Calculated spin-spin coupling surfaces in the water molecule; prediction and analysis of J (O,H), J (O,D) and J (H,D) in water isotopomers

• Published in: Molecular physics Vol. 94; no. 5; pp. 851 - 862
• Main Author: R. D. WIGGLESWORTH W. T. RAYNES S. P. A. SAUER J. ODDERSHEDE
• Format: Journal Article
• Language: English
• Published: Taylor & Francis Group 01-08-1998
• ISSN: 0026-8976; 1362-3028
• DOI: 10.1080/002689798167700

Ab initio symmetry and internal valence coordinate oxygen-proton and proton-proton spin-spin coupling surfaces for the water molecule have been computed for the first time. Calculations have been performed at the SOPPA (CCSD) level using a large basis set and a grid of forty-nine geometries on the two surfaces. Equilibrium values differ significantly from some other calculated values especially for the Fermi contact terms. The bond length dependence of J(O, H) is 'normal' i.e. J(O, H
1
) is much more sensitive to stretching the O-H
1
bond than the O-H
2
bond. This contrasts greatly with the corresponding situation in methane.
The surfaces have been averaged over the nuclear motion using a recent highly accurate force field to give values of J (O, H) and J (O, D) for H
2
17
O, HD
17
O and D
2
17
O and J(H, D) for HD
16
O, HD
17
O and HD
18
O over a range of temperatures. For J (O, H) and J (O, D) bond stretching at first order is the dominant part of the nuclear motion correction with second order bending making an important contribution. For J (H, D) the second order bending is by far the largest contribution to the nuclear motion corrections although the other terms partially cancel this contribution. Non-additivity can be largely attributed to the bending term for J (O, H). As expected, the bending terms also contribute relatively more to the temperature dependence of the couplings for J (O, H), J (O, D) and J (H, D). Our calculated J (O, H) in H
2
17
O of -77.22Hz at 293K is in very good agreement with Wasylishen and Friedrich's observed value of -78.70 (±0.02) Hz in cyclohexane at this temperature. Our calculated J(H, D) in HD
16
O at 323K of -1.233Hz is close to a recent experimental value of -1.114 (±0.003) Hz in nitromethane-d
3
observed by Sergeyev et al. at that temperature.