RU2155830C2 - Process of preparation of charge to produce solid solutions of chalcogenides of lead and tin by para-phase methods - Google Patents

Abstract

FIELD: technology of production of semiconductor materials and devices. SUBSTANCE: charge for production of solid solutions of chalkogenides of lead and tin is prepared by way of synthesis of charge material from elements in melt, standard directed crystallization of melt, removal of part of ingot enriched with own excessive components and impurities, grinding and processing of re-sublimations in dynamic vacuum. Then homogenizing firing is conducted at constant temperature, fraction of grains with diameter of 200-500 μm is isolated and preliminary evaporation at temperature equal to evaporation temperature of growing of layers and crystals is carried out. All mentioned operations are needed in combination and are sufficient for realization of proposed process of preparation of charge to produce layers of chalkogenides of lead and tin with improved reproducibility of composition and electrophysical properties. EFFECT: improved reproducibility of composition and electrophysical properties of layers and crystals of solid solutions of chalkogenides of lead an tin. 1 dwg, 4 tbl

Description

 The invention relates to the field of technology of semiconductor materials and devices, and more particularly to a technology for the preparation of a mixture for producing solid solutions of lead and tin chalcogenides by vapor-phase methods.

 The technical problem solved by this invention is the development of a method for preparing the initial mixture for growing single crystals and layers of solid solutions of lead and tin chalcogenides by vapor-phase methods.

There are a number of ways to prepare the initial mixture to obtain lead and tin chalcogenides and their solid solutions. According to the method described in [1], a thin layer of lead or tin chalcogenide is applied to the inner surface of the reactor by a single sublimation-condensation under gas-dynamic flow conditions. Next, the reactor is evacuated and a long (100 hours) annealing is performed under conditions close to equilibrium. During the annealing process, material components that are redundant with respect to the composition corresponding to the condition of the lowest total pressure in the system at the annealing temperature p _min (T) are discharged to the colder zone of the reactor. At the end of annealing, the material is collected in an ingot.

The described method provides a mixture with a fixed composition corresponding to the condition of the lowest total pressure p _min at the annealing temperature for a given material. This allows one to ensure high reproducibility of the properties of single crystals and layers of binary compounds PbSnX (X is chalcogen) [1, 6].

The disadvantages of this method include:
- the duration of the annealing process;
- the difficulty of deep cleaning from residual impurities [1];
- the impossibility of preparing large volumes of the charge in one technological cycle; this disadvantage stems from the need to apply the processed material on the walls of the reactor in the form of a thin layer;
- in addition, according to the provisions of chemical thermodynamics [4], in quasibinary systems based on chalcogenides of lead and tin for a given temperature there can be no more than three points corresponding to the lowest total pressure p _min , and no more than one point proper for Pb _1-x solid solutions Sn _x X and no more than one - for each of the binary components PbX or SnX; therefore, according to the described method, at a given annealing temperature T _{anne, a} solution with a single molar ratio x ^* can be prepared, and the preparation of a solid solution with other x values is impossible;
- the method is unsuitable for producing a mixture of compounds and solid solutions for which the conditions of minimum total pressure do not exist, since during the annealing process, the homogeneity of the processed material may be violated.

 Overcome the first three of these disadvantages allows the method of multiple cyclic sublimation-condensation (MCCS) [1].

 The MCSC method is implemented as follows.

 The material to be processed is placed in an evacuated ampoule, the Sami ampoule is placed in a three-zone heater so that the end of the ampoule with the material is in the hottest part. The processed material sublimates in the hottest zone and condenses in the zone with an intermediate temperature, and excess intrinsic components and impurities are discharged to the coldest end of the ampoule. The heater is pushed onto the ampoule at a constant speed. In this case, the process of sublimation-condensation is repeatedly cyclically repeated.

However, the disadvantages of this method still include:
- the impossibility of obtaining for a given annealing temperature more than one fixed composition of Pb _1-x Sn _x X solid solutions for systems in which there is a point of least total pressure p _min ;
- the inability to use this method for preparing a charge of materials for which there is no condition p _min .

