The H-1 nuclear magnetic resonance spectra of benzyl silane and benzyl trichlorosilane, obtained in CS2 and benzene-d6 solutions, are analyzed. The long-range coupling constants between the methylene and para ring protons are used to derive apparent twofold barriers about the Csp2-Csp3 bonds of 7.4 +/- 1.6 and 8.1 +/- 1.1 kJ/mol for the silane and the trichlorosilane, respectively. These are higher than that for ethylbenzene and are attributed mainly to the stabilization of the perpendicular conformer, that with the C-Si bond in a plane perpendicular to the phenyl plane, by sigma-pi conjugation (hyperconjugation) of the C-Si bond and the pi electron system. Molecular orbital computations confirm the predominantly twofold nature of the internal barrier in benzyl silane and also for benzyl germane and stannane. The calculated barriers for the silane derivatives are rather higher than the experimental values. The computed barriers have magnitudes that appear to change with X in much the same order as do the hyperconjugative interactions deduced in other ways for CH2X(CH3)3 groups (X = Sn, Ge, Si, C). The angles CCX in benzyl-X (X = CH3, SiH3, SiCl3, GeH3, SnH3) are all computed to decrease smoothly as sin2-psi, where psi is the angle by which the C-X bond twists out of the phenyl plane.

The H-1 nuclear magnetic resonance spectrum of 2-phenyl-1,3-dithiane, as a dilute solution in a CS2-C6D12-TMS solvent mixture at 300 K, is analyzed to yield 8 chemical shifts and 22 distinct coupling constants, (n)J(H,H), n = 2-6. The coupling constant between H-2 and the para proton indicates, first, that the bisected conformer (phenyl plane perpendicular to the pseudo plane of the dithiane ring) is most stable and, second, that the apparent twofold barrier to rotation about the Csp(2)-Csp(3) bond is 9.6 kj/mel. The AM1, STO-3G* and STO-3G* computations confirm the twofoldedness of the barrier, the AM1 barrier is 9.4 kJ/mol. The empirical equation, (3)J(H,H) = 13.36 cos(2) phi + 9.59 cos phi - 0.43, reproduces the vicinal coupling constants of the CH,CH,CHI fragments and implies puckering angles phi(ee), phi(ea), and phi(ae) of 54 degrees, 61 degrees, and 64 degrees, respectively. It is implied that (3)J at phi = 0 degrees is larger than at phi = 180 degrees. This results is discussed in terms of the latest theoretical approach to (3)J in the HCCH fragment. The (4)J(H,H) signs and magnitudes for the CH2CH2CH2 fragment agree reasonably well with theory. For the CH2SCH fragment, (4)J(H,H) values are positive, in contrast to corresponding numbers in the propanic fragment, perhaps the first experimental values for certain rigid orientations about a heteroatom. INDO MO FPT computations on propane, dimethyl ether, and dimethyl sulfide confirm the experimental trend in (4)J(H,H). (2)J(H,H) and (5)J(H,H) values are compared to those in related molecules. The striking differential shifts of the axial and equatorial protons are attributed to differential van der Waals interactions with the 3p lone-pair orbital on sulfur. A comparison of the ring proton chemical shifts with those in phenylcyclohexane and isopropylbenzene implies that C-S bonds are weaker net electron donors by hyperconjugation than are C-C bonds. It is also proposed that the ortho protons are deshielded by intramolecular van der Waals interactions with the 3p orbitals on the sulfur atoms.

