J. Org. Chem., Vol, 38,No. i6, 1973 2917
COMMUNICATIONS SCEEME I1
N=N
/
\
PhCHZN,
/S 'C
-
-x*
I
I
S02C,H,CH3.p
N
7,R'=Ph; R2-H 8, R I P R2 = Me
I
EtzN
Me
Lc' /
\ /S
PhCH,N\
C
N-N
/
RN
c''
\
NH
-
+ p-XCGH,SO,CI ether
N=N \
1
NEt
RN\C/
NSOZ CgH, X-p
1
II
S
S
5
4
R = n-Bu or PhCH,; X = H, CH, or C1
A2-tetrazoline-5-thiones (4) ,6 showed different physical and spectroscopic characteristics and a much higher thermal stability. It is reasonable to assume that an external stabilized 1,3 dipole (e.g., 6) is the intermediate in the thermal conversion of 2 t o 3. This has been confirmed by carrying out the decomposition of 4-benzyl-5-tosylimino-A2-1,2,3,4-thiatriazoline in the presence of electron-rich dipolarophiles.' Thus, thiazolidine 7 (mp 161-162") was obtained in 73% yield when 2 (R = PhCH2, X = CH,) was decomposed in the presence of an equimolar amount of /?-trans-N,N-dimethylaminostyrene in CC1, a t 60". The structure of 7 was deduced from microanalysis, ir (1530 ~ m - 9 nmr , ~ [ring protons a t r 5.55 and 5.68 ( J = 2.5 Hz), two nonequivalent benzyl protons a t 4.72 and 5.81 (J = 14.5 He)], and mass spectra (14. + at m/e 465, M e + - HX'i\!Iea a t m/e 420, and PhCHC(NMe2)HS.+ a t m/e 179). Similarly, when 2 (R = PhCH2,X = CHI) was heated with an equimolar amount of N,N-dimethylaminoisobutene in benzene for 3 hr, a 1:1 adduct (mp 132133') was obtained in 55% yield, corresponding to structure 8 on the basis of ir (1530 cm-l), nmr [ring proton a t T 6.12, ring methyls a t r 8.66 and 8.77, two nonequivalent benzyl protons at r 4.55 and 6.08 (J = 14.5 Hz)], and mass spectra a t m/e 4.17, (
M
a
+
M . + - NMe2, Me2CC(NMe2)HS.+ at m/e 131). The stereochemistry of 7 'cvas deduced from the C-4C-5 hydrogen coupling constant ( J = 2.5 Hz), whereas the indicated regiochemistry rests upon the observed chemical shift values of the C-N and C-S absorptions in the I3C nmr spectra of 7 and 8. (See Scheme 11.) (6) E . Lieber and J. Ramaohandran, Can. J. Chem., 87, 101 (1969). (7) Preliminary experiments indicate t h a t electron-poor olefins &re not suitable dipolarophiles for 6.
Ynamines also proved to be suitable dipolarophiles for 6. For instance, when 2 (R = PhCH2, X = CHs) was heated with 1 equiv of N,N-diethylaminopropyne in benzene for 4 hr, thiazoline 9 (mp, 118-119") was obtained (50-60001, by nmr, 21% isolated). The adduct exhibited ir (C=N a t 1500 cm-l), nmr [benzyl protons a t r 4.90 (s), ring methyl at T 7.853, and mass spectra (M. + at 429, Me + PhCHz at m/e 338, M . + - Tos a t m/e 274) consistent with the structure.
