The reaction of o -phenylenediamine with α , β -unsaturated carbonyl compounds

The structures of the products obtained by the reaction of o -phenylenediamine and two isomeric chalcones have been identified as 1,5-benzodiazepines. A 1 H, 13 C and 15 N NMR study in solution combined with B3LYP/6-31G** calculations allowed to determine the conformations present in solution.


Introduction
The reaction of binucleophiles like o-phenylenediamine 1 with α,β-unsaturated carbonyl compounds 2 can afford seven-3, six-4 and five-membered rings 5 (that in some cases can be oxidized to benzimidazoles 6) (Scheme 1).Benzodiazepines 3 correspond to the attack on the CO and the terminal carbon of the olefin, quinoxalines 4 correspond to the attack on the CO and the α carbon of the olefin, and the benzimidazole derivatives 5 to a double attack on the carbonyl group.o-Aminothiophenol has been used instead of 1.The literature abounds in assignment errors that have been solved only recently by NMR spectroscopy and X-ray crystallography.An example of this is compounds 3 and 4 1 and the corresponding sulfur derivatives of 3 and 4 (NH replaced by S). 2,3 Vinyl (R' = H) and ethenyl (like styryl, R' = Ph) benzimidazoles 6 are usually prepared using other ways. 4,5he problem of differentiating between structures 3 and 4 has conclusively been solved by a rigorous analysis of their NMR spectral characteristics. 1,2However, in this report, a case that shows interesting variations in the NMR characteristics making the assignment more complicated will be described.

Scheme 2
The reaction affords in both cases only one compound.The main features of 1 H NMR spectra in CDCl 3 of the isolated compounds are reported in Figure 1.It appears that the spectrum of the 4-NO 2 derivative is consistent with a 1,5benzodiazepine (or thiazepine) 1,2 thus it corresponds to 8a.However, that of the 2-NO 2 derivative is different, not only the chemical shifts but also the Karplus-type coupling constants are clearly different.Our first hypothesis is that in the latter case the compound could have the structure 9b.To verify this assumption, the 13 C and 15 N NMR spectra of 8a and that of the unknown 2-NO 2 derivative b have been recorded (Table 1).It is clear that the chemical shift differences ∆δ between benzodiazepines 8 and quinoxalines 9 reported in the literature 1 (similarly for benzothiazepines), 2 are much larger (see for instance N5) than those observed between the 4-NO 2 a and 2-NO 2 b derivatives.Highlighted in red are the most important differences related to the position of the nitro group which affect δC3, δC4 and the  Following the assignment of both 8a and 8b 1,5-benzodiazepines, the differences in 1 H NMR data (Figure 1) are next given in Table 2.The assignments of Tables 1 and 2 are based on ( 1 H-1 H) gs-NOESY experiments as well as on gs-HMQC and gs-HMBC heteronuclear ( 1 H-13 C and 1 H-15 N) correlations..00 (d), 3 J H3"H4" = 8.1, 4 J H3"H5" = 1.4 H4" (NO 2 ) 7.43 (ddd), 3 J H4"H5" = 7.7, 4 J H4"H6" = 1.5 H5" 8.19 (m) 7.52 (ddd), 3 J H5"H6" = 7.9 H6" 7.61 (m) 7.83 (dd) a J = 1.7 Hz The analysis of the AMX system corresponding to the protons at positions 3 and 4 leads to the couplings represented in Figure 1 (the experimental spectra are reported in Figure 2).
The geminal coupling constants are not useful but the two vicinal coupling constants, through the Karplus relationship, 6 allow the inference of certain conclusions about the conformational changes introduced in the seven-membered ring by the 2-NO 2 group.Since the ethane fragment belongs to a seven-membered ring and there is an N atom in one of the extremities, we have modified the original Karplus equation to fit our values: 3 J HH (Hz) = 7.76 cos 2φ -1.10 cos φ + 1.40 (1) Using eq.[1], the compounds should have the conformation represented in Figure 3.We have assumed a perfect ethane geometry with angles of 60º and 180º.In the case of the 4-NO 2 derivative 8a, angles of 170º and 50º (with regard to H M a gauche -10º) led to couplings of 10.0 Hz (instead of 9.5 Hz) and 3.9 Hz (instead of 3.3 Hz).In the case of the 2-NO 2 derivative 8b, there is not a single conformation that could explain the measured coupling constants.We have to assume two conformations of similar energy in rapid interconversion leading to average signals.With regard to H M , the H X atom occupies in one case a gauche conformation (+15º) and in the other an anti conformation (-15º).The average couplings would be 4.5 Hz (instead of 4.3 Hz) and 5.7 Hz (instead of 6.3 Hz).The validity of the conformations assigned to derivatives 8a and 8b were tested by carrying out Density Funtional Theory (DFT) calculations at the B3LYP/6-31G** level (see experimental part) corresponding to the seven-membered ring inversion that, due to the 10,11fused phenyl ring and the exocyclic double bond on C2, are described as affecting essentially C3 which can be up or down (methylene flip).
In the case of the 4-NO 2 derivative 8a, there are two minimum energy conformations with relative values of 0.0 and 4.7 kJ mol -1 , thus, we can assume that only the most stable one (Figure 4) is present in solution.This structure has HCCH dihedral angles of 179.7º and 61.4º (the C2-C3-C4-C1" angle amounts to 176.7º), close to those calculated in Figure 3 (170º and 50º, respectively).The 2-NO 2 derivative 8b presents two conformations of similar energy, the 8b1 (E rel = 0.0 kJ mol -1 ) and the 8b2 (E rel = 2.5 kJ mol -1 ).The 8b1 (Figure 5) has HCCH dihedral angles of 65.4º and 52.3º (C2C3C4C1" = 71.3º)while those of 8b2 (Figure 6) are 177.1ºand 56.9º (C2C3C4C1" = 176.0º).These angles can be compared with those of Figure 3  Other relevant features of these conformations are in 8b1 the proximity of the nitro group to H4 and to one of the H3 protons; on the other hand, in conformation 8b2 the nitro group is close to H4 and to the N5-H5.A mixture of both conformations can explain why δC3, δC4 and the 1 J NH coupling of N5 are the most affected properties in Table 1.It also provides an explanation why in Figure 1, the H X proton in 8b is deshielded compared with that in 8a (Table 2).

