Substituent effect on the energy barrier for σ-bond formation from π-single-bonded species, singlet 2,2-dialkoxycyclopentane-1,3-diyls

Background: Localized singlet diradicals are in general quite short-lived intermediates in processes involving homolytic bond-cleavage and formation reactions. In the past decade, long-lived singlet diradicals have been reported in cyclic systems such as cyclobutane-1,3-diyls and cyclopentane-1,3-diyls. Experimental investigation of the chemistry of singlet diradicals has become possible. The present study explores the substituents and the effect of their substitution pattern at the C(1)–C(3) positions on the lifetime of singlet octahydropentalene-1,3-diyls to understand the role of the substituents on the reactivity of the localized singlet diradicals. Results: A series of singlet 2,2-dialkoxy-1,3-diaryloctahydropentalene-1,3-diyls DR were generated in the photochemical denitrogenation of the corresponding azoalkanes AZ. The ring-closed products CP, i.e., 3,3-dialkoxy-2,4-diphenyltricyclo[3.3.0.02,4]octanes, were quantitatively obtained in the denitrogenation reaction. The first-order decay process (k = 1/τ) was observed for the fate of the singlet diradicals DR (λmax ≈ 580–590 nm). The activation parameters, ΔH‡ and ΔS‡, for the ring-closing reaction (σ-bond formation process) were determined by the temperature-dependent change of the lifetime. The energy barrier was found to be largely dependent upon the substituents Ar and Ar’. The singlet diradical DRf (Ar = 3,5-dimethoxyphenyl, OCH2Ar’ = OCH2(3,5-dimethoxyphenyl)) was the longest-lived, τ293 = 5394 ± 59 ns, among the diradicals studied here. The lifetime of the parent diradical DR (Ar = Ph, OCH2Ar’ = OCH3) was 299 ± 2 ns at 293 K. Conclusion: The lifetimes of the singlet 1,3-diyls are found to be largely dependent on the substituent pattern of Ar and Ar’ at the C(1)–C(3) positions. Both the enthalpy and entropy effect were found to play crucial roles in increasing the lifetime.