In an ABX high-resolution NMR spectrum the detection of combination peaks in the X region yields only the chemical shift of X, JAX + JBX and the positive quantities [Formula: see text]. However, the presence of an additional isotopic perturbation on νA − νB, the difference between the resonance frequencies of A and B, yields two X spectra; therefore four quantities, C. Hence all quantities in the surd become available from the composite X spectrum. A generalization to ABMRX spin systems, applicable to the 1H and 19F NMR spectra of the two isotopic molecules of 1-chloro-2,4-difluorobenzene, is possible. It turns out that, from the H-5(X) spectrum alone, the following spectral quantities are extractable; the two values of νA − νB where A and B are the 19F nuclei; JAB; JAX and JBX with their relative signs; JMX and JRX where M and R are H-3 and H-6. No 37Cl/35Cl isotope effect is detectable on the 1H shielding nor on any coupling constants. F-2 undergoes an isotope shift of −1.64(3) ppb. The isotope shift, over five formal bonds, of F-4 is −0.54(3) ppb (larger shielding in the presence of 37Cl). This magnitude is three times larger than that over four formal bonds in another molecule, 2,6-dichloro-4-fluorophenol. Key words: 1H NMR, of 1-chloro-2,4-difluorobenzene, isotope effects in NMR, reflection of 37Cl/35Cl isotope effects on 19F shielding in the 1H NMR spectrum of 1-chloro-2,4-difluorobenzene; isotope effects in NMR, over three and five bonds by 37Cl/35Cl on 19F shielding in 1-chloro-2,4-difluorobenzene; 19F NMR, 37Cl/35Cl isotope effects over three and five bonds in 1-chloro-2,4-difluorobenzene.

The internal rotational potential for benzal fluoride is computed at various levels of molecular orbital theory, including correlation-gradient, MP2 (frozen core) methods. The perturbations of the potential caused by solvents are calculated with the Onsager model (ellipsoidal cavity with l = 6 in the multipole expansion) as well as with the self-consistent isodensity – polarizable continuum model (SCI–PCM). Analysis of the 1H and 19F nuclear magnetic resonance spectra in cyclohexane-d12 and acetone-d6 solutions yields long-range spin–spin coupling constants from which the expectation values of [Formula: see text] can be derived. These expectation values can be compared with those calculated from the theoretical internal rotational potential. Reasonable agreement is found for potentials obtained from MP2/6-31G* approaches in both solvent models. Long-range coupling constants between 19F and 13C nuclei are also reported and provide very rough checks of the [Formula: see text] values. For the isolated molecule an additivity scheme based on the potential for benzyl fluoride reproduces much of the potential for benzal fluoride except for a deviation caused by the rather larger relative magnitude of the fourfold component in the latter. The minimum in the potential for benzal fluoride occurs for a torsional angle, [Formula: see text] of 90° corresponding to a conformation in which the C—H bond of the side chain lies in a plane perpendicular to the phenyl plane and is rationalized on the basis of electrostatic forces. The conformations of minimum energy for the benzyl and benzal fluorides and chlorides are compared and contrasted. The magnitudes of the internal potentials of the fluorides are only a little larger than thermal energies at 300 K and can become smaller than the latter in soludon. Key words: NMR spectroscopy, of benzal fluoride; spin–spin coupling constants, long range in benzal fluoride; solvent effects, on internal rotational potential in benzal fluoride; molecular orbital computations, structure, internal rotational potential, and its solvent perturbations in benzal fluoride; benzal fluoride, 1H, 19F, and 13C NMR on, internal rotational potential, MO computations.

An excellent linear correlation (r = 0.9999) exists between the spin–spin coupling constants 1J(1H,13C), in benzene dissolved in four solvents (R. Laatikainen et al. J. Am. Chem. Soc. 117, 11006 (1995)) and Ando's solvation dielectric function, ε/(ε – 1). The solvents are cyclohexane, carbon disulfide, pyridine, and acetone. 1J(1H,13C)for gaseous benzene is predicted to be 156.99(2) Hz at 300 K. Key words: spin–spin coupling constants, 1J(1H,13C) for benzene in the vapor phase; spin–spin coupling constants, solvent dielectric constant dependence of 1J(1H,13C) in benzene; benzene, estimate of 1J(1H,13C) in the vapor; nuclear magnetic resonance, estimate of 1J(1H,13C) in gaseous benzene.

