E / Z Conformational equilibrium of N-substituted 2 H -yran-2-imines

Prevailing Z -conformations were found in N -aryl(or alkyl)substituted 2 H -pyran-2-imines 4a - e on the NMR time scale. Conformational assignment was based on 1 H and 13 C NMR chemical shifts as well as on induced shifts by LSR. The barrier of E / Z interconversion ( ) was calculated from variable temperature


Results
In the series of pyran 2-imines 4a-e, the 1 H and 13 C NMR spectra recorded at room temperature displayed two sets of signals (except for 4c), assigned to E, Z-conformers at the exocyclic C,N bond.With E/Z ratio far from 50:50 for all five compounds investigated, the two sets of chemical shifts are readily separated.For each conformer, the individual 1 H and 13 C resonances were unambiguously assigned using 2D experiments (COSY, HETCOR and COLOC) and are displayed in Tables 1 and 2 The stereochemical assignment was performed as follows: in each pair of conformers, the upfield signal for the C-3 atom of the pyran ring (Table 2) was assigned to E-conformation, according to the general trend for α-carbons in compounds with C,N double bond. 19The chemical shift difference between C-3 signal in E-4 and in Z-4 is almost constant in the series (8−10 ppm) and agrees with the difference between C-3 and C-5 signals in the pyran 4-imines 5, 16 accounting for the same effect of the nitrogen substituent R. The H-3 signals (Table 1) supported this assignment: in the N-phenyl substituted imines 4a, 4b this signal is shifted upfield in the E-conformer by 0.27 ppm, while much smaller differences (up to 0.06 ppm) are observed in the E, Z pairs of the N-alkyl substituted imines 4d, 4e.In E-4a (4b) the H-3 atom is shielded by the aromatic ring on nitrogen, non-coplanar with the pyran ring because of steric crowding.
An independent conformational proof was provided by the shifts induced with lanthanide shift reagents (LSR).In Fig. 1  The results for Z-4b (prevailing conformer) were rationalized by assuming the paramagnetic center of the shift reagent complex to be on the axis of the lone-pair orbital on nitrogen. 20This renders the substituent on nitrogen and the H-3 atom as closest neighbours to the lanthanide, in agreement with these groups having the highest MIS values.The results for the minor E- conformer were less clear-cut, meaning that the geometry of complexation is not obvious (contribution of the oxygen lone-pairs is not precluded).Still, the MIS value for the ortho- hydrogens of 6-Ph substituent are higher in E-4b than in Z-4b, suggesting reversal of the nitrogen lone-pair.Similar observations were made with E, Z-4d as substrate.In conclusion, the results obtained with LSR are meaningfully correlated with the conformational assignment based on chemical shift values.
The E, Z equilibrium composition in CDCl 3 (see Table 1) is strongly biased towards Zconformer, compound 4c appearing as single Z-conformer.
The free energy of activation at the coalescence temperature ( ) for E, Zinterconversion was determined from variable temperature ≠ ∆ Tc G 1 H-NMR spectra (Table 3).The spectral parameters of the signals monitored for coalescence given in the table were measured at 20ºC.Corrections for temperature dependence of δν values have been applied.The precision in coalescence temperature T c measurement was ± 2K.The rate constants k were evaluated from equations for unequally populated doublets 21a and the values were calculated from Eyring equation, assuming a value of unity for the transmission coefficient K.   a Experimental error range ±0.5 The data in Table 3 showed that measured at two different temperatures were quite close within experimental errors range, and also that it was not sensitive to changing the solvent from C ≠ ∆G 5 D 5 N to C 2 D 6 SO.The 4-R' substituent had negligible effect on .However, the nitrogen substituent was found to have a significant effect: the barrier is higher by 11 kJ•mol the N-iPr imine 4d than in the N-Ph analog 4b.

