COMMUNICATIONS T O THE EDITOR
5314
Val,
73
tion was accomplished most conveniently by oxidiz- filament lamp, ordinarily used as the comparison ing the crude product with chromium trioxide to radiator for measurements of this type, could not 32-isoallospirostan-3,ll-dione (m.p. 236-237", be employed due to the fact that the flame tem[ c u ] ~ ~-17'; D found: C, 75.16; H, 9.22), followed peratwe is above the melting point of tungsten by Raney nickel hydrogenation (room tempera- (3640"K.) ; the same is true of the ordinary carbon ture) to 22-isoaIlospirostan-:3p-ol-lI-one (V) l 3 (tt1.p. arc (sublimation point of carbon 3810°K. a t atm. 223-226", [ C Y ] ~ ~-29'; D found: C , '73.28; H, 9.93. pressure). Since no simple man-made radiator seemed to be available which had the desired charThe acetate14 showed 1n.p. 322-22'24', [ C U ] ~ O D -31'; found: C, 73.92: H, 9.60). Alternately, themixture acteristics the optical train was modified to use the of I11 and IVa was converted with ethanedithiol sun as the comparison source. A two-mirror helioin the presence of zinc chloride to 22-isoallospiro- stat was used in keeping the sun's image focused on st an-Xp-01- 11,12-dione 12-cycloethylenemercaptol the spectroscope slit. A neutral, variable absorb(ni.p. 289-291'; found: S, 11.98), which upon ing agent of good optical quality was used in Raney nickel desulfurization led directly to the 11- adjusting the brightness of the sun so that line-reketone V. This substance (V) has already been versa1 could be readily obtained and easily reprotransformed14 into allopregnan-3p-ol-11,20-dione duced. An optical pyrometer reading of the sun through the same filter then gave the flame temperand thence3 to cortisone. ature. The maximum flame temperature was (13) The 3.11-dione as well as T proved to he identical with specimens prepared from diosgenin by the performic acid method (cf. '2 measured a t 4300°K. using stoichiometric quantiStork, J. Romo, C. Rosenkranz and C. Djerassi, ibid., 73, 3646 (1951)) ties of hydrogen and fluorine. The agreement be(14) E. M. Chamherlin, \T, V. Ruyle, A. E. Erickson, J. M. Chemtween this value and the theoretical temperature is erda. L. A t , Aliminosa, R . L. Erickson, G. E. Sita, and M . Tishler, admittedly fortuitous, the accuracy of our measureibid., 73, 2396 (1931), reported m.p. 224-229'. [ e ] * ' o -39.4" for a sample prepared from diosgenin. ments being approximately f 150'. RESEARCH LABORATORIES CARLDJERASSI In this particular line-reversal application it is SYNTEX,S. A . HOWARD J. RINGOLD not possible to employ the sodium lines for reversal LAGUNAMAYRAN 413 G . ROSENKRANZ due to the occurrence of sodium in the solar spech.IEXIC0 CITY 17, D. F. trum. Lithium was used in the present work RECEIVED OCTOBER1 , 1951 since, for all practical purposes, it does not appear THE TEMPERATURE OF TRE HYDROGEN-FLUORINE in the solar spectrum. Complete details of our theoretical calculation FLAME and experimental work will be published elsewhere. Sir: INSTITUTE OF R. H. WILSON,J R . The reaction Hz(g) f F&) + 2HF(g) is highly T H E RESEARCH UNIVERSITY J. B. CONWAY exothermic and in addition gives rise to a reaction TEMPLE PENNSYLVANIA A . ENGELBRECHT product possessed of exceptional thermal stability. PHILADELPHIA, A. T'. GROSSE These factors combine to produce a very high RECEIVED OCTOBER1, 19,51 flame temperature which heretofore has not been measured. Theoretical calculations based on the method de- THE SYNT-SIS OF THE BLOOD SERUM VASOCONSTRICTOR PRINCIPLE SEROTONIN CREATININE scribed in Wennerl have produced the following SULFATE flame temperature-composition relationships : Sir: Theoretical flame temperature at 1 atm. Feed composition The existence of a potent vasoconstrictor agent in Mole Vr, E'? ? H: total pressure, OK. the sera of mammals was reported as early as 1868 10.0 90.0 1920 by Ludwig and Schmidt' and since that time has 33.3 66.7 3550 been the . subject of over eighty publications.* 50.0 50. 0 4300 Experiments with sera obtained under various con66.7 33 3 3800 ditions and allowed to stand for variable periods beIn these calculations the latest thermodynamic fore use have produced a maze of conflicting biologdata were employed.2,3,4*5 ical data. This problem has been clarified only reThe construction and operation of the hydrogen- cently by Rapport3>'who isolated from beef serum fluorine torch has already been described by Priest a very active vasoconstrictor substance in crystaland Groese.6 This torch produces a diffusion-type line form. Rapport was able to show that the crysflame due to the present inability to premix hydro- talline material was a complex, or mixed salt, comgen and fluorine. posed of equimolecular proportions of sulfuric acid, The flame temperature was measured by means of creatinine and an indole base.5 On the basis of the line-reversal technique.7 The tungsten ribbon- chemical tests as well as analytical and spectral data (1) R. R. Wenner, "Thermochemical Calculations," McGraw-Hill Rapport has proposed that the indole base is 5-hyBook Co., New York, N. Y., 1941. droxy-3-beta-aminoethylindole(I) (5-hydroxytryp(2) G. M . Murphy and J. E. Vance, J. Chem. Phys., 7 , 8 0 6 (1939). ,
( 3 ) M. G. Evans, E. Warhurst, and E. Whittle, J . Chcm. Soc., 1624 (1950). (4) W. M. Latimer, MDDC-1462, United States Atomic Energy Commission, Declassified Document, November (5) A.P.I. Research Project 44, National Bureau of Standards. (6) H. F. Priest and A. V. Grosse, 2nd. Eng. Chcm., SO, 431 (1947). 17) See B. Lewis and G . v. Elhe's chapter in "Temperature, Its Measurement and Control in Science and Industry," Reinhofd Publ. Corp., New York, N. Y., 1940: and C. W. Jones, B. Lewis, J . B. Friauf. and C. St. J. Perrott, THISJOURNAL, 68, 869 (1931).