The 1H, 19F, and 13C {H} nuclear magnetic resonance spectra at 300 K of 4-fluoro- and 3,5-difluorobenzyl fluoride, dissolved in CS2–C6D12 and acetone-d6, are fully analyzed. Spin–spin coupling constants over four, five, and six formal bonds are used to derive expectation values of sin2θ and [Formula: see text] and the apparent twofold internal rotational potentials; [Formula: see text] is the angle by which the α C—F(C—H) bond twists out of the ring plane. The conformation of lowest energy has [Formula: see text] for the 3,5-difluoro compound in the polar and nonpolar solutions, whereas it has [Formula: see text] for the 4-fluoro derivative. The magnitudes of the potentials lie between 2 and 4 kJ/mol, that is, comparable to thermal energies. These data are compared with previous results for the parent compound and its 3,5-dichloro derivative. Geometry-optimized molecular orbital computations, including some correlation-gradient procedures, for benzyl fluoride and the two fluoro derivatives have [Formula: see text] for the conformations of highest energy of the free molecules. However, geometry-optimized SCFRF calculations of the solvent perturbations of the potential (dipole terms are insufficient) are in semiquantitative agreement with experiment in the sense that both solvents are predicted to destabilize the conformation with [Formula: see text] For example, the predominant twofold component in the computed (6-31G*) potential is 3.4 (free), −0.7 (CS2), and −3.3 (acetone-d6) kJ/mol for benzyl fluoride, a negative number indicating [Formula: see text] for the stable conformer; the experimental values are −0.8(2) (CS2) and −2.7(2) (acetone-d6) kJ/mol. The agreement between experiment and theory is of a similar quality for the fluoro derivatives. The stabilization of the conformer with [Formula: see text] for the 4-fluoro derivative is tentatively attributed to hyperconjugative electron acceptance by the α C—F bond, enhanced by the π-electron donor at the para position. A number of coupling constants are discussed in terms of the possible mechanisms of their transmission. Future experiments are indicated. Keywords: 1H, 19F, 13C NMR of 4-fluorobenzyl fluoride and 3,5-difluorobenzyl fluoride; MO calculations and internal rotational potentials in benzyl fluoride, 3,5-difluorobenzyl fluoride, and 4-fluorobenzyl fluoride; solvent effects and experimental and theoretical approaches to internal rotational potentials in 3,5-difluorobenzyl fluoride and 4-fluorobenzyl fluoride.

The1H nuclear magnetic resonance spectra of phenylallene, diluted in acetone-d6and benzene-d6, yield long-range coupling constants over as many as eight formal bonds between the ring and side-chain protons. These are discussed in terms of σ- and π-electron spin–spin coupling mechanisms, which are sensitive to the torsion angle between the allenyl and phenyl fragments. The torsion angle is assessed by means of molecular orbital computations of the internal rotational potential, whose height is calculated as 16.0 kJ/mol at the MP2/6-31G* level of correlation-gradient theory. Comparison with experimental and theoretical internal rotational potentials for styrene suggests that steric repulsions in the planar form of styrene amount to about 4 kJ/mol. In a field of 7.0 T, phenylallene is partially aligned, entailing a positive dipolar coupling constant between the methylene protons, from which absolute signs of the spin–spin coupling constants involving these protons can be inferred. Such coupling constants over seven and eight bonds, to the meta and para protons, are taken as being mediated by the extended π-electron system, providing a measure of π-electron contributions to coupling constants between meta protons and those in side chains (spin correlation). Some coupling constants between protons and13C nuclei in the side chain, as well as between ring protons and these13C nuclei, are also discussed in terms of spin coupling mechanisms. Solvent perturbations of one-bond proton–carbon coupling constants in the allenyl group do not follow the usual pattern in which an increase in polarity of the solvent is associated with an increase in the magnitude of the coupling constant. Keywords:1H NMR, phenylallene;1H NMR, long-range spin–spin coupling constants in phenylallene; phenylallene, internal rotational potential, molecular orbital computations; molecular orbital calculations, an internal rotational potential in phenylallene.

