J. Chem. Research (S), 1997, 42–43 J. Chem. Research (M), 1997, 0401–0429 Stereochemistry and X-ray Crystal Structure of ‘Pyrrole Trimer’: Synthesis of cis-2,5-Di(pyrrol-2-yl)pyrrolidine (cis Pyrrole Trimer) and X-ray Crystal Structure of cis- 1-(4-Methylphenylsulfonyl)-2,5-di(pyrrol-2-yl)pyrrolidine (Monotosyl cis Pyrrole Trimer) Yuekun Zhao, Roy L. Beddoes and John A. Joule* Chemistry Department, The University of Manchester, Manchester M13 9PL, UK Pyrrole reacts with aqueous hydrochloric acid at 0 °C to give a 2 :1 mixture of trans(1) and cis(2) isomers of 2,5-di(pyrrol- 2-yl)pyrrolidine which, by conversion into a mixture of the corresponding monotosyl derivatives then exposure of these to sodium hydroxide under phase-transfer conditions, are converted completely into the tosyl derivative (4) of the cis isomer, from which the tosyl group can be cleaved to produce pure cis ‘pyrrole trimer’.‘Pyrrole trimer’ [2,3-di(pyrrol-2-yl)pyrrolidine], produced in moderate yield by the brief treatment of pyrrole with 20% aqueous hydrochloric acid at 0 °C, was first described in 18881 but its structure was not established until 1957.2 Later it was assigned3 the stereochemistry shown in 1 on the grounds of the isolation of trans-pyrrolidine-2,5-dicarboxylic acid from oxidative degradation,4 though this work has never been described in full.Requiring the cis isomer 2, we sought confirmation of the trans stereochemistry assigned to pyrrole trimer thinking that it would be necesary to bring about isomerisation.Careful TLC and 1H NMR (in CDCl3) analyses seemed to show that ‘pyrrole trimer’ was homogeneous; however a small crystalline sample was subjected to X-ray analysis from which it was clear (Fig. 1) that the unit cell contained three molecules, two trans, but one cis. Careful HPLC then confirmed the composition of the material and allowed separation of small quantites of the two pure stereoisomers. The mixture of isomers reacted with toluene-p-sulfonyl chloride in the presence of diisopropylethylamine giving a 2:1 mixture of mono-tosyl derivatives, 3 and 4, in which only the central, pyrrolidine nitrogen had reacted.When the mixture of monotosyl derivatives was treated with sodium hydroxide under phase-transfer conditions, quantitative conversion into the pure cis isomer 4 took place. We envisage this as an equilibration, occurring via reversible pyrrole- N·H deprotonation then C·NTs cleavage (Scheme 1).An X-ray crystal-structure determination (Fig. 2) on 4 showed that the two pyrrole rings on one side of the pyrrolidine are well away from the tosyl substituent on the opposite side providing an explanation for the thermodynamic preference for the cis isomer. To conclude the synthesis of pure cis pyrrole trimer 2, sodium–ammonia treatment cleanly and quantitatively converted 4 into 2. 42 J. CHEM. RESEARCH (S), 1997 *To receive any correspondence (e-mail: j.a.joule@man.ac.uk).Fig. 1 ORTEP plot of the structure of ‘pyrrole trimer’, 1+2 Scheme 1 Fig. 2 ORTEP plot of the structure of the monotosyl cis pyrrole trimer 4X-ray Crystallography.·Measurements were made on a Rigaku AFC5R diffractometer with graphite-monochromated MoKa radiation. Data were collected at 19�1 °C. Structures were solved by direct6 (1+2) and heavy-atom Patterson7 methods and expanded using Fourier techniques.8 For 1+2, non-hydrogen atoms were refined anisotropically; for 4, some were refined anisotropically and others isotropically.Hydrogen atoms were included but not refined. Data for 1+2.