Sept. 20, 1957
COMhIUNICATIONS TO T H E
them as deoxy-CDP-X and CDP-X, but the nature of X remained uncertain although it was noted that it was basic. I am indebted to Dr. Reiji Okazaki of Nagoya University for preparing lyophilized samples of unfertilized sea urchin eggs. RESEARCH LABORATORIES OF TAKEDA PHARMACEUTICAL INDUSTKIES, LTD. OSAKA, J.4PAN
YUKIO S U G X S O
EDITOR
5075
eters determined from reactivities in the aromatic series are used t o correlate the effects of corresponding substituents in the aliphatic series. This usage completes a cycle wherein inductive effects from the aliphatic series (01 parameters) have been used to evaluate from aromatic series reactivities (equation ( l ) ) , the resonance parameter, UR-. This parameter is now applied to the correlation of resonance effects in the aliphatic series.
RECEIVED JULY 17, 1957 1 CORRELATION O F CARBANION REACTIVITIES BY U R - PARAMETERS’
Sir: The definition of nucleophilic resonance parameters] UR-, has been made recently according to the equation2 UR- =
U-
-
(1)
UI
where U- is the dual Hamniett sigma value which pertains to the reactions of p-substituted derivatives of aniline and phenol,3and UI is the inductive substituent constant.2 The UR - values provide a scale of the powers of substituent groups to delocalize negative charge by conjugation. Useful applicability of the scale to certain nucleophilic reactivities is suggested by the correlation of the rates of carbanion formation of substituted methanes in water a t 25’ by the equation log kl
5
(26.0)uk-
+ (4.0)(Zu1) - 21.8 + log
~ Z H
- IO 0
(2)
The term log n~ is a statistical correction term, n~ being the number of ionizable H atoms. The rate constants, k l , are those tabulated by Pearson and Dil10n.~ The correlation, shown in Fig. 1, holds over a spread of rates of eight powers of ten with an average deviation of 0.25 log unit. This is quite satisfactory since the precision of UR- values is no better than 10.03. Figure 1 shows a plot of log ( k l / n ~ us. ) the quantity (26.0)OR(~.O)~UI. The full line is one of unit slope. The term (4.O)z1~1presumes t o measure the inductive contribution, and the term (26.0)UR-, the resonance contribution to the logarithm of the ionization rate. The enormous (and therefore rather crude) constant for susceptibility to resonance interaction, 26.0 (compared to 4.0 for susceptibility t o inductive interaction), indicates that a much greater stabilization by resonance than inductive interaction results from the substitution of conjugated groups directly a t the carbanion carbon compared to the corresponding interactions of these groups acting through the ring system of p-substituted benzene derivatives. Equation (2) is unique in that resonance param-
+
(1) This work was supported in part by the Officeof Naval Research Project NR055-328. Reproduction in whole or in part is permitted for a n y purpose of the United States Government. (2) R. W. Taft, Jr., THIS JOURNAL, 79, 1045 (1957): c j . also M. S. Newman, ”Steric Effects in Organic Chemistry,” John Wiley and Sons, Inc., N. Y., 1956, p. 578. 594, and J. D. Roberts and W. T. Moreland, Jr., THISJ O U R N A 76, L , 216 (1953). (3) Cf. (a) reference ( 2 ) : (b) L. P. H a m m e t t , “Physical Organic Chemistry,” McGraw-Hill Book Co., Inc., h-ew York, N. Y., 1940. p. 184: (c) H . H. Jaffe, Chem. Revs., 53, 191 (1953). (4) R. G. Pearson and R. L. Dillon, THISJ O U R N A L , 76, 2439 (1953). These authors have noted qualitatively the correlation given by equation (2) cf. footnote (35).
I
j
I
-150 (260)U;
f
-20 0 ( 4 0 ) Z Ur.
Fig. 1.-Correlation by equation (2) of rates of carbanion HnO [XIXZXICIformation: XlXzXsCH HIO+, where X, is as indicated and XZ = Xa = H unless otherwise indicated. A line of unit slope is shown.
