Edited by AND JOURNAL OF PHYSICAL SCIENCE sir Wm. Crookes, F.R.S. Published Weekly. Annual Subscription, free by post, £1. Entered at the New York Post Office as Second Class Mail matter. PAGE 13 14 16 Registered as [PRICE 44 Chemist, highly trained in the Analysis of all classes of Steel, Ferro alloys, Copper and Nickel alloys, Oils, Fats, and general Analytical work, desires position in works.-Address, Chemist," CHEMICAL NEWS Office, 16, Newcastle Street, Farringdon Street, London, E.C. 19 Wanted a young Chemist for Explosives 17 19 21 22 23 ..... State age, wages required, and give particulars of training. Address, E. W., CHEMICAL NEWS Office, 16, Newcastle Street, Farringdon Street, London, E.C. CH HEMICAL LABORATORY TO BE LET in Westminster. Well fitted Top light, Balance Room, &c.23 Apply, Julius Bertram, 14, Suffolk Street, Pall Mall. FOREIGN SOURCES .......... 23 C. PASS AND SON, LIM., BRISTOL, Are Buyers of PARTNER or DIRECTOR required with £3000 working capital to further extend well-established and progressive Business, North of England, Manufacturing Chemical specialities used by Engineers. Audited accounts show over £1000 per annum net; every investigation.- Particulars from Messrs. Arnold and Co., 60, Queen Victoria Street, London. CITY AND COUNTY BOROUGH OF BELFAST. The Library and Technical Instruction Com- Commencing salary £350 per annum, increasing to £400 per annum. Particulars of the Duties and Conditions of Appointment may be obtained from the undersigned, with whom applications, on the special forms provided for the purpose, must be lodged not later than noon on WEDNESDAY, JULY 20th, 1910. Canvassing is forbidden, and will disqualify. MUNICIPAL SCHOOL OF TECHNOLOGY, MANCHESTER. The Committee of the Municipal School of Technology, Manchester, offer FIVE RESEARCH STUDENTSHIPS, each of the value of £40 with fees remitted, tenable for one year in the Faculty of Technology of the Unversity of Manchester. The Studentships are open only to Graduates of Universities, and applications (on forms to be obtained), giving full particulars of qualifications and previous training, should be addressed to the undersigned not later than JULY 14, 1910. J. H. REYNOLDS. GRIFFIN KINGSWAY, LONDON, W.C. KAHLBAUM'S PURE CHEMICALS. Price List & Particulars on application. HAR CHEMICAL NEWS, July 8, 1910 THE CHEMICAL VOL. CII., No. 2641. pound would not be unlikely to appear at some stage during NEWS. the reduction of the pentabromide in hydrogen. At first HALIDES OF TANTALUM.* By WALTER K. VAN HAAGEN. THE chloride alone of all the possible halides of tantalum has received more than passing attention. A bromide has been recorded, but the iodide is absent from our literature. H. Rose (Pogg. Ann., 1856, xcix., 87) claimed to have obtained the bromide by strongly igniting a mixture of tantalic oxide and carbon in a current of carbon dioxide laden with bromine vapour. He failed to give any analytical results. He surmised that his product was the bromide. Moissan believed it was produced upon heating tantalum metal in a stream of bromine vapour. With pure materials the conditions essential for the satisfactory preparation of tantalum pentabromide may be briefly summarised:: 1. Tantalic oxide, previously strongly ignited, should be intimately mixed with an excess of pure carbon, an equal weight of the latter being a convenient and quite sufficient amount. Starch carbon answers well, but it usually leaves an appreciable ash, hence pure sugar carbon is preferable. The excess of carbon indicated makes the mass sufficiently permeable to the bromine. 2. Air should be completely expelled. Raise the charge to a full red heat in a current of carbon dioxide to remove the last traces of moisture. 3. Phosphorus pentoxide is essential as the final drying agent for the carbon dioxide. 4. A high temperature must be maintained during the passage of the bromine, otherwise nearly all of it will escape unchanged. 5. The resulting pentabromide should be fused from time to time in order that the combustion tube may not be obstructed. It is easy to get a yield of about 70 per cent of the theory. The product from the above procedure was re-sublimed in an atmosphere of carbon dioxide. Upon analysis it showed : Theory. 