S. (WITH WHICH IS INCORPORATED THE "CHEMICAL GAZETTE”). Sixty-eight Years. Published Weekly. Annual Subscription free by post £1. Entered at the New York Post Office as Second Class Mail Matter.

Friday, July 14, 1916.

AMATEUR PHOTOGRAPHER LE

hydroxide and the thorium precipitated with recrystallised oxalic acid. The thorium oxalate after complete washing was dried at 110° C. and the sample preserved. Of the oxalate thus obtained 8.5 grms. were digested with 50 cc. of sulphuric acid (made by diluting acid of sp. gr. = 1.84 with an equal volume of water), adding a little nitric acid to oxidise traces of organic matter which discoloured the

THE SEPARATION OF THORIUM FROM IRON liquid, and warming until nitric acid could no longer be

WITH THE AID OF THE AMMONIUM SALT
OF NITROSOPHENYLHYDROXYLAMINE
("CUPFERRON ").

By WILLIAM M. THORNTON, JUN.

THE ammonium salt of nitrosophenylhydroxylamine, which was first introduced into analytical chemistry by O. Baudisch (Chem. Zeit., 1909, xxxiii., 1298; 1911, xxxv., 913; Baudisch and King, Journ. Ind. Eng. Chem., 1911, iii., 629), has been made a subject of study by several chemists (Nissenson, Zeit. Angew. Chem., 1910, xxiii., 969; Chem. Zeit., 1910, xxxiv., 539; Biltz and Hödtke, Zeit. Anorg. Chem., 1910, lxvi., 426; Hanus and Soukup, Ibid., 1910, lxviii., 52; R. Fresenius, Zeit. Anal. Chem., 1911, 1., 35; Bellucci and Grassi, Gazz. Chim. Ital., 1913, xliii., 1, 570; Rodeja, Anal. Fis. Quim., 1914, xii., 305; 1914, xii., 379; Ferrari, Annali Chim. Appl., 1914, ii., 276; 1915, iv., 341; Turner, Am. Journ. Sci., 1916, xli., 339). Because of its selective action as a precipitant many clean-cut separations have been effected, thus solving a variety of analytical problems which without the use of the reagent would involve much difficulty. Bellucci and Grassi (loc. cit.) have shown that in solutions decidedly acid with either sulphuric or hydrochloric acid the sub. stance precipitates quantitatively titanium, and that under like conditions titanium can be completely separated from aluminium in one precipitation. Following the work of Bellucci and Grassi the author (Am. Journ. Sci., 1914, xxxvii., 173; Ibid., 1914, xxxvii., 407) has demonstrated that after throwing down the iron as ferrous sulphide from a solution containing sufficient ammonium tartrate to hold up titanium, and after acidifying the iron-free filtrate, the titanium can be quantitatively precipitated by the "cupferron" reagent notwithstanding the presence of tartaric acid, and further, that if the above-mentioned filtrate be strongly acidified with sulphuric acid and contain also a sufficient quantity of tartaric acid titanium can be quantitatively separated from both aluminium and phos. phoric acid in one operation. Pursuing a similar technique E. M. Hayden, jun., and the author (Am. Journ. Sci., 1914, xxxviii., 137) succeeded in separating zirconium from both iron and aluminium. During the same year Ferrari (loc. cit.) by means of the "cupferron" reagent separated zirconium from aluminlum, but did not consider the more complicated case of iron being present as a third ingredient. Owing to the fact that thorium bears a marked resemblance to zirconium in its chemical relations, the author has seen fit to study the former element with respect to the "cupferron" reagent. The outcome of this investigation has been to establish conditions under which thorium is quantitatively precipitated by the reagent, and also to accomplish the indirect separation of thorium from iron.

