The study of an entirely new series of alloys may be undertaken from a desire to obtain knowledge applicable to the perfection of industrial alloys, or merely to test certain theoretical considerations. In the case of antimony & tellurium alloys, the theoretical side only is at present important; but it is hoped that the results of this investigation may have some industrial value, at least indirectly, by furnishing one more example for comparison when studying commercial alloys.

The statement is found in various places that antimony is isomorphous with arsenic, bismuth and tellurium; but very little experimental evidence has been offered in support of this assumption. All four metals crystallize in rhombohedra, and the ratio of the axes a: c is approximately the same—which seems to indicate, but does not prove, their isomorphism. The close association of these metals in minerals also seems to indicate that they are isomorphous. There are, however, certain differences in their properties which render this conclusion somewhat doubtful, and it is therefore desirable that independent evidence be found to decide the question. In a paper on lead-tellurium alloys, it has been shown that two metals which are isomorphous give, as a fusibility-curve, a straight line connecting their melting-points. The best examples of this class of alloys are the alloys of gold and silver, gold and platinum, and bismuth and antimony. In each of these cases the properties of the alloys are a mean between the properties of the two metals. The isomorphism of gold and silver and bismuth and antimony might be confirmed by means of the microscope, but, so far as we know, no work has been published on this side of the problem. The microstructure of binary alloys of isomorphous metals should show an intimate mixture of the two metals without any evidence of compounds or of a eutectic mixture. In connection with some work on the isomorphism of selenium and tellurium we had occasion to prepare several bismuth-antimony alloys, in which we found the microstructure to consist of a very homogeneous mixture of the two metals.

It was hoped that by taking advantage of the fusibility-curve which could be established by determining the freezing-points of a series of alloys, and by a study of the microscopical appearance of etched specimens of these alloys, we might be able to form some idea in regard to the isomorphism of tellurium and antimony.

With regard to the more practical side of the problem, it might be said that tellurium and antimony are so similar in their physical properties that the former might replace antimony in some of its alloys. If this should be found possible, it would open up a field of usefulness for the large quantities of tellurium which now go to waste. The effect of even small quantities of antimony on malleable metals, such as copper and gold, is most injurious, making them hard and brittle; but, on the other hand, it gives the necessary hardness to the lead of type- metal, and produces valuable alloys with tin in Britannia metal for decorative objects, and, with tin, zinc and copper, Babbitt metal for bearings. With lead, its 12 per cent, alloy, the eutectic mixture of these two metals, possesses valuable properties, and is used, on account of its power to resist the action of sulphuric acid, for making the so-called “ lead chambers ” for the manufacture of that acid.

Tellurium likewise possesses the power of hardening other metals. This hardening power is very marked in the alloys with lead, and Roberts-Austen has found that it diminishes the tenacity of gold and copper. So far as we know, no ternary alloys containing tellurium, and corresponding to Britannia or Babbitt metals, have been prepared.

Methods and Apparatus

In order to study the properties of the tellurium-antimony alloys, sufficient of each metal was taken to form a button weighing about 20 grammes. In all, fifteen alloys were made, varying from 5 to 95 per cent, of tellurium.

The antimony used was the so-called chemically pure metal, which was found to contain traces of lead and sulphur. Its freezing-point was 624° C., and its specific gravity 6.693.

The tellurium was obtained from the residue of the Baltimore Copper Works, and was purified by the method described (ante, p. 532) in the paper above referred to. Its freezing-point was 446°, and its specific gravity 6.243.

The mixed metals were put in a porcelain crucible, covered with powdered charcoal, fused over the blast-lamp, and allowed to stand until cold. No heat-phenomena were noticed during the melting of the two metals.

From the button thus formed a section was sawed out for microscopic examination, and the larger of the remaining portions was used for a specific-gravity determination. The two larger parts were then united and used for the freezing-point determination.

Freezing-temperatures of the alloys were measured by means of a thermoelectric junction in connection with a galvanometer of the d’Arsonval type, made by Keiser and Schmidt, of Berlin, and so arranged that readings could be taken by means of a needle passing over a graduated scale. This type of instrument was found to be not nearly so satisfactory for experimental work as a delicately adjusted reflecting galvanometer.

There was considerable lag in the instrument, which could only be obviated by constant tapping ; and even by this method we were not sure that our results were correct within five degrees. The pyrometer was calibrated against the following substances:

In order to protect the junction from alloying with the vapors of antimony and tellurium, the two wires were passed through two very small hard glass tubes, and these tubes were then placed in a larger piece of hard glass tubing of about 5 mm. diameter.

Freezing-Points

To determine the freezing-points, the alloys were placed in a Battersea annealing-cup, and were heated to about 100 degrees above their melting-point. The molten mass, protected from oxidation by a layer of finely divided charcoal, was thoroughly stirred, to insure complete admixture of the constituents. The junction was placed vertically in the alloy, and readings of the galvanometer were taken every ten seconds, until the temperature had fallen to 100°. The beginning of the various points of retardation was taken as the freezing-point. The results of these determinations are given in Table I., and are shown graphically in Fig. 1.

