FR2802535A1 - Synthesis of indium phosphide for production of e.g. optoelectronic instruments comprises use of direct process from indium and phosphorus in completely closed reaction system - Google Patents

Abstract

Indium phosphide is synthesized from indium and phosphors by a direct process in a completely closed reaction system using a reactor in which at least two containers are used one inside the other. Indium phosphide is synthesized from indium and phosphors by a direct process in a completely closed reaction system sing a reactor in which at least two containers are used one inside the other. The temperature and the pressure is brought to a maximum value of 1070-1250 (preferably 1100-1200) deg C and 1850-2000 bars respectively with a constant temperature increase w.r.t time according to the formula y = kx, where y is the temperature in deg C, x is the time in minutes and k is a constant with a vale of 5-20 deg C/minute.

Description

The present invention relates to a direct synthesis process of indium phosphide.

Indium phosphide is of increasing interest for the manufacture of optoelectronic devices (lasers, photodetectors) or microelectronic components (HEMT, JFET and HBT transistors).

Progress has been made recently: it has been possible to improve the purity of this material in the monocrystalline or polycrystalline state and to reduce the density of dislocations (EPD, pitting density) to a value of less than 104 cm. -2 by controlling the conditions of crystalline growth (growth of phosphide under controlled pressure and reduction of thermal gradients).

At the present time, particular attention is given to InP semi-insulating substrate wafers. It is in the field of optoelectronics that sub strats are predominantly used in, but the semi-insulating material InP is becoming more and more important in electronics, for high-frequency, and which form the backbone of the transfer-service systems found in the field of telecommunications, a market in continuous expansion. Unfortunately, InP techniques are not yet sufficiently developed to warrant large-scale production.

At present, the LEC ("liquid encapsulated Czochralski") drawing technique, which involves iron doping, is used to prepare semi-insulating InP on an industrial scale. The polycrystalline or pre-draw material usually found on the market is an n-type semiconductor, and it is therefore essential to use an acceptor doping agent, namely 99.999% very pure metal iron. The amount of doping agent needed to obtain a semi-insulating material is directly proportional to the amount of residual impurities still present. Since platelets of semi-insulating InP substrate with a high iron concentration (high density of precipitates) have a bad influence on the performance of the devices, it is obligatory to use very pure raw materials. One of the essential conditions to be respected in order to obtain single crystals of very good quality which have suitable electrical properties is to have very pure polycrystalline indium phosphide whose stoichiometry is well adjusted. Special attention must be paid to mastering the purity of the starting materials, the products of synthesis, the stoichiometry of these products and the production costs.

The specific conditions which must be observed in preparing ultra-pure and stoichiometric indium phosphide are the following: the overall proportion of impurities present in the starting materials used, In and P, must be less than 1 ppm; the synthesis operation must be carried out under an inert atmosphere (argon, nitrogen) and in crucibles which are neither reagents nor contaminants; - The reaction system must be closed or pressurized to prevent phosphorus from evaporating.

The synthesis of indium phosphide is an extremely delicate operation, because of the high vapor pressure prevailing above the liquid phosphorus and the highly flammable character of the yellow phos phore formed during the synthesis or present in small quantities. in the red phosphorus which serves as the starting material. The synthetic procedures usually employed to produce polycrystalline indium phosphide are as follows: Bridgman high pressure process (HB); the solute diffusion method (DS); - the phosphorus injection process (IP). According to the high pressure Bridgman method (JA Adamski, "Synthesis of Indium Phosphide" ("Synthesis of indium phosphide"), <I> J. </ I> Crystal Growth <I> 64, </ I> 1 (1983), H. Temkin, "Preparation and Characterization of High Purple Bubble InP" ("Preparation and Characterization of Massive and Very Pure InP"), Cr7-stal Growth <I> 64, </ I-10-14 (1983)), indium phosphide is synthesized in a high-pressure oven to prevent the bulb from exploding. Indium is contained in a graphite tube, closed by plugs of the same material and supported by a quartz tube. On the outside is another quartz tube, where pieces of phosphorus are placed, as well as a disc of quartz wool, which separates the phosphorus from the graphite tube.

The quartz bulb is inserted into a steel casing and pressurized to 20 to 30 atm. The system consists essentially of a three zone furnace. During the operation, the quartz tube is moved at a speed of 6 cm / h, by means of a coil.

The main disadvantage of this process is the presence of impurities from the graphite container. It is possible to avoid contamination of the product by using pBN (pyrolysis boron nitride) crucibles, but this could cause adhesion phenomena, the indium does not react completely and therefore adheres to the walls of the container. . To avoid the use of graphite containers, a system with pressure equalization was subsequently developed. In this case, the reaction is carried out in recipents, nacelles and crucibles, quartz. The polycrystalline in dium phosphide synthesis takes place in a horizontal cooling gradient furnace disposed inside a high pressure autoclave.