The H-1, 'F-19 and C-13{H-1} nuclear magnetic resonance spectra of 1,1,1-trifluoro-2-phenylethane, 1, in CS2-C6D12, acetone-d(6), and benzene-d(6) solutions, on analysis, yield long-range coupling constants from which are derived the apparent twofold barriers to rotation about the Csp(2)-Csp(3) bonds. The twofold barrier is 9.0(2) kJ/mol, independent of solvent, 4.0 kJ/mol larger than that of ethylbenzene, also independent of solvent. The theoretical barrier heights for the free molecules at the post-Hartree-Fock level of molecular orbital theory (frozen-core MP2/6-31G*) also differ by 4 kJ/mol, but are about 1 kJ/mol higher than the experimental estimates. The,perpendicular conformer is the most stable for both molecules. Comparisons are made with the benzyl halides, in which the internal barriers are remarkably sensitive to solvent. A spin-spin coupling constant over five formal bonds, (5)J(H, F) involving the ortho protons in 1, is +0.74(2) Hz and is discussed in some detail in terms of its dependence on internuclear distances (possible through-space interactions). The solvent perturbations of (3)J(H, F) and of (2)J(C, F) are of opposite sign. Other long-range coupling constants or their absence are also pointed out. For example, those between F-19 and C-13 nuclei or protons at the meta position are effectively zero; at the para position they are significant.

The free energies of activation at 110 K for rotation about the exocyclic C—C bonds in 2,6-difluorobenzaldehyde and 2,4,6-trifluorobenzaldehyde, in dimethyl ether solutions, are 18.8 ± 0.5 and 20.0 ± 0.5 kJ mol−1, respectively, as determined from 19F{1H} dynamic nuclear magnetic resonance measurements. For the parent compound ΔG≠ is 32.2 kJ mol−1 in the same solvent. These free energy barriers, the lowest available for benzaldehyde derivatives, are likely a result of steric and electrostatic repulsions between the C+—O− and C+—F− bonds. Computations of the spectroscopic barrier in the 2,6-difluoro compound at various levels of molecular orbital theory imply that the barrier is predominantly twofold, with a fourfold component of opposite sign, whose magnitude is about 10% of the twofold component. A correlation-gradient computation, MP2/6-31G*, finds a barrier height of 18.6 kJ mol−1 for this compound, lower by 3.0 kJ mol−1 than found with the 6-31G* basis and 2.9 kJ mol−1 with 6-31G**. Similar computations are compared for the parent compound and the 4-fluoro, 2,4,6-trifluoro, and 3,5-difluoro derivatives. Linear relationships exist between the computed spectroscopic barriers (ΔE values at absolute zero for the free molecules) and the free energy barriers for benzaldehyde and the four fluoro derivatives; the theoretical barriers utilize 6-31G** and correlation-gradient MP2/6-31G* procedures. For the 2,6-difluoro derivative, the computed frequencies of the torsional motions about the exocyclic C—C bond yield spectroscopic twofold barriers. These barriers are much lower than the computed energy differences between the planar and perpendicular conformers, perhaps because the negative fourfold components flatten the potential at its minimum. A rough estimate of the relationship between ΔG≠ and ΔE0 for the 2,6-difluorobenzaldehyde suggests that the solvent increases the internal barrier by only about 3 kJ mol−1. By way of contrast, the AM1 barriers, scaled by a factor of 1.9 (as previously recommended) range from 17.3 to 22.6 kJ mol−1, the ΔG≠ values from 18.8(5) to 34.4 kJ mol−1, and the MP2/6-31G* (correlation-gradient) barriers span 18.6 to 36.8 kJ mol−1 for benzaldehyde and the four fluorine derivatives. It seems likely that the internal barrier in benzaldehyde is considerably larger than that modeled on torsional frequencies. Keywords: Free energies of activation, internal rotational barriers in 2,6-difluoro- and 2,4,6-trifluorobenzaldehyde; molecular orbital computations, internal rotational barriers in 2,6-difluoro- and 2,4,6-trifluorobenzaldehyde; correlation gradient computations on internal barriers in benzaldehyde and four of its fluorine derivatives.