-
Acknowledgment. The authors are indebted to the IWONL for postdoctoral (E. V. L.) and doctoral (J. M, V.) fellowships and to the Centrum voor Hoogpolymeren (IWONL-Agfa-Gevaert) for financial support. DEPARTMENT OF CHEMISTRY EMIELVAN LOOCK LABORATORY OF MACROXOLECULAR JEAN-MARIE VANDCNSAVEL AND ORGANIC CHEMISTRY GERRITL ' A B B ~ * UNIVERSITY OF LOUVAIN GEORQE SYETG B-3030HEVFGRLEE, BELGIUM RECEIVED APRIL9, 1973
Correlation between Proton Magnetic Resonance Chemical Shift and Rate of Polar Cycloaddition
Summarg: A significant correlation has been found between the chemical shifts of the proton a t position 6 of 9-substituted acridizinium perchlorates and the log of the ratio of rate constants (klko) for the cycloaddition of 9-substituted acridizinium salts with styrene; the chemical shift data likewise give a significant correlation with Hammett substituent constants.
Sir: It was shown earlier that the rate of cycloaddition of the 9-substituted acridizinium cation (1) with
I
IO
11
I
R
2918
1.2
-
1.0
-
0.8
-
0.6
J. Org. Chem., Vol. 58, No. 16, 1973
COMMUNICATIONS
(Hz)
-
0-
1 D)
-4
0.4
-
A0
0.2
-
-a I
-0.4
I
I
I
0
0.4
0.8
OP 0.0
-
-0.2
-
Figure 2.-Leasbsquares
TABLE I COMPARISON OF CHEMICAL SHIFTDATAWITH up AND WITH THE RATEOF ADDITIOX OF STYRENE TO 9-SUBSTITUTED ACRIDIZINIUM PERCHLORATES
'4
-0.4}
L
plot of A 8 us. Hammett c,,.
R
I
I
I
-4
0
4
I
a
I
12
A6D
Figure 1.-Least-squares
plot of log k/ko 08. A6O.
styrene' or acrylonitrile2 was related to the electron deficiency a t position 6. Like the pmr spectrum of other aromatic quaternary cations3 that of the acridizinium ion shows the protons flanking the quaternary nitrogen to be strongly deshielded. Of these two strongly deshielded protons, that a t position 6 gives resonance (isolated singlet) a t the lower field, the chemical shift (10.6-11.0 ppm) varying with the nature of the 9 substituent. Proton magnetic resonance spectra of the acridizinium perchlorates (1) were obtained a t 39 f 1' using a Varian A-60 spectrometer operating at 60 MHz. Chemical shifts of the proton a t position 6 were measured from an internal benzene (Spectrograde) standard. Shifts were obtained a t four concentrations in the range of 2.0-3.5 mol % in freshly distilled dimethyl sulfoxide. Each sample was scanned a minimum of five times a t a rate of 1 Hz/sec. The standard deviation in 6 varied from 0.08 to 0.26 Hz; the estimated average uncertainty is f 0 . 2 Ha. Dilution plots were made and extrapolated to infinite dilution to give 6O. The data are recorded in Table I. A least-squares plot of log k / k o for the addition of (1) I . J. Westerman and C. K. Bradsher, J . OTQ.Chem., 86, 969 (1971). (2) C. K . Bradsher, C. R. Miles, N. A. Porter, and I. J. Westerman, Tetrahedron Lett., 4969 (1972). (3) E . g . , (a) H. Diekmann, G. Englert, and K . Wallenfels, Tetrahedron, 80, 281 (1964); (h) I. C. Smith and W.G. Schneider, Can. J . Chem., 39, 1158 (1961); (c) W. W. Paudler and T. J. Kress, J . Heterocycl. Chem., 6, 561 (1968).