 Closest to the technical nature of the claimed is the method [2], selected for the prototype.

 According to this method, the material is synthesized from elements in the melt, then normal directed crystallization of the melt is carried out, the final part of the ingot enriched with impurities and excess intrinsic components is removed, crushed and processed by sublimation in a dynamic vacuum. Re-sublimation is carried out in a sealed ampoule, one end of which is placed in the furnace, and the other is brought out and immersed in a Dewar vessel with liquid nitrogen. In this case, the processed material condenses on the walls of the ampoule in the place where the temperature corresponds to the condensation temperature of the material vapor, and the excess intrinsic components and impurities in the cooled end of the ampoule.

 This method of preparation of the mixture is characterized by high productivity, provides deep cleaning from impurities and excess of its own components.

 However, the presence of an inhomogeneous thermal field in the furnace will lead to the fact that in the condensation zone of the processed material there will be some distribution of the molar ratio along the length of the condensed ingot. That is, the resulting material will be heterogeneous in magnitude x, which makes it impossible in the future to obtain reproducible in composition of layers and crystals.

In addition, the method is still unsuitable for the preparation of materials for which there is no condition for a minimum of total pressure p _min , since with this treatment it is possible to form new phases.

 The aim of the present invention is to improve the reproducibility of the composition and electrophysical properties of layers and crystals (or increase the yield of usable layers and crystals) of solid solutions of lead and tin chalcogenides.

 This goal is achieved in that the preparation of the mixture to obtain solid solutions of chalcogenides of lead and tin is carried out by synthesizing the material of the mixture from the elements in the melt, normal directed crystallization of the melt, removing part of the ingot enriched with excess intrinsic components and impurities, grinding and processing by sublimation in a dynamic vacuum, moreover, according to the claims, homogenizing annealing is performed at a constant temperature, then a grain fraction with a diameter of 200-500 microns and so forth is isolated plaguing prior evaporation at a temperature equal to the evaporation temperature during the growth of crystals in the layers and during the time necessary for the output system on stabilizing composition vapor mode.

 All of the above characteristics in the aggregate are necessary and sufficient for the implementation of the proposed method of preparation of the mixture to obtain layers of chalcogenides of lead and tin to achieve improved reproducibility of their composition and electrophysical properties.

 The need to introduce operations of homogenizing annealing at a constant temperature is due to the fact that the ingot obtained after processing by sublimation in a dynamic vacuum has an inhomogeneous distribution of the molar ratio along its length. Homogenizing annealing ensures that the ingot is uniform in size x.

 The necessity of isolating the grain fraction with a diameter of 200-500 microns is due to the following reasons.

As was experimentally shown in [3], the time dependence of the molar ratio x _v in vapors over Pb _1-x Sn _x X solid solutions (where X is chalcogen) has the characteristic form shown in the drawing.

 Section II corresponds to the stabilization of the molar ratio in vapors, namely, layers reproducible in composition can be obtained in that section. This section also corresponds to the constancy of the content of the excess component (metal or chalcogen) in the vapors [3].

 The length of section II is determined by the grain size of the initial charge. For grains having a diameter of less than 200 μm, a stabilization section does not exist, therefore, obtaining layers and crystals reproducible in composition and properties is impossible. The stabilization area appears when the grain diameter of the mixture is about 200 μm and its length increases with increasing grain diameter.

 However, with an increase in grain diameter, the growth rate of layers and crystals decreases due to a decrease in the total surface area with which evaporation occurs. In view of this, an increase in grain diameter above 500 μm is impractical.

