The effects of σ_C—Se–π hyper­conjugation in 1-meth­oxy- and 1-nitro-4-[(phenyl­selan­yl)meth­yl]benzene

Phenyl selenium substituents have been observed to accelerate the rates of unimolecular solvolyses in systems having a leaving group at the β-position with respect to the selenium substituent. For example, the conformationally biased phenyl­selenyl­cyclo­hexyl tri­fluoro­acetate (3) (Scheme 1) solvolyses at a rate 10⁸ times faster than the corresponding selenium-free system (4) (White et al., 2002). The origin of this rate enhancement is believed to be from a combination of `conventional' neighbouring-group participation, in which a selenium nonbonded pair of electrons provides the nucleophile to displace the tri­fluoro­acetate leaving group, with the formation of the seleniranium ion (5), and stabilization of the inter­mediate carbenium ion (6) by σ_C—Se–π hyperconjugation. These two modes of participation can be differentiated as being examples of nonvertical and vertical participation (Hanstein et al., 1970; Lambert, 1990). Application of the variable oxygen probe to β-phenyl­selenyl-substituted esters with varying electron demand provides strong evidence for the importance of vertical participation in the ground-state structures of these esters (White et al., 2002), while calculations and structural studies on systems in which the C—Se bond inter­action with orbitals with much higher electron demand suggests that nonvertical participation becomes of increasing importance (Lim et al., 2011; Harris & White, 2013) as the electron demand increases.

As part of our continuing investigations into the structural effects of C—Se hyperconjugation we have determined the structures of 1-meth­oxy-4-[(phenyl­selanyl)methyl]­benzene, (1), and 1-nitro-4-[(phenyl­selanyl)methyl]­benzene, (2), which were prepared according to Scheme 2. In both (1) and (2), the PhSe—CH₂ bond can inter­act with the π-system of the aromatic ring; however, the strength of the resulting σ_Se—C–π hyperconjugation should differ between the electron-rich anisole ring in (1) and the electron-poor nitro­benzene ring in (2).

Reagents were obtained from Aldrich and used without further purification.

Preparation of compound (1). To a stirred solution of di­phenyl diselenide (0.22 g, 0.70 mmol) in ethanol (20 ml) was added sufficient sodium borohydride until the solution was colourless and was then treated with an ethanol solution (2 ml) of 4-meth­oxy­benzyl bromide (0.23 g, 1.67 equivalents). The resulting solution was stirred for 1 h, then diluted with water (50 ml) and the mixture extracted with di­ethyl ether (3 × 20 ml). The combined ether extracts were dried (MgSO₄) and evaporated under reduced pressure to (1) as a colourless solid (yield 0.29 g, 91%). Recrystallisation from hot ethanol afforded (1) as white blocks (m.p. 340–343 K). ¹H NMR (CDCl₃): δ 7.47 (2H, m), 7.33 (2H, d, J = 7.9 Hz), 7.28 (3H, m), 6.89 (2H, d, J = 7.9), 4.57 (2H, s), 4.09 (3H, s); ¹³C NMR (CDCl₃): δ 158.4, 133.3, 130.5, 130.0, 129.8, 127.1, 113.8, 55.2, 31.6; ⁷⁷Se NMR (CDCl₃): δ 329.4.

Preparation of compound (2). Compound (2) was prepared in 88% yield using the same general procedure as described for compound (1). Recrystallization from hot petroleum gave (2) as yellow plates (m.p. 327.5–329.5 K). ¹H NMR (CDCl₃): δ 7.64 (2H, m), 7.39 (2H, d, J = 7.5 Hz), 7.34 (3H, m), 6.96 (2H, d, J = 7.6 Hz), 3.93 (2H, s); ¹³C NMR (CDCl₃): δ 146.7, 134.2, 131.9, 129.9, 129.8, 129.3, 127.1, 113.8, 31.9; ⁷⁷Se NMR (CDCl₃): 359.1.