80 (Ha)
A80
(Hz)
up
k X 108 min-la
-0.170b 2.0 f 0.1 -5.9 187.5 188.0 -5.4 -0.151b 2.8 I 0.1 193.4 0.0 0.000 5 . 0 f 0.2 F 193.5 0.062b 5.4 f 0.2 0.1 c1 1.1 0.227b 10.1 I 0 . 5 194.5 Br 194.1 0.7 0.232* 11.2 I 0 . 8 C02H 3.5 0.406O 18.1 I 0 . 7 196.9 0.77gb 105 & 5 10.3 NO2 203.7 a Reference 1. D. H. McDaniel and H. C. Brown, J . Org. Chem., 23, 420 (1958). H. Van Bekkum, P. E. Verkade, and B. M. Wepster, Red. Trav. Chim. Pays-Bas, 78,815 (1959). CHa i-Pr H
styrene to 9-substituted acridizinium derivatives vs. A6O (change in chemical shift from R = H) is shown in Figure 1. The correlation factor of 0.98 is quite satisfactory . As follows from the earlier observation that log k / k o gave a significant linear free-energy plot with the Hammett up, a plot (Figure 2) of Au0 vs. up gave a significant correlation of 0.97 (for all values or for primary values This correlation of proton chemical shifts with Hammett substituent constants can be interpreted as arising from the polarization of the C-H bond at position 6 which must in turn arise from the density of ?r electrons at that The slopeS of the line (Figure 2) is 15.8 f 1.3 Hz/sigma. While there has been an increasing number of attempts to relate the pmr of aromatic ring hydrogens to (4) An equally "significant" correlation (0.98) may be obtained by the use of
Vf.
( 5 ) T. K. Wu and B. P. Dailey, J . Chem. P h y s . , 41, 2796 (1964). (6) H. Spiesecke and W.G. Schneider, J . Chem. Phys., 36, 731 (1961). (7) M. T. Tribble and J. G. Traynham, "Advances in Free Energy Relationships," N. B. Chapman and J. Shorter, Ed., Plenum Press, London, 1972, Chapter 4. (8) The fairly common use of the symbol p to designate this slope may confuse the casual reader since i t is not the dimensionless p of the Hammett relationship.
COMMUNICATIONS
J . Org. Chem., Vol. 38, No. 16, 1973 2919
electron density at the carbon to which they are attachedl7l9no one previously appears to have related the rate of cycloaddition of such systems to the pmr of a proton at a carbon atom which would be involved in the creation of a new u bond. This is surprising in that Hobgood and Goldsteinlo nearly a decade ago demonstrated an “approximately linear” relationship between the chemical shift of the proton a t the 4-trans position of substituted butadienes and the log of the rate constant for cycloaddition with maleic anhydride. We feel that this correlation of chemical shift and cycloaddition rates will prove to be particularly important in the study of steric vs. electronic effects in polar cycloaddition. (9) E.g., T. Schaefer and W. G. Schneider, Can. J . Chem., 41, 966 (1963); P . J. Frank and H. S. Gutoasky, Arch. Scz., 11, 215 (1958); A. Veillard, J . Chzm. Phys., 59, 1056 (1962);A. Veillard and B. Pullman, C . R. Acad. Sci., 263, 2418 (1961); K. T. Potts and J. Bhattacharyya, J. Org. Chem., 37, 4410 (1972). (10) R. T. Hobgood Jr., and J. H. Goldstein, J. Mol. Spectrosc., l a , 76 ( 1964).
GROSSCHEMICAL LABORATORY DUKE UNIVERSITY DURHAM, NORTHCAROLINA 27706
T . G. WALLXS N. A. PORTER C. K. BIZADSHER*
RECEIVED MAY30, 1973
Two-step Synthesis of a Triketone of the endo-Tetracyclo[5.5.1.0z~6.010~13]tridecane1 Series. X-Ray Crystallographic Proof of Its Structure and Stereochemistry
A compound (6),obtained by reaction of glyoxal with dimethyl 3-ketoglutarate in aqueous solution at room temperature and subsequent treatment of an intermediate 0-keto ester ( 5 ) with acid, is shown by X-ray crystallography to be endo-tetracyclo [5,5.1.02,6.0 lo,13]tridecane-4,S, 12-trione.