The dependences of the molar ratio in pairs on the evaporation time (see drawing) have an initial section. In this section, the composition of the vapor changes over time, therefore, to ensure reproducibility of the composition of the obtained samples, preliminary evaporation is necessary during the time determined by section I [3]. Strictly speaking, the reproducibility of the composition and electrophysical properties of the layers can be ensured by choosing a longer duration of preliminary evaporation. However, in this case, after the expiration of time τ, the unproductive consumption of the charge material will occur. Therefore, a longer preliminary evaporation is impractical. The stabilization level of the composition in section II x _v ^o is determined by the evaporation temperature. If the pre-evaporation temperature differs from the evaporation temperature during the growth of layers and crystals, then due to the difference in x ^o values for these temperatures, the composition of the vapor during the growing process will change. That is, a process similar to section I in the drawing will be observed. Accordingly, reproducibility of the composition and properties of the obtained samples cannot be ensured. Therefore, preliminary evaporation must be carried out at a temperature equal to T _isp , when growing crystals and layers.

Let us prove the significance of the differences. In the claimed combination of features, the isolation of a grain fraction with a diameter of 200-500 μm, preliminary evaporation during the time calculated according to [3], and also annealing at a constant temperature of 800-1000 ^o C are new features, first proposed by the authors. As for the preliminary evaporation at a temperature equal to the evaporation temperature in the process of growing layers and crystals, this feature is, in principle, known, but in combination with the features listed above, leads to completely different positive effects.

 New signs in the presence of the totality of the characteristics indicated in the claims lead to a completely new positive effect: achieving the highest reproducibility of the composition and electrophysical properties of the layers and crystals obtained by vapor-phase methods, compared with other existing methods.

 This allows us to conclude that the proposed method meets the criterion of materiality of differences.

The preparation of the mixture to obtain layers and crystals of compounds A ⁴ B ⁶ or their solid solutions by the proposed method is as follows.

 Lead or tin chalcogenide or solid solutions are synthesized in a melt in evacuated quartz ampoules from elements at a temperature of 1350-1400 K. Then, the melt is crystallized according to the vertical version of the Bridgman method. In this case, the excess of intrinsic components (metals or chalcogen) is forced out by the crystallization front to the upper part of the ingot. Enriched with an excess of its own components, part of the ingot is cut off.

The remaining ingot is ground and placed in a quartz reactor. The reactor is evacuated to a pressure of no higher than 10 ^-3 Pa and installed in the furnace so that the end of the reactor containing the material is in the high-temperature zone of the furnace and the free end leaves the furnace. Processing in a dynamic vacuum is carried out at a furnace temperature selected from the condition T _furnace = 0.75 T _S , where T _s is the solidus temperature for a given composition x of solid solutions Pb _1-x Sn _x X (X is chalcogen).

Upon completion of processing in dynamic vacuum, the material is sealed in a quartz ampoule at a residual pressure of the atmosphere in the ampoule not higher than 10 ^-3 Pa, placed in the isothermal zone of the furnace, and annealed at a temperature of 800-1000 ^o C. Next, the mixture is successively sieved through sieves with a mesh size 500 microns and 200 microns. At the same time, a grain fraction with a diameter of 200-500 microns remains on the second sieve. Next, this fraction is placed directly in the evaporation device and carry out preliminary evaporation according to the claims.

 Example 1

The mixture was prepared to obtain a solid solution of Pb _0.93 Sn _0.07 Se by the present method. The initial batch was synthesized from lead of the C-000 grade, tin of the OSCh grade and selenium of the ChDA grade.

 The composition of the material was set by weighing on an analytical balance with an accuracy of ± 0.1 mg.

The synthesis was carried out in evacuated quartz ampoules with double walls at a temperature of 1400 K for 3 hours. The residual pressure in the ampoules at the time of desoldering was less than 5 • 10 ^-3 Pa. The synthesized material was crystallized according to the vertical version of the Bridgman method with an ampoule pulling speed of 1 mm / h. The upper part of the ingot was cut off. The residue was crushed in an agate mortar and placed in a container for processing in dynamic vacuum. The container was installed in the furnace so that the end containing the processed material was placed in the hot zone of the furnace. The temperature of the hot zone was maintained equal to 810 ^o C. the Processing was carried out for 10 hours.

