NOTES
758
Doull' who measured (e.g., for PbC12) volume increase of anolyte, when a known quantity of electricity was passed through the melt which mas divided into two compartments by a sintered disc, These compartments terminated in capillaries in which mere situated molten lead electrodes which were open to the atmosphere. The apparatus was made horizontal to prevent gravitational flow. Lorenz and Janz2have shown that liquids which do not wet glass (e.g., molten lead), required considerable pressure to cause their displacement in a glass capillary. To eliminate the possible sticking at the interfaces between molten met,al and air, the present authors have modified the apparatus so that only the molten salt in the capillaries is in contact with air. In the present method, there are advantages in using solid metal electrodes, e.g., silver. It was decided, therefore, to investigate molten silver nitrate, this being one of the few suitable silver salts which wets glass. Experimental The vitreous silica apparatus consisted of a short tube (bore 0.6 em.) divided into two compartments by an ultrafine (porosity 4) disc and terminating a t each end in a uniform capillary tube (bore 0.1 em.). Current from an electronically stabilized power supply was introduced by platinum wires sealed through a small hole in the side of each compartment into large beads of pure silver which form the electrodes. The cell was heated in a furnace capable of rotation. It was filled with melted analytical grade AgNO3 in the vertical position and then rotated into the horizontal position, where the boundaries of molten AgN03 in the capillaries were observed using a cathetomet>er. After levelling the apparatus to prevent gravitational movement of the melt, current was passed for measured time intervals and the resulting movement of the boundaries measured. A range of current from 26 ma. (for 15 min.) to 100 ma. (for 3 min.) was used. Silver "trees" were formed at the cathode, cf. Aziz and Wetmore,3 but electrolysis was interrupted before they extended t o the disc. Reversal of polarity caused the trees to break. Temperature was kept constant to 3zO.l' during each run.
Results and Discussion The transport number of the anion is given by 2-
96,500 Vp = _____
qE
where V in cc. is the Golume increase of t,he anolyte ( i e . , the volume increase due to movement of electrolyte boundary plus the volume of silver anodically dissolved); p in g. c ~ . - ~the , density of the molten salt; q in coulomb, the quantity of electricity passed; and E , the equivalent weight of the electrolyte. For 25 runs using different cells and currents from 26 to 100 ma., t = 0.251 f 0.014 (95.5% confidence limits) at 232 f 4". I n contrast with Bloom and Doull's method, ultrafine discs must be used as the prevention of gravitational mass transfer through the disc is difficult with coarser discs even with careful levelling of the furnace. Lorenz and Janz2 likewise found that melts such as AgN03, which wet, glass, can be displaced in glass capillaries by very small pressure. The present value of t is in good agreement with the approximate value predicted by Aziz and Wetmore$ for the same salt and, within experimental error, is identi(1) H. Bloom and N. J. Doull, THISJOURNAL, 60, 620 (1956). (2) M. R. Lorenz and G . J. J a m , ibzd., 61, 1683 (1957). (3) P. M. Aaiz and F. E. W. Wetmore, Canadian J . Chem., 30, 779
(1952).
Vol. 63
cal with the value of 0.25 which Laity and Duke4 obtained by a bubble cell method. I n each case the only moving interface is between air and molten AgN03. Assistance from the University of New Zealand Research Fund for the purchase of apparatus is gratefully acknowledged. (4) R. E. Laity and F. R. Duke, J . Amer. Electrochem. Soc., 106. 97 (1958).