13,1947.
(1) C . Ludwig and A. Schmidt, Arb. physiol. Anstall (1868).
Lcipsig, 1
(2) For literature reviews see (a) T. C. Janeway. H. B. Richardson, and E. A. Park, Arch. Inl. Mcd., 21, 565 (1918); (b) G. Reid and M. Bick, Ausfrai. J. E x p . Bioi. Med. Sci., 20, 33 (1942). (3) M. M . Rapport, A. A. Green and I. H. Page, Science, 108, 329 (1948). (4) M . M. Rapport, A. A. Green and I . H . Page, J . Biol. Chem., 176, 1243 (1948). (6) M. M . Rapport. ;bid. 180, Re1 (1949).
COMMUNICATIONS TO THE EDITOR
Nov., 1951
v
2I O 0]
$
V Noturol Serotonin Creotinine Sulfate
.e
.3
E 3
8
1600 1500 1400 1300 1200 1100 1000 900 800 Frequency in ern.-'. absorption spectra of natural and synthetic serotonin creatinine sulfate complex (Nujol mull).
3200 Fig. 1.-Infrared
2800
2400
1800 1700
tamine) to which he has given the name "serotonin."
9.20. Found: C, 67.26; H, 6.34; C1,11.78; N, 9.20.) The 5-benzyloxytryptamine base obtained from the above salt was catalytically debenzylated using HO ,/ --CHt-CH1-NH? 10% palladium-charcoal catalyst. To the alcohol solution obtained on removing the catalyst was added one equivalent of sulfuric acid. ConH centration of the solvent gave the sulfate as a heavy I a m . The cum was dissolved in water a t 60' and We wish to report a t this time confirmation of the i n equivaleLt of creatinine sulfate (neutral salt) proposed structure for serotonin. 5-Hydroxytryp- and several volumes of hot acetone added. On tamine has been synthesized in these Laboratories cooling the solution the complex crystallized as miand converted to the creatinine sulfate salt. Com- croscopic plates indistinguishable from the natural parison of the latter with a sample of the natural product. The crystals melt somewhat higher [214product kindly made avaiiable to us by Dr. Rap- 216' (Kofler)] than the value given by Rapport port has established the two to be identical as shown [212-214' (Kofler)]. Analytical data indicate that by the infrared spectra reproduced in Fig. 1.6Thus the synthetic compound crystallizes as a monohyin serotonin we have what appears to be the first drate as does the natural product. (Anal. Calcd. instance in which the 5-hydroxyindole ring system for C14H23NS08:C, 41.47; H, 5.72; N, 17.28. hag been demonstrated to exist in the mammalian Found: C,41.41; H,5.52; N, 17.45.) Synthetic serotonin creatinine sulfate has been organism. The compound was prepared from 5-benzyloxy- foundgto produce an increase in blood pressure and indole.' This intermediate was converted to 5- a contraction of intestinal smooth muscle in agreebenzyloxyindoleacetonitrile by the general method ment with the findings of Rapport3with the natural of Majima and Kotake.* Lithium aluminum hy- product. These and other biological effects of serodride reduction of the nitrile gave 5-benzyloxytryp- tonin and its analogs are being investigated and detamine which was isolated as the hydrochloride tailed reports will be published a t a later date. LABORATORIES MERRILL E. SPEETER salt; m.p. 248-250' (uncor.). ( A d . Calcd. for RESEARCH UPJOHN COMPANY RICHARD V. HEINZELMANN C17HpClN20: C, 67.42; H, 6.32; C1, 11.71; N, THE KALAMAZOO, MICHIGAN DAVIDI. WRISBLAT
-g\J
(6) We wish t o thank Dr. J. L. Johqsorr of our Physics Department for the infrared data and Mr. W. A. Struck and associates of our Microanalytical Laboratory for the aaalyticsl dats. (7) H. Burton and J. L. Stoves, J . Chem. Soc., 1786 (1942). ( 8 ) R. Majimn w d M. Kotake, Bcr., 68,2042 (1923).
RECEIVED SEPTEMBER 20, 1951 (9) We wish to acknowledge the cooperation of Dr M. J VanderBrook and his staff of our Department of Pharmacology for these observations.