Next, the processed material was placed in an ampoule, evacuated to a pressure of 10 ^-3 Pa, soldered and installed in the isothermal zone of the furnace. Homogenizing annealing was carried out for 64 hours at a temperature of 810 ^° C. The material thus obtained was sieved successively through sieves with a mesh diameter of 500 μm and 200 μm.

The grain fraction remaining on a sieve with a mesh diameter of 200 μm (i.e., the fraction with a grain diameter of 200-500 μm) was placed in a device for growing layers of A ⁴ B ⁶ compounds by the hot wall method. The design of the device is described in detail in [7]. Preliminary evaporation was carried out for 63 minutes at a temperature of 805 K (the evaporation time was calculated according to [5]). Next, a substrate of monocrystalline barium fluoride chipped along the (111) plane was brought to the evaporating device and the film was grown for 60 min. The temperature of the substrate T _vap = 630 K.

 At the end of the growing process, the following substrate was let down. The growing process was repeated for 10 substrates with identical technological parameters. The total evaporation time (excluding preliminary evaporation) was 10.5 hours. The layer growth rate is 2.3 μm / h.

The composition x _{1 of the} obtained layers was monitored by x-ray spectral microanalysis (PCMA). The composition was calculated from the measured intensities of the analytical lines of the characteristic x-ray emission of lead PbL and SnL by the ZAF method [8]. The relative error in determining the composition is not more than 2 rel.%. In addition, the Hall factor R _H and thermoEMF coefficient α were controlled in the obtained layers.
The measurement errors R _H and α were estimated by carrying out a series of 20 measurements on one sample. The relative errors R _H and α were 5% and 3% (relative), respectively.

The measurement results x ₁ , R _H and α of the obtained layers are summarized in table 1.

 As can be seen from table 1, the parameters characterizing the composition and electrophysical properties of the obtained layers remain constant within the accuracy of the measurements. This result indicates the achievement of the effect of ensuring the reproducibility of the composition and properties of the obtained layers.

 Example 2

Grain fraction smaller than 200 microns, obtained by sieving through a sieve with a mesh size of 200 microns when preparing the charge Pb _0,93 Sn _0,07 Se was loaded into the evaporation device for producing layers by the method of "hot wall" and pre-evaporation was carried out for 63 minutes at T _evap = 805 K. The films were grown in the same way as in Example 1. The growth time of each layer was 60 minutes. T _sub = 630 K, the total evaporation time of the charge in the growth process is 5.2 hours.

5 layers were obtained. The values of x ₁ , R _H and α were controlled analogously to example 1.

 The measurement results of these parameters are summarized in table 2. As can be seen from table 2, the properties of the layers change during evaporation. Thus, when using a grain fraction with a diameter of <200 μm, satisfactory reproducibility of the composition and electrophysical properties of the obtained layers cannot be ensured.

 Example 3

The mixture was prepared to obtain a solid solution of Pb _1-x Sn _x Se by the proposed method. The composition of the initial load corresponded to the formula Pb _0.9 Sn _0.1 Se. The charge preparation operations corresponded to those described in Example 1. After sieving on sieves, the grain fraction was placed in the reactor to obtain layers by the “hot wall” method and preliminary evaporation was performed at a temperature of 843 K for 57 minutes (calculated according to [3]).

Next, layers were grown. The evaporation temperature was 843 K, T _vap = 650 K, the growing time of one layer was 30 minutes. A total of 6 layers were obtained; the total evaporation time was 3.5 hours. The molar ratio x ₁ , the values of R _H and α were controlled. Control methods and measurement errors are similar to example 1. The measurement results are summarized in table 3.

 As can be seen from table 3, within the accuracy of the analysis, the composition and properties of the layers remain constant throughout the process of their growth. Thus, a positive effect is achieved - the reproducibility of the composition and electrophysical properties of the resulting layers.

 Example 4

The mixture obtained in accordance with example 3, was pretreated at a temperature of 843 K for 20 minutes (less time than that calculated by [3]). The layers were grown at T _evap = 843 K, T _vap = 650 K, the growth time of each layer was 30 minutes. It was obtained 5 layers, the total evaporation time of 3 to 5 hours. The measurement results x ₁ , R _H and α are summarized in table 4.