A standard solution of thorium sulphate was employed for these experiments. This was prepared by dissolving the Welsbach Light Company's thorium nitrate in boiling water and precipitating the thorium with a boiling solution of sebacic acid according to the method of Smith and James (Journ. Am. Chem. Soc., 1912, xxxiv., 281). The thoroughly washed precipitate was dried and ignited to thorium oxide in a platinum dish. The residue was then subjected to a prolonged digestion with hot sulphuric acid. After cooling, the semi-solid mass was poured into cold water and the solution filtered from an insoluble residue of unattacked thorium oxide. The filtrate was made nearly neutral with redistilled ammonium

detected by its odour. On pouring the residue into cold water the thorium sulphate dissolved completely and the solution was made up to a volume of one litre. Two experiments were made in order to set the standard of this solution. Weighed portions were treated with redistilled ammonium hydroxide at the boiling temperature, the resulting thorium hydroxide ignited to the oxide, and the latter brought to constant weight over the blast lamp. Duplicate determinations gave the following results :

Thorium sulphate solution. (a) 25 cc. 25 740 grms. (b) 25 cc. 25 757

Thorium oxide.

o'0922 grm. 0.3582 per cent
0'0925
0'3592

The mean of these two values was taken as correct. Preliminary experiments soon revealed the fact that even with very small concentrations of free sulphuric acid the precipitation of thorium by the "cupferron" reagent is incomplete.

The author therefore resorted to the ex

pedient of throwing out the thorium from a medium containing acetic as the only free acid. Accordingly weighed portions of 25 cc. of the standard thorium sulphate solution, containing also about 125 cc. of sulphuric acid (1:1), were taken and treated with 15 grms. of ammonium acetate in the form of a strong solution and the volume made up with water to 500 cc. A 5 per cent "cupferron" solution was then added gradually with constant stirring till present in some excess, 15 cc. being the volume actually used. The precipitate, after having been thoroughly coagulated by stirring, was thrown on to a paper filter and washed with a 1 per cent solution of ammonium acetate. The moist paper with its contents was then placed in a tared platinum crucible, dried at 100-110° C., and ignited first with the Bunsen burner and then with the blast lamp to constant weight. In this way the results of Table I. were obtained, which are within the limit of error for ordinary analytical work.

TABLE 1.-The Estimation of Thorium by Means of the "Cupferron" Reagent.

No.

ThO2 found. Grm. 0'0924

Volume of Error. CaH3O2NH, solution. Grm. Grm. Cc. -0.0001 15 500 15 500

ThO2 taken. Grm. I. 0'0925 2. 0'0923 0'0917 -0'0006 The thorium salt of nitrosophenylhydroxylamine[C6H5(NO).N.O] 4Th, assuming a formula analogous to the one proposed by Bellucci and Grassi for the titanic derivative-differs a good deal in properties from the corresponding compound of either titanium or zirconium. In the case of the last two elements a high concentration of free sulphuric acid is consistent with total precipitation, while in the case of thorium extremely small concentrations of the same acid exert a marked solvent effect on the precipitate. Although very similar in appearance to the zirconium precipitate the thorium precipitate is rather different in texture; whereas the former permits filtration by suction the latter passes through the paper in small quantities under the influence of very light pressure. This is unfortunate from a manipulative standpoint, since the precipitate cannot be drained at the pump, necessitating the removal of included water by slow drying.

In the second series of experiments thorium was separated from iron. Known quantities of iron were taken by weighing off portions of pure dry ferrous aminonium sulphate. The solution (about 150 cc.), containing sufficient tartaric acid to hold up the bases in ammoniacal solution, was made slightly alkaline with ammonium

NEWS

hydroxide and colourless ammonium sulphide was added | and thus through a relay operate a mechanism for in moderate excess. After settling, the ferrous sulphide was filtered off and washed ten times with water containing a little colourless ammonium sulphide. Five cc. of sulphuric acid (1 : 1) were then added and the hydrogen sulphide thus liberated removed by boiling. After cooling to room temperature 25 grms. of ammonium acetate were added, the volume made up to 400 or 500 cc., and a 5 per cent "cupferron" solution added in decided excess. From this point on the determination was made just as in the case of thorium alone. Table II. contains the results of three experiments :

A separation of thorium from iron with the aid of the "cupferron" reagent has been satisfactorily worked out and the experimental data show a fair degree of accuracy. Let the reader distinctly understand, however, that the above process is not offered as an analytical method for practical purposes. The well-known oxalate (see E. Benz, Zeit. Angew. Chem., 1902, xv., 297) precipitation is satisfactory and separates thorium from nearly all the common elements with which it is likely to be associated. But another link has been added to the chain of "cupferron" results, and some comparative data on titanium, zirconium, and thorium with respect to this remarkable reagent have been brought to light which it is hoped will prove of interest.

