NOTES
Jan., 1942
175
point 146-149’ of the triphenylbismuth dichloride, neglect of atomic polarization. The conclusion for which the literature values range from 136 to to be drawn from these data is, therefore, that the 150°, would not rule out such a possibility for the molecular structures of triphenylantimony dibismuth compound, but the fairly sharp melting chloride and triphenylbismuth dichloride are point of the antimony compound is not indica- slightly unsymmetrical or that the specimens here tive of a mixture. measured were mixtures of a symmetrical with an Another possibility to consider is that the dif- unsymmetrical structure. It is to be regretted ferences of 29 and 28 cm.a between the molar that it has thus far been impossible to obtain a refraction and the total polarization from which sufficient concentration in solution of the ethylthe small moment values have been calculated are bismuth dichloride and dibromide very kindly not dipole polarizations but atomic polarizations. sent to us by Dr. George Calingaert of the Ethyl Antimony pentachloride, which might be expected Gasoline Corporation for measurement. to have this trigonal bipyramidal structure, which has been found also for phosphorus pentachloride, Summary has been reported to have a polarization12of about The dipole moments of triphenylgermanium 54 in the pure liquid and about 49 in carbon tetrabromide, triphenyltin chloride, benzylmercury chloride solution. The molar refraction MRD calchloride, triphenylantimony dichloride, and triculated for it by the method used in obtaining phenylbismuth dichloride have been measured in the refraction in Table I1 is 47 which indicates benzene solution. Minimum values have been an atomic polarization not greater than 8 or much calculated for the metal-to-halogen bond moments less than 3. As the atomic polarization of triin the first three compounds and used to estimate phenylantimony dichloride would, probably, arise the minimum amounts of ionic character in the largely from its two Sb-C1 dipoles, it should not be three bonds. The small moments found for trigreater than that of antimony pentachloride. It, therefore, appears extremely improbable that phenylantimony dichloride and triphenylbismuth triphenylantimony dichloride should have an dichloride axe smaller than would be expected for atomic polarization of 29, a value altogether ab- unsymmetrical arrangements at the apices of a normal in any case. It seems reasonable, there- trigonal bipyramid, but argue against the exfore, to regard these two moment values as real, pected symmetrical, trigonal bipyramidal arrangealthough they may be somewhat high because of ment as the only structure for these molecules. PRINCETON, NEWJERSEY RECEIVED NOVEMBER 3, 1941 (12) Simons and Jessop, THISJOURNAL, 53, 1263 (1931).
NOTES Note on the Specific Gravity of Sodium Dichromate Solutions BY DAVIDF. ALTIMIER
The “International Critical Tables” and similar reference handbooks have based their tables for the density of sodium dichromate solutions on the erroneous work of Stanley.’ The specific gravity of sodium dichromate solutions has been more accurately determined in the low concentration ranges by Jones and Bassett.2 (1) Arthur Stanley, Chemical News and Journal of I n d u s t i d Science, 54, 194-195 (1886). (2) H. C. Jones and H. P. Bassett, Am. Chrm. J . , 84, 290-349 (1906).
The results of recent determinations in this Laboratory are presented in the accompanying table. Experimental.-The customary experimental procedure was followed in determining the specific gravity. The sodium dichromate, granular dihydrate, was prepared by recrystallization of the technical grade and the impurities were less than those specified by the American Chemical Society for c. P. potassium dichromate.8 The solutions were made by dissolving weighed amounts of sodium dichromate in distilled water and then deaerating the solutions. The temperature of the bath was maintained at 15.6 0.1”C. (60 0.2’F.). After weighing, athe di(3) Committee on Andytical Reagents, American Chemical ScEng. Chem., 17, 756 (1926); Ind. Eng. Ckem., Anal. Ed.,
aety, Ind.
1, 171 (1929); 8, 221 (1931); 19, 689 (1940).