 As can be seen from table 4, the properties of samples N 1 and N 2 significantly differ from the properties of samples NN 3, 4, 5. Moreover, the properties of the latter do not differ within the accuracy of the analysis. It should be noted that the sum of the pre-evaporation time and the growing time of the first layer is less than the estimated pre-evaporation time (57 minutes). According to the drawing, during the preliminary evaporation, the molar ratio in the vapor changes. For this reason, reducing the pre-evaporation time is impractical, since the properties of the first samples of the series obtained then will not be reproducible until the sum of the actual pre-evaporation time and the evaporation time during the cultivation exceeds the calculated pre-evaporation time. As applied to the process of growing single crystals of lead-tin chalcogenides, an insufficient duration of preliminary evaporation will lead to the fact that in the initial section the resulting crystals will be heterogeneous in composition x.

 Example 5

A mixture of Pb _0.93 Sn _0.07 Se was prepared analogously to example 1. The grain fraction remaining on a sieve with a mesh diameter of 500 μm (that is, the fraction with a diameter of over 500 μm) was placed in a device for growing layers of compounds A ⁴ B ⁶ by the "hot" method walls. " Pre-evaporation was carried out at a temperature of 805 K for 63 minutes.

The growth of the films was carried out analogously to example 1. The growing time of each layer was 60 minutes. The temperature of the substrate T _vap = 630 K.

 6 layers were obtained. The layer growth rate was 0.31 μm / h. It should be noted that for practical purposes, layers of a thickness of at least 1.5 microns are required. When using a grain fraction with a diameter> 500 μm, more than 5 hours are required to obtain one layer of the required thickness. In view of this, the use of this fraction of grains is impractical.

List of references
1. Paramonov V.I. The effect of the kinetics of sublimation on the electrophysical and luminescent properties of lead sulfide and selenide. / Abstract. diss. Cand. physical -mat. sciences. - M., 1988.

 2. Preier H. "Recent Advances in Lead-Chalcogenide diode Lasers." - Applied Physics, 1979, v. 20, No. 3, p. 189-206.

3. Makhin A. V. Physicochemical laws of crystallization of the Pb _1-x Sn _x Se and PbTe layers doped with gallium. / Abstract. diss. Cand. physical mat. sciences. -L., 1990.

 4. Prigozhin I.P., Defey R. Chemical thermodynamics. - Novosibirsk, Science, 1966.

 5. Kovalev A. N., Medvedev S. A., Paramonov V. I. Investigation of the effect of repeated freeze-drying on the properties of lead selenide. - Electronic equipment. Ser. 6. Materials. - 1979, N 4, p. 70-74.

 6. Kovalev A.N., Paramonov V.I. The formation of photosensitive layers of lead sulfide. - The surface. Physics, chemistry, mechanics. - 1985, N 4, p. 96-101.

 7. Makhin A.V., Yaskov D.A. Growing layers of solid solutions of lead and tin selenides by the hot wall method. / Izv. LETI .: Sat. scientific tr Leningrad electrical engineer institute - L., 1989, p. 97-100.

 8. Duguzhev Sh. M., Makhin A.V., Moshnikov V.A., Yaskov D.A. Features of x-ray microanalysis of semiconductor solid solutions. / Proceedings of the 2nd Int. conf. on electron beam technologies "ELT-88". - Varna, 1988, p. 829-833.

Claims (1)

 A method of manufacturing a mixture to obtain layers of solid solutions of lead and tin chalcogenides by vacuum evaporation by synthesizing the material of the mixture from elements in the melt, normal directed crystallization of the melt, removing part of the ingot enriched with excess intrinsic components, grinding and processing in a dynamic vacuum, characterized in that after processing in a vacuum, homogenizing annealing is carried out at a constant temperature, then a grain fraction with a diameter of 200-500 microns is isolated and preliminarily produced evaporation at the same temperature as that when growing layers.

Images