T H E INFRARED SPECTRA OF GASEOUS MAGNESIUM CHLORIDE, MAGNESIUM BROMIDE AND NICKEL CHLORIDE AT ELEVATED TEMPERATURES1 BY STERLING P. RANDALL, FRANK T. GREENEA N D JOHN L. M ARQR AVE Department of Chemistry, University of Wisconsin, Madison 6 , tvisconsin Received October 11, 1.968
The infrared absorption spectra of gaseous MgClz, MgBrzand NiC12 have been observed by means of a Beckman IR 2 infrared spectrophotometer, using a Vycor tube (2.5 cm. diameter, 70 cm. long) fitted with potassium bromide windows as the gas cell. A nichrome wound resistance furnace was used for heating this cell. The temperature was measured and controlled by means of a chrome1 Palumel thermocouple in conjunction with an Amplitrol and potentiometer. The temperatures at which absorption was observed mere: MgC12, 1000°; MgBr2, 100OO; NiC12,850". Electron diffraction measurements of tthe vaporized group I1 halides2 indicate that these molecules are linear. MgClz and MgBr2 should therefore exhibit two infrared active fundamental frequency absorptions3 at frequencies v 3 for the antisymmetric stretching vibration and v2 for the bending vibration. The one absorption maximum observed for each salt undoubtedly is due to the antisymmetric stretching vibration. Since the Beckman I R 2, with potassium bromide optics, has a long wave length limit of about 400 cm.-l, one does not expect to observe the bending absorption which occurs at still longer wave lengths. The structure of NiCl, in the gas state has not been determined, but one may tentatively assign a linear structure to this molecule. If NiC1, possessed a bent symmetrical structure, it wou!d have three infrared active fundamental frequencies. The symmetric and antisymmetric stretching vibrations should have fairly similar magnitudes, while the bending vibration would absorb a t longer wave lengths. Only one absorption maximum for NiClq was observed, and this is ascribed to the antisymmetric stretching vibration of a linear molecule. This assumpt,ion of linearity is tentative since there is a possibility that a bent NiCl2 molecule might have symmetric and antisymmetric stretching frequencies differing considerably in magnitude so that one absorption might fall outside the (1) Paper presented at the fall meeting of the American Chemical Society a t Chicago, Illinois, September 11, 1958. (2) P. A. Akishin and V. P. Spiridonov. Kristallograjiya, 2, 475
(1957).
( 3 ) G. Hersberg, "Infrared and Raman Spectra," D. Van Nostrand Co., Inc., New York, N. y . , 1945.
NOTES
May, 1959
wave length region studied, or the bent molecule might possess frequencies so similar t'hat, especially a t high temperatures, the absorptions might superimpose to give one resulting maximum. The three absorption maxima observed were : MgC12, 588 cm.-'; MgBr2, 490 cm.-l; NiCL, 505 cm.-l. Buchler and Klemperer have recently published work on the infrared spectrum of vaporized MgC12,4and report a value of 597 cm.-l for the antisymmetric stretching frequency. The frequency of the symmetric bond stretching vibration for these molecules is not found in the literature. This precludes the calculation of the interaction force constant. In the few examples for inorganic molecules of this type where the bondbond interaction force constant is k ~ i o w n ~the .~ value is quite small. If one assumes a value of zero for this interaction constant, the stretching force constant, k,, may be obtained from valence force field equation^.^ Also, the infrared inactive frequency due to the symmetric stretching vibration may be calculated. The calculated force constants for the XY2 molecules are shown in Table I. The force constants for the diatomic X Y molecules as obtained from known vl values6 are included for comparison. The calculation of the force constant, k, for the diatomic species is based on the,O-l transition, i.e., the effect of anharmonicity is included in the calculation. The observed frequency for the antisymmetric stretching vibration and the calculated frequency for the symmetric stretching frequency for the three compounds also are shown.
759
reaction mechanism was however not understood. It was not clear whether the reduction of optical activity was due to racemization, or simple destruction of the mandelic acid. This paper reports on the y-radiolysis of aqueous solutions of mandelic acid. Both racemization and destruction were found to occur. These phenomena may be explained by conventional radical mechanisms. Experimental Irradiation Samples.-The samples were prepared by introducing aqueous solutions of optically active mnndelic acid3 into Pyrex sample tubes, which were subsequently frozen, evacuated and sealed. Thcse samples were irradiated at the Gamma Irradiation Facility of the Argonne National L ~ b o r a t o r y . ~ Measurements.-Optical activity measurements were made with the sodium-D line using a Fric polarimeter. The irradiated mandelic, acid was isolated by first treating a portion of the irradiated sample with dilute sodium hydroxide solution and extracting some of the decomposition products four times with diethyl ether. The aqueous phase then was acidified with dilute hydrochloric acid and the acidic ingredients were extracted four times with ether. This latter ethereal phase was evaporated to dryness and the mandelic acid so recovered was recrystallized twice from benzene. Finally, the acid was dissolved in water, and the optical activity was measured.