It appears from Fig. 1, in which ordinates represent temperatures and abscissae percentage-composition, that the antimony-tellurium alloys belong to the class in which occur one or more definite chemical compounds, and in this respect are similar to the lead-tellurium alloys. The compounds in this class of alloys are always indicated by a maximum point in the fusibility-curve, and by a uniform field when examined under the microscope. In this case it will be seen that the maximum point in the curve corresponds to a freezing-point of 629°, and a composition of 61.37 per cent, of tellurium and 38.63 of antimony, which indicates the compound Sb2Te3. This compound, antimony telluride, forms with tellurium a eutectic alloy containing 87 per cent, of tellurium and melting at 421°, and is isomorphous with antimony, consequently does not form a eutectic with it. This latter fact places the antimony-tellurium alloys in the sub-class of the third general class of alloys in which one of the compounds is isomorphous with one of the elements. The portion of the curve which connects the freezing-points of antimony and antimony telluride is approximately a straight line, and, considered by itself, represents an isomorphous mixture.

In that part of the curve, which is included between 38 and 10.0 per cent, of antimony, we should expect to find, and do find, homogeneous mixed crystals of the pure antimony and antimony telluride. The microscopic field between these two points for all alloys examined is very similar, except where the percentage of antimony is very high, in which case there is some tendency for the telluride to segregate out in masses.

The isomorphism of these two substances seems all the more probable, when we consider the fact that the mineral tetradymite, Bi2Te3, crystallizes in the rhombohedral system. It is more than likely that the antimony telluride, if it occurred as a mineral, would likewise be found to crystallize in that system. Although tellurium crystallizes in the hexagonal system also, its cleavage indicates that it is not isomorphous with antimony, and consequently not with antimony telluride.

The question might arise as to the existence of a compound of the composition corresponding to the formula SbTe. All we can say at present is that there is no evidence in favor of such a compound. There are no a priori reasons why it should not exist, and it is entirely possible that it should exist and form an isomorphous mixture with antimony and antimony telluride Sb2Te3. If it were formed in this mixture, we should expect a more marked evolution of heat in the alloys corresponding to this composition; but the amount of heat evolved in the alloy corresponding to this composition was not above the average for the other alloys.

Oppenheim makes the statement that antimony forms either iron-gray SbTe, or tin-white Sb2Te3. The only evidence, however, in favor of the compound was that the iron-gray mass was homogeneous and had cleavage-planes. The assumption seems to have been based on a very small amount of evidence; for the result of our work shows that with these proportions an alloy is formed which, although brittle, homogeneous and crystalline, is merely an isomorphous mixture of antimony and antimony tritelluride.

From an inspection of the freezing-point curve it seems to be evident that antimony and tellurium are not isomorphous, as has been generally supposed, and that consequently they do not mix at the temperature of fusion as such; On the contrary, tellurium and antimony telluride form a series of alloys which in all respects is similar to the class of alloys of which the lead-tin and silver-copper alloys are good examples. In other words, the freezing-point of either antimony tritelluride or of tellurium is lowered by the presence of the other, no matter which one we consider as the solvent. They are mutually soluble, and mix in all proportions.

Percentage-Composition of the Constituents

To express approximately the composition of any particular alloy, a diagram, Fig. 2, has been constructed, which will show at a glance the percentage of each constituent. The abscissae represent the percentages of antimony and tellurium, and the ordinates the division of the total 100 per cent, of the alloy into percentages of tellurium and eutectic alloy, antimony telluride and eutectic alloy, or isomorphous mixture of antimony and antimony telluride. For instance, an alloy whose chemical composition is 40 per cent. Te and 60 per cent. Sb is made up of a mixture of 85 per cent, antimony and 65 per cent, antimony telluride. Again, if we wish to obtain an alloy containing 84 per cent, of eutectic and 66 per cent, of antimony telluride, it is readily seen that it must consist of 70 per cent, of tellurium and 30 per cent, of antimony.

Specific Gravities

In view of the nature of these alloys, it was thought desirable to study the specific gravities; but from the results obtained no conclusion could be drawn. It seems to be true that nearly all the physical properties of alloys are additive, and give no clue to the nature of the constitution. Various values for the specific gravity of tellurium have been given. Rammelsberg found for the amorphous variety the value 5.93 and for the crystalline 6.38 to 6.42. At 0°, Spring found for the uncompressed, 6.2322, and for the compressed, 6.2549. Later, he reported the value 6.22. Klein and Morel found values varying from 6.204 to 6.215 ; and recently Lenher and Morgan have reported the value 6.1993. Priwoznik obtained the value 6.2549 on a specimen which had been carefully prepared, and which had been fused in a current of hydrogen.

