3706
COMMUNICATIONS TO THE EDITOR
VOl. 74
Of the radicals HO, HOz and HOa, rapid exchange seems possible only for HO. More complete experimental results will be presented in a later report, containing also data on related syr;tems.
be introduced into the thio-S of the thiosulfate by side-reactions during its formation. If tetrathionate was prepared from thiosulfate labeled a t the thio-S, 2-3% of the total activity was always found in the solid sulfur; if inactive tetrathionate was reduced with active sulfide, 97% JONES HERBERT JONESLABORATORY UNIVERSITY OF CHICAGO OTTO L.FORCHHEIMERwas found in the sulfur. CHICAGO 37, ILLINOIS HENRYTAUBE The results indicate that, save for a minor sideRECEIVED JUXE 7, 1952 reaction, the reduction of tetrathionate with sulfide proceeds according to TRACER STUDIES ON SOME REACTIONS OF THIOSULFATE AND TETRATHIONATE Sir :
The oxidation of thiosulfate to tetrathionate with iodine, the reduction of tetrathionate to thiosulfate with sulfide, and the decomposition of both thiosulfate and tetrathionate with mercuric chloride in presence of bisulfite and excess formaldehyde have been studied with the aid of S%. In order to separate the products of the reactions with mercuric chloride, three successive precipitations were made: HgC12.2HgS (from thiosulfate) and HgC12.2HgS S (from tetrathionate) were precipitated in the cold by means of a large excess of concentrated buffered mercuric chloride and addition of some ammonia after one hour (I), sulfate was precipitated, also in the cold, with acetic acid and barium chloride (11), finally the protected bisulfite was exidized with potassium hypobromite and precipitated as barium sulfate (111). Radioactive contamination of I11 was excluded by an intermediate scavenging operation which consisted of the addition of inactive thiosulfate or tetrathionate, mercuric chloride and barium chloride. These intermediate precipitates were checked to be practically inactive. If thiosulfate labeled a t the central S-atom was treated in this way the activity distribution was: lyOin I, 95% in 11, none in 111. If the same thiosulfate was titrated to tetrathionate with iodine, and analyzed in the same manner, again 1% was found in I, and 95yo in 11, but this time 2-3% entered into 111. Save for the infrequent side-reaction which caused the formation of radioactive sulfite during the decomposition of tetrathionate, the reactions may be assumed to proceed according to
+
+ 3HgC12 + 2Hz0 +HgC12.2HgS + 2S*O4" + 4C1- -+ 4H' 2SS*Oa' --+OaS*SSS*Os"+ 2e2OsS*SSS*Os' f 3HgClz + 4H20+ HgC12.2HgS f 4S*Od' + 4C1- + 8H+ + 25 2SS*Oa'
If tetrathionate, obtained by titration of the same thiosulfate with iodine, was reduced immediately with inactive s a d e (in presence of inactive bisulfite and excess formaldehyde), the sulfur formed was inactive. The atrates from this reaction were analyzed with mercuric chloride both directly, and after they had been titrated back to tetrathionate. In the first case the activity distribution was found to be: 1% in I, 96% in 11, 1-274 in 111; in t4e second case: 1% in I, 95% in I1 and 3% in 111. The activitier fomd in all fractions I may well
03S*StStS*03'
+ s- +2StS*O,' + s
This investigation represents part of the research program of the Foundation for Fundamental Research of Matter (F. 0. M,). It was performed with the financial aid of the Netherlands Organization for pure Research (2. W. 0.). INSTITWTVOOR KERNPHYSISCH ONDERZOEK 18 OOSTERRINGDIJX, AMSTERDAM HERMAN B. v. D. HEIJDE A. H. W. ATEN,JR. THENETHERLANDS RECEIVED JUNE9, 1952