176
NOTES
Vol. 64
chromate content of each solution was checked by electrometric titration according to the method described by Kelley.' All weighings were made in air and calculated to weights in vacuo. It will be noted that the specific gravity determinations made with technical NazCrzOr2HzO in tap water' are in good agreement with the data €or c. P. NatCr*0,.2H& and consequently the data may be used for determining the dichromate content of solutions of technical sodium dichromate without introducing any appreciable error.
added to the xylene solution of the two components, do not influence the results of the condensation in the instances investigated so far. (c) Dicyclohexenyl and maleic anhydride are easily condensed in nitrobenzene with simultaneous aromatization (I). However, with cinnamic or crotonic acid the tetrahydro products (11) were obtained. It is evident that nitrobenzene acts only on systems capable of double ortho-enolizaTABLEI tion, as in the adducts with quinones or maleic Nmfixo?, Specific gravity wt. per cent. 15.6°/16.6aC.(8O0/6O0F.) anhydrides. In these cases the di-enolic form is 1.50 1.0105 dehydrogenated and one new double bond es4.50 1.0344 tablished in the newly formed six-membered ring. I . 1076 13.41 The removal of the remaining two "hydroaro22.61 1.1921 matic" hydrogen atoms now depends on the sta31.44 1.2833 bility of the intermediary dihydrobenzenenucleus, 39.48 1.3763 44.10 1.4356 which in most cases undergoes spontaneous de47.38 1.4790 hydrogenation. A t times the intermediary stage 48.10 1.4886 can be isolated, as in the condensation of dicyclo1.5683" 53.07" hexenyl and phenylquinone (III).2 The possi56.28" 1.6200" bility of double enolization is, however, not sufI . 6259" 56.68" 1.6352 57 * 23 ficient for the aromatizing effect of nitrobenzene, 1.6849" 60.25" as anthracene endosuccinic anhydride is not af61.40" 1.7056" fected by this solvent. Here only the two hydro63.92 1.7464 " These values obtained by using technical grade sodium gens in a-position to the carboxyls would be removable for steric reasons and the intermediary dichromate in tap water. IV would not be stable under the experimental (4) Kelley and co-workers, Ind. Eng. Chem., 9,780 (1917). conditions, but would be converted into V.3 (5) These determinations were made by W. H. Hartford of this Laboratory. The only other case recorded in literature where RESEARCH LABORATORY nitrobenzene failed to dehydrogenate the addition MUTUAL CHEWCAL Co. OF AMERICA product, is the work of Weidlich4 on b k ( 1 , l ' RECEIVED SEPTEMBER 19, 1941 MD. BALTIMORE, dialin). This latter reaction is now under investigation. On the Mechanism of "Aromatizing" Diene As a final proof for the above assumed reaction Reactions in Nitrobenzene mechanism, we have now found that meso-diBY F. BERGMANN phenylsuccinic acid (VI), when heated for twelve Diels-Alder reactions in nitrobenzene allow one hours in nitrobenzene, yields diphenylmaleic anto obtain in a single step an aromatic condensa- hydride (VII)in about 40% yield. The di-enolic tion product.1 The mechanism and the general system involved in all these reactions, -C(OH)= I 1 applicability of this modified method of diene C-C=C(OH)-, may be regarded as the vinylog synthesis have now been investigated with the following results. (a) If the isolated tetrahydro product, which is obtained in the usual solvents, is boiled in nitrobenzene for difierent periods, aromatization occurs. Therefore, no active intermediary stage, as has beeo observed in diene reactions, is responsible for the easy dehydrogenation with nitrobenzene. I I1 (R CH8; C&,) 111 (b) Other nitro compounds, e. g., p-chloro-, (2) Ch. Weizmann, E. Bergmann and T . Berlin, THIS JOURNAL, p-bromonitrobenzeneand m-dinitrobenzeoe, when 80, 1351 (1938).
(1) E. Bergmann, L. Hnskelbergand
in press.
F.Bergmann, J . olg. Chem.,
(3) Diels and Friedrichsen, Ann., 613, 145 (1934). (4) Wddlich, Bur., 71, 1203 (1938).
