Nov., 1948
3965
C O M M U N I C A T I O N S T O T H E EDITOR STUDY OF SOAP MICELLES IN NON-AQUEOUS micelle formation. For calcium xylylstearate SOLVENTS USING A FLUORESCENT DYE’ this concentration is 1 X mole/liter and for Sir : calcium xenylstearate it is 8 X mole/liter. A number of dyes may serve as indicators for the presence of soap micelles in aqueous systems.2 We have found rhodamine B peculiarly suited to a similar detection of soap micelles in non-aqueous solvents such as benzene, cyclohexane, cetane, and &-(2-ethyIhexyl)-sebacate. This dye is slightly soluble but almost non-absorbing and nonfluorescent in benzene but fluoresces strongly upon the addition of as little as mole per liter of an oil-dispersible metal soap such as the calcium or 10 20 30 sodium salt of an arylstearic acid, a petroleum Square root of soap concentration X 103. sulfonic or naphthenic acid or of an alkyl sulfonic Fig. 1.-Fluorescence intensity of 2 X lo-’ molar rhoacid. damine B in benzene as a function of soap concentration: That :tdsorption of the dye on the soap micelle 0, calcium xenylstearate; 0 , calcium xylylstearate. is essential to the fluorescenceobserved is confirmed DIVISION LORRAINE ARKIN by this fact; the light emitted when a dilute so- CHEMISTRY NAVALRESEARCH LABORATORY C. R. SINGLETERRY lution containing dye and soap is illuminated with WASHINGTON, D. C. plane polarized green light is 28% polarized, RECEIVED SEPTEMBER 30, 1948 whereas that from a similar concentration of the dye in methyl alcohol (with a comparable visHYDROCARBON OXIDATION INITIATED BY cosity) is only 2.3y0 polarized. This difference ATOMIC HYDROGEN’ is to be expected from Perrin’sa relation for the Sir: depolarization resulting from Brownian rotation Mechanisms for the gas-phase oxidation of during the excited period hydrocarbons have been proposed by many in1lP = 1/Po f U/A - 1/31. RTIVq vestigators. The most widely accepted of these if instead of taking V equal to the hydrodynamic have postulated attack on a primary C-H bond volume of the dye molecule one takes V to be the with resultant degradation to formaldehyde. volume of the micelle to which the former is ad- Many have been based on the observation that sorbed. V is estimated from osmotic pressure aldehydes other than formaldehyde are absent in measurements to be 14,000 cc. per “mole” in this the products. Walsh2has recently reviewed these case.* Taking for 7 the maximum and minimum schemes and concluded that they were all unsatisvalues of the average life of the excited rhodamine factory and has postulated that initial attack of an B molecule reported by Szymanowski6we calcu- oxygen-containing radical should occur at tertiary late polarizations, p , of 20 and 26%, respectively. in preference to secondary and secondary in prefThese observations furnish a n obvious basis for a erence to primary C-H bonds. Although this is to new method of determining the average volume of be expected, experimental evidence is largely misssoap micelles or other suitable colloids. ing, since aldehydes and ketones occurring in If the concentration of dye in benzene is held early stages of the oxidation are conceivably deconstant while that of the soap is varied from graded to formaldehyde. to 0 mole/liter and the fluorescence is measured This communication reports some results of exunder constant illumination it is found that the periments in which oxidation of Cthydrocarbons emission becomes undetectable a t a very low but has been initiated by atomic hydrogen. This finite soap concentration, as appears in Fig. 1. method of oxidation, which has only been briefly The plot of intensity against the square root of previously reported for methane and acetylene by the soap concentration approximates linearity a t Geib and Harteck3supplies evidence regarding the low concentrations and permits extrapolation to (1) The work described in this paper was done in connection with determine the soap concentration at zero inten- Contract NOrd 7020 with the United States Naval Bureau of Ordsity, which we take as the critical concentration for nance, as colirdinated by the Applied Physics Laboratory, The Johns (1) The opinions or assertions contained in this communication are the authors’ and are not to be construed as official or re0ecting the views of the Navy Department. Not copyrighted. 1%)Corrin and Harking, THISJOIJRNAL, 69, 679 (1947). (3) Perrin, J . de Physique, [VI]7 , 390 (1926). (4) Unpublished measurements in this Laboratory. (6) Srymanowski, Z . Phrsik, OS, 460 (19363.
Hopkins University and with Contract N6-06106 with the Bureau of Aeronautics and O5ce of Naval Research, as co6rdinated by Princeton Univerdty. Acknowledgement is due Dean Hugh S. Taylor, who ham general supervision of thia project, and Professor Robert N. Pcene. (2) A. D. Wnlsh, Trass. Faraday Soc., 49, 289 (1946). (3) K. R. Geib and P. Hnrteck, 2 . Physik. Chrm.. A170, 1 (1934).