Finally the author wishes to state that the experimental part of the work on thorium was carried out in the laboratory of the College of the City of New York, and to thank Prof. Charles Baskerville for fostering the investigation. Wilmington, Delaware, June 6, 1916.

equalising the temperature, but it would be difficult to provide an adequate relay. A differential mercury-, or, better, a differential gas-thermometer might be used to attain the object sought, and this seemed much more promising. Because the relay seemed to be the crucial part of the contrivance, and because the last-named method requires a less sensitive relay than the others, it was tested first, and is described herewith. With a differential gas thermometer, in which contact is made between a very fine platinum point and a small meniscus, a fairly strong electric current may be sharply made and broken, and this, through a common long-distance tele

A SYNTHERMAL REGULATOR, A DEVICE FOR
AUTOMATICALLY MAINTAINING AN
ADIABATIC CONDITION IN CALORIMETRY.
By THEODORE W. RICHARDS and GEorge d. osGOOD

THE method of adiabatic calorimetry, as recently developed in the Harvard Chemical Laboratories, demands that the bath surrounding the calorimeter should be changed in temperature at the same rate as the calorimeter itself, so that no heat should be lost or gained during the calorimetric determination. The outside bath has therefore been heated or cooled either by a suitable ⚫ chemical reaction, or by hot or cold water, or by electricity (Richards, Fourn. Am. Chem. Soc., 1909, xxxi., 1280; see also Benedict and Higgins, Ibid., 1910, xxxii., 462), so as to keep pace with the inside. Heretofore this quantitative identity of temperature has usually been established from moment to moment by the experimenter, who has observed both temperatures and acted accordingly. That this technique is feasible and accurate has been abundantly proved, but nevertheless with quick reactions the method makes considerable demands upon the operator; accordingly it seemed worth while to arrange an automatic device for relieving him of strain. Such a device might be called a "synthermal regulator."

Obvious methods for accomplishing this end will occur to any one familiar with this kind of problem. A multiple theremocouple or a pair of resistance thermometers might be connected with a delicate galvanometer in such a fashion that any inequality in the temperature of the two baths would cause a deflection in a galvanometer,

NEWS

graph relay, can govern a powerful current capable either of directly heating the outside bath or of operating any desired mechanism for this purpose.

Summary.

In brief, a delicate differential hydrogen thermometer the help of a relay, enabled a stronger current to operate with a sensitive mercury contact was devised. This, with a heating or cooling mechanism for causing one bath to follow the temperature of another within o'03. The apparatus, which may be called a "synthermal regulator," is of service in adiabatic calorimetry, or in other cases where identical but changing temperatures are desired in Chemical Society, xxxvii., No. 7. two contiguous vessels. Journal of the American

The differential thermometer took the form illustrated in Fig. 1. Two rounded copper or silver cylinders, holding each about 28 cc., were attached by de Khotinsky cement or sealing-wax (afterwards covered with paraffin to protect from caustic liquids) to a very narrow glass U tube (15 mm. diameter), with bulbs of familiar form. Suitable stopcocks were provided above, so that the whole contrivance could be filled with hydrogen (chosen because of its great conductivity) and conveniently adjusted. A finely pointed platinum wire, containing 8 per cent of iridium, was sealed into one of the bulbs, so as to make contact with mercury at the widening of the capillary A REVISION OF THE ATOMIC WEIGHT OF U-tube. This differential thermometer is highly sensitive, quickly making and breaking contact within 1/100°. It should be set up with the platinum wire outward, so as to make the current when the inner vessel is the warmer. The electromotive force of the current employed should be moderately low, to diminish sparking, and the hydrogen must be pure to avoid combustion, because the removal of hydrogen in this way may cause changes in the setting. The current thus made and broken is sent through a good relay as already indicated.

With a regulating current thus made and broken it is possible, as stated, to operate a variety of mechanisms for regulating the temperature of the outer bath of the calorimeter. In our first experiments we used the strong current for turning on and off the flow of sulphuric acid into the outer bath, which contained alkali. The current was led through an electro-magnet, which operated a plunger-valve in the delivering vessel. For cooling reactions ice-water could obviously be delivered in a similar way. Unless the size of the jet is rather carefully proportioned to its task this device was found at times to admit too much liquid at once, overstepping the mark. We overcame this difficulty with a constant jet by cutting off the current every few seconds with the help of a constantly rotating circular key.

