Structure of a rearrangement product: 1-methyl-3-(5-amino-1-benzylimidazol-4-yl)-1,2,4-triazole, C13H14N6

Journal of Crystallographic and Spectroscopic Research, Vol. 17, No. 4, 1987

Structure of a rearrangement product: 1-methyl-3-(5amino-l-benzylimidazol-4-yl)-l,2,4-triazole, C13H14N6 CAROL AFSHAR, HELEN M. BERMAN,* and PATRICIA SAWZIK

The Institute .for Cancer Research Fox Chase Cancer Center Philadelphia, Pennsylvania 19111 LESLIE LESSINGER 1

Chemistry Department Barnard College New York, New York 10027 BENJAMIN B. LIM and RAMACHANDRA S. HOSMANE

Chemistry Department The University of Maryland Baltimore County Campus Catonsville, Maryland 21228 (Received January 1 7, 1987)

Abstract

The title compound was prepared during the course of an attempt to synthesize 8-amino-3-benzyl-6-methylimidazo[4,5-e][1,2,4]triazepine. The crystals are monoclinic, space group P 2 J c , a = 6.2375(5), b = 9.1070(8), c = 22.182(2) A_, /3 = 91.797(7) ~ Z = 4. The structure was solved by direct methods and refined by least squares to R = 0.063 on all 2142 measured reflections. The imidazol-4-yl-triazole system is planar, conjugated, and aromatic. One hydrogen atom o f the exocyclic amino group is involved in a bifurcated hydrogen bond, of which one branch is an intramolecular attraction to a triazole ring nitrogen atom, while the other branch is intermolecular. The crystal is held together by two N - H - - - N hydrogen bonds between molecules related by Visiting scientist, The Institute for Cancer Research. 533

0277-8068/87/0800-0533505.00/09 1987PlenumPublishingCorporation

Afshar et aL

534

translation along the a axis and by 7r-7r stacking interactions of the imidazol4-yl-triazole ring systems and the phenyl tings across centers of inversion.

Introduction

In our research program to synthesize and investigate the properties of a wide variety of little-explored 5 : 7-. 5 : 8-. and 5 : 9-fused novel heterocyclic systems of potential chemical, biochemical, and pharmaceutical interest (Hosmane et al.. 1986)o we desired to synthesize 8-amino-3-benzyl-6-methylimidazo[4,5-e][1,2,4]triazepine (I). Our synthetic approach (Fig. 1) involved the ring closure of N-amino-N-methyl-N'-(1-benzyl-4-cyanoimidazol-5-yl)formamidine in refluxing toluene-methanol, catalyzed by tfifluoroacetic acid. This reaction indeed provided a product whose spectral and microanalytlcal data were compatible with structure I~ The proton NMR spectrum of the product in deuterated dimethyl sulfoxide exhibited a singlet at 6 3.82 corresponding to the methyl group, two CH singlets at 6 8.35 and 7.28, respectively, a D20 exchangeable NH2 group at ~ 5.41, a benzyl CH2 at 6 5.09. and the corresponding phenyl group at 6 '7.28. The mass spectrum (70 eV) of the product revealed the molecular ion at m/e 254. The infrared spectrum showed the absence of nitfile absorption in the 2200 c m - 1 region, thus indicating that the ring closure had indeed taken place. The elemental microanalyses were in excellent agreement with the calculated values for structure L In spite of strong apparent corroboration by physical and microanatytical data, care needs to be exercised in assigning a structure such as I. Compound I, a 5 : 7-fused heterocycle with eight 7r-electrons in its triazepine ring, is antiaromatic by HiJckel standards (4n + 2 rule) and is thus prone to opportunisac rearrangements to a more stable 5: 6- or 5 : 5-fused system. Rearrangement to the 5:6-fused II or III (Fig. 2) can be envisioned as proceeding through an initial electrocyclic transformation of I to the tficyclic intermediate Ia. Compound II can arise by the facile nng opening of Ia. while III would result from

NH2

%~%

4r ~ Toluene/{VleOH Me-N HzN I~N ~i ~ CF3CO'I'" ,H \....~/LL.N/ Me CH=Ph CH2Ph 3: Fig. 1. Proposedsynthetic scheme.