31.39 68.61 Та = 32.21 32'03 31'79 Br 68.70 68.68 68.38 Tantalum pentabromide consists of yellow elongated lamellæ, curving or clinging to the tube in_beautiful arborescent forms resembling frost flowers. They fuse easily to a transparent ruby-coloured liquid. Their colour is suggestive of that of potassium bichromate. They may be sublimed without melting. The vapour of the bromide is yellow in colour, somewhat resembling that of chlorine. The bromide melts at about 240°, and begins to boil at 320°. It fumes strongly in the air. It dissolves rapidly in absolute methyl or ethyl alcohol, forming at first an amber-yellow coloured liquid which soon becomes colourless. Usually the heat generated causes the alcohol to boil. These reactions certainly point to tantalic esters. Anhydrous ethyl bromide is an interesting solvent for tantalum pentabromide. When the latter is brought into this liquid heat is evolved, and a reddish coloured solution results. If the latter be cooled in water or evaporated in a vacuum desiccator golden-yellow coloured crystals separate. The solution fumes strongly in the air. Tantalum pentabromide may be sublimed in at atmosphere of hydrogen. This is possible at a temperature just sufficient for the sublimation. At more elevated temperatures a partial reduction to the metallic state occurs. If a lower bromide of tantalum should exist such a com * An abstract frrom the author's doctorial thesis (1909). From the Journal of the American Chemical Soci ty, xxxii.. No. 6. there seemed no evidence of this. At times, however, the sublimate appeared to be different. Indeed, it had been noticed that toward the posterior end of the tube, i..., beyond the metallic deposit, there was a slight greenish partly almost black film. It dissolved in water with an intense green colour, and in methyl and ethyl alcohol with the same colour. Its analysis indicated a tantalum tribromide, but Chapin, in the John Harrison Laboratory of Chemistry, has since demonstrated that it is not this, but that it is in reality a bromo-tantalum bromide, (Ta6Br12) Br2 (Journ. Am. Chem. Soc., xxxii., 323). An oxybromide of tantalum was not observed. Efforts were made both by Rose (Pogg. Ann., 1856, xcix., 593) and by Moissan (Comptes Rendus, cxxxiv., 211) to obtain an iodide of tantalum, but without avail. So the query presented itself: is it not possible to transpose tantalum bromide by means of a suitable iodide? Silver iodide suggested itself for this purpose. Accordingly, tantalum pentabromide was distilled through a column of granular well-dried silver iodide in a current of carbon dioxide. A brown sublimate resulted. It contained con siderable free iodine, which was carefully expelled, and the residue analysed. This analysis indicated a pentabromide with unmistakable evidences of combined iodine. How could there be free iodine unless there had been a reduction of the bromide ? Finally, it was found that the potassium iodide used in the preparation of the silver iodide contained some iodate. Most likely then the reaction had proceeded as follows:7TaBr5+5AgIO3 = TaI5+3Ta205+5AgBr + 15 Br2. The liberated bromine set free iodine from the silver iodide and probably from the tantalum iodide, allowing only a small amount of the latter to escape, while by far the greater portion of the bromide distilled over unchanged. This view is further supported by the fact that iodine separates when tantalum pentabromide is distilled through a layer of potassium iodate. Ac The next thought was to try hydrogen iodide. cordingly, tantalum pentabromide was slowly distilled in a steady stream of anhydrous hydriodic acid gas. Soon the reddish colour of the bromide changed to brown, while the escape of hydrobromic acid, together with the excess of the hydriodic acid, could be proved at the exit of the tube. Further, an analysis of the dark brown product showed that only one-third of the bromide had been converted into an iodide. Therefore the experiment was repeated, about 3 grms. of the pentabromide being distilled as slowly as possible in a brisk current of hydrogen iodide for about four hours. The product was brownish black in colour, and showed much iodine. Its tantalum content was found to be 22.98 instead of 22:37 per cent as required by TaI5. The remainder of the preparation was subjected to another distillation in hydrogen iodide, and analysed with these results :— ON THE CALCULATION OF OPTICO-CHEMICAL fundamental constants relating to various other groups, nC+2(n+1 − a)H + (n − a− 1)L1+aL2 = =nA-(n-a-1)B -ar, so that knowing the values of A and B, the values of r can at once be obtained from the molecular refractivities of such hydrocarbons. Thus, amylene is C5H10; i.e., Μα 5A-3B-г, Therefore 24.65, My 25'45. ra =5Aa - 3Ba - 24.65-3'00, гy = 5AY - 3By -25°45=2:59. Mean values:-гa = 2'94, гy = 2.62 (see Table V.). = Using these values, and the values of A and B already obtained, the molecular refractivities (for both Ha and Hy lines) of all the ethylene hydrocarbons considered have been calculated, and are exhibited, together with the dif ferences between them and the experimental, and the percentage errors in Table V. It has been found, using the old method of calculation, that in certain compounds (namely, those containing the we are enabled to employ data relating to substituted linking >C - c_c - C<) the influence of the unsaturated Inay links is far greater than expected from theory. The consideration of this point we must postpone for the present; it must be understood that the above constants refer only to the simple influence of the ethene links. With regard to A, the data relative to acetylene hydrocarbons is too scanty to allow of the correct value of this constant being thereby obtained; so that we must postpone the calculation of its value until, with the determination of r'ry-ra = 2·62 - 2'94 = −0·32, and also A' - Ay - Aa, B' By - Ba as before. Moreover, if it be admitted that the influence on molecular dispersion of the hydrogen atom and the Calculation of Optico-chemical Constants. 