The pressure of the phosphorus in the quartz ampoule is balanced by the pressure of the inert gas introduced into the autoclave, so that the pressure difference between the two compartments is close to zero. The system includes a transducer that measures the differential pressure and sends this information to a servomechanism that performs the necessary corrections to the pressure in the autoclave. During the reaction, the pressure inside the system reaches about 30 atm. The technique of synthesis by diffusion of solutes (DS) (E. Kubota and K. Sugii, "Preparation of High purity InP by the Synthesis, Solute Diffusion Technique" ("Preparation of very pure InP by synthesis, technique of diffusion of solutes Physics 52, 2983-2986 (1981)) is one of the processes for obtaining growth crystals from solutions which can be used to prepare polycrystalline indium phosphide. .

Red phosphorus is placed in the bottom of a quartz ampoule, and a crucible containing indium is placed at a certain distance from the bottom.

For a few hours, the indium is distilled under vacuum, in order to drive off the indium oxides therein, in the same synthesis furnace and according to a suitable temperature program, and then evacuate the ampoule (pressure lowered to a value of 1.333.10-4 Pa (10-6 mmHg)) and sealed.

Ingots consisting of aggregates with small grains (from 2 to 10 cnm 3) are obtained, for an average temperature of synthesis of 900 C and thermal gradient within the molten indium of 20 C / cm, and these solid bodies are formed. at a rate of 3 to 4 mm per day.

This DS method is simple and inexpensive, but it takes a very long time and therefore can not be implemented in the industry.

In the phosphorus injection (IP) process, a reactor operating at high pressure is used (JP Farges, "A Method for the" in-situ "Synthesis and Growth of Indium Phosphide in a Czochralski Puller" (" Synthesis Method and In Situ Growth of Indium Phosphide in a Czochralski Drawing Apparatus, <i> J. </ I> Crystal Growth <I> 59, </ I> 665-668 (1982) SB Hyder and CJ Holloway Jr., "In-situ Synthesis and Growth of Indium Phosphide" ("Synthesis and In Situ Growth of Indium Phosphide"), J. Electron. <I> Mater. </ I> 12, 575-585 (1983)). Phosphorus is contained in a bulb, the gap of in dium. To come into contact with the molten indium, the phos phore vapors pass through a molten B203 boron oxide layer. In the chamber where the reaction takes place a pressure of 3 to 6.1 MPa (30 to 60 atm) of argon or nitrogen. Once indium and boron oxide have melted, the crucible is moved vertically until the end of the appendage of the quartz ampoule is immersed in the molten indium. Inside the bulb, the phosphorus is at a temperature of 520 to 570 C.

By doing so, one can obtain 1 to 2 kg of polycrystalline in dium phosphide, where the indium is in slight excess on the external faces of the crystals and in the final solid body. If the bulb is not coated with pBN (pyrolysis boron nitride), the product may be contaminated with silicon.

Initially, when this process was developed, two main objectives were aimed at - producing phosphorus-rich indium phosphide, with the hope of obtaining undoped semi-insulating material; - Thermally synthesize the polycrystalline material and draw a single crystal, in a single operation and in the same reactor, as in the case of gallium arsenide.

But neither of these objectives has been fully achieved, it seems at present that only a few manufacturers, of which only one is present on the market, use this process to prepare a polycrystalline product by synthesis and subsequent growth by limiting the contamination of the product with silicon. This polycrystalline material is then subjected to a pulling operation to make it a monocrystal, according to the usual techniques.

found in the literature a description of a high pressure autoclave non-conventional (HPDS) process at 186.3 MPa (1840 atm) (R.0 Savage, <B> LE. </ B> Anthony, TR AuCoin, Ross RL, W. Harsh and HE Cantwell, "High Pressure Direct Synthesis of Bulk Indium Phosphide", in "Semi-insulating III-V Materials ", edited by DC Look and JS Blakemore, Shiva Publishing Limited, 171- (1984)).

These authors work according to a temperature / pressure cycle which includes operations for evacuating a fraction of the gases, argon and phosphorus, present in the reaction environment; these operations, carried out at temperatures exceeding 750, are essential so that the maximum operating pressure of the equipment is not exceeded. It can therefore be considered that the reaction system is open. In this process, the stoichiometric ratio can not be accurately set, since phosphorus vapor is removed during evacuation operations, and there are more problems with the safety of the installations. In addition, a single unsealed container is used, that is, a lid crucible.

The Applicant has now succeeded in developing an improved process compared to the method currently used in industry, namely the Bridgman high-pressure process. In this new process, the reaction system is completely closed, and there is therefore no evacuation operation, because the maximum operating pressure of the reactor never reached.

