CONTRIBUTIONS TO OUR KNOWLEDGE

ANTIMONY PERCHLORIDE.*

By RICHARD ANSCHÜTZ and P. NORMAN EVANS.

SOME months ago+ we showed that antimony pentachloride can be distilled, undecomposed, under much diminished pressure; our next step was the attempt to determine the vapour-density under similar conditions. The fact that the boiling point of antimony pentachloride lies much lower than that of the trichloride would seem to show that the vapour-density of the pentachloride, as in the case of the trichloride, corresponds to the simpler formula. Nevertheless, on account of the fundamental importance which the establishment of the simple formula SbC15 would have for the valence of antimony, it seemed indispenably necessary to make a determination of the vapour-density. We will preface our further observations with the remark that we have not yet succeeded in determining the vapour-density of antimony pentachloride under diminished pressure; however, in the course of many unsuccessful attempts which we have made to this end, we had one point thrust on our notice, which on investigation led to some important results concerning our knowledge of the chemical nature of antimony pentachloride.

Antimony pentachloride is well known to be an extremely hygroscopic body. Consequently in the application of La Coste's modification of Victor Meyer's method for the determination of vapour-densities under diminished pressure, it was necessary to substitute liquid paraffin for water; yet, in spite of numerous careful experiments, we could not attain our object, the liquid paraffin offered too great a resistance to the air, and before all the air had been driven out the antimony pentachloride distilled into the upper and cooler parts of the apparatus. During these experiments it was proved to us that it is next to impossible to prevent the formation of traces of the white substance, which is the result of the action of water on antimony pentachloride. We were accordingly constrained to put this method aside.

Before attempting the determination of antimony pentaIchloride by another method, we deemed it best first to find out the nature of the error caused by the formation of the minute amount of the product of water reacting on antimony pentachloride.

After careful consideration we were struck by the contradiction between the conclusions of Daubrawat and those of R. Weber, § concerning the behaviour of antimony

* A Paper read before the Royal Society, June 16, 1887.

+ Chem. Soc. Journ., vol. xlix., 186, p. 708.
Liebig's Annalen, vol. clxxxvi., 1887, p. 118.

§ Poggendorf, Annalen, vol. cxxv., 1865, p. 86.

pure state.

Reaction of Water on Antimony Pentachloride. Daubrawa allowed 1 part by weight of distilled water to drop from a pipette into a flask, surrounded by ice, containing 16 parts by weight of antimony pentachloride. He found the decomposition accompanied by hissing, and by the formation of vapour, giving no white pulverulent precipitate, but forming a yellowish distinctly crystalline mass, which adhered to the flask. This crystalline mass,

the behaviour of which with water Daubrawa describes in detail, is, according to his analysis, antimony oxychloride, SbOCl3, and he expresses its formation from antimony pentachloride and water by the following equation ;

We proceeded exactly as above described, and joined to our flask a receiver containing a weighed quantity of water in which to collect the liberated hydrochloric acid. To 215 grms of well-cooled antimony pentachloride, which had been purified by distillation under diminished pressure, we added drop by drop 12 grm., the calculated amount, of water, and found to our surprise that no hydrochloric acid was liberated; also that the yellowish crystalline product was equal in weight to the sum of the weights of the water and antimony pentachloride originally taken.

The resulting product was only partly soluble in chloroform, and was therefore a mixture of different

substances.

We therefore altered the conditions of the experiment. In order to moderate the reaction we dissolved 2014 grms. of pure antimony pentachloride in about the same volume of chloroform, and added drop by drop to the cooled solution the calculated amount, II grm., of water, shaking constantly. The formation of the new body was soon apparent by the separation of almost colourless crystals; the yellow colour of the chloroform, caused by

the antimony pentachloride (which according to our experience is never colourless, but always a bright yellow liquid), gradually disappeared, and at the end of the reaction the crystals which had separated were covered by colourless liquid. Under these changed conditions there was again no trace of hydrochloric acid set free. We next heated our product nearly to the boiling point of chloroform, adding sufficient dry chloroform to dissolve the crystals which had separated. From this solution a compound was deposited in feathery crystals. After decanting the mother-liquor, these were washed with a small quantity of chloroform and placed to dry in a vacuum desiccator on a porous plate that had been previously heated. This body is so extraordinarily hygroscopic that the analyses were not very easy; however, the results leave no doubt that SbC15H2O is the formula :

tube to 100° on a water-bath. The crystals of the monohydrate which were first formed went into solution, and on opening the tube before a lamp at the end of six hours we found much pressure; a gas smelling like phosgene was liberated. The tube was again closed and heated afresh to 100°. The opening and closing was repeated, first at intervals of six hours, and later, as the pressure diminished, of twelve hours, until after about fourteen days the pressure was no longer noticeable.

