detonation velocity in equimolar cyanogen-oxygen mixtures with the measurements of Dixon.8 Serious objections can be raised to those calculations and the measurements, but the finding that the dissociation energy is higher than 7.383 ev. is correct. Our measurements, made with piezoelectric gates,9 since improved, included tubes of 1.2 to 10 cm. diameter in 180 cm. lengths. No change of detonation velocity with the distance was apparent within the high precision of these measurements ( * 5 in./sec.), ruling out effects of rarefaction, which tends to lower detonation velocity. Figure 1 shows the effect of tube diameter and makes clear that infinite wave has been closely approximated.
3
2973
COMMUNICATIONS TO THE EDITOR
June, b951
1
28001
1 2670 26901 26501,
, 7 ,
,/
1
0.08 0.10 0.12 0.14 1/D~2. Fig. 2.-Comparison of calculated and measured detonation velocities: 0,calculated assuming D C ~ N=Z 140 kcal.; 0 , calculated assuming D C Z N =~101 kcal.; V, interpolated for D N ~= 8.573 ev.; a), interpolated taking DCZNZ= 114 kcal.'; dotted lines indicate limits of experimental uncertainty.
tion velocities for three total pressures and for a mixture containing additional nitrogen. These data eliminate the two lower values of the dissociation energy. A clear-cut decision between 0 0.5 1.0 1.5 2.0 2.5 9.764 and 11.8 ev. is not yet possible, but Herzberg" has adduced such strong evidence against the higher l/d in inches. value that the 9.764 ev. (225 kcal.) value appears to Fig. 1.-Detonation velocity plotted against the rebe proven. The 170 kcal. for the sublimation enciprocal of pipe diameter: 50% GNz 50% 0 2 at 1 atm. ergy of carbon12 is indirectly supported by these initial pressure. data since the dissociation energy of nitrogen must Calculations of thermodynamic equilibria used the be raised by a commensurate amount, if a lower new heat of combustion of cyanogen'O and Bureau figure is taken for carbon, to bring calculation and of Standards Thermodynamic Tables, extrapolated experiment into agreement again. Further data pertinent to energies of nitrogen by us to 5600-6400°K., the temperatures involved dissociation and sublimation of carbon will be pubin the detonation of equimolar cyanogen-oxygen mixtures. We allowed for equilibria between CO, lished later. (11) G. Herzberg, paper delivered before Am. Phya. Society MeetN2, CN, C, N, 0, since concentrations of NO, CzN2, 02,CZ,C02 were shown by rougher calculations to ing, Pittsburgh, March, 1951. L. Brewer, P. W. Gilles and F. A. Jenkins, J . Chest. P h y s . , 16, be too low to affect the results significantly. The 797(12) (1948). values in Fig. 2 are based on 170 kcal. for the subliG. B. KISTIAKOWSKY CHEMICAL LABORATORY mation heat of carbon and on 101 or alternately on GIBBS HERBERTT. KNIGH r HARVARD UNIVERSITY 140 kcal. for the reaction C2N2 2 CN. Recent CAMBRIDGE, MURRAY E. MALI\ MASSACHUSETTS measurements2 indicate an intermediate value. RECEIVED MAY4, 1951 The velocity for 8.573 ev. has been interpolated on the curves and, the slope of the upper curve is estimated from rougher calculations. Dotted lines RESOLUTION INTO OPTICAL ISOMERS OF SOME give total error limits for the experimental value. AMINO ACIDS BY PAPER CHROMATOGRAPHY Table I compares calculated and measured detona- Sir:
+
OBSERVED
Initial conditions
V meas. meters/sec. V calcd. meters/sec.a
AND
TABLE I DETONATIO?; VELOCITIES
CALCULATED
% % % % CaNz 49.9 CzNz 49.9 CzNa 49.9 CzNz 42.3 Os 49.9 Oa 49.9 01 49.9 Oa 42.3 A 0.2 A 0.2 A 0.2 A 0.2 NI 15.2 P = 0 . 5 atm. P = 1 atm. P 3 1 . 5 a t m . P = 1 atm.
I
2745
2742
*6
2773 2769
*5
2780 2780
*5
2678
*5
2678
Using DN, = 225 kcal.; Xo = 170 kcal. : C ~ N4 Z 2CN 140 kcal.
