conditions yielded a mixture of the 5,f and 5a epimers fromwhich the latter was ... epimeric diols in approximately equal amounts: the 3a-hydroxy compound, m.p..
Proceedings of the NATIONAL ACADEMY OF SCIENCES Volume 53 * Number 5 * May 15, 1965
AN EXPERIMENTAL METHOD FOR DETERMINING A-VALUES: THE METHYL GRROUP* BY J. F. BIELLMANN AND WILLIAM S. JOHNSON INSTITUT DE CHIMIE, STRASBOURG, FRANCE, AND DEPARTMENT OF CHEMISTRY, STANFORD UNIVERSITY
Communicated March 22, 1965
Some time ago we undertook a study with the view to designing a general method for determining A-values of various groups. The idea in principle was to employ an equilibration technique of a 3a-substituted 6-keto steroid, I = II. Since it is known' that when R = H, the equilibrium lies largely, but not entirely, in favor of the 5a (A/B trans) isomer I (R = H), we felt that such a system would in principle be particularly suitable for the measurement of the A-values (defined as the freeenergy difference between an equatorial and axial substituent) of fairly large R groups since the equilibrium I =. II would move toward the less abundant species II as the group size of R is increased. This expectation is a result of the fact that R assumes the stabilizing equatorial conformation in II and the destabilizing axial conformation in I. The present communication constitutes a report of the application of this principle to the determination of the A value of the methyl group. Recently, Neville Jones and Kine' have reported the measurement of A-values for OH, OCH3, OAc, and Cl by a similar technique, but with the 3-substituent in the 0- rather than the a-configuration. In this 303 system the equilibrium is pushed even further toward the side of the 5,3 (A/B trans) isomer, and the method can therefore not be employed for accurate determination of A-values of large groups. We elected to work in the cholestanone series. The previously known3 reference compounds I (R = H) and II (R = H) were prepared by a new approach: 6,3acetoxy-53-cholestan-3-one4 was reduced by the Caglioti procedure5 to 6f0-acetoxy5f3-cholestane which, on saponification followed by oxidation with sodium dichromate in acetic acid, was converted into 5(3-cholestan-6-one (I, R = H), CD. in ethanol [0]294-29 - 14058. Equilibration of this last substance under basic conditions yielded a mixture of the 5,f and 5a epimers from which the latter was easily separated because it was present in preponderance (see below), CD. in ethanol [0]296 - 5377. The 3a-methylcholestan-6-ones, I (R = CHR) and II (R = CHO), were obtained as follows. The reaction of excess methylmagnesium iodide with 6,8-acetoxy-513-cholestan-3-one4 yielded a separable mixture of C-3 epimeric diols in approximately equal amounts: the 3a-hydroxy compound, m.p. 113.5-116°, [a]D + 200(c 1, CHCl3) [Found: C, 80.2; H, 11.9 ] and the 3,6-hydroxy isomer, m.p. 144-1450, [a]D + 230 (c 0.5, CHCl3) [Found: C, 80.2; H, 11.9]. The assignments of configurations at C-3 were made on the basis of infrared spectral properties in the hydroxyl stretch region and on the basis of the behavior 891
892
CHEMISTRY: BIELLMANN AND JOHNSON
PROC. N. A. S.
H
R (equatorial)
FIG. 1.