On working up the products of the reaction we found, besides antimony tri- and penta-chlorides, phosgene in solution in the chloroform, from which we prepared diphenyl carbamide, m.-p. 239°. For the formation of phosgene gas it is not necessary to work in close tubes. A chloroform solution of the monohydrate, heated to its boiling point on a water-bath, yields a steady stream of phosgene mixed with hydrochloric acid. A similar decomposition takes place when carbon tetrachloride is heated with antimony pentachloride monohydrate in a closed tube to 100°. We will communicate the quantitative relations of this reaction after we have verified our first observations by repeated research.

One can make the tetrahydrate of antimony pentachloride even more easily than the monohydrate. For this purpose we dissolved 29'3 grms. of antimony pentachloride in about twice its volume of chloroform, and after

We provisionally designate this body as antimony pentachloride monohydrate, without expressing any opinion concerning the constitution of this addition product of equal molecules of water and antimony pentachloride. From chloroform, the monohydrate crystallises in leafy or feathery crystals resembling sal-ammoniac, having a melting point lying between 87° and 92°. If exposed to the air it deliquesces to a clear liquid, which over sulphuric acid gradually crystallises again in broad needles, described by Daubrawa as a property of his supposed oxychloride. We have not further examined this body.

When one tries to distil the antimony pentachloride monohydrate under diminished pressure-under 20 m.m. with the bath at 105°-a mobile yellow liquid distils over; this, after two rectifications, boiled constantly at 73° under 17 m.m. (bath 90°). Two chlorine determinations gave, as was expected, results for antimony pentachloride.

I. O'2510 grm. substance gave 0'5965 AgCl.
2. 0*2879

0.6885 AgCl.

....

II. 59'15

58.80 There separated from a fraction boiling somewhat higher crystals of antimony trichloride, leaving a residue of a waxy consistency which could not be distilled.

surrounding the flask containing this solution with ice, let fall drop by drop 7 grms., the calculated quantity, of water. Again, in this experiment, no hydrochloric acid was given off. The mixture remained at first liquid, but when placed in a vacuum desiccator over sulphuric acid and paraffin, there separated slowly a hard crystalline mass, which was quite insoluble in chloroform. Two chlorine determinations made after washing the substance with chloroform agree with the formula SbC15H8O4.

1. 0'3963 grm. substance gave o'7696 AgCl.
2. 0'3414
0'6530 AgCl.

Found.

Calculated for SbCl, H2O4. 47'77

48.04 47'31

CI .. Thus besides the monohydrate of antimony pentachloride a tetrahydrate also exists, the latter being easily prepared from a chloroform solution of the former by the action of water, and differing from the monohydrate among other things in being insoluble in chloroform.

From the foregoing evidence it appears that antimony oxychloride, SbOCl3, is certainly not formed by the action of water on antimony pentachloride; but if equal molecules of these substances are allowed to react, the compound SbC15H2O is formed, which we have called antimony pentachloride monohydrate. This product is most conveniently formed when water acts on chloroform solution of antimony pentachloride, without setting free hydrochloric acid, according to the equation

When one brings together weights equivalent to 1 mol. of antimony pentachloride and 4 mols. of water, the tetrahydrate of antimony pentachloride is formed, as described by R. Weber. We must put aside Daubrawa's contradiction of the facts of R. Weber, and strike Daubrawa's hypothetical antimony oxychloride, SbOCl3, from the list of the known antimony compounds.

(To be continued.)