Scheme 1:
Alkoxy group effect on the lifetime of π-single-bonded species DR.
In the present study, the effect of the bulky 3,5-dimethoxyphenyl group substituent was investigated on the lifetime of the localized singlet diradicals. Thus, the aryl substituent was introduced at C(1), C(2), or/and C(3) positions of the diradicals DRd-g, and the substituent effects on the lifetime of the singlet diradicals were compared with the lifetime of a phenyl-groupsubstituted diradical DRc and the parent diradical DRa. The laser flash photolysis technique was used for the generation of DRc-g from the corresponding azoalkanes AZc-g (Scheme 2).
The steady-state photolyses of AZc-g in benzene solution were performed with a Xenon lamp (500 W) through a Pyrex filter (hν > 300 nm). The ring-closed compounds CPc-g were quantitatively obtained in the denitrogenation reaction (Scheme 2). The quantum yields of the denitrogenation of AZc-g were determined to be ≈0.90 by comparison with those reported for similar azoalkanes [33]. The quantitative formation of CPc-g and the high quantum yield of the denitrogenation process suggest the clean generation of DRc-g in the photoirradiation reaction of AZc-g.
Detection of singlet diradicals DRc-g. The detection of singlet diradicals DRc-g was examined by the photochemical denitrogenation of azoalkanes AZc-g in a glassy matrix of 2-methyltetrahydrofurane (MTHF) at 80 K, [AZ] ≈ 4 × 10 −3 mol/L, and by the laser flash photolysis experiments of AZc-g at room temperature in benzene solution. First of all, the MTHF matrix solution of AZ was irradiated with a 500 W Xenon lamp through a monochromator (λ irr = 360 ± 10 nm). A strong absorption band, which corresponds to DRc-g, was observed in the visible region at 80 K (570-590 nm, Table 1), as exemplified for the photoirradiation of AZe in Figure 2a. The strong absorption bands are quite similar to those of singlet diradicals DRa,b with λ max = 574 nm and 572 nm [1,28], respectively. The assignment of the strong band to the singlet diradical is further supported by the following facts: (a) The absorptions obtained on photolysis in a MTHF glass were thermally persistent at 80 K and resembled that of the transient absorption spectra in solution (for example, DRe, λ max = 590 nm, Figure 2b); (b) the species were ESR-silent in the MTHF-matrix at 80 K; (c) the lifetime of the transient was insensitive to the presence of molecular oxygen (decay trace at 580 nm, Figure 2c); and (d) the activation parameters (Table 1) are similar to those for the decay process of DRa, in particular, the high (ca. 10 12 s −1 ) pre-exponential Arrhenius factors (logA) are indicative of a spin-allowed reaction to the ring-closed products CPc-g [34]. Lifetime of singlet diradicals DRc-g and activation parameters for the ring-closing reaction. The decay traces of the intermediary singlet diradicals DRc-g at 293-333 K were measured in a benzene solution by the laser flash photolysis technique (λ exc = 355 nm). The lifetime (τ = 1/k) was determined by the first-order decay rate constants (k) of DRc-g at 580 nm, e.g., Figure 2c for DRe. As shown in Table 1, the lifetime of the singlet diradical was largely dependent on the substituents Ar and Ar'. The activation parameters, ΔH ‡ , ΔS ‡ , E a , logA, were determined from the Eyring plots and Arrhenius plots, which were obtained from the temperature-dependent change of the lifetime (Table 1). For comparison, the lifetime of diradical DRa (Table 1, entry 1) was also measured under similar conditions, and determined to be 299 ns at 293 K. The obtained lifetime was nearly the same as that obtained previously by us (292 ns) [28].
The lifetime of DRc (Ar = Ar' = Ph) was found to be 1305 ns at 293 K (Table 1, entry 2), which was ca. 4.5 times longer than the parent DRa. On introduction of a 3,5-dimethoxyphenyl ring at C(2) position of the 1,3-diradical, i.e., DRd (Ar = Ph, Ar' = 3,5-dimethoxyphenyl), a further increase of the lifetime at 293 K was observed to be 1933 ns (Table 1, entry 3). The result clearly indicates that the steric bulkiness plays an important role in increasing the energy barrier for the ring-closing reaction. Indeed, the activation enthalpy (ΔH ‡ = 36.6 kJ mol −1 , Table 1, entry 3) for DRd was found to be higher than that for DRa (ΔH ‡ = 32.7 kJ mol −1 , Table 1, entry 1). Interestingly, the effect of an aryl group substituent at C(1) and C(3) positions on the lifetime was found to be larger than that at C(2); compare the lifetime of DRe (4001 ns, Ar = 3,5-dimethoxyphenyl, Ar' = Ph, Table 1, entry 4) with that of DRd (1933 ns, Table 1, entry 3). When the 3,5-dimethoxyphenyl group was introduced at all of the C(1), C(2), and C(3) positions, the lifetime of the diradical DRf (ΔG ‡ = 42.2 kJ mol −1 , Table 1, entry 5) was dramatically increased to 5394 ns at 293 K. The activation entropy (ΔS ‡ = −27.8 and −19.4 J mol −1 , Table 1, entries 4 and 5) also plays an important role in increasing the lifetime of the singlet species. A much shorter lifetime was found for the diradical DRg (Ar = 3,5-dimethoxyphenyl, Ar' = H). Thus, the introduction of the bulky substituents is needed at all positions C(1)-C(3) of the 1,3-diradicals to increase the lifetime. The repulsive steric interactions of the Ar group with the Ar' group are suggested to play important roles in increasing the energy barrier of the reaction from the diradicals to the ring-closed compounds CP. The results clearly indicate that the substituent effect using the sterically bulky group is effective to prolong the lifetime of the singlet diradicals.