Structure of C13H14N6

535

r-

NH,

\~..~L..~I"

~H 2

Electrocycllc

CHzPh

I

CH2Ph

la

NH-.-NHMe

NH

CHzPh

III

]

CHzPh

II

Fig. 2. Pathway for rearrangementof I to form II and III. the acid-catalyzed methanol-assisted Dimroth rearrangement (Lister, 1971) of II. Rearrangement to the 5:5-fused IV (Fig. 3), on the other hand, would involve an acid-catalyzed addition of methanol to the triazepine ring system of ! to form Ib, followed by sequential ring opening (to Ic), ring closing (to Id), and aromatization with loss of a molecule of methanol. Compounds I - I V all possess the same molecular formula, CI3H14N6,and the acquired physical and microanalytical data were consistent with any one of the four structures, albeit 1H NMR of the product rendered structures II and III less likely as compared with I or IV. Therefore, this crystal structure analysis was undertaken to assign the chemical structure of the product unambiguously.

Crystal structure determination: Experimental Colorless needles, elongated along a, were obtained by recrystallization from 50 % methanol. The space group was determined from the systematically absent reflections 0k0, k = 2n + 1 and hOI, l = 2n + 1. Unit cell parameters at 21 ~ were determined from least-squares fitting of the diffractometer angles for 25 centered reflections evenly sampling reciprocal space, with 30 ~ < 20 <

536

Afshar e~ aL

~

H2

NeOH/H|

OMe~H ' r

CHzPh

mb

CHzPh

OMe~ ~-'~.. NH2

CH

Me'~~~~ " CHzPh

Id

]c

"~OH

MoJ< CH2Ph iV Fig. 3. Pathway for rea~angemem of I to form IV.

43~ The density was determined by flotation in a C6HsC1/CC14 mixture, and clearly indicated one molecule o f formula weight 254.3 in the asymmetric Llnit, and no solvent incorporated into the crystal. Crystal data and information concerning intensity data collection and least-squares refinement are given in Table I. The crystal used for intensity measurements had approximate dimensions 0.5 • 0.1 • 0.1 mm. Data were collected on an Enraf-Nonius C A D - 4 diffractometer with graphite-monochromated Cu Kc~ radiation, X = 1.54178 ~ No signs of decay were observed in the intensities o f three standard reflections or

537

Structure of C13H14N6 Table 1. Crystal data, intensity data collection, and refinement Molecular formula Formula weight (g/tool) Melting point (~ Space group (monoclinic, No. 14) Unit cell a (6) b (A) c (A) ,8( ~) V (~3) Z (formula units per unit cell) D~ (g/cm 3) D,,, (g/cm3) F(000) (electrons) (absorption coefficient, em ~) Data collection scan method 20 range (deg) Standard reflections (hkl) Variation in standard intensities Minimum relative transmission T Number of intensities measured Merging R ~mFequiv/]~fequlv Number of unique intensity data Number with I > 2o(I) Range of Miller indices h, k, l Number of data used in refinement Number of parameters refined Functional minimized Weights used in final cycles Final R = E(IFol - If~l)/Zlfol Final wR = [~;w(I Fol - [Fcl)Z/~w[ Fo121'/2 Final variance ("goodness of fit' ') V = [~w(IFo] -]F~[)~/(2142 - 179)] '/2 Final difference density (e/A 3) =

C13HI4N6

254.297 254-255 (d) P2~/c 6.2375(5) 9.1070(8) 22.182(2) 91.797(7) 1259.43 4 1.341 1.336(3) 536 7.15 0/20 3.0-130.0 (035), (236), (306) +3% 0.817 2391 0.029 2142 1928 - 7 to 7, 0 to 10, 0 to 26 2142 179 ~w(I Fo[ - ]Fct) 2 w = 1/((~2(Fo) + 0.0005F 2) 0.063 0.064 0.62 -0.24 to +0.25

in the crystal itself during the data collection. Intensities w e r e corrected for Lorentz and polarization effects, and an e m p i r i c a l absorption correction was applied, using C-scan intensity variation data. N o correction was m a d e for extinction.