·Crystal size 0.60Å0.13Å0.13 mm; 5992 unique (Rint=0.021) reflections in the 6389 collected. An empirical absorption correction, using the program DIFABS,9 was applied resulting in transmission factors from 0.80 to 1.00. The data were corrected for Lorentz and polarization effects. A correction for secondary extinction was applied (coefficient=0.13510Å10µ4).The final cycle of fullmatrix least-squares refinement was based on 2397 observed reflections [Ia3.00s(I)] and 406 variable parameters and converged (largest parameter shift was s0.01 times its esd) with unweighted and weighted agreement factors of R=0.089 and Rw=0.064. The standard deviation of an observation of unit weight was 3.31. The weighting scheme was based on counting statistics and included a factor (p=0.005) to downweight the intense reflections.Crystal data for 1+2. Clear, prismatic, primitive monoclinic, Mr 603.81; V=3272(1) Å3; a=9.226(3), b=18.901(3), c=18.786(3) Å, b=92.67(2)°; space group P21/n (No. 14); Z=4; Dcalc=1.23 g cmµ3; F(000)=1296; h, 1 to 10; k, 0 to 22; l, µ22 to 22. Data for 4.·Crystal size 0.13Å0.28Å0.48 mm; 1826 reflections were collected. The data were corrected for Lorentz and polarization effects. The final cycle of full-matrix least-squares refinement was based on 1044 observed reflections [Ia3.00s(I)] and 118 variable parameters and converged (largest parameter shift was s0.01 times its esd) with unweighted and weighted agreement factors of R=0.061 and Rw=0.055.The standard deviation of an observation of unit weight was 3.06. The weighting scheme was based on counting statistics and included a factor (p=0.008) to downweight the intense reflections. The maximum and minimum peaks on the final, difference Fourier map corresponded to 0.33 and µ0.28 e ŵ1, respectively. Crystal data for 4.Clear, prismatic, primitive orthorhombic, Mr 355.45; V=1762.4(6) Å3; a=17.340(4), b=12.302(2), c=8.262(2) Å; space group Pnma (No. 62); Z=4; Dcalc=1.34 g cmµ3; F(000)=752; h, 0 to 20; k, 0 to 14; l, µ9 to 0. Y. Z. is an EPSRC-funded post-doctoral assistant: we thank the EPSRC for their support for this work and the SERC for funds for the purchase of the Rigaku AFC-5R diffractometer. Techniques used: IR, 1H NMR, mass spectrometry, X-ray crystallography References: 9 Schemes: 1 Tables 1–3: For ‘pyrrole trimer’, 1+2: positional parameters and B(eq); intramolecular distances (non-hydrogen atoms); intramolecular bond angles (non-hydrogen atoms) Tables 4–6: For monotosyl cis pyrrole trimer, 4: positional parameters and B(eq); intramolecular distances (non-hydrogen atoms); intramolecular bond angles (non-hydrogen atoms) Appendix: Anisotropic displacement parameters for 1+2 and 4 Received, 19th August 1996; Accepted, 30th October 1996 Paper E/6/05759D References cited in this synopsis 1 M.Dennstedt and J. Zimmerman, Chem. Ber., 1988, 21, 1478. 2 H. A. Potts and G. F. Smith, J. Chem. Soc., 1957, 4018. 3 G. F. Smith, Adv. Heterocycl. Chem., 1963, 2, 287. 4 Personal communication from R. Huisgen and V. Vossius, quoted in ref. 3. 6 R. Miller, S. M. Gallo, H. G. Khalak and C. M. Weeks, J. Appl. Crystallogr., 1994, 27, 613. 7 F. Hai-Fu, Structure analysis programs with intelligent control, Rigaku Corporation, Tokyo, Japan, 1991. 8 DIRDIF94: P. T. Beurskens, G. Admiraal, G. Beurskenes, W. P. Bosman, R. de Gelder, R. Israel and J. M. M. Smits, Technical report of the crystallography laboratory, University of Nijmegen, The Netherlands, 1994. 9 N. Walker and D. Stuart, Acta Crystallogr., Sect. A, 1983, 158. J. CHEM. RESEARCH (S), 1996