+
(8) H. E. Dixon, PhiZ. Trans., AlM, 97 (1894). (9) D. J. Berets, E. F. Grecne and 0. B. Kirtiakowsky, Tarn 72, 1080 (1050). (10) Private communication from Dr. E. J. Proscn of the Bureau of Stsndardr. JOURNAL,
We have attempted to resolve several amino acids by means of paper chromatography, using I-methyl-(P-phenylisopropyl)-amine-([.ID - 17.8") as a solvent. The results are shown in Fig. 1. While the d- and 2-forms of leucine gave the same Rf values (a), those of the acidic amino acids, glutamic acid (b), and tyrosine (c) gave differences in Rf values from 0.02 to 0.03. In the case of dand 1-isomers of tyrosine-3-sulfonic acid (d, e), the difference in Rfvalues was 0.21, and the racemic mixture was also completely resolved. We had anticipated that the Rf values of dand ,?-aminoacids should be reversed when the dsolvent was used. However, identical results were observed using d-, 1- and dl-solvents. The tendency to resolve was also observed when in-
297-1
y01, 73
~OhIMUNICATIONSTO THE EDITOR
active solvents such as n-butanol and acetic acid (h) or lutidine mixture (i) were used. Thus it became clear that the resolution was not due to the optical character of the solvent. For confirmation, a chromatopile separation was carried out in which a solution of 250 mg. of dl-tyrosine3-sulfonic acid was dried on 20 sheets of 9 cm No. 2 (Toyo) filter paper rhc 20 disks were incorporated a t 100 disks frorri the top of 21 $00 sheet pile packed iu 15 cni. After 27 hours, the solvent (n-butanol: water : acetic acid :l-inethyl-(P~~hen~lisopropvl)-amine, 15: 5 : 1 : 1) reached the
FORMATION OF FORMALDEHYDE FROM GLYCEROL-C1' BY PROPIONIBACTERIUM'
Sir : It has been previously reported that following the fermentation of glycerol as well as other substrates by P. arabinosum in the presence of C14 formaldehyde, the isolated propionate was completely labeled, the highest activity occiirring in the carboxyl position. Using glycerol l-C14 as a substrate in a similar fermentation it has now been possible to isolate labeled formaldehyde from the mediums3 Bighty micromoles of unlabeled formaldehyde and 2.4 millimoles of w glycerol l-C14 per 100 ml. were fermented with resting cells using phos283 280 phate-bicarbonate buffer. At the 270 269 _. end of 18 hours the formaldehyde was separated by neutral distilla222 tion. Twenty-eight micromoles OD$ 194 were recovered as determined by O@j the chromotropic acid method.4 To this 124 micromoles of carrier formaldehyde were added and the dimedon derivative was made and found to have a constant specific .72,.67 M,M .IO..I2 .78,.75 activity when recrystallized from aqueous acetone and from aqueous .OL .03 .21 alcohol. The specific activity was a 14 *0° 000 1560 cts./min./mM. of carbon or 8,600 cts./min./mM. carbon in the cccc-n m ddl I d dl 1 ad1 I original undiluted formaldehydeob_. tained from the fermentation. The average activity of the 1- and 3Leucine Glutamic Tyrosine -# TYrOSina 3-sultonic ocid ocld carbons of the added glycerol-CI4 22 24 I20 215 48 25 41 24 hr. 26.5 was 13,000. Similar results have ' 6 23-24 17-16 29-25 21-23 19-21 18-19 19-20 11-13 11-12 been obtained in a second experi8elrent ( 1 ) (2) (1) (2) (51 (4) 6) 16) (7) ment. Formaldehyde added to glyFig. 1.-Uni-dimensional resolution of amino acids ; solvent and compound cerol-l-C14and isolated without ferexcursion distances are recorded in mm.; solvent: (1) l-amine:AcOH:HzO:nmentation was found to be inactive. BuOH, 1:l:l:l; (2) l-amine:water, 4 : l ; (3) l-amine:AcOH:HZO: n-BuOH, The propionate formed by resting 1:1:2:6; (4)d-amine: AcOH:H20:n-BuOH, 1:l:l:l; (5) dZ-aminr:AcOH:H20:cells of P.arabilzosum from glycerol n-BuOH, 1:1:1:1; (6) n-RuOH .AcOH:H@. 4:l :I ; (7) lutidiiie .AcOH :HzO:1-C14 also has been dearaded. It n-BuOH, 1:1:2fl was separated by steam udistillation, bottom arid 5 further 24 hours were required for a chromatographed on a silica gel column and degradedta6 after conversion to lactate through total of 733 g. of solvent. After drying at room temperature, the approxi- bromination. The distribution of activity exmate location of the two isomers was determined pressed as cts./rnin./mM. was : CH8-(4760)-CHzwith ninhydrin. The amino acid was found in (4840)-COOH-(8100). The activity of the COz two parts, (I) a t disks 180-253 and (11) a t disks and bicarbonate of the buffer was 1630. This 275-330. distribution is strikingly similar to that found Each group of disks was eluted, and the amino when formaldehyde CI4 was fixed in the propionacids were obtained as mercury salts which were ate.2 decomposed with hydrogen sulfide to give crysFrom the results of the above experiments it talline products. The crystals from (I) tasted seems that formaldehyde is formed largely from bitter and (11) very sweet. The rotation of (I) was [cx]'~D -4.14' (2N NaOH, c = 1.69%): 11, the 1,3-carbons of glycerol. Whether or not the 2 ' 1 ) This work was supported by grant AT-30-1-1050 from the [ ( r ] l 3 4-4.47' ~ (2NNaOH, c = 1.11%). .\tomic Energy Commission under the auspices of the Office of Naval It is most reasonable t o consider that these Research Arknowledgment of many helpful suggestions is due t o resolutions are due a t least in part to the asym- nr. If. G.Wood. (2) F. W.Leaver, THISJOURNAL, 78, 5326 (1950). metric character of the cellulose.
-
-
rvy
I
LABORATORY OF ORGANIC CHEMISTRY INSTITUTE OF POLYTECHNICS OSAKACITYUNIVERSITT MINAMI-OHGIMACHI, KITA-KU OSAKA,JAPAN
RECEIVEDAPRIL 25, 1951
-
(3) The l-C'a glycerol was kindlv furnished by sky of Harvard University.
nr. M. I,.
Karnov.
M. KOTAKE (4) B. Alexander, 0.Landweler and A. M. Seligman, J . Biol. Chem., T. SAKAN 160,51 (1946). N. NAKAMURA (5) H. G. Wood and C. H. Werkman. J . B a d . , SO, 332 (1925); S. SBNOH Riochcm. J . , SO, 618 (1936). Ifi)1,.
F
GQndwin,THISTOIJRNM,, 4%, 89 (1920).