3966
COMMUNICATIONS TO THE EDITOR
Vol. 70
phthalate (11)by adding a benzene solution of the potassium salt of I to a benzene solution of phthalic anhydride according to the general procedure of Fuller and KenyonSa The liquid I1 was purified by crystallizing its barium salt, m. p. 84-86', from ether-pentane. Anal. Calcd. for Ca~H420~Ba"20: C, 54.13; H, 6.25; Ba, 19.35. Found: C, 54.38; H, 6.54; Ba, 18.94. The brucine salt from 139 g. of 11, m. p. 139.5140.5' (A), [ C Y ] ~ ~-7.8' D [c = 6.3%],4 crystallized in heavy prisms from acetone. After a fourteenstep systematic fractional crystallization, the brucine salt, m. p. 141-142' (B), comprising the head fraction crystallized as ti ht, hemispherical pellets, [a]"D -4.8' [c = 9.S$].4 Anal. Calcd. for C39H480sN~: C, 69.62; H, 7.19; N, 4.16. Found: C, 69.91, 70.32; H, 7.04, TABLEI 7.15; N, 4.59, 4.11. Hydrocarbon oxidized Main carbonyl products On regeneration the brucine salt (B) gave 11, n-Butane Acetaldehyde [aIo6 4-7.3" ~ [c = 8.00/0,4 a = 0.59'1. The curve Isobutane Acetone, formaldehyde obtained by melting mixtures of A and B showed 1-Butene Propionaldehyde, formaldehyde a minimum a t 136' indicating that A is a 1:l 1,3-Butadiene Acrolein, formaldehyde, glyoxal compound of brucine -(+)-I1 and brucine (Each product was definitely identified by precipi- - (-)-Ira tating with 2,4-dinitrophenylhydrazine,separaThe recovery of I was effected by adding an tion by fractional recrystallization and determina- ethereal solution of 9.0 g. of 11, [(Y]~'D $5.5' tion of melting points and mixed melting points [c = 8.5%],4 to six equivalents of ethylmagneswith pure compounds.) The aldehyde or ketone ium bromide. The optically active I (2.0 g.) had first listed for each particular hydrocarbon was the following properties: b. p. 78-80' (52 mm.); present in largest amount; some formaldehyde 152' (760 mm.); n18D 1.4259; [aI2'D -2.6' was also present in the n-butane products. Per- [C = 5.1%,4 a = -0.13'1. oxides and other products were also present. Anel. Calcd. for CSHISO:C, 73.78; H, 13.93. Insufficient evidence is available to determine Found: C, 73.55; H, 14.02. the exact mechanism but the results clearly show During the preliminary distillation of the that oxidatioin initially occurs a t tertiary in pref(-)-I there was formed some olefin, b. p. 112erence to seclondary and secondary in preference to primary C!-H, bonds. Oxidation occurs a t a 115', which showed no optical activity. A sample of 11, 13.7 g., [ a ] " ~-2.7' [c = 5.1%14, from the double bond if present. tail fraction gave (+)-I (2.7 g.), [ a ] 2 8+1.5' ~ FRICKCHEMICAL LABORATORY [c = 5.9%,4 cy = 0.09O1; b. p. 152' (760 mm.); ELMER J. BADIN PRINCETON, NEWJERSEY n2% 1.4234. RECEMCDSEPTE~WER 7, 1948 The infrared spectra of racemic I, partially resolved (+)-I and partially resolved (-)-I were identical in all details between 700 and 4000 wave RESOLUTION OF AN ALIPHATIC TERTIARY numbers. We are indebted to Professor Ralph S. ALCOHOL Halford and to Mr. Abraham Savitzky for these sir: measurements. We wish to report the partial resolution of (3) Fuller and Kenyon, J . Chem. SOC.,126, 2304 (1924). 2,4-dimethyl-4-hexanol (I, methylethylisobutyl(4) All optical rotations were measured in 95% ethanol. carbinol), apparently the first successful resoluOF CHEMISTRY tion of an aliphatic tertiary alcohol in which the DEPARTMENT WILLIAM VON E. DOERXNG UNIVERSITY hydroxyl group is attached to the asymmetric COLUMBIA NEWYORK27, N. Y. H. ZEISS HAROLD carbon at0m.l RECEIVED NOVEMBER 1, 1948 The racemic alcohol I, b. p. 152-153' (760 mm.) ; n% 1.4277; n2% 1.4234, was synthesized according to Clarke2 and converted to its hydrogen PENICILLIN AMIDE
products occurring in initial stages of oxidation of paraffins and olefins. The results seem significant in that they indicate the "initial point of oxygen attack" in the gas-phase oxidation of hydrocarbons. Experiments were carried out by mixing atomrich hydrogen from a Wood's discharge tube with a hydrocarbon mixture. The gases reacted a t 20' and products were condensed a t liquid nitrogen temperature. Results (from experiments with hydrogen, hydrocarbon and oxygen in the ratio of 3 :1: 1; hydrogen about 60% dissociated; total pressure, 0.4 mm. ; total gas velocity, 300400 cm./sec.) in Table I show the main carbonyl products present from oxidation of C4 hydrocarbons. Ratios of aldehyde or ketone to hydrocarbon as high as 0.20 resulted.
(1) Paolini and Divizia, Atti m a d . Linwi, (6) 23, 171 (1914), have claimed the partisl resolution of dl-linalool ( al.7' instead of *20° for the natural product) without furnishing evidence that the o p t l d activity is due to resolved linalool rather than to a resolved impurity. (2) Clarke, T"9 JOWXNAL, 80, 11U (lW8), reported b. p, 1Sl0 (168 rnrn.); a*% 1.4288.
Sir:
In considering methods for obtaining the amide of penicillin, both the symmetrical anhydride' and a mixed penicillin anhydride were considered as (1)
Carpenter. THIOJ o o i a ~?O,~ ,2964 (1948).