Discussion
The stereochemical analysis gave prevailing 4-Z conformation by two independent procedures.The consistency of the data in the series is reassuring, particularly since the stereochemical analysis in opened-chain related systems was not devoid of controversy.Imidate esters 8 had been initially reported 22 as stable Z-isomers ), but later studies found much lower barriers for E, Z interconversion and established prevailing E-conformation. 23,24This was explained by preference in antiperiplanar E-ap conformation, in which partial cancellation of dipoles occurs (Scheme 4).It is interesting to note that E-ap geometry was indeed found in imidates 8 by NOE experiments. 25Obviously, this cannot be the driving force in the cyclic pyran 2-imines 4.
For explaining the prevalence of Z-4 conformation, both electronical and steric factors are taken into account.The electrostatic repulsion between nitrogen and oxygen lone pairs should disfavour the E-conformer.The steric effect operates in the same sense, since the interference between the substituent R on nitrogen and the H-3 atom of the pyran ring disfavours also the E- conformer.Indeed, the molar fraction of the E-conformer in CDCl 3 was 0.25 in 4a and 0.35 in 4b when R = Ph, but decreased to 0 when R was as bulky as tBu (4c).The aromatic ring may easily release the steric strain by torsion, shielding the 3-H atom as observed experimentally.
The free energy of activation of E, Z-interconversion in imines 4 decreased significanly when the substituent on nitrogen changed from alkyl to aryl, but presented negligible solvent and temperature effects (Table 3).These findings are in agreement with an "in plane" inversion (sp nitrogen) mechanism for interconversion.≠ ∆G 26 The aromatic ring on nitrogen, orthogonal to the inversion plane, gives a better conjugation with the occupied p orbital in the transition state as compared to the conjugation with the occupied sp 2 orbital in the ground state, decreasing the interconversion barrier.
Finally, a comparison with topomerization of pyran 4-imines 5, occurring by nitrogen inversion or a mechanism intermediate between inversion and rotation, 18 seems appropriate (barriers obtained in solvents of similar polarity are compared).For the N-phenyl pyran 4-imine 5 (R = Ph), the barrier in CD 3 NO 2 was found to be 81.3 ± 0.5 kJ•mol -1 .This is quite close to the barrier for Z-E isomerization found for 4a in DMSO-d 6 (80.7 ± 0.5 kJ•mol -1 ) but higher than the barrier for the reversed E-Z process (75.0 ± 0.5 kJ•mol -1 ).Since the difference between ground state and transition state in 5 and in Z-4a is the same, a destabilizing effect should operate in E-4a and the interorbital repulsion between non-bonding electrons on nitrogen and the endocyclic oxygen atom may well account for it.

Conclusions
The stereochemical analysis based on 1 H and 13 C NMR spectroscopy in the series of N-aryl (or alkyl) substituted 2H-pyran-2-imines 4a-e established, by two independent procedures, prevalence of Z-conformation.The barriers determined by ≠ ∆ Tc G 1 H-DNMR are in agreement with a planar nitrogen inversion mechanism for E, Z interconversion.To our knowledge, this is the first stereochemical description of N-substituted pyran 2-imines.

Experimental Section
General Procedures.The NMR spectra were recorded with a Varian Gemini 300 BB instrument operating at 300 MHz for 1 H and 75 MHz for 13 C.The chemical shifts are given in ppm, from TMS as internal standard.The temperature controller in the DNMR experiments was calibrated against ethyleneglycol, the measurement accuracy being ± 2K.Preparation of compounds 4a-4d were published earlier. 1

Table 1 .
1 H-NMR Data: δ H [ppm], coupling constants [Hz] and molar fraction of E, Z-4 aromatic protons in ortho-position; b obscured by signals of the prevailing conformer.
a a C q ; b 2CH-ortho; c 2CH-meta; d CH-para; e Me 3 C; f Me 3 C; g Me 2 CH; h Me 2 CH are given the molar induced shifts (MIS in ppm, slopes in representation of induced shifts vs molar ratio between LSR and substrate) for imine 4b, with Eu(fod) 3 as LSR.