Substrate for growing epitaxial layers of gallium arsenide - Google Patents

Abstract

FIELD: chemistry.

SUBSTANCE: invention involves use of substrates made from intermetallic compounds to grow epitaxial layers of GaAs, said compounds having a strict stoichiometric composition, and specifically from gallium lanthanides GaLa₃ and Ga₃La₅, gallium zirconides Ga₃Zr and Ga₃Zr₅, aluminium zirconide Al₃Zr, aluminium ceride CeAl₂, palladium beryllide BePd, magnesium lanthanide MgLa, aluminium lanthanide Al₂La, platinum stannide Pt₃Sn, indium lanthanide InLa, tin zirconide SnZr₄, platinum plumbide Pt₃Pb.

EFFECT: invention enables to significantly improve electrophysical parameters of gallium arsenide by preventing diffusion of substrate components into the epitaxial layer.

1 tbl

Description

The invention relates to electronic equipment, namely to materials intended for the manufacture of semiconductor devices of a wide class of application using epitaxial layers of gallium arsenide.

At present, the high quality of auto-epitaxial layers of gallium arsenide by electrophysical properties has been achieved. Despite this, the epitaxial layers of gallium arsenide still cannot displace silicon layers in the mass production of electronic equipment. This is due to the fact that modern technology for their production requires substrates of gallium arsenide single crystals with a diameter of 250 mm or more. However, the current capabilities of industrial technology for growing gallium arsenide single crystals are limited by a diameter of about 100-150 mm. An increase in the diameter of single crystals over this size leads to the appearance of a radial inhomogeneity in the composition of gallium arsenide single crystals.

Multiple attempts have been made to grow high-quality layers of gallium arsenide on other alternative substrates (substrates for hetero-epitaxial layers of gallium arsenide), uniform in composition with a large diameter single crystals, for example, on silicon or sapphire. These attempts were unsuccessful primarily due to the large difference in the crystal lattice periods of gallium arsenide on the one hand and silicon and sapphire, on the other.

Attempts are also made to apply transition layers between gallium arsenide and, for example, silicon to reduce the difference in the periods of the crystal lattices between adjacent epitaxial layers, which turned out to be unsuccessful [1].

In [2], a new approach was proposed for choosing substrates for growing epitaxial layers of gallium arsenide, namely, using substrates made of intermetallic compounds (which include gallium arsenide itself), but with a metallic type of conductivity. In this patent, for the first time, it was proposed to use single crystals of intermetallic compounds of binary alloys NiAl, CoAl, AlTi, NiGa as substrates for epitaxy of gallium arsenide. The advantage of these substrate materials is the very low vapor pressure of the components, which allows you not to worry about exceeding the critical vapor pressure of the components during the synthesis of the starting material of the intermetallic compound and to prevent (significantly reduce) the evaporation of one of the components during the growth of single crystals, which usually leads to a violation of stoichiometry in grown single crystal. From this point of view, in contrast to the synthesis of gallium arsenide, the technology for growing these single crystals is simpler and the cultivation of high-quality, fairly uniform in composition single crystals of large size was not a big problem. The disadvantages of these materials include primarily the relatively high melting point (for example, in NiAl, the melting point is ~ 1660 ° C). This requires the use of special high-temperature equipment for growing single crystals of these compounds. The need to use such equipment makes it difficult to use the simpler horizontal Bridgman method, which is widely used in laboratory practice.

A significant decrease in the temperature of growing single crystals for substrates compared to [2] was achieved using intermetallic compounds described in the patent [3], which is a prototype of the alleged invention. The intermetallic compounds proposed in this patent have significantly lower melting points (nickel stannate - Ni ₃ Sn ₂ -1138 ° С, nickel aluminate - Ni ₂ Al ₃ -1264 ° С, cobalt germanate - Co ₂ Ge-1200 ° С, nickel germanate - Ni ₂ Ge-1200 ° С, palladium antimonide - PdSb-805 ° С, manganese semi-animonide - Mn ₂ Sb-948 ° С, tin arsenide - SnAs-605 ° С. Such relatively low melting points greatly simplify the growth of these single crystals. were grown both by the Czochralski method and by the simpler and more affordable Bridgman method.

An X-ray microanalysis of substrates made from grown single crystals showed that these substrates, within the sensitivity of an X-ray microanalysis, are fairly uniform in composition over their entire surface. The epitaxial layers grown on these substrates were characterized by a fairly high crystallographic perfection. However, the measurement of the electrophysical properties of the epitaxial layers of gallium arsenide grown on these substrates showed that even at sufficiently low growth temperatures (300-350 ° C), the electrophysical properties of these layers (aging) gradually decrease compared with the properties of layers grown on gallium arsenide substrates. This leads to a severe limitation of the use of such structures in the creation of components of microelectronic devices. Our studies on the temperature dependence of the electrophysical properties (electron mobility in the epitaxial layer on the substrate temperature at the time of growing) show that, in contrast to ordinary cases, when an increase in the substrate temperature increases the electron mobility, in our case, an increase in temperature leads to a deterioration in their electrophysical properties . In our opinion, this is due to the fact that with an increase in the substrate temperature, the process of diffusion of one of the components of the substrate material into the epitaxial layer of gallium arsenide growing on it is activated.