The object of the present invention is therefore a process for direct synthesis of indium phosphide from indium and phosphorus. What characterizes this process, called "VHPS" is that the synthesis is carried out in a totally closed reaction system, using a reactor where there are at least two containers, arranged one or more in the other or others, in order to reduce the amount of phosphorus released in the reaction environment. This system is carried at a maximum temperature of 1070 to 1250 ° C., preferably 1100 to 1200 ° C., at a maximum pressure of 185 to 200 MPa (1.85 to 2 kilobars), with the result that the increase in temperature is constant over time, according to the following formula y = kx wherein y represents temperature expressed in C, x represents the time expressed in minutes, and k represents a constant which is from 5 to 20 ° C / min.

From the point of view of production, one of the major problems encountered is the obligation to carry out the operations of synthesis and drawing of single crystals in two different apparatuses. Indeed, with the standard equipment, it is not possible to react directly with gaseous phosphorus and liquid indium in the same reactor where the single crystal is drawn, because at temperatures close to the melting point In dium phosphide, the phosphorus vapor pressure is very strong. Therefore, in the synthesis process of the invention, it is necessary to employ a reactor capable of withstanding very high pressures of inert gas, such as 200 MPa (2 kilobars) of argon or nitrogen.

It is to limit the unavoidable leakage of phosphorus during operations that it is necessary to operate under such high pressure, and therefore in containers of appropriate geometric shape.

In the appended figures, the diagrams of the temperature and pressure conditions observed in the closed system according to the invention are shown (FIG. 1a) and, by way of comparison, in the open system adopted by Savage et al. and described above (Figure 1b).

From these diagrams, it can be seen that in the technique proposed by the Applicant, there is no degassing phase, since the maximum operating pressure of the apparatus has never been reached. On the contrary, the diagram of Figure lb shows that when the temperature exceeds 750 C, it is necessary to lower the internal pressure, and thus evacuate, through a degassing pipe, a mixture of argon and phosphorus vapor. The fact of no longer having to perform such an operation makes it possible to better control the stoichiometric ratio of the reaction and to have fewer problems with regard to the pollution of the environment.

The advantages provided by the process of the invention are indicated below.

In the first place, this process makes it possible to prepare ingots whose dimensions allow them to be introduced as such into crucibles of 12.7 cm (5 inches) which are used to produce indium phosphide monocrystals of 5.1. cm (2 inches). Compared to the Bridgman process implemented in industry, this offers considerable advantages, including the elimination of two stages of handling of the material, hence a lower risk of contamination, improved product stoichiometry, since employing one or more discs of polycrystalline material, of suitable diameter, reduces the phosphorus leakage that occurs during the heating step in the LEC drawing process. In the second place, the process of the invention makes it possible to achieve, for the concentration of charge carriers, values of less than 1 atoms per cubic centimeter. This can be achieved by eliminating all quartz switchgear. In the past, various experimenters have attempted, in the context of the HB process, to replace the quartz of the bulb with other materials, in order to eliminate any contamination with silicon, which makes it obligatory to pre-draw the material. polycrystalline obtained. But they found themselves faced with various problems that the materials they had at the time hardly allowed them to solve. On the contrary since operating according to the method of the invention can directly produce a material which contains very little silicon, this process makes it possible to solve this problem of contamination.

We will now give an example of embodiment of the invention, but it should be understood that this example has no limiting character.

Example The raw materials, namely indium 6N and red phosphorus 6N, are introduced under a well-defined atmosphere (class 100) into quartz crucibles each provided with a lid, previously cleaned with aqua regia for at least one hour. minus two hours and rinsed with ultrapure water (18 MS2). These crucibles are placed in a sup port of graphite crucibles, also closed by a lid, and the whole is put in a reactor.

The system is placed under a pressure of MPa and heated to 1150 ° C (see Figure <B> 1 </ b> a).

The polycrystalline indium phosphide thus obtained is treated first with ethanol to remove any phosphorus residue possibly present on the surface, and then with a 1/1 mixture of hydrochloric acid and nitric acid. to eliminate any excess indium.

Claims (2)

1. A method for the direct synthesis of indium phosphide from indium and phosphorus, characterized in that the synthesis is carried out in a totally closed reaction system, employing a reactor where it has at least two containers arranged one in the other, and one carries system at a maximum temperature of 1070 to 1250 C, under a maximum pressure of 185 to 200 MPa, making sure that the rise temperature is constant over time, according to the following formula y = kx wherein y represents the expressed temperature C, x represents the time expressed in minutes, and k represents a constant which is from 5 to 20 C / min. Process according to the claim, characterized in that the maximum temperature is from <B> 1100 </ B> to 1200