on dehydration with phosphorus oxychloride in pyridine. The 6,3 hydroxyl group was reacetylated with pyridine and acetic anhydride, and the resulting hydroxy acetate was dehydrated by treatment with perchloric-acetic acid. Both of the C-3 epimeric hydroxy acetates gave an excellent yield of the same product which, as shown by nmr spectroscopy, was 6f3-acetoxy-5j3-3-methyl-cholest-3-ene, m.p. 109.5110.50, [aID + 230 (c 1, CHC13). Hydrogenation of the olefinic bond was expected to occur from the # face of the molecule, since the a side is severely hindered due to the cis A/B ring fusion. The product of hydrogenation was saponified and the noncrystalline 6$ hydroxy compound oxidized with Jones reagent to give 3a-methyl5,B-cholestan-6-one (II, R = CH3), m.p. 107-108° [Found: C, 83.9; H, 12.2], CD. in ethanol [01294 -j16335. EquilibrationlofjII (R = CH3) with methanolic sodium hydroxide gave a mixture containing preponderantly starting material which was removed in part by crystallization. Preparative thin-layer chromatography of the mother liquors on silicagel gave a small yield of 3a-methylcholestan6-one (I, R = CH3), m.p. 108-108.5°, CD. in ethanol [01292 - 5115. The highresolution mass spectrum of this material gave a molecular ion peak at 400.3693 i 0.0024 (calcd., 400.3705). The equilibration of the two pairs of ketones was carried out at 200 in ethanol containing 10 per cent potassium hydroxide. Circular dichroism was used to follow the processes, and the equilibria were approached from both sides. Our results showed that the equilibrium I =l II (R = H) was 92 i 3 per cent in favor of I (R = H), which is to be compared with the reported value of 88 A 2 per cent obtained in methanol at 250 by O.R.D. measurements.' The equilibrium I =r II (R = CH3) was similarly determined to be 30 i 3 per cent in favor of I (R = CH3). From these data the A-value for the methyl group at 200 is calculated to be 2.0 ± 0.35 kcal/mole, which is in agreement with previously estimated values.6 We feel that the applicability of our method is thus established. It is expected that somewhat better accuracy may be obtainable by the use of optical rotatory dispersion techniques for measuring the equilibria. * This program was initiated in the Department of Chemistry, University of Wisconsin, under the partial sponsorship of the Government of France and the French Alliance of New York who awarded fellowship grants to J. F. B. in 1959. Support was provided by the U.S. Public Health Service, the National Science Foundation, and the Centre National de la Recherche Scientifique. 1 See Allinger, N. L., M. A. Darooge, and R. B. Hermann, J. Org. Chem., 26, 3626 (1961), and references cited therein. 2 Jones, D. Neville, and D. E. Kine, Proc. Chem. Soc., 1964, 334 (1964). 3 See Henbest, H. B., and T. I. Wrigley, J. Chem. Soc., 1957, 4596 (1957). 4 Jones, D. N., J. R. Lewis, C. W. Shoppee, and G. H. R. Summers, J. Chem. Soc., 1955, 2876
(1955).
VOL. 53, 1965
CHEMISTRY: RAHN, SHULMAN, AND LONGWORTH
893
6 Caglioti, L., and P. Grasselli, Chem. Ind. (London), 1964, 153 (1964). 6Eliel, E. L., Stereochemistry of Carbon Compounds (New York: McGraw-Hill Book Co., Inc., 1962), p. 236.
THE UV-INDUCED TRIPLET STATE IN DNA BY R. 0. RAHN, R. G. SHULMAN, AND J. W. LONGWORTH BELL TELEPHONE LABORATORIES, MURRAY HILL, NEW JERSEY
Communicated by W. 0. Baker, March 30, 1965
Previous investigations have shown that the phosphorescence from native DNA is not the sum of the emission from the individual nucleotides.1 . 2 In this note we shall present evidence from electron spin resonance (ESR) and optical emission studies which shows that the observed emission from UV-irradiated DNA is from the thymine moieties. Because thymine only phosphoresces when it loses its N1 proton (pK = 9.8), these results are consistent with an effective transfer of this proton from thymine to adenine during emission. All measurements reported were made at 770K on samples dissolved in a 1:1 ethylene glycol-water glass. Concentrations used for making optical measurements were in the range 10-3-10-4 M. For ESR measurements the concentrations ranged from 3 X 10-4 M to 3 X 10-2 M. Optical emission measurements were made using Bausch and Lomb 0.5-meter grating monochromators for both excitation and emission. The light was modulated at 150 cps and phase-sensitive detection was used to observe the signal. A Varian Associates X-band spectrometer with 100 kc/s modulation was used for the ESR measurements. For excitation in the microwave cavity an Osram 200-watt high-pressure Hg lamp was used with an aqueous NiSO4 filter. With the exception of poly rAU and poly dAT, at least three different samples from different sources were used for all of these measurements. At 770K, neutral pH, the purine nucleotides adenosine monophosphate (AMP) and guanosine monophosphate (GMP) have long-lived triplet states which phosphoresce.3 4 The pyrimidine nucleotides thymidine monophosphate (TMP) and cytidine monophosphate (CMP) show no phosphorescence at neutral pH. In DNA, the observed phosphorescence bears little resemblance to either of the two purine nucleotides for, as shown in Figure 1, the DNA emission is less intense, redshifted, and less structured than the purines. Measurements of the triplet-state lifetimes show that native DNA has a 0.3-sec decay time compared with 2.3 sec for AMP and 1.3 sec for GMP. From Figure 1 we estimate the quantum yield of phosphorescence of DNA to be --0.2 per cent compared with -2 per cent for GMP. As part of an attempt to understand these observations, measurements were carried out on synthetic polynucleotides which serve as models for DNA and RNA. Among the systems studied was poly rAU which is a double-stranded polymer with Watson-Crick hydrogen bonding between the adenine and uracil bases. In this system, where the adenine and uracil bases alternate in a single strand, the observed emission resembled that of uridine at pH > 10 where the N1 proton is removed
