US3622405A

US3622405A - Method for reducing compositional gradients in{11 {11 {11 {11 {11 {11 {11 {11 {11 {11

Info

Publication number: US3622405A
Application number: US48361A
Authority: US
Inventors: Joseph L Schmit
Original assignee: Honeywell Inc
Current assignee: Honeywell Inc
Priority date: 1970-06-22
Filing date: 1970-06-22
Publication date: 1971-11-23
Anticipated expiration: 1988-11-23

Abstract

Compositional gradients in a body of Hg1 xCdxTe can be removed by annealing the body at a temperature which is greater than the solidus temperature and less than the liquidus temperature for the average composition of the body.

Description

United States Patent [72] inventor [21 Appl. No. [22] Filed [45] Patented [73 Assignee 54 METHOD FOR REDUCING COMPOSITIONAL GRADIENTS IN Hg1 Cd Te [50] Field ofSearch 148/13, 13.1, 20.3, 1.6, 3

, [56] References Cited UNITED STATES PATENTS 3,468,727 9/1969 Gross 148/13 3,537,912 11/1970 Aven et a1 148/1.6

Primary Examiner-Richard 0. Dean Attorneys-Lamont B. Koontz and Omund R. Dahle 2 Claims, 1 Drawing Fig.

8 ABSTRACT: Compositional gradients in a body of Hg; Cd, [52] U. Cl 148/13, Te can be removed by annealing the body at a temperature 7 148/ l which is greater than the solidus temperature and less than the [51 1 Int. Cl C22 U16 liquidus temperature for the average composition f the body TL}. L3 84 si S l T 4 53 TEMPE RATURE 0 X 5 X| X4 X 2 MOLE RATIO X, OF CdTo PATENTED 23 3.622.405

TE M PE RATURE MOLE RATIO, X, OF CdTe IN VliN'l ()R,

JOSEPH L. SCHMIT ATTORNEY.

METHOD ron REDUCING COMPOSITIONAL GRLIENTS IN H .-.cd.;re

BACKGROUND OF THE INVENTION The development of solid-state detectors of wave lengths within the infrared portion of the electromagnetic spectrum has led to the use of semiconductor alloys having the proper energy gap for intrinsic photoconductivity at wave lengths within the range of 1.7 to 30 microns. One successful intrinsic detector material that has been developed for photoconductive detectors is mercury cadmium telluride, a semiconductor material which is an alloy of a semimetal, mercury telluride, and a semiconductor, cadmium telluride. The mole ratio, x, of cadmium telluride in the alloy determines the energy gap and therefore the optical and semiconducting properties of the alloy. Since the energy gap of mercury cadmium telluride is determined by the composition of the alloy, it is highly desirable to produce an ingot of mercury telluride having a uniform composition.

It has been extremely difficult to produce an ingot of Hg Cd Te havinga constantcomposition. This difficulty arises from the tendency of the compounds to segregate when cooling, giving different compositional regions within the ingot that has been formed. The reason for this tendency to segregate can be explained graphically by a phase diagram, FIG. I, which has as its abscissa the composition, or mole ratio, of the material and as its ordinate the temperature.

The phase diagram forHgi-rcdzTe shows three regions, or phases: The liquid phase, which is the region above the liquidus line 10, the solid phase, which is the region below the solidus line 11, and the liquid plus solid, or slush phase, the lens-shaped region between the liquidus and solidus lines. When a certain mole ratio, X of CdTe within the liquid phase is cooled to a temperature T located on the liquidus line, material having a mole ratio x is precipitated. This is caused by the fact that the liquidus and solidus lines are separate, and for a given temperature T there is a point Ll on the liquidus line which corresponds to a mole ratio x, and a point S2 on the solidus line which corresponds to a mole ratio x Therefore, the first precipitated lig C d Te has a mole ragg x As the CdTe rich material is precipitated out, the amount CdTe remaining in the liquid is depleted so as to relocate the composition of the liquid to a point at the left of point Ll. Upon further cooling, the liquid having composition x reaches point L3 and a solid is precipitated out of composition x For this reason an ingot of ;ilg, ,CdITe exhibits compositional gradients over its length.

Growth of mercury cadmium telluride can be accomplished by the use of the modified Bridgman method described by E. L. Stelzer et al. in the IEEE Transactions on Electron Devices, pages 880-884, Oct. 1969. In this method stoichiometric amounts of the three elements, mercury, cadmium, and tellurium, plus some excess mercury are loaded into a thick wall quartz capsule, which is then evacuated and sealed off. The sealed capsule is heated in a furnace and rocked back and forth to insure mixing, after which it is solidified from one end by sequentially cooling the three zones of the furnace. Rapid solidification reduces thermal segregation, resulting in a largely single crystal ingot containing a dendritic structure, with altemating regions of high and low mole ratios. As the cooling rate is increased, the dendritic structure becomes finer. A subsequent high-temperature anneal, at a temperature below the solidus line is then used to remove the dendrites and a low temperature anneal is used to adjust stoichiometry. While it is possible to remove the dendritic microscopic compositional gradients, which are generally less than 1 mm., the high-temperature annealing step takes weeks and leaves macroscopic x gradients on the order of several millimeters unaffected. Removal of macroscopic .r gradients by this method takes months or even yours and is therefore impractical.