Summary:
Reaction of glyoxal (1) with dimethyl 3-ketoglutarate (2) in aqueous solution at room temperature and pH 5.0 has been found2to give the ester 3,3which yields cis-bicyclo [3.3.0]octane-3,7-dione (4)334 on treatment with acid. Compound 4 was accompanied2 by another ketone C13H1403 (mp 148-151O)j having spectroscopic properties very similar to those of 4; it is undoubtedly derived from a &keto ester analogous to 3 which, however, was not isolated. We now wish to report the isolation of this intermediate, apd the elucidation of structure and stereochemistry of the C13 compound by X-ray crystallography. Formation of a compound ClaH1403 through reaction of 1 with 2, followed by treatment with acid, could be rationalized by assuming that 3 forms initially and subSir:
(1) Dr. K. L. Loening, Director of Nomenclature, Chemical Abstracts Service, has advised us that compound 6 can be correctly designated either as octahydro-lH-dicyclopentaIa,cdlpentalene-1,4,6(2H,4aH)-trione or as tetracyclo[5.5.l.0*~~.0~~~~~ltridecane-4,8,12-trione. We wish t o thank Dr. Loening for his helpful interest. (2) J. M. Edwards and U. Weiss, Tetrahedron Lett., 4885 (1968). (3) G. Vossen, Dissertation, Bonn, 1910; P. Yates, E. 8. Hand, and G. R . French, J. Amer. Chem. Soc., 82, 6347 (1960). (4) H.W.Wanzlick. Chem. Ber., 86, 269 (1953). (5) It has been observed recently that recrystallization from methanol raises the melting point to 160O. The sample used for X-ray crystallography has this melting point.
Figure 1.
sequently reacts with one molecule each of 1 and 2 in an aldol reaction analogous to the one taking place in its own formation; for the resulting p-keto ester, structure 5 appears logical.6 Assuming the usual cis stereochemistry a t the junction of two cyclopentane rings,’ formula 5 represents two stereoisomers with ring D in syn or anti relationship to rings A and B. Treatment of 5 with acid would then give 6, C13H&, which could again be the syn or anti isomer. An ester 5 , mp 173-176”, having the expected compositions and spectroscopic properties, was obtained in 10% yield when 1 and 2 were allowed t o react in the required molecular ratio 2:3 for 1 week in aqueous solution at room temperature and pH 3 instead of the pH 5 used in the earlier work;2 trituration of the resulting precipitate with methanol and recrystallization from the same solvent gave pure 5 . Treatment of 5 with hot 22% HC12 yielded 6. An X-ray crystallographic investigation has now shown that structure 6 is indeed correct and that the compound has the all-cis stereochemistry (6a), The crystals wece monoclinic, P21/n, a = 6.299 (1) A, b = 15.511 (1) A, c = 10.943 (1) A, 0 = 105.78 (l)O, 2 = 4. A total of 1933 independent X-ray intensities (328, unobserved) were measured by means of an EnrafNonius CAD-4 diffractometer. The structure was solved by direct methods using our own semiautomatic program. With anisotropic thermal parameters for the C and 0 atoms and isotropic parameters for the hydrogen atoms, the structure has been refined by fullmatrix least-squares to an R factor of 0.036. Estimated standard deviations C-C and C--0 bond lengths are typically 0.003 A. An ORTEP drawingg of 6 (Figure 1) shows its conformation and demonstrates that the molecule is chiral, lacking the mirror plane which the conventional structural formula would indicate. The observed conformation is very reasonable if intramolecular interactions are taken into consideration. Since the crystals are centrosymmetrical, (6) Reaction of 3 with 1 and 2 could also take place a t positions 2 and 4, or 2 and 6. However, the resulting p-keto esters isomeric with 6 could undergo decarboxylation t o a Cia compound only with violation of Bredt’s rule and are thus quite unlikely. This point was brought t o our attention by a referee on our eariier paper.2 (7) Cf.E. L. Eliel, “Stereochemistry of Carbon Compounds,” McGrawHill, New York, N. Y.,1962,pp 273-274. (8) Satisfactory analytical and mass spectrometric data were obtained for 5. (9) C. K. Johnson, O R T E P Report ORNL-3794 (2nd revision), 1970. Oak Ridge National Laboratory, Oak Ridge, Tenn.
