The gas-phase molecular structure and quadratic force field of bis(dimethylamino)hydroborane, [(CH3)2N]2BH, determined by electron diffraction and MO-calculations

Abstract

The molecular structure of bis(dimethylamino)hydroborane, [(CH3)2N]2BH, has been determined by gas-phase electron diffraction at 298K and by MO-calculations using the 6-311++G∗∗ basis set with precursory 6-31G∗ computations at the HF, MP2, B3LYP, and B3PW91 levels of theory. All MO-results are consistent with C2 molecular symmetry with slightly non-planar B–NC2 groups. HF and DFT vibrational analyses were carried out using literature data with some reassignments to obtain refined SQM force fields. The B3PW91/6-311++G∗∗ results were used for calculations of vibrational amplitude quantities and geometrical constraints needed in the GED structural analysis that was based on C2 molecular symmetry. The two types of carbon atoms are specified by E and Z subscripts, referring to C–N–B–N. Shrinkage corrections were implemented for the (C2N)2BH skeleton in the final refinement, and the structural parameters with effects from correlation in the data and uncertainty in the s-scale (0.1%) included in the standard deviations, are: ra(B–H)=121.7(fixed) pm, ra(B–N)=142.5(4) pm, ra(N–CE)=145.6(2) and ra(N–CZ)=ra(N–CE)−0.16pm giving 