The position of the proton resonance signal of a molecule in a gas is found to depend upon the gas pressure. Four distinct phenomena are considered to contribute to the displacement which the signal undergoes as the density is changed: (a) bulk susceptibility, (b) van der Waals interactions, (c) electric fields, and (d) neighbor-molecule magnetic anisotropy. Theoretical expressions are presented for each of these effects. Experiments were performed on CH4, C2H6, and HCl as pure gases and as mixtures with a variety of foreign gases. For HCl at 30°C, the chemical shift changes by —0.41 ppm as the pressure is raised to 55 atm. Good correlation between theory and experiments is obtained in all cases. Parameters describing the influence of an electric field on the proton screening constant are deduced for C–H bonds and for HCl. The bulk susceptibility contribution is large, van der Waals and neighbor-molecule anisotropy effects are small, and the electric fields from permanent dipoles and quadrupoles and from induced dipoles are all important in some instances.

Articles you may be interested inCurrent density maps, magnetizability, and nuclear magnetic shielding tensors of bis-heteropentalenes. I. Dihydro-pyrrolo-pyrrole isomers Natural chemical shielding analysis of nuclear magnetic resonance shielding tensors from gauge-including atomic orbital calculations Ab initio, symmetry-coordinate and internal valence coordinate carbon and hydrogen nuclear shielding surfaces for the acetylene molecule are presented. Calculations were performed at the correlated level of theory using gauge-including atomic orbitals and a large basis set. The shielding was calculated at equilibrium and at 34 distinct geometries corresponding to 53 distinct sites for each nucleus. The results were fitted to fourth order in Taylor series expansions and are presented to second order in the coordinates. The carbon-13 shielding is sensitive to all geometrical parameters and displays some unexpected features; most significantly, the shielding at a carbon nucleus ͑C 1 , say͒ is six times more sensitive to change of the C 1 C 2 H 2 angle than it is to change of the H 1 C 1 C 2 angle. In addition, for small changes, ͑C 1 ͒ is more sensitive to the C 2 H 2 bond length than it is to the C 1 H 1 bond length. These, and other, examples of ''unexpected differential sensitivity'' are discussed. The proton shielding surface is much more as expected with ͑H 1 ͒ being most sensitive to the C 1 H 1 bond length, less so to the CC bond length and hardly at all to the C 2 H 2 bond length. The surfaces have been averaged over a very accurate force field to give values of ͑C͒, ͑H͒, and ͑D͒ for the ten isotopomers containing all possible combinations of 12 C, 13 C, 1 H, and 2 H nuclei at 0 K and at a number of selected temperatures in the range accessible to experiment. For the carbon shielding the dominant nuclear motion contribution comes from the bending at ''the other'' carbon atom with the combined stretching contributions being only 20% of those from bending. For the proton shielding it is the stretching of the CH bond containing the proton of interest which provides the major nuclear motion contribution. For ͑C͒ in H 13 C 13 CH at 300 K our best result is 117.59 ppm which is very close to the experimental value of 116.9 ͑Ϯ0.9͒ ppm. For ͑H͒ in H 13 C 13 CH at 300 K we obtain 29.511 ppm which is also in very close agreement with the experimental value of 29.277 ͑Ϯ0.001͒ ppm. Calculated values are also very close to recent, highly accurate carbon and proton isotope shifts in the ten isotopomers; carbon isotope shifts differ by no more than 10% from the measured values and proton isotope shifts are generally in even better agreement than this. The observed anomaly whereby the 13 C isotope shift in H 13 C 12 CD is greater than that in D 13 C 12 CH both with respect to H 13 C 12 CH is explained in terms of the bending contribution at ''the other'' carbon. The observed nonadditivity of deuterium isotope effects on the carbon shielding can be traced to a cross term involving second order bending.

Matrix elements are presented for the Hamiltonian of a nonlinear, nonrigid polyatomic molecule in a multiplet electronic state. Their use is only appropriate for electronic and vibrational spectra since hyperfine interactions involving nuclear spins and nuclear quadrupole moments are not considered. For the most general case, nine parameters are required to take full account of spin—rotation interactions, and five are required for spin—spin interactions. For molecules of orthorhombic symmetry only three spin—rotation parameters and two spin—spin parameters are nonzero. For nonlinear molecules in doublet and triplet electronic states, explicit formulas are presented for (a) the rotational term values of symmetric rotors and (b) spin splittings of asymmetric rotors possessing orthorhombic symmetry. All these formulas reduce to well-known expressions for diatomic molecules in 2Σ and 3Σ states when K-dependent terms are ignored. Application of the above formulas to the results of Dressler and Ramsay on the 2B1 ground states of NH2 and ND2 permits the determination of the spin—rotation parameters of these molecules. All five spin parameters of formaldehyde in its lowest 3A2 state are given together with curves of spin splittings in the lower K levels. The spin parameters of HCHO, NH2, and ND2 are compared with those of NO2 and ClO2 found by recent microwave studies. For a triplet state of an orthorhombic molecule, the spin—spin constants determined by band spectroscopy are simply related to the spin constants D and E determined from zero field splittings in electron spin resonance spectroscopy. The surprisingly small value of D=0.42 cm—1 for the lowest triplet state of formaldehyde is briefly discussed in terms of a breakdown of the orbital approximation for this prototype ``n—π* state.''

No abstract

No abstract