Crystal data, data collection and structure refinement details are summarized in Table 1. All H atoms were refined using a riding model, with C—H = 0.95 Å and U_iso(H) = 1.2U_eq(C) for aromatic, C—H = 0.99 Å and U_iso(H) = 1.2U_eq(C) for methyl­ene, and C—H = 0.98 Å and U_iso(H) = 1.5U_eq(C) for methyl H atoms.

The structure of (1) (Fig. 1) reveals that the meth­oxy substituent is coplanar with the aromatic ring [C8—O1—C1—C2 = 0.4 (3)°], a conformation which allows delocalization of the p-type lone pair of the meth­oxy O atom into the aromatic ring. The C1—O1 bond length (Table 2) is comparable to the distance of 1.369 Å obtained from the Cambridge Structural Database (CSD, Version ????; Allen 2002). This result was obtained using a search fragment consisting of a meth­oxy-substituted aromatic ring having no ortho substituents, and restricting the search to organic compounds with R factors < 5%. Of these, 5065 hits were obtained of which 5049 had a CH₃—O—C(ipso)—C(ortho) torsion angle within the range ±30°. Inter­estingly, the few structures with CH₃—O—C(ipso)—C(ortho) torsion angles in the range 60–120°, for which lone-pair delocalization is absent, had a mean C—OCH₃ bond length of 1.387 Å, which illustrates the structural effects of delocalization of the lone pair into the aromatic ring. Thus, the meth­oxy group in (1) satisfies the geometrical requirements to act as an electron-donating group. The coplanar conformation of the meth­oxy substituent, although electronically preferred, does occur at the cost of significant steric inter­actions between the meth­oxy group and the ortho H atom (H2), nonbonded inter­actions are relieved by the opening up of the O1—C1—C2 angle [124.5 (2)°] and the closing of the O1—C1—C6 angle [115.64 (19)°].

The nitro group in (2) (Fig. 2) is also essentially coplanar with the aromatic ring [O1—N1—C1—C6 torsion angle = -5.9 (5)°]. The C1—N1 bond length in (2) (Table 3) is essentially identical with that obtained from a similar CSD search for nitro-substituted aromatic compounds, i.e. 1.467 Å. This search resulted in 4450 hits, which, with one exception the nitro group was coplanar with the aromatic ring. This conformation allows for delocalization of electron density from the aromatic ring onto the nitro group. Thus, (2) contains an electron-deficient aromatic ring substituted with a phenyl­selenyl­methyl substituent. The ipso-nitro angle [C2—C1—C6 = 122.4 (3)°] shows a deviation from the idealized value of 120° which is consistent with the expected structural effects of an electron-withdrawing group (Domenicano et al., 1975). Inspection of the aromatic C—C distances within the substituted C1–C6 aromatic ring in (1) and (2) reveals no significant differences in the pattern of bond lengths which may have arisen due to different electronic characters of the meth­oxy substituent in (1) and the nitro substituent in (2).

For both (1) and (2), the C7—Se1 bond is close to orthogonal to the plane of the C1–C6 aromatic ring; the Se1—C7—C4—C5 torsion angle is 96.4 (2)° in (1) and 76.4 (3)° in (2). This conformation which is no doubt favoured due to steric effects, also satisfies the stereoelectronic requirements for stabilising σ_C—Se–π hyperconjugation between the C—Se bond and the aromatic ring. This inter­action which can be represented by the resonance forms illustrated in Scheme 3, might be expected to increase the Se1—C7 bond length and decrease the C7—C4 bond length, which gains some double-bond character as a result of this inter­action, furthermore hyperconjugation is expected to be stronger in (2) which has an electron deficient ring than in (1) which is electron rich. Differences in the degrees of hypercongation between the C—Se bond and the aromatic ring in (1) and (2) are suggested by the ⁷⁷Se NMR chemical shifts; this is 329.4 p.p.m. in (1), while in (2) this is significantly deshielded at 359.1 p.p.m., suggesting some transfer of electron density from the Se atom. However, comparison of the Se1—C7 and C7—C4 bond lengths between the two structures reveals that these parameters do not differ significantly, thus any structural effects arising from the differing degrees of σ_C—Se–π hyperconjugation in (1) and (2) are clearly too small to be dete­cta­ble by this technique.