RU2344209C2 - Method of producing mono-crystals of indium antimonide alloyed with tin - Google Patents

Abstract

FIELD: metallurgy, crystal growing.

SUBSTANCE: invention refers to process of production of A^IIIB^V semi-conducting compositions. Mono-crystals of indium antimonide alloyed with tin are produced by means of bombardment with a full specter of reactor neutrons with successive heating, annealing and cooling. Heating is carried out at the rate of 20÷40 deg/min to temperature of annealing, defined by the formula T_anneal=450+(tgN_sn-14)-7 [°C], where N_sn is concentration of introduced alloying addition of tin [cm^-3]; annealing is performed during 20 minutes, while the successive cooling is carried out at the rate of 5-10 deg/min to the temperature of 350÷400°C, and further at the rate of 20-40 deg/min to an ambient temperature.

EFFECT: alloying of indium antimonide plates with tin to high concentrations, also upgraded uniformity of tin distribution and electrone mobility.

2 ex, 1 tbl

Description

The invention relates to a technology of semiconductor compounds of type A ^III B ^V and can be used to obtain single-crystal tin doped indium antimonide.

Single crystals of indium antimonide are traditionally doped during melt growth by the Czochralski method. This method has several disadvantages and, first of all, a high heterogeneity of the distribution of the dopant in the bulk of the material (up to 15 ÷ 20% in diameter and up to 50 ÷ 100% along the length of the ingot); low stability of parameters after heat treatment; high degree of compensation. In addition, the doping of indium antimonide single crystals with tin during the growing process cannot be realized to concentrations above 5 ÷ 7 × 10 ¹⁶ cm ^-3 due to the low distribution coefficient of tin. The widespread use of indium antimonide in the manufacture of electronic devices, the high degree of integration of devices put forward stricter requirements for the quality and geometric dimensions of single crystals. Improving the parameters of the material by metallurgical methods in the process of growing single crystals is currently practically impossible.

The aim of the invention is the doping of indium antimonide plates with tin to high concentrations, increasing the uniformity of tin distribution and electron mobility.

This goal is achieved by the fact that when implementing the method of doping indium antimonide single crystals with tin, including the introduction of tin into a solid sample, heat treatment and cooling, tin is introduced by irradiation with a stream of particles containing thermal neutrons, heating is carried out at a rate of 20 ÷ 40 deg / min, heat treatment is carried out for 20 min at a temperature determined by the dependence

T _anne = 450 + (logN _Sn -14) 7

where N _Sn is the concentration of the introduced dopant of tin [cm ^-3 ].

Moreover, the cooling is carried out first at a speed of 5 ÷ 10 deg / min to a temperature of 350 ÷ 400 ° C, and then at a speed of 20 ÷ 40 deg / min to room temperature.

Irradiation of indium antimonide with thermal neutrons leads to the conversion of the In ¹¹⁵ isotope, which is 96% of the natural mixture, to tin by the reaction

In ¹¹⁵ (n, γ) In ¹¹⁶ → Sn ¹¹⁶ .

The resulting tin atoms at the sites of the cationic sublattice manifest themselves as donors.

The depth distribution of the introduced impurity (in the direction of the particle flow) has the following character:

N _Sn (x) = Ф _t N ₀ Kσexp (-N ₀ Kx),

where Ф _t - thermal neutron fluence, cm ^-2 ; N ₀ is the concentration of atoms of the substance, cm ^-3 ; K is the content of the required isotope,%, σ is the isotope activation cross section, cm ² .

The magnitude of the thermal neutron fluence is determined by the specified value of the tin concentration according to the dependence

Ф _t = N _Sn / N ₀ Kτ,

where N _Sn is the concentration of tin introduced as a result of nuclear reactions.

With a sample thickness d << (N ₀ Kσ) ^-1 , a uniformly doped sample is obtained.

Annealing the samples under these conditions removes radiation defects that arise in the material during irradiation.

At a heating rate of less than 20 deg / min, partially coagulated thermostable clusters are formed in the material as a result of diffusion of point structural defects, the annealing of which requires higher temperatures.

The upper limit of the heating rate of 40 deg / min ensures the complete destruction of various kinds of clusters of point defects. Exceeding this limit leads to the appearance of new defects that worsen the uniformity of the distribution of the introduced impurity.

Since the heat treatment temperature, at which the complete decay of radiation defects occurs, depends on the neutron fluence, an equation was derived that allows for a given heating rate to accurately determine the optimal heat treatment temperature depending on the flux of thermal neutrons. The duration of the heat treatment, as experience shows, should be at least 20 minutes. When the exposure time is less than 10 minutes, they do not have time to completely annihilate radiation defects.

The optimal mode of crystal cooling after annealing of radiation defects is stepwise. In this case, at the first stage, the cooling rate should be within 5 ÷ 10 deg / min to a temperature of 350 ÷ 400 ° C. The need for such a speed is caused by the fact that the crystal must pass the lower plastic limit with an optimally low speed in order to exclude the generation of additional defects formed during cooling in the quenching mode. The upper limit of the cooling rate at which additional defects begin to occur is determined experimentally and is equal to 10 deg / min.