Experimental Section
General Procedures.Melting points were determined in open capillaries and are uncorrected. 1H NMR spectra for analytical purpose were recorded on a Bruker 300 MHz instrument using TMS as an internal standard.IR spectra were recorded on a Buck Scientific IR M-500 spectrophotometer.Elemental analyses were carried out in a Perkin Elmer-2400 instrument and mass spectra were recorded on Kratos MS-50 mass spectrometer.Most of the common chemicals such as dehydroacetic acid (DHA), aldehydes, and o-phenylenediamine, were obtained from commercial suppliers.3-Cinnamoyl-4-hydroxy-6-methyl-2-pyrones (chalcone analogs of DHA, 7a-b) were prepared according to literature procedure. 7

General method
To a solution of 7a (0.602 g, 2 mmol) in ethanol (30 ml) a few drops of piperidine and ophenylenediamine (0.21 g, 2 mmol) were added.The mixture was heated under reflux for 3-4 h and then AcOH (1 ml) was added.Refluxing was continued for another 3-4 h.About half of the solvent was distilled off under reduced pressure and the oily residue was allowed to stand at room temperature overnight.The crystalline solid product 8a thus separated was filtered, washed with cold aqueous ethanol (2-3 ml, 50: 50 by v/v) and dried.Compound 8b was prepared similarly starting from 7b.
Selected parameters for ( 1 H-1 H) gs-NOESY were spectral width 8000 Hz, the acquisition data size was 1024 points and 16 transient was accumulated per increment, with a 1 s relaxation delay, 850 ms for the mixing time, for a total of 256 experiments, data processing using zero filling in the F1 domain and shifted sine-bell apodization of factor 0 in both dimensions.2D ( 1 H-13 C) gs-HMQC, ( 1 H-13 C) gs-HMBC and ( 1 H-15 N) gs-HMBC, were acquired and processed using standard Bruker NMR software and in non-phase-sensitive mode.Gradient selection was achieved through a 5% sine truncated shaped pulse gradient of 1 ms.Selected parameters for ( 1 H-13 C) gs-HMQC and gs-HMBC spectra were spectral width 3000 (gs-HMQC) or 8000 (gs-HMBC) Hz for 1 H and 12.0 kHz for 13 C, 1024 x 256 data set, number of scans 2 (gs-HMQC) or 4 (gs-HMBC) and relaxation delay 1s.The FIDs were processed using zero filling in the F1 domain and a sine-bell window function in both dimensions was applied prior to Fourier transformation.In the gs-HMQC experiments GARP modulation of 13 C was used for decoupling.Selected parameters for ( 1 H-15 N) gs-HMBC spectra were spectral width 8000 Hz for 1 H and 12.5 kHz for 15 N, 1024 x 256 data set, number of scans 4, relaxation delay 1s, 40 ms delay for the evolution of the 15 N-1 H long-range coupling.The FIDs were processed using zero filling in the F1 domain and a sine-bell window function in both dimensions was applied prior to Fourier transformation.Density funtional theory (DFT) calculations.The optimization of the structures of all compounds discussed in this paper was carried out at the hybrid B3LYP/6-31G** level 9,10 with basis sets of Gaussian type functions using Spartan '02 for Windows. 11 Scheme 1

Table 2 .
1H chemical shifts (δ in ppm) and 1 H-1 H coupling constants (J in Hz) of compounds 8a and 8b in CDCl 3