Structure solution and refinement T h e structure was s o l v e d by direct phase d e t e r m i n a t i o n methods using the p r o g r a m system MULTAN 80 (Main et al., 1980). Since the m o l e c u l a r structure was uncertain, care was taken to assign C and N a t o m identities correctly on the basis o f p e a k electron densities, sensible patterns o f thermal m o t i o n parameters, bond lengths, and total v a l e n c e , including b o n d i n g to H atoms (all o f

Afshar et aL

538

which were initially located in difference electron density Fourier syntheses). Only the structure IV gives complete consistency on all measures. The structure was refined using full-matrix least squares, with all C and N atoms assigned anisotropic thermal motion parameters. Hydrogen atoms were included in calculated positions, 1.08 A, from C and 1.00 ]~ from N, riding on the atoms to which they are bonded, and in the final refinement cycles the C H 3 and NH2 groups were allowed to move as rigid bodies, with angles H - C H and H - N - H fixed at 109.5 ~ All H atoms were assigned a single isotropic thermal parameter, which refined to U = 0.0861(25) ~2. Weights were chosen so as to minimize the variation of the variance as a function of the magnitude of F o, assuming a form w = 1/(o 2.... ting + gl Fo 2)~ All 2142 measured unique F o were used in the refinement, which converged to a final residual R = 0.063 and final weighted residual wR = 0.064. For the 179 parameters refined, the ratio of shift to estimated standard deviation had an average value of 0.014 and a maximum value of 0.064 (except for the methyl group rotation parameters. for which the ratios were 0.12. 0,18. and 0.41) in the final refinement cycle. The final difference electron density map had no distinct features. Least-squares refinements and geometric calculations were performed using the SHELX program system (Sheldrick, 1976). Scattering factors for C and N were taken from Cromer and Mann (1968)~ with A f ' and zXf" from Cromer and Liberman (1970). The scattering factor used for H was that for spherical bonded H given by Stewart et al. (1965).

Results and discussion Final atomic coordinates and thermal motion parameters are listed in Table 2. Bond lengths and bond angles are given in Table 3. A view of the molecule, with the atom numbering scheme we have used, is shown in Fig. 4. (Formal single and double bonds are indicated in Fig. 3.) Comparison of the bond lengths within the imidazol-4-yl-triazole ring system and to its exocyclic substituents with standard values (International Tables for X-Ray Crystallography, Vol. III, 1983) for carbon-carbon, carbon-nitrogen, and nitrogen-nitrogen bonds clearly indicates that this is a conjugated aromatic system. Such a system should be essentially planar, and it is. The values in Table 4 show that no atom of the ten atoms in the 2-ring system deviates by more than 0.08 A from the best plane through those ten atoms; the exocyclic methyl carbon atom, amino nitrogen atom, and methylene carbon atom deviate by a few tenths of an A.ngstrom from the same plane. Considered separately, the 5-membered imidazole and triazole rings are even more rigorously planar, as shown in Table 4. The planes through these two rings make an angle of 715 ~ with each other. Just as the carbon-carbon bond lengths in the phenyl ring

S t r u c t u r e o f C~aHi~N6

N~

9

~ N N I

I l i l

539

N N ~ II

NN I

III

I

II

~ i

I

I

cq

§ • 4-

I II

x

Z

9

3 | ~ / 9

m @

540 Table 3. Bond lengths (A) and bond angles (deg) .or pamntheses Atoms N(1)-C(2) N(1)-C(13) C(2)-N(3) N(3)-C(4) C(4)-C(51 C(4)-C(7) C(5)-N(1) C(5)-N(6) C(7)-N(8) N(8)-N(9) N(9)-C(10)

estimated standard devia.'~on~ in

Distance

Atoms

Distance

1.363(2) 1.456(2) 1.306(3) 1.395(2) 1.371(31 1.443(3) 1.380(21 1.373(2) 1.324(2) 1.362(2) 1.322(3)

N(9)-C(12) C(10)-N( 1i) N(I 1)-C (7) C(131-C(141 C(141-C(15) C(151-C(161 C(!6)-C(17) C(171-C(18) C(181-C(191 C(19)-C(14)

1.449(3~ 1.330, 3) I. 364(3) t.520(3t [.388(3) 1,385(31 !.367(41 1.374tza~ [ .401 (4) 1.37943)

Atoms

Angle

Atoms

Angie

C(2)-N(1)-C(51 C(5)-N(t)-C(13) C(2)-N(1)-C(13) N(1)-C(2)-N(3) C(21-N(31-C(41 N(3)-C(4)-C(5) N(3)-C(4)-C(7) C(5)-C(4)-C(7) N(1)-C(5)-N(6) C(4)-C(5)-N(6) N( 1)-C(5)-C(4) C(4)-C(7)-N(8) C(4)--C(7)-N(11) N(8)-C(7)-N(111 C(7)-N(8)-N(91

106.7(21 125.5(21 126.6(2) 112.7(2) 104.9(2) 110.0(21 t23.8(2) 125,9(21 121.5(2) 132.6(2) 105,7(2) 124.1(2) 121.7(2) 114.1(21 102,8(2)