With apparatus thus constituted it was found that equality in temperature between the inside and outside vessels could be usually maintained within o'02°, or at most o'03°, even when the rise in temperature in the baths was rapid. With slow temperature changes the adjust ment is even better. Thus it is capable of rendering real service in adiabatic calorimetry, and might also be useful in regulating a bath around apparatus for determining freezing points, or in other similar exigencies.

This apparatus was devised in the autumn of 1912. More recently one of us, with the collaboration of Dr. S. Tamaru, has used it in a different way; the stronger current of electricity from the relay was employed directly for heating the outside bath. Although not so suitable for very rapid reactions as the method of adding sulphuric acid to alkali, this method is very convenient for moderately slow ones, and we found that the regulator worked sufficiently well for many purposes when used in this way. In the course of this later work the importance of equality of pressure in the two metal bulbs at the moment of electrical contact became evident. This is easily adjusted by the stopcocks and slight tilting of the apparatus. Otherwise, of course, the quantity Ap/At may be perceptibly different on the two sides, and therefore a progressing deviation of the two temperatures may

occur.

The chief difficulty with very rapid reactions seems to be the lack of equable distribution of the heat, especially in the inner calorimeter vessel. If the different parts of this vessel are unequal in temperature of course the action of the differential thermometer will be irregular. Evidently very efficient stirring is necessary in both the inner and the outer vessel.

We are indebted to the Carnegie Institution of Washington for much of the apparatus used in these experiments.

|

NEODYMIUM.
BY

GREGORY PAUL BAXTER, WILLIAM HENRY WHITCOMB,
OLUS JESSE STEWART, and HAROLD CANNING CHAPIN.

In an earlier investigation on the atomic weight of neody. mium by Baxter and Chapin (Proc. Amer. Acad., 1911, xlvi., 213; Fourn. Am. Chem. Soc., xxxiii., 1; Zeit. Anorg. Chem., lxx., 1) the material examined was purified first by crystallisation of the double ammonium nitrate and then by crystallisation of the nitrate from concentrated nitric acid. The first process was found particularly successful in eliminating samarium, the second in freeing the neodymium from praseodymium. The fractions in the final series were converted to chloride, which after very careful dehydration was analysed by comparison with silver, with the result that the atomic weight of neodymium was found to be 144 27 (Ag=107.880).

During the purification of the neodymium material as described above attempts were made to free the neodymium from its companions by other methods. One of these was the fractional crystallisation of the chloride. This crystallisation proceeds less readily than that of either the double nitrate or nitrate, and after the process had been carried on for ten series of crystallisations little evidence of separation could be observed. At the same time another portion of the original material was converted to nitrate and fractionally crystallised from concentrated nitric acid. As the chief impurity, praseodymium, seemed to accumulate rapidly in the more soluble fractions, the portion which had been crystallised as chloride also was converted to nitrate and fractionally crystallised from nitric acid. Fractional crystallisation of these two portions was carried on separately for some time, for seventy-five series with one portion, for seventy-six with the other; then the two portions were combined and further crystallised in the same way. Both the least soluble and the most soluble fractions frequently were rejected, the least soluble fraction to eliminate samarium, gadolinium, &c., the most soluble to remove praseodymium, cerium, and lanthanum, for Demarçay has shown that the separation occurs in this order (Comptes Rendus, 1896, cxxii., 728; 1900, CXXX., 1021). The least soluble fraction was thus forty times discarded, the most soluble eighty-four times. After 158 series of crystallisations in all had been carried out the purity of the fractions of the last series, seventeen in number, was investigated spectroscopically. As all of the fractions seemed to be fairly pure and the intermediate fractions very pure, it was decided, for the purpose of comparison with the material purified for the earlier investigation, to examine quantitatively the new material. Since at the outset only one and one-half kilogrms. of fairly pure double ammonium nitrate was used the final fractions were not large, and therefore instead of analysing each fraction separately they were combined in pairs, except that the three least soluble fractions-2590, 2591, and 2592-were combined. The combinations actually analysed were:2590+2591+2592, 2593 +2594, 2595+2596, 2597+2598, 2599 +2600, 2601 +2602, 2603+2604.