+
+
It has been assumed in using equation ( 2 ) that values for the substituents Br, C1, and CH3 (in the presence of a single conjugating group such as NO,, CH3C0, CH3S02,or CN) are zero, and that the former substituents contribute only to the ZUI term. If equation ( 2 ) is applied to polysubstitution of conjugating substituents (using ZUR-) substantial deviations are obtained (Table I lists some typical deviations). UR-
TABLE I DEVIATIONS FROM EQUATION (2) OF RATES OF IOSIZATIOS OF POLYSUBSTITUTED METHASES WITH MORE THAN OKE CONJUGATING GROUP Substituted methane
log(k/na) exptl
CNCH2CN CH3COCH2COCHa CHsCOCH(Br)COCH3 CHaCOCH(CH3)COCHa CH~COCHZNOZ NOZCHINOI
-0.3 -0.3 -0.3 -2.8 0.0 +1.4
log(k/*H) calcd. eqn. (2)
+ 1.2 + 8.6 f10.4 + 8.4 +11 0 f13.5
Deviation log units
+ 1.5 4- 8 . 9 4-10.7
+11.2 f11.0 +12.1
It is apparent from Table I that steric inhibition of resonance (and possibly other steric effects) of the second by the first conjugated substituent (and b y bulky substituents, e.g., Br, CH3) contributes appreciably to the failure of equation ( 2 ) . For example, the rate of ionization of dinitromethane is twelve powers of ten slower than predicted by equation (2). On the other hand, the
COMRIUNICATIONS TO WE EDITOR
5076
sterically more conipact CN groups of CH2(CN)s lead to a deviation of only 1.5 powers of ten greater rate, according to equation (a),than that observed. The results listed in Table I also suggest that saturation of resonance stabilization in the ionization state contributes to the failure of equation (2) since the deviations increase with increasing values of 2 U R - . 5
Vol. 79
which gives a positive Morgan--Elson test for 2acetamido aldoses, reduces Fehling solution inore readily than 111, and consumes one mole of base to yield formic acid and the 2-acetamid0-2~3-dideoxypentose, V (m.p. 128-130", [ a I z 3-SI" ~ (c, 1.0 in ethanol); calcd. for CiHlnN04: C, 47.99; H, 7.48; S , S.00. Found: C, 47.86; H , i.41; N , S.09).
( 5 ) T h e logarithms of t h e ionization constants of substitutcd methanes (reference 4) also appear to follow equation (2) with ,OR 7, a n d log K, of methane C -40. However. t h e ionization 30, P I constants for CHaSOzCHa and C H a C S are so crude t h a t the quantitative significance of t h e relationship is uncertain. T w o points arc worthy of comment. Nitromethane deviates (greater acidity) by about six log units, which is t h e same magnitude a s t h e deviation of this compound from t h e norm in a plot of log A1 2's. log R a (cf. Fig. 1, reference 4). T h u s t h e deviation apparently can he attributed t o t h e abnormally slow recombination rate of "nitrocarbanion" with hydrogen ion. On t h e other hand, t h e relationship is followed reasonahly well by both CHz(CN)l and C H ( C S ) 3 . (CHCO)?CW2 and (CHaCO)\C H deviate substantially (weaker acids) in t h e direction expected for steric inhibition of resonance in t h e carbanion ion.