The intermetallic compounds declared in [2, 3] were non-stoichiometric compounds. As our studies have shown, in these intermetallic compounds there is a noticeable diffusion of one of the components (usually excessive) both over the entire single crystal substrate and diffusion into the gallium arsenide layer growing on this substrate. This diffusion of components from the substrate into the growing layer at elevated temperatures occurs almost constantly, which leads to a deterioration in the stability of the electrophysical parameters of the grown layer.

The technical problem solved in the proposed invention is the exclusion of harmful diffusion of the substrate components into the epitaxial layer of gallium arsenide. At the same time, a positive technical result is achieved, namely, the stability of the electrophysical parameters of the grown layer is increased.

This technical problem is solved by the fact that it is proposed to use intermetallic compounds from the group of gallium lanthanides GaLa ₃ and Ga ₃ La ₅ , gallium zirconides Ga ₃ Zr and Ga ₃ Zr ₅ , aluminum zirconide Al ₃ Zr, aluminum ceride CeAl ₂ , beryllide as the substrate material palladium BePd, magnesium lanthanide MgLa, aluminum lanthanide Al ₂ La, platinum stannide Pt ₃ Sn, indium lanthanide InLa, tin zirconide SnZr ₄ , platinum plumbite Pt ₃ Pb.

The proposed intermetallic compounds consist mainly of elements of the third group: lanthanum, gallium, indium, aluminum or elements of the fourth group - zirconium, lead, tin, which, if they enter gallium arsenide, create only small levels that do not capture the main carriers. This distinguishes these materials from the materials used in [3] - metals of the transition group — nickel, cobalt, manganese, which usually create extremely undesirable deep levels in gallium arsenide.

These intermetallic compounds not only have good agreement between the crystal lattices and the gallium arsenide layer grown on them, but also ensure the stability of the electrophysical parameters of the active layers grown on them (Table 1 shows some properties of this group of intermetallic compounds).

Table 1 Compound Al ₃ Zr Ga ₃ Zr Bepd Al ₂ ce Pt ₃ sn Al ₂ la GaLa ₃ The periods of the crystal lattices, Å a _t = 4.013 a _t = 3.963 a _k = 8.19 a _k = 8.055 a _k = 4.004 a _k = 8.145 a _k = 5.66 Melting point, ° C 1580 1275 1455 1480 1402 1405 615 The mismatch of the periods of the crystal lattices, in%: +0.389 -0.85 +0.265 +0.75 +0.18 +1.88 +0.12

Compound Ga ₃ La ₅ Ga ₃ Zr ₅ Inla MgLa Pbpt ₃ Snzr ₄ The periods of the crystal lattices, Å a _t = 8.056 a _g = 8.080 a _k = 3.985 a _k = 3.970 a _k = 4,058 a _k = 5.65 Melting point, ° C 825 1600 1125 745 916 1327 The mismatch of the periods of the crystal lattices,%: +0.763 +1.06 -0.318 -0.69 +1.5 -0.05 a _k is the cubic lattice constant, a _t is the tetragonal lattice constant, a _g is the hexagonal lattice constant

We have grown single crystals of these intermetallic compounds, prepared substrates for epitaxial growth from them, and carried out the process of epitaxial growth by molecular epitaxy using known methods described in particular in the patent [2]. Our studies have confirmed that the quality of the GaAs epitaxial structures grown on the surface of these intermetallic compounds is distinguished by a sufficiently high perfection suitable for instrumentation. Their main advantage is the stability of electrophysical properties over time: control measurements of such important parameters of the epitaxial layers as mobility and electron concentration during the year did not reveal any noticeable changes.

BIBLIOGRAPHY

1. Kadowa et al. Study of initial buffer in GaAs-on-Si grows Journal of crystal grows 115 (1991), 128-132.

2. RF patent No. 2267565, IPC С30 В 19/12 “Substrate for growing epitaxial words of gallium arsenide”.

3. RF patent No. 2308784, IPC HO1L 21/203 “Substrate for growing epitaxial words of gallium arsenide”.

Claims (1)

A substrate for growing epitaxial layers of gallium arsenide made of a single crystal of an intermetallic compound, characterized in that the intermetallic compound is selected from the group of stoichiometric intermetallic compounds, namely gallium lanthanides GaLa ₃ and Ga ₃ La ₅ , gallium zirconides Ga ₃ Zr and Ga ₃ Zr ₅ , aluminum zirconide Al ₃ Zr, aluminum ceride CeAl ₂ , palladium beryllide BePd, magnesium lanthanide MgLa, aluminum lanthanide Al ₂ La, platinum stannide Pt ₃ Sn, indium lanthanide InLa, tin zirconide SnZr ₄ and platinum platinum Pt ₃ Pb.