TERRAMYCIN. V. STRUCTURE OF TERRINOLIDE. AN ACID DEGRADATION PRODUCT OF TERRAMYCIN
Sir: Among the products formed by the degradation of terramycin' in dilute hydrochloric acid at elevated temperatures is terrinolide (I), P K a , = 4.6, F K a g = 7.5 (dimethylfQnnamide-water) : [Q]D - 16.0' ( G 1% in 1: 1 methanol41 N hydrochloric acid). Anal. Calcd. for C20H16N08:C, 60.45; H, 3.81; N, 3.53. Found: C, 60.46; H, 4.10; N, 3.52. On hydrolysis in hot 12 N sulfuric acid, terrinolide loses ammonia and carbon dioxide to yield a nitrogen-free, optically-inactive compound, decarboxamidoterrinolide (11),2 PKa, = 4.7, PKaS = 10.2 (dimethylforrnamide-water). Anal. Calcd. for CleH1407: C, 64.41; H, 3.98; C-methyl, 4.25. Found: C, 64.10; H, 4.41; C-methyl, 3.82. Pentamethyldecarboxamidoterrinolide: m.p. 152153O, Anal. Calcd. for CUHZ4O7:C, 67.91; H, 5.70; CH80, 36.55. Found: C, 67.85; H, 5.74; CH1O, 35.95. Terrinolide and decarboxamidoterrinolide have been assigned structures I and 11, respectively. OH
k
OH OH 0 I, R CONHz 11, R = H
.OH
OH OH I11
Alkali fusion of I1 yields l18-dihydroxy-4-methyl3-naphthoic acid (III).* I, I1 and I11 enhance the (1) P. P. Regna, I. A. Solomons, K. Murai, A. E. Timreck, K. J. Brunings and W. A. Lazier, TBIBJOURNAL, 78, 4211 (1951). (2) T d n o l i d c and decarboxamidoturinolide were originally assigned the formulae CBHBNOIand C l B l h , respectively, in our &st Communication (R. Pasternack, P. Regna, R. Wagner, A. Bavley, F. Hochsrein, P. Gordon and K. Brtmlnga, T E JOUBNAL, ~ 78, 2400 (1951)). The formation of &able lolvntes and a tendency of these compounds to decompose under conditiotu of moleeulu wdght determination complicated the assipnment of the molecular farmular. (a) F. A. Aochitds, #I el., a9 bo publimlmd.
July 20, 1952
COMMUNICATIONS TO THE EDITOR
3707
acidity of boric acid to the marked degree charac- 55.17; H, 4.84; N, 5.85; C1, 7.40. Found: C, 54.85; H, 5.13; N, 5.97; C1, 7.19. P-Apoterrateristic of 1,8-naphthalenediols. Lithium aluminum hydride reduction of penta- mycin hydrochloride (I)2, [a]'% -28" (c, 1% in methyldecarboxamidoterrinolide consumes 0.5 mole ethanol), PKa, = 3.6, P K a , = 5.2, P K a , = 7.8 of hydride and yields no hydrogen; the reaction (dimethylfonnamide-water). Anal. Calcd. for CZZproduct is a dialcohol (IV), m.p. 114-115', anal. H22Nz08.HCl.H20: C, 53.17; H, 5.07; N, 5.64; Calcd. for C24H2a07:C, 67.27; H, 6.59. Found: C1, 7.14. Found: C, 52.90; H, 4.76; N, 5.65; C, 67.09; H, 6.69, which is readily dehydrated in C1, 6.90. We consider a-and P-apoterramycin to be stereodilute mineral acid to an ether, m.p. 141-142'. Anal. Calcd. for C24H2608: C, 70.23; H, 6.38. isomers of structure I. The two compounds are Found: C, 70.06; H, 6.06. That the lactone interconvertible in acid and alkaline solution, and structure shown by this reduction is a phthalide is their ultraviolet spectra are virtually identical. indicated by the extreme resistance of decarbqxCHI amidoterrinolide to hydrolysis, This assignment is in agreement with the carbonyl band at 5.73 p (dioxane) in the infrared absorption spectra of I and 11. The marked similarity of the ultraviolet spectra of I, I1 and 1,8-dihydroxynaphthalene-2carboxylic acid8 determines the orientation of the 11 OH OH 0 phthalide ring on the naphthalenediol system. I Further, the acidity of terrinolide, @ K a , = 4.5, is in good agreement with the acidity of 1,8-dihydroxy-2Alkali fusion of the apoterramycins yields 1,8naphthaldehyde, pKa = 4.5 (dimethylformamide- dihydroxy-4-methylnaphthalene-3-carboxylicacida water). and 2,5-dihydroxybenzoquinone. Both isomers The presence of five phenolic hydroxyl groups in lose carbon dioxide, ammonia and dimethylamine I and I1 is shown by the formation of pentamethyl in hot concentrated hydrochloric acid, and are and pentaacetyl derivatives. The stability of I1 converted to decarboxamidoterrinolide (11).4 in strong acid, and the marked susceptibility of I Pentamethylterrinolide is formed by treatment of and I1 to air oxidation suggest that the C8H6Oa a-apoterramycin with methyl iodide and potassium moiety not accounted for by the dihydroxybenzo- carbonate in acetone. The presence of the diphthalide system is a trihydroxybenzene. ComOH parison of the ultraviolet spectrum of the dialcohol (IV) to the composite curves derived from 3hydroxymethyl-4-methyl-1,8-naphthalenedi01~ and the three isomeric trihydroxybenzenes indicates that the C6H608 group is hydroxyhydroquinone. The presence in terrinolide of a carboxltmide OH OH OH 0 group which is lost by hydrolysis in sulfuric acid is supported by the infrared absorption spectra of 11, R = H 111, R = CONHZ I, I1 and their derivatives. The second acid constant of I, p K a 2 = 7.5 (compare p K a l = 10.2 for hydroxybenzophthalide system in the apoterra11) requires that the carboxamide group in terrino- mycins and texrinolide (111) is evident since the lide (I) be attached to the hydroxyhydroquinone absorption spectra of these compounds are pracring between two phenolic groups. tically superimposablein the 330-420 mp region of (4) J. Btieseken, J. de Bruin and W.van Rijswijk de Jong, Rec. frau. the ultraviolet spectrum, and all three compounds chim., 68, 3 (1939). show the strong enhancement of the acidity of RESEARCH LABORATORIES F. A. HOCHSTEINboric acid characteristic of 1,8-naphthaIenediol~.~ CHAS.PFIZERAND Co., INC. P. P. REGNA Thus, the apoterramycins differ from terrinolide BROOKLYN 6, N. Y. K. J. BRUNINGS (111) only in the structure of the isolated sixCONVERSE LABORATORY HARVARD UNIVERSITY R. B. WOODWARDmembered ring. CAMBRIDGE, MASSACHUSETTS The isolation of 2,5-dihydroxybenzoquinonefrom RECEIVED JUNE25, 1952 a-apoterramycin suggests the relative positions of two carbonyl groups, a hydroxyl and a dimethylamine group in the isolated ring. The stability of TERRAMYCIN. VI. THE STRUCTURE OF 4- AND the apoterramycins under the conditions of their 6-APOTERRAMYCIN, ACID REARRANGEMENT formation from terramycin excludes a- or yPRODUCTS OF TERRAMYCIN diketone structures within the carbocyclic ring since Sir : (2) This compound was described by R. Pasternack, P. P. Regna, In 1.5 N aqueous hydrochloric acid a t 60°, terramycinl loses a molecule of water and re- R. L. Wagnu. A. Barley, P. A. Hochstein, P. N. Gordon and K. J. Brunings, ibid.. 71,2400 (1951). as CaHuNxOrHCL It has since been arranges to form two closely related optically active found that Karl Fisher reagent shows the presence of one molecule compounds: a-Apoterramycin hydrochloride (I), of water, and that recrystallization from methanol displaces the water [ c Y ] ~ ~ +123O D (c 1% in ethanol), PKa, = 4.0, with a molecule of methanol. (8) F. A. Hochstein, e1 at.,to be published. pKa* = 5.1, *pKa, = 8.4 (dimethylfonnamide(4) F. A. Hochstdn, P.P.Regna, K. J. Bruninga and R. B..Woodwater). Anal. Calcd. for CaHttNnO,.HCl: C, ward, T E J ~O ~ A L 74, , 8706 (1952).
x;
@&$pCO..; ' I/
r [ n
(1) P. P.Regna, I. A. Solomons, K. Mud,A. E. Timreck, P.J, Bcuninga and W. A. L d a , T E I ~ J o w u .TS, 4211 (1061),
(6) J. B(LcKLen,J, do Bruin and W.van Rijswiik de Jour, ILc. taw. shim., I, I (1989).