N(8)-N(9)-C(10) N(8)-N(9I-C(12) C(101-N(9)-C(12) N(9)-C(10)-N(111 C(7)-N(11 l-C(10) N( 1)-C( 13)-C(141 C(13)-C(14)-C(15) C(13 I-C(14)-Cf 191 C(151-C(141-C(19) C(141-C(151-C(16) C( 151-C( 16)-C(17) C(I6)-C(171-C(18) C(171-C(18)-C(19) C(14J-C(191-C(181

109.7(2t 12 t.2(21 129.0(21 110.9(21 i02.5(21 I l 1.9C2' [21.5(21 1 t9.8(2) 11~7 7(2) 120.8,2) 120.2(3 ) 120.1(31 120.0(31 120,3(31

N(3)

Fig. 4. 1-Methyl-3-(5-ammo-!-benzylmUdazol-4-yl)-I,2,4-triazole, showing the atomic numbering scheme.

Structure of C13H14N6

541

Ii

I

I

9

o 0

~

0

0

0

0 0

0

~

0

E~ -4-

o

9

~ 0 0 0 II

~

0

0

I I I

~

0 ~

e~

I

~6

e~

542

Afshar el aL

provide an internal calibration standard by which to judge the accuracy of all the bond lengths determined in this structure, the planarity of the phenyl nng provides an internal standard for any least squares planes, so this data is also provided in Table 4. There is an intramolecular hydrogen bond. N(6)-H(61) - - 9N( 11L essentially in the plane of the imidazol-4-yl-triazole ring system. This is one branch of a bifurcated hydrogen bond. the other branch of which is intermolecular, N(6)-H(61) 9 9 9 N(8) at 1 ~ x , y , z . The other hydrogen atom on the exocyclic amino group is involved in a simple intermotecular hydrogen bond. N(6)H(62) ' 9 9 N(3) at 1 -,- x. y, z. The detailed geometry of these hydrogen bonds is given in Fig. 5. Apart from the hydrogen bonding, the mmrmolecular interactions stabilizing the crystal are primarily hydrophobic, and include ~r-Tr stacking of the irnidazol-4-yl-triazole ring systems and the phenyl nngs across centers of inversion. All intermolecular interactions are shown in the stereo packing diagram (Fig. 6).

N(6) ,,, N(3) 3.093 N(6) ,,, N(8) 3.085 N ( 6 ) - , , N(11) 2.992 "7"1I 8"6,'116.5

~

-

-

~

" - -

/ 2.11,5

13o.8

165.5

1 1 0 " 2 ' ~ 119.6

N(6) """ H(61) N(6) --- H(62)

1.00 1,00

Fig. 5. Hydrogen bonding geometry (only selected portions of two molecules are shown).

Structure of C13HI4N6

543

7X..

/

/

,

/

:

Fig. 6. Stereo packing diagram.

Acknowledgments This work was supported by NIH research grants GM21589 (H.M.B), CA06927 (H.M.B), RR05539 (H.M.B), and CA36154 (R.S.H.). We wish to thank Dr. S. L. Ginell for helpful discussions.

References Cromer, D. T., and Liberman, D. (1970) J. Chem. Phys. 53, 1891-I898. Cromer, D. T., and Mann, J. B. (1968) Acta Crystallogr. A 24, 321-324. Hosmane, R. S., Bhan, A., and Rauser, M. E. (1986) Heterocycles 24, 2743-2748; Bhan, A., and Hosmane, R. S. (1986) Abstr. Am. Chem. Soc. 192nd Natl. Mtg. (American Chemical Society, Washington, D.C.), p. ORGA 184. Imernational Tables for X-Ray Crystallography (1983) Vol. III (Kynoch Press, Birmingham), pp. 270, 276. Lister, J. H. (1971) Fused Pyrimidines. Part lI: Purines (Wiley-Interscience, New York), p. 314 and references therein. Main, P., Fiske, S. J., Hull, S. E., Lessinger, L., Germain, G., DeClercq, J.-P., and Wootfson, M. M. (1980) MULTAN80, a system of computer programs for the automatic solution of crysta! structures from X-ray diffraction data (University of York, England). Sheldrick, G. M. (1976) SHELX-76, a program for crystal structure determination (University of Cambridge, England). Stewart, R. F., Davidson, E. R., and Simpson, W. T. (t965) J. Chem. Phys. 42, 3175-3187.

British Lending Library Division Supplementary Publication No. 67013 contains 9 pages of structure factor tables.