Conclusion
We have succeeded in generating long-lived singlet diradical species DRc-g, τ 293 = 580-5394 ns, which were much longerlived species than DRa (τ 293 = 299 ns). It was found that the lifetimes are largely dependent on the substituent pattern of Ar and Ar' at the C(1)-C(3) positions of the 1,3-diyls. Thus, both the enthalpy and entropy effect were found to play crucial roles in increasing the lifetime.

Experimental
All reagents were purchased from commercial sources and were used without additional purification, unless otherwise mentioned. Azoalkanes AZc-g were prepared according to the methods described previously (Scheme 3) and were isolated by silica gel column chromatography and GPC column chromatography. 1 H and 13 C NMR spectra were reported in parts per million (δ) by using CDCl 3 or C 6 D 6 as internal standards.
Assignments of 13 C NMR were carried out by DEPT measurements. IR spectra were recorded with a FTIR spectrometer. UV-vis spectra were taken by a JASCO V-630 spectrophotometer. Mass-spectrometric data were measured by a Mass Spectrometric Thermo Fisher Scientific LTQ Orbitrap XL, performed by the Natural Science Center for Basic Research and Development (NBARD), Hiroshima University.
Preparation of diazenes AZc-g 3,6-Diaryl-1,2,4,5-tetrazine 1. 3,6-Diphenyl-1,2,4,5-tetrazine was purchased and directly used. The preparation of 3,6-(3,5-dimethoxyphenyl)-1,2,4,5-tetrazine (Ar = 3,5-dimethoxyphenyl) is as follows: In a 50 mL round-bottom flask, benzonitrile (3.7 g, 22.7 mmol) was dissolved in 10 mL of absolute ethanol. Hydrazine (3.6 mL, 90 mmol) and sulfur (0.43 g, 13.5 mmol) were quickly added, and the solution was stirred at room temperature for 1 h and then heated under reflux for 3 h. The remaining orange cake was solidified further in an ice bath. The solid was vacuum filtered, and washed with cold ethanol (3 × 10 mL) giving the crude dihydrotetrazine. The crude orange solid was then placed in a 50 mL beaker and dissolved in 20% acetic acid (15 mL) and 10 mL ether at room temperature with stirring. An aqueous solution of 10% NaNO 2 (20 mL) was added to the solution in an ice bath. The immediate purple cloudiness signifies the completion of the reaction, as well as the evolution of brown nitric oxide gas. Vacuum filtration and washing with hot methanol (3 × 10 mL) gave the tetrazine as a red solid (3.07 g, 81%

4,4-Dimethoxy-3,5-bis(3,5-dimethoxyphenyl)pyrazole (3g).
To a solution of 1,3-bis(3,5-dimethoxyphenyl)-2,2dimethoxypropane-1,3-dione (2.8 g, 6.92 mmol) in chloroform (10 mL) was added dropwise NH 2  General procedure. To a solution of cyclopentadiene (1 mL) and pyrazole (2 mmol) in CH 2 Cl 2 (2 mL) was added dropwise trifluoroacetic acid (1 mmol) in an ice bath under nitrogen. The reaction was traced by TLC analysis. After stirring for about 15 min, the reaction was quenched with 10% aq NaHCO 3 until the pH of the solution reached 8. After washing with water and brine, the organic phase was dried with MgSO 4 , then filtered and concentrated. The [4 + 2] cycloadduct was dissolved in benzene (2 mL), and 5 mg of PtO 2 was added as a catalyst. The mixture was stirred under a hydrogen atmosphere for 24 h at room temperature. After stirring, the catalyst was removed by filtration over Celite, and the solvent was evaporated under reduced pressure. The product was purified by column chromatograph to give the product as colorless liquid (ca. 60%). The endo configuration was determined by NOE measurements.