At a cooling rate below 5 deg / min, a set of point structural defects has time to undergo restructuring, for example, with the formation of irreversible clusters, which include tin atoms that do not exhibit donor properties. This process takes place with a sufficiently long exposure of the material at temperatures below 350 ÷ 400 ° C; therefore, cooling the irradiated crystal from 350 ÷ 400 ° C to room temperature should be carried out at a higher speed, of the order of 20 ÷ 40 deg / min. The upper limit of the cooling rate is determined by the mechanical strength of the heat-treated material.

Such a heat treatment mode allows one to get rid of radiation defects, improve the uniformity of the electrophysical properties of the crystal, and increase the mobility of charge carriers. The optimal annealing temperature was calculated by the formula depending on the fluence of thermal neutrons.

Example 1

Unalloyed indium antimonide single crystal obtained by the Czochralski method with a diameter of 76 mm with electron concentration and mobility of 6 × 10 ¹³ cm ^-3 and 5 × 10 ⁵ cm ² V ^-1 s ^-1, respectively, at 77 K and a nonuniform distribution of charge carriers in diameter 23% cut into plates up to 2 mm thick, which are introduced into the channel of a nuclear reactor and irradiated with a full spectrum of reactor neutrons at a thermal neutron flux intensity of 1 × 10 ¹² cm ^-2 s ^-1 to a set of thermal neutron fluence of 2 × 10 ¹⁷ cm ^-2 . After deactivation, the samples are sealed in a vacuum quartz ampoule for subsequent annealing.

Knowing the neutron fluence, according to the proposed formula, the annealing temperature of radiation defects is preliminarily calculated:

T _anne = 450 + (logN _Sn -14) · 7 [° С].

Then, the prepared ampoule with irradiated samples is placed in a resistance furnace and heated at a speed of 20 deg / min to a temperature of 470 ° C, maintained at this temperature for 20 min, then cooled to a temperature of 350-400 ° C at a speed of 10 deg / min, then to room temperature at a speed of 20 deg / min. After heat treatment, the Hall coefficient is measured, which confirms the presence of an electron concentration in the crystal at a level of 4 × 10 ¹⁷ cm ^-3 . In this case, the electron mobility is μ = 50,000 cm ^-2 V ^-1 s ^-1 , and the heterogeneity of the distribution of the concentration of electrically active tin along the plate diameter is Δ = 1.7%.

Example 2

An undoped single crystal of indium antimonide obtained by the Czochralski method with a diameter of 76 mm with an electron concentration and mobility of 6 × 10 ¹³ cm ^-3 and 5 × 10 ⁵ cm ² V ^-1 s ^-1, respectively, at 77 K and a nonuniform distribution of charge carriers along the diameter of 23% cut into plates up to 2 mm thick, which are introduced into the channel of a nuclear reactor and irradiated with a full spectrum of reactor neutrons at a thermal neutron flux intensity of 1 × 10 ¹² cm ^-2 s ^-1 to a set of thermal neutron fluence of 2 × 10 ¹⁸ cm ^-2 . After deactivation, the samples are sealed in a vacuum quartz ampoule for subsequent annealing.

Knowing the neutron fluence, according to the proposed formula, the annealing temperature of radiation defects is preliminarily calculated:

T _anne = 450 + (logN _Sn -14) · 7 [° С].

Then the prepared ampoule with irradiated samples is placed in a resistance furnace and heated at a speed of 40 deg / min to a temperature of 480 ° C, kept at this temperature for 20 min, then cooled to a temperature of 350-400 ° C at a speed of 20 deg / min, then to room temperature at a speed of 40 deg / min. After heat treatment, the Hall coefficient is measured, which confirms the presence of an electron concentration in the crystal at a level of 4 × 10 ¹⁸ cm ^-3 . In this case, the electron mobility is μ = 30,000 cm ² V ^-1 s ^-1 , and the heterogeneity of the distribution of the concentration of electrically active tin along the plate diameter is Δ = 2.1%.

Table 1 shows the data characterizing the effect of heat treatment temperature, heating and cooling rates on the electrophysical parameters of tin-doped material by neutron irradiation.

Compared with the base object, which is the method for doping indium antimonide during crystallization from tin-doped melt, the proposed method allows to obtain a material having approximately an order of magnitude lower inhomogeneity in electron concentration and mobility greater by 10-30% for a specific carrier concentration charge.

Information sources

1. A.Ya. Nashelsky. The production of semiconductor materials. Moscow, Metallurgy, 1989, 271 pp.

2. N.G. Kolin, D.I. Merkurisov, S.P. Soloviev. Electrophysical properties of nuclear doped indium antimonide. FTP, t. 33, issue 7, p. 774, 1999

Claims (1)

A method for producing tin-doped indium antimonide single crystals by irradiating a full spectrum of reactor neutrons, followed by heating, annealing and cooling, characterized in that, in order to increase the mobility of charge carriers and the electrical activity of the dopant in single crystals, heating is carried out at a rate of 20 ÷ 40 degrees ./min to the annealing temperature determined by the formula
T _anne = 450 + (logN _Sn -14) · 7 [° С],
where N _Sn is the concentration of the introduced dopant of tin [cm ^-3 ], annealing is carried out for 20 minutes, and subsequent cooling is carried out at a speed of 5 ÷ 10 deg./min to a temperature of 350 ÷ 400 ° C, and then at a speed of 20 ÷ 40 deg./min to room temperature.