Results The mandelic acid solutions became turbid upon irradiation and water-insoluble products were observed a t higher irradiation dosages. Invariably, the specific rotations of the solutions were lowered. Benzaldehyde was produced (identified by the mixed melting point method as the 2,4-dinitrophenylhydrazine derivative). Phenylacetic acid, though a possible product, was not positively idenTABLE I tified by either infrared spectrometry or wet analyVIBRATIONAL FREQUENCIES I N C M . - ~ A N D FORCE CONSTANTS sis. I N lo6 DYNES/CM. FOR TRIATOMIC A N D DIATOMIC HALIDES Within the range of the irradiation dosages used, va VI kl k the optical activity decreased steadily with dose (obsd.) (ealcd.) XYz XY (Table I). In addition, measurements of the MgClz 588 297 1.83 1.77 MgBrz NiC12
490 505
178 341
1.48 2.41
1.48 2.26
It is interesting to note, especially for the magnesium compounds, that the force constants for the triatomic and diatomic species are of nearly equal magnitude, on the assumption that the interaction constant in the XY2 molecule is zero. Additional work is in progress in an effort to determine the ex:mt value of this constant. (4) A. Buchler and W. Klemperer, J . CRem. P h y s . , 29, 121 (1958). ( 5 ) W. Klernperer and L. Lindeman, ibid., 25, 397 (195G). (6) G . Herzberg, "Spectra of Diatomic Molecules," D. Van Nostrand Co., Inc., New York, N. Y., 1950, Second Edition.
RADIATION INDUCED RACEMIZATION OF I-MANDELIC ACID I N AQUEOUS SOLUTION' BY PAUL Y . FENG A N D STEPHEN W. TOBEY Phuaica Research Department, A m o u r Research Foundation, Chicago, Illinois Receiued October lY, 1068
Wright2 has found that the optical activity of solid mandelic acid is reduced by pile irradiation, and that benzaldehyde is a reaction product. The (1) This work was supported b y the ARF reactor research program. (2) J. Wright, Disc. Faraday Sac., 12, 64 (1952).
TABLE I RADIATION INDUCED RACEMIZATION OF !-MANDELIC ACID IN
Sample
;.y1nb
lrradiation dosage
rnl.) 1.662 1.662 1.662 1.862 9.85 9.85 9.85 9.85
AQUEOUSSOLUTION
e.v./g.)
Initial optical rotation (degrees)
Final optical rotation (degrees)
Indirect, action C-values for the disappearance of opticully act.ive inandelic acid
5.53 4.95 4.08 3.96 5.8 4.8 3.4 3.2
-4.73 -4.73 -4.73 -4.73 -28.0 -28.0 -28.0 -28.0
-2.77 -8.00 -3.20 -3.21 -25.G -28.5 -27.3 -27.4
4.9 f 0 . 2 4.7f .2 5 , 2 f .2 5.3 f .2 G f l 5 f l 3 f 2 3f2
(1020
mandelic acid recovered from the irradiated samples showed that both destruction and racemization of the mandelic acid molecules occurred. In n sample where the total reduction of optical activity amounted to 30y0,the fraction of optical activity lost by destruction of the mandelic acid molecules was twice as large as the fraction of optical activity lost by racemization. (3) Prepared according to the procedure of R. H. Mrtnske and
T.B. Johnson, J . Am. Chem. Sac., b l , 1909 (1929). (4) The cooperation of Miss Gladys Swope is gratefully acknowledged.