Using a specimen of tellurium which had been prepared from recrystallized basic nitrate, we found the value 6.243. This value represents several actual determinations on specimens known to be pure and free from blow-holes, and is also the average of several other determinations on other specimens. The most accurate determination was made from a button weighing from 30 to 40 grammes, which had been fused several times under charcoal. This was split in half, and the outside surfaces were smoothed on an emery wheel. It was suspended in boiling water for some time; the water was allowed to cool, and this button of 18.64 grammes was weighed in water and in air. As great care had been taken to remove all traces of silica and selenium, these two sub-

stances could not have affected the results. The greatest sources of error are blow-holes, to which the extreme crystalline character of the metal makes it liable, the inclusion of oxide, and the presence of heavier metals. We have reason to believe that by our method of preparation these factors were avoided.

The specific gravity of antimony telluride is given by Bodeker and Giesecke as 6.47 to 6.51 at 13°. For the alloy containing 60 per cent, of tellurium we obtained 6.46. The results for the other alloys were, as a whole, very unsatisfactory, and the values given in the following table are only approximate. Some of the alloys were so crystalline that it was impossible to obtain them free from air-spaces. In certain cases, large drusy cavities were found in the center of an ingot after weighing. This tendency to form inter-crystalline cavities was especially marked in the alloys containing from 0 to 40 per cent, of tellurium. Beyond this point there was a tendency for the eutectic alloy to fill up these spaces, as is shown in Fig 5, where the eutectic has flowed in between the long, parallel crystals of antimony telluride. The crystals in this case are colored dark on account of superficial oxidation, but ordinarily are almost silver-white.

As the percentage of tellurium approaches the amount necessary for the eutectic alloy, the structure is somewhat micaceous, some of the planes of fracture showing a matte surface. This micaceous structure is responsible for a turning over of edge, and is very evident when one attempts to pulverize a piece in a mortar. Most of the other alloys show a remarkable crystalline cleavage, the surfaces of which are very brilliant.

Both antimony and tellurium, and all their alloys, have approximately the same degree of hardness, somewhat above 2.5 (mica) and decidedly lower than 4 (fluorite). From 0 to 20 per cent, of tellurium, the alloys closely resemble antimony.

Microscopical Examination of the Alloys

The preparation of satisfactory samples for microscopic examination proved to be very difficult. On account of the highly crystalline character of some of the specimens, there was a great tendency for small fragments to split out along cleavage-planes, thus leaving a pitted surface. By careful treatment, a fairly satisfactory surface could be obtained by first rubbing on an oil-stone under water, then polishing lightly on a wheel with a mixture of rouge and stearic acid, and finally on a piece of chamois stretched on a wooden block.

The process of etching was likewise difficult. Hydrochloric and nitric acids and iodine were all used without any appreciable effect. The best results were finally obtained by electrolysis in either dilute hydrochloric acid or ammonia. The specimen to be etched was attached to one of the poles of a battery, and a current of about 0.1 ampere was allowed to pass for a few minutes. The structure was brought out by this means; but there was considerable superficial oxidation which produced beautiful colored effects on the surface. This darkening effect is well shown in Fig. 5, where the crystals of Sb2Te3 have branched out through the eutectic alloy. Before etching, the whole surface was brilliantly white; but after etching, the crystals were colored a beautiful blue, which in some cases revealed the structure to the naked eye. The alloys from 0 to 60 per cent, of tellurium were all richly colored, and showed under the microscope a uniform crystalline field with the single exception of the alloy containing 5 per cent, of tellurium which was composed of rounded granules of antimony telluride, imbedded in pure antimony. A cross-section of any of these alloys showed branching crystals extending from the bottom to the top of the ingot.

The alloy containing 61.37 per cent, of tellurium presented an absolutely uniform appearance, being composed entirely of the compound Sb2Te3, and is shown in Fig. 4. The dark spots in the photograph show the cavities produced as a result of the crystallization. With an increase in the percentage of tellurium beyond this point, the eutectic alloy of antimony telluride and tellurium began to show. In alloy No. 11, Fig. 5, this is shown very clearly. In this alloy long crystallites of antimony telluride appear throughout the mass, and the space between the separate crystals is filled with the eutectic. In this case the crystals are dark, on account of the superficial oxidation produced in etching.

Figure 6 shows the appearance of the eutectic alloy, as it was usually found in the alloys containing from 60 to 90 per cent, of tellurium. In slowly cooled alloys the structure was invariably the same, but more rapid cooling produced an entirely different appearance. Figure 7 represents a photograph of the pure eutectic, prepared from several other alloys by pouring off the still liquid mass after a part had solidified. The photograph represents a surface which had been cast on glazed porcelain, and was taken just as it appeared after cooling, without any further treatment. The light background is tellurium, and the dark portions are antimony telluride. The latter was probably brought out in such great contrast on account of its extreme tendency to oxidize.

Figure 8 shows the usual appearance of the alloys consisting of the eutectic and an excess of tellurium.
It is proposed to study in the same manner other binary alloys of tellurium, and subsequently some of the ternary alloys.