NICKEL

3 of nickel ores, &c., when it is necessary to estimate that metal only. The solution of the two metals is freed, as far as possible, from acids, glacial phosphoric acid (sodic pyrophosphate may also be employed) is added until the precipitate first formed commences to dissolve, and then an excess of potassic cyanide, which will dissolve the remainder and form a red or yellow solution according to the amount of phosphoric acid added. Heat the solution and keep boiling for a minute or two, adding more cyanide at intervals until a drop of potassic hydrate ceases to give a precipitate. After the solution has become cold render very distinctly alkaline with potassic hydrate, then throw down the nickel by the addition, in considerable excess, of a strong solution of bromine in potassic hydrate, then warm the liquid to facilitate the precipitation. Filter off the black precipitate and wash well out, dissolve it off the filter in warm dilute sulphuric acid, and after saturating with ammonia, deposit the nickel electrolytically from the warm solution. Any cobalt present remains with the iron, but manganese will be found in the nickel solution completely or partly deposited on the opposite pole as few of the results obtained. Beneath are a oxide according to the amount present.

Found. Ni.

o'0995 grm.

0'0985

0'1320

c. Flakes thoroughly freed from adhering regulus. d. Regulus surrounding the imbedded flakes. e. Regulus suddenly cooled to prevent formation of flakes.

Now, the strange thing in connection with this is that an absolutely sulphur-free alloy of nickel and iron should separate out from, and exist still free from sulphur in a metal bath containing at least one fourth of its weight of that element, in spite of powerful affinities, and this singularity is the more striking when it is remembered that one hundredth of a per cent of sulphur will render nickel useless as far as malleability is concerned.

I should just like to add that under favourable conditions the regulus may contain 10 per cent of the alloy. Erdington, June 25, 1887.

DIRECT PRECIPITATION OF NICKEL OXIDE IN PRESENCE OF IRON.

By THOMAS MOORE.

It seems quite useless to attempt to precipitate oxide of nickel free from iron from solutions of the simple salts of these metals, for the simple reason that oxide of nickel is capable of precipitating ferric hydrate from such solutions. If, however, both metals are converted into the double cyanides, the iron into ferro- or ferricyanide of potassium, and the [nickel into the nickelo-cyanide of potassium, then it is possible, by the action of bromine and potassic hydrate, to convert the iron into soluble ferricyanide of potassium, whilst the nickel, being incapable of forming an analogous compound, is precipitated as the insoluble black nickelic hydrate. By the simple addition of potassic cyanide to a ferric solution it is exceedingly difficult to obtain complete solution, owing probably to the formation of different cyanide compounds; if, however, a little glacial phosphoric acid be previously added a clear solution is at once obtained, which, after boiling, is not precipitated by potassic hydrate, and consists chiefly of potassic ferrocyanide, and some other red compound which as yet I have not been able to determine. Based on the above facts, I adopted the following process for the assay

Erdington, June 25, 1887.

ANALYSIS OF VINEGAR.

By B. F. DAVENPORT, late Vinegar Inspector for Boston.

THE following detailed practical method of determining whether a sample of "cider vinegar or apple vinegar conforms to the requirements of the statute relating thereto, which requires that it should be not only the legitimate and exclusive product of pure apple juice or cider, but also that it should not fall below the quality of possessing an acidity equivalent to the presence of not less than 4 per cent by weight of absolute, that is monohydrated, acetic acid, and should yield upon full evaporation at the temperature of boiling water not less than 2 per cent by weight of cider vinegar solids, may prove of interest to those dealing in the article. As the limits set by the statute are in per cents by weight, the portion of vinegar taken for the tests should, for perfect accuracy, be also taken by weight, that is the quantities of 6 and of 10 grammes are to be taken for the tests of strength and of residue, but as taking it by measure, if of about the ordinary atmospheric temperature of 60° to 70° F., will make the apparent percentage at most only 1 to 2 per cent of itself greater than the true, that is will make a true 5 per cent vinegar appear to be, say, from 5'05 to 5'10 per cent, measuring proves in practice to be accurate enough for all common commercial purposes, and therefore the quantities of 6 and of 10 cubic centimetres by measure may be taken in place of as many grammes.

All the measuring apparatus necessary for making the legal tests is one of the measuring tubes, called burettes. It is most convenient to have this of a size to contain 25 to 50 c.c., that is cubic centimetres, and have these divided into tenths. The best form of burette is the Mohr's, which is closed by a glass stop-cock. Besides this only a dropping-tube, called a pipette, graduated to deliver 6 and 10 c.c., will be needed. These tubes are to be obtained of any philosophical or chemical apparatus dealer, being articles generally kept in stock by them for common use, like yard sticks.