A process known as zone leveling can be used to remove macroscopic compositional gradients. Zone leveling entails heating a small region of the ingot to above the melting temperature, and then advancing the molten region, or zone through the length of the ingot at a rate which is slow enough for thermal equilibrium to be approximately maintained. As the zone advances, the material left behind solidifies and new material is melted. The first material to resolidify has a greater concentration of CdTe than the material which has yet to be melted. As the zone progresses, the concentration of the liquid changes until it attains a concentration which produces, upon solidification, material which has a concentration which is equal to the concentration of the region about to be melted. When this condition is attained, the concentrations of the solids entering and leaving the zone are equal, and hence no further change of concentration occurs in the zone or in the solid freezing from it until the zone reaches the end of the ingot.

The disadvantage of the zone-leveling process is that the movement of the zone is very slow, since the ingot must be in near thermal equilibrium. An additional disadvantage is that it is difficult to control the liquid-solid interface and therefore compositional gradients may still occur.

SUMMARY OF THE INVENTION In the present invention, compositional gradients in an ingot o f Hg ,Cd Te can be removed by annealing at a tempeg ture which is greater than the solidus temperature and less than the liquidus temperature for the average composition of the ingot. Due to the greatly reduced annealing time with this m ho tafieriC r oi nifp os on an A s produced by forming ingots quickly, removing the compositional gradients by annealing the body at a temperature above the solidus temperature and below the liquidus temperature for the average composition of the body, further annealing the body at a temperature near but below the solidus temperature, and adjusting stoichiometry with a low-temperature anneal.

BRIEF DESCRIPTION OF THE DRAWING FIG. 1 is a (temperature, mole ratio) phase diagram for Hit-4 5 i- BRIEF DESCRIPTION OF THE PREFERRED EMBODIMENT In the present invention, an ingot HgI-ICdITe having dendrites and compositional gradients is heated to a tempera ture above the solidus line and yet below the liquidus line for the average composition x, of the ingot. This shifts the composition of the precipitated material within the dendritic regions closer and closer to the average composition x,. This process occurs at a much faster rate than in a similar process employing heating below the solidus line because the material has a greater diffusibility when heated above the solidus line. The greater diffusibility is due both to the higher temperature of the annealing and also to the fact that for a given temperature T between the solidus and liquidus temperatures T and T of average composition X that material having a smaller concentration of CdTe than x is liquid, thus allowing it to flow and mix. Due to this higher diffusibility and mixing, it is possible to remove macroscopic compositional gradients as well as microscopic ones. By this method, one can obtain an ingot of Hg1-ICdITe which is closer to the desired uniformity over greater distances in a much shorter time than has been possible by prior art techniques.

It can be seen that the higher the temperature is within the specified range, the greater the difiusibility of the material is, and therefore annealing time is at a minimum when the temperature is just below the liquidus temperature.

In one preparation, an ingot having a composition ranging from x=0.l30 to x=0.3l8 over l3 mm. was annealed for 10 days at a temperature T=75S C., which is greater than the solidus temperature and less than the liquidus temperature for the average composition, M270. The resultant ingot had a composition x=0.270i0.02 over the same distance.

3 4 A second ingot annealed at 750 C. for 8 days was found to property or right is claimed are defined as follows: have a composition x=0.5 61:0.0l4 across the diameter of 13 1. A method for reducing compositional gradients in a body mm. of Hg, ,Cd Te wherein the body is annealed at a tempera- Due to the greatly reduced annealing time obtained by this ture which is greater than the solidus temperature and less method Hg, ;Cd.rTe of uniform composition which i uit than the liquidus temperature for the average composition of ble for detectors can be produced by forming ingots quickly, Said yremoving macroscopic compositional gradients by annealing The method accordancc wlth clalm 1 and further the body at a temperature which is greater than the solidus Pflsmgi temperature and less than the liquidus temperature of the annefilmg the y at a temperature near h average composition of the ingot, removing microscopic com- Sohdus temperature for the average Composltlo" of said positional gradients with further annealing at a temperature lf and near but below the solidus temperature, and adjusting 'f g the body at a lower temperature to adlust stoichiometry with a low-temperature anneal. stolchlometry' The embodiments of the invention in which an exclusive l

Claims

2. The method in accordance with claim 1 and further comprising: annealing the body at a temperature near but below the solidus temperature for the average composition of said body, and annealing the body at a lower temperature to adjust stoichiometry.