The 1H nuclear magnetic resonance spectra of 2-formylstyrene, from dilute solutions in CS2–C6DI2 and acetone-d6, are analyzed to yield precise chemical shifts and spin–spin coupling constants. The long-range coupling constants imply a conformational distribution in which the O-trans conformer is 55% abundant in both polar and nonpolar environments. They also imply that the vinyl group, on average, is twisted out of the aromatic plane to a much larger extent than in styrene. The 6-31G* basis set gives an ab initio potential for the torsion of the vinyl moiety with a relatively deep minimum at 38° out-of-plane, for the O-cis conformer. For the O-trans conformer, two minima are found, one at 45° and another at 129.6°. Essentially the same potential is obtained with the 6-31G** basis. The latter corresponds to a close approach of the hydrogen atom of the formyl group and π orbitals or the β-carbon atom of the olefinic side chain. This local minimum is interesting in terms of a hypothesis used to explain the photochemistry of the molecule. The long-range coupling constants are consistent with the conformational properties calculated for the free molecule; they also indicate no significant difference between the conformational behaviour of the molecule in the two solvents. A proximate coupling constant of −0.16 Hz exists between the formyl and methine (α) protons. The latter is strongly deshielded in the presence of the formyl group, so that it becomes even less shielded than some of the aromatic protons. Keywords: 1H NMR, 2-formylstyrene (o-vinylbenzaldehyde); long-range spin–spin coupling constants, 2-formylstyrene; conformations, three nonplanar of 2-formylstyrene; molecular orbital calculations, conformations of 2-formylstyrene.

The 1 H, 19 Fand 13 C{ 1 H} nuclear magnetic resonance spectra of 1,1,1-trifluoro-2-phenylethane, 1, in CS 2 –C 6 D 12 , acetone-d 6 , and benzene-d 6 solutions, on analysis, yield long-range coupling constants from which are derived the apparent twofold barriers to rotation about the Csp 2 —Csp 3 bonds. The twofold barrier is 9.0(2) kJ/mol, independent of solvent, 4.0 kJ/mol larger than that of ethylbenzene, also independent of solvent. The theoretical barrier heights for the free molecules at the post-Hartree–Fock level of molecular orbital theory (frozen-core MP2/6-31G*) also differ by 4 kJ/mol, but are about 1 kJ/mol higher than the experimental estimates. The perpendicular conformer is the most stable for both molecules. Comparisons are made with the benzyl halides, in which the internal barriers are remarkably sensitive to solvent. A spin–spin coupling constant over five formal bonds, 5 J(H, F), involving the ortho protons in 1, is +0.74(2) Hz and is discussed in some detail in terms of its dependence on intenuclear distances (possible through-space interactions). The solvent perturbations of 3 J(H, F) and of 2 J(C, F) are of opposite sign. Other long-range coupling constants or their absence are also pointed out. For example, those between 19 F and 13 C nuclei or protons at the meta position are effectively zero; at the para position they are significant. Keywords: 1,1,1-trifluoro-2-phenylethane; 1 H, 19 F, and 13 C NMR; long-range spin-spin coupling constants; through-space 1 H, 19 F spin–spin coupling constants; internal rotational potential; molecular orbital computations of internal potential.

The proton chemical shifts and the proton spin–spin coupling constants are reported for 1-phenyl-1-butyne and 1-phenyl-1-pentyne dissolved in CS2/C6D12 and acetone-d6. The long-range coupling constants between the methylene and ring protons are used to derive the twofold barriers to internal rotation in these molecules. They are 0.5 ± 0.1 kJ/mol; the perpendicular conformer is the most stable, the one in which the C(3)—C(4) bond of the side chain lies in a plane perpendicular to the phenyl group. This energetic preference is assigned to the difference between C—C and C—H hyperconjugative interactions with the aromatic π electron system, the C—C interaction being larger. Comparison of the twofold components of the internal rotational barriers in biphenyl and diphenylacetylene, that in the latter amounting to 40% of that in the former, implies that the hyperconjugative component of the internal barrier in ethylbenzene is 1.2 ± 0.3 kJ/mol, a minor component of the total magnitude. Molecular orbital computations of the conformational energies of 1 -phenyl-1-butyne all agree that the perpendicular conformer has the lowest energy but only to the extent of 0.1 kJ/mol at most.

