Hideyuki
Tukada
Graduate School of Integrated Science, Yokohama City University, Kanazawa-ku, Yokohama, 236-0027, Japan.. E-mail: htht@yokohama-cu.ac.jp
First published on 6th January 2000
EPR observation of N-(2-methoxycarbonyl-2-yloethyl)phenylaminyl at cryogenic temperature shows that the diradical has a triplet ground state, and is thermally labile for ring-closure reaction with low and distributive activation energies centered around 3.2 kJ mol−1.
Photolysis (330 ± 20 nm) of methyl 1-phenyl-4,5-dihydro-1H-1,2,3-triazole-4-carboxylate 25 in a 1∶1 mixture of C2D5OD–CD3OD (50 mmol dm−3) was performed in an EPR cavity (X-band) with a 500 W super-high-pressure mercury lamp and glass filters at 5 K. The EPR spectrum after irradiation for 5 min persisted for at least 1 h at this temperature, and showed a set of signals characteristic of a triplet biradical (half-field signal; 0.1569 mT) with zero-field-splitting (ZFS) parameters of |D| = 3|E|6 (Fig. 1).
![]() | ||
Fig. 1 X-Band EPR spectrum of 1 obtained after UV irradiation (330 ± 20 nm) to methyl 1-phenyl-4,5-dihydro-1H-1,2,3-triazo-4-carboxylate 2 in a 1∶1 mixture of C2D5OD–CD3OD matrix for 5 min at 5 K (ν = 9.0067 GHz, power = 0.2 mW, modulation = 1 mT). Asterisk denotes doublet impurities. |
The signals at 2516 and 3953 mT were assigned to overlapping signals due to transition of Z and Y with the ZFS parameters |D/hc| = 0.06716 cm−1 and |E/hc| = 0.02239 cm−1. The D value is close to those of cyclobutane-1,3-diyls7 (|D/hc| = 0.050–0.11 cm−1) and cyclopentane-1,3-diyls8 (|D/hc| = 0.041–0.084 cm−1), and is fully consistent with a localized 1,3-diradical structure with an average distance of 0.34 nm between two unpaired electrons.8
EPR measurements were then performed on warming from 5 to 50 K in the dark. At each temperature (Ti) the signal intensity (I) stabilized within 3 min. Re-cooling from Ti to 5 K did not regenerate the initial intensity, which indicates that the biradical was irreversibly diminished by the reactions. However, no further decrease in signal intensity was observed between Ti and 5 K. The temperature dependence of the signal intensity for the annealed sample obeyed the Curie law (IT = constant) between 5 and Ti.
Consequently, the IT vs. T plot for the warming process did not become flat [line (i) in Fig. 2], whereas the plot for the post-annealed process showed a constant value [line (ii) in the Fig. 2]. The latter indicates that the triplet is the ground state of biradical 1.
![]() | ||
Fig. 2 IT vs. T plots, where I is the EPR signal intensity of the |ΔMs| = 2 transition of 1 in a 1∶1 mixture of C2D5OD–CD3OD, and T is the absolute temperature. (i) Open circles indicate a heating process from 5 to 35 K, and closed circles show the temperature dependence of the signal intensity between 5 and 18 K after annealing at 18 K. |
It seems reasonable that the initial decrease in signal intensity of the triplet can be attributed to decay through a ring-closure reaction to give methyl 1-phenylaziridine-2-carboxylate, which is the main photo-product at room temperature in solution (Scheme 1).5 Decay by hydrogen abstraction from the matrix can probably be excluded because no new signals and no increase in the doublet species (indicated as * in Fig. 1) were observed up to 50 K. At 35 K, the triplet EPR signals originating from the biradical dissappeared immediately and simultaneously.
![]() | ||
Scheme 1 |
The same thermal behaviour of 1 was observed using 2-methyltetrahydrofuran as the matrix, indicating that the hardness of the surroundings of the diradical had little influence on ring-closure.
The former behavior [line (i)] can be interpreted as decay through transition states with distributive activation energies due to matrix site effects and conformers of 1, rather than classical or tunneling processes having a single activation energy. Such a mechanism with distributive activation energies was successively analyzed by a ‘distribution slicing method’ in the case of cyclobutane-1,3-diyls by Dougherty’s group.7 Applying this method to our case (50% of 1 decayed at 16 K) gave an average activation energy, Ea = 3.2 kJ mol−1, under the assumption that A = 108 and t = 180 s, where A is a frequency factor in the Arrhenius equation and t is the slicing time. The activation energy, which possibly corresponds to that required for spin inversion (ISC), is slightly smaller than that in the case of cyclobutane-1,3-diyls (Ea = 4.0–9.7 kJ mol−1) or cyclopentane-1,3-diyls (Ea = 8–>20 kJ mol−1).
The low activation energy in the case of 1 is attributable primarily to the flexible open chain structure and secondly to the interaction between a lone-pair on the aminyl radical and the p-orbital of the counterpart carbon radical.
According to density functional theory calculations (U-B3LYP/6-13G*)9 for the biradical 1, the triplet has a (0,0) conformation (following the notation for trimethylene)3 and lies below the singlet by 1.3 kJ mol−1 (vertically; the singlet biradical has no minimum on the energy surface, and the optimization of the structure gives methyl 1-phenylaziridine-2-carboxylate). The computed energy gap ΔES-T is consistent with the result of the EPR experiment. Most of the spin populations of triplet 1 are distributed on both the formal radical centers (spin/charge = 0.63/−0.47 on the N atom, and 0.90/−0.16 on the C2′ atom), and agree with the average distance between unpaired electrons as estimated from the experimental D parameter.
In conclusion, we have been able to detect triplet diradical 1, a heteroatom analogue of trimethylenemethane, and found that the diradical is thermally labile above 5 K with low and distributive activation energies.
This journal is © The Royal Society of Chemistry 2000 |