90.0 + 90.0 + L1.0 01.6t I. 2. 3. 4. 5. 6. 7. 8. 9. 10. II. 12. 13. 14. 15. 16. Octane . C8H18 38.95 38.77 38.96 0.18 39'90 39'73 39.87 -0.03 £6.0 96.0 16.0 Decane C10H22 48.21 C5H10 47.94 49'37 49'27 -0.27 24.65 24.69 24'71 C6H12 29.34 29.26 29.28 C8H16 38.55 38.40 28.12 38.42 9. Decylene C10H 20 47.38 47:55 47.56 +0'17 Valerylene C3H8 24.16 24°32 24'17 C6H8 -0.13 S1.0 81.0+ 10.0+ 91.0+ Ɛ0.0 +0'12 O'147 0'075 II. Diallyl.. 15 C.H link is negligible (see §4), La' must equal 0.32, and the value of L3' can be similarly obtained. In this case, the general formula for open chain ethene hydrocarbons becomes (n - I a)Lı', aL2'. Whether we use this formula, or the non-simplified one, or calculate molecular dispersions as the difference between the calculated - and a-molecular refractivities, the same values will of necessity be obtained. We give both sets of formulæ in Table VI., together with the experimental and calculated molecular dispersions, and the differences between them. The agreement is very satisfactory. §7. Criticism of Brühl's Method of Calculating Opticochemical Constants. We have already indicated one error in Brühl's method of calculating optico-chemical constants; namely, that due to the ignoring of the influences of the ethane and certain other links. This very common error we have dealt with in general terms in the opening chapter of our monograph, "On the Calculation of Thermo-chemical Constants," and There nothing further need be said on this matter. is, however, another fundamental error in Brühl's calculations. He bases his constants on the results obtained for aldehydes and ketones, which he treats as if they were true carbonyl-oxygen compounds-otherwise the method adopted would not hold good. But chemists are inclined nowadays to regard such bodies as mixtures (under ordinary conditions) of the two dynamic isomers the enol and the ketol form-and hence Brühl's calculations are fallacious. Now, the first of these errors is of such a nature that it is not indicated by a comparison of the theoretical and experimental results, since the same errors are (and must of necessity be) made in applying the constants as are committed in obtaining them; this, however, affords no excuse for employing the method in question. The other error, however, should be indicated by a disagreement between the results obtained experimentally for some simple substances (which were not made the basis of the method of calculation in question) and the calculated values. Evidently the hydrocarbons would constitute such a criterion in question. In Table VII. we give a comparison of the experimental molecular refractivities and dispersions of the hydrocarbons considered in this paper with the theoretical values calculated according to Brühl,* and according to the method outlined herein, respectively. In using Brühl's constants we have necessarily assumed with Brühl that the value of the C.H and C.C links is zero, so that the values obtained are not affected by this error of Brühl's. Brühl's values for My have been obtained as the sum of Ma and My- Ma. The average error for each set of results is obtained by dividing the sum of the individual errors (neglecting signs) by the number of individual errors, and gives a relative criterion of the accuracy of the method of calculation. With regard to both a- and y-refractivities the average error in the case of Brühl's theoretical values is almost twice that given by our method. And with regard to molecular dispersions, the average error in the case of Brühl's theoretical values is also greater than that given by our theoretical values. Brühl did not obtain a constant value for the influence of the ethene link on molecular dispersion, so no comparison is given in the case of My and My Ma for ethene hydrocarbons. Table VIII. gives, in a convenient form, the values of the "fundamental optico-chemical constants" obtained in this paper. ERRATUM.-P. 5, col. I, line 21 from bottom, for "C+2H-L1," read "C+2H+L1." The Polytechnic, Regent Street, London, W. * Brühl's constants are as follows:-Refractivity constants, C=2'365, H=1103; ethylene bond, = 1.836; dispersion constants, Co'039, H=0'036 (see Zeit. Phys. Chem., 1891, vii., 191). |