The 1H nuclear magnetic resonance spectra of ethylbenzene-β-13C in CS2/C6D12 and acetone-d6 solutions yield long-range 1H,1H and 1H,13C coupling constants. The 13C {1H} NMR spectra yield 13C, 13C couplings. The conformational dependence of some of these coupling constants is compatible with two values of the barrier to internal rotation about the exocyclic Csp2—Csp3 bond. If the fourfold component of the internal rotational potential is not larger than about 20% of the twofold component and the perpendicular conformer is most stable, then the barrier height is probably less than 6 kJ/mol. However, if the stable conformer has a torsion angle of 60° for the exocyclic C—C bond, then the coupling constants are consistent with a twofold barrier of about 23 kJ/mol. Experimental values of [Formula: see text] and [Formula: see text], where ψ and θ are the torsion angles for the exocyclic C—C and C—H bonds, respectively, are compared to those obtained from INDO MO FPT computations of the angular dependence of nJ(H,C), and nJ(H,H), nJ(C,C) for n ≥ 3. For example, the computations very likely give the correct qualitative ψ dependence of 5J(C,C), yet overestimate its extremum by about a factor of two, either because of an overestimate of the σ–π exchange integrals or because of too large a valence orbital density at the carbon nucleus. Other nJ values are discussed in a similar manner and, because optimized geometries are used in the computations, a somewhat more reliable treatment arises for some coupling constants; an example is 3J(C,C) at small torsion angles.

The analyses of the 1H nuclear magnetic resonance spectra of 2-(diphenylphosphino)benzaldehyde in CS2/C6D12 and acetone-d6 solutions yield stereospecific coupling constants from which the populations of the O-cis and O-trans conformers are derived. The free energy differences favouring the O-trans conformer at 300 K are 2.7 and 0.9 kJ/mol, in the polar and nonpolar solutions, respectively; in the crystal only the O-cis conformer exists. The coupling constant, 4J(CHO, P), is estimated as −7.1(2) Hz in the O-trans confomer and 3J(CHO, P) as +29.4(1.3) Hz. Their magnitudes depend on the proximity of the C—H bond to the lone pair on phosphorus. nJ(C, P) are reported for triphenylphosphine and for the benzaldehyde derivative as dilute solutions in the two solvents, demonstrating a significant solvent dependence for some of these coupling constants. Some simple relationships are proposed between nJ(C, P) and the torsion angle about the C—P bond, estimates of the latter coming from AM1 and STO 3G MO computations. nJ(C, P) are also sensitive to intrinsic ring substituent perturbations, as are the nJ(H, P); for example, 5J(H, P) is negative in the disubstituted ring of 2-(diphenylphosphino)benzaldehyde but positive in the phenyl groups. The nJ(H, P) are also discussed with respect to their dependence on the torsion angles about the C—P bonds. It appears that the conformational properties of the aromatic rings in triphenylphosphine and its formyl derivative are very similar. Further, the phosphorus atom is polarized such that the carbonyl bond is attracted towards the positive region near phosphorus, and the C—H bond of the formyl group more towards the lone-pair region; the actual torsion angles represent a compromise between these attractive forces and the repulsive forces between bonds on neighbouring aromatic moieties. CNDO/2 MO and INDO MO FPT computations of nJ(C, P) and nJ(H, P) are of mixed utility, although the former bear out the idea that the proximate [Formula: see text]lone-pair interaction dominates 3J(CHO,P) and 4J(CHO,P).