CCL: 2J NMR spin-spin coupling involving P and H

 Sent to CCL by: "Tobias   Kraemer" [t.kraemer=-=hw.ac.uk]
 Dear colleagues,
 I am currently working on a problem related to indirect NMR spin-spin
 coupling within a bis-ortho-carborane phosphine compound. More precisely,
 the compound is (bis-o-CB)P-CH2CH3, containing an ethyl group bonded to
 the P atom. Now, what is interesting about this compound is that in the
 experimental 1H NMR spectrum the coupling between the CH2 (methylene)
 protons and P is absent (or very small and beyond instrument sensitivity)
 while the coupling to the CH3 (methyl) protons is clearly seen. Hence,
 the CH3 peak is split into a doublet of triplets due to 3J(PH)
 and 3J(HH) coupling, while the peak for CH2 is a regular quartet from
 3J(HH) coupling. For such systems empirical Karplus-type relationships
 have been published in the literature (e.g. Gagnaire et al., Chem.
 Commun. 1968, 1469), relating the P(lonepair)-P-C-H dihedral angle to the
 2J(PH) coupling constant.
 In the present case I find very good agreement between the experimental
 structure and the DFT-optimised structure including, importantly, the
 orientation of the Et group. According to  Gagnaires work, in this
 orientation there are two lp-P-C-H torsion angles, 83 and 162 deg, both
 of which would be expected to result in near-zero 2J(PH) couplings (as
 observed experimentally). Whilst the 3J(PH) and 3J(HH) coupling constants
 are reproduced within 1 Hz by my DFT calculations (B3LYP/aug-cc-pVTZ-J
 [coupling atoms]/cc-pVDZ), the average 2J(PH) coupling constant is
 notably overestimated (~6 Hz). The individual computed 2J(PH) coupling
 constants are 1.5 and 10.9 Hz, respectively. Hence, the average value
 should be around 6.2 Hz, based on the fact that the (partial) rotation
 around the PC(H2) vector is not hindered, rendering these two methylene
 protons equivalent (very much like the 3 protons in the methyl group).
 Things are further complicated by the fact that the experimental 2J(PH)
 coupling is restored (~10 Hz) when additional atoms are bonded to the P
 atom (such as Se), which clearly points to an involvement of the lonepair
 (change of hybridisation). There is a possibility of the involvement of
 the lonepair in through-space coupling. We have also considered the idea
 of direct dipolar coupling, but I am not convinced by this argument.
 I wonder if anyone has dealt with similar systems and seen comparable
 effects (P lonepair...)? How realistic is it to think that I can capture
 this effect in the DFT calculations?
 I would appreciate any ideas, before I embark on method testing. The
 overall agreement of other individual PH coupling constants is
 reassuring, when compared with results I have seen in the literature
 for e.g. (tert-butyl)PCl2. I am somewhat puzzled by the above result.
 Thanks for your help
 Regards
 Tobias
 Dr. Tobias Kraemer MRSC
 Research Associate
 Institute of Chemical Sciences
 School of Engineering & Physical Sciences
 Heriot-Watt University
 Edinburgh EH14 4AS
 United Kingdom
 email: t.kraemer|hw.ac.uk
 phone: +44 (0)131 451 3259