I
COLLEGE OF CHEMISTRY ASD PHYSICS THEPENNSYLVANIA STATEUNIVERSITY ROBERT IV. TAW,JR. UNIVERSITY PARK,PENNSYLVANIA RECEIVED XLXXJST 7 , 1957 THE STRUCTURE OF MYCOSAMINE
Sir: We wish to report in this communication on the structure of an amino-sugar mycosamine, which represents the nitrogen-containing moiety of the j. antifungal antibiotics, nystatin,2 C4fiH77N019,3s4 1,5 This amino and amphotericin B, C46HT3N020. sugar was isolated in the form of its tetraacetatel from the mixture of products resulting from the sulfuric acid catalyzed acetolysis of either the antibiotics or their hydrogenated derivatives. Structure I is assigned t o the parent amino sugar on the basis of the following evidence: Chromatographic fractionation on alumina of the chloroform soluble portion of the acetolysis products yielded tetraacetylmycosamine (TI m.p. j. 159-161°, [ c Y ] ~ ~+39' D (c, 1.0in ethanol); calcd. I-i I-I 110 0 1I for CtjHgN04.4CH3CO: 50.75; H, 6.39; hT, o=c c-0 cod 4.23; acetyl (total), 32.0. Found: C, 50.40; + H, 6.44; K j4.26; acetyl, 50.9),hydrolyzable by .H20 barium niethoxide in methanol to N-acetylinycosPO\ CII, CHI CI310 amine' (111, m.p. 191-192", [ a I z 3 D -46' (c, 1.0 CI-IJO VIIIa 1 7 I I 11) in ethanol) ; calcd. for C6H12N04.CH3C0: C, 46.82; H, '7.37; K, 6.83; N-acetyl, 21.0. Found: C, Furthermore, methyl N-acetylmycosaniinide V I 46.58; H, 7.22; N, '7.07; N-acetyl, 21.2): which reduces hot Fehling solution only slowly and gives (m.p. 168-170", [ f f I z 3 D +47" (c, 0.9 in ethanol); a positive iodoform test. Treatment of I11 with calcd. for C9HI7NO5: C, 49.30; H, 7 3 2 ; N, (i.39; periodic acid (consumption 1 mole) yielded 2-acet- OCH3, 14.1. Found: C, 49.00; H, 7.56; K, 6.13; ainido-3-hydroxy-4-formoxypentanal1 1 7 (amor- OCH3, 13.5), obtained from 111 with methanolic phous, calcd. for C8Hl3NO5: C, 47.29; H , fi.45; hydrogen chloride, was reduced TT. ith lithium N, 6.85. Found: C, 47.22; H , 6.84; N, 6.93, aluininuni hydride to methyl N-ethyliiiycosaininide, Y I I , (m.p. 90.5-92.5', [Cylz3D +25" (c, 1.0 (1) J. D . D u t c h e r t M . B. Young. J. H . Sherman, W.E. Hihhits and in n a t e r ) ; calcd. for C9Hl9KO4. C, ,5266; H, D. R. Walters, "Antibiotics Annual," 1966-1957, Medical Encyclupedia, Inc., New York, N. Y., 1956, p. 866. 9.33; N, 6.83; mol. weight, 205. Found: C. (2) T h e E. R.Squibb and Sons trademark for nystatin is "5lyco52.51; H, 9.19;X, 6.84; neut. eq. (perchloric acid statin." titration), 211), and this base was degraded with (3) J. D . Dutcher, G. Boyack a n d S. FOX."Antibiotics Annual," sodium periodate (2 moles) t o the known6 D'1953-1956, Medical Encyclopedia, Inc., New York, 19.53 11. 191. ( 4 ) J. D. Dutcher, D. R. Walters and 0 . P. Wintersteinrr, " ' I ' h ~ ~ r n i ~ v ~ n c ~ l i ~ ~ ~ v - ~ - n i c t ~ i y l d i g aldehyde l y c o l i c l r I I I (111p.
c,
I
of I'ungiis Disenses," T,ittlr, Flrown :ii1(1 C < > n n p i i i y ,l { o , t o ~ i , \ I u i lKh5, 1'. 168. ( 5 ) J . Vandrlintte, J. I,. Wnclitel : I I I < I 1,). '1'. Stiller, ' ' A i 1 1 .\iiii~iiil," 1!155-I!l,76, h'ledical EncvvIoI~~ili,i, l n c . , l!l.*)A, 1 3 6 i 7
,
11')
IO2
i
I 1
f-
[o]? I) -tI;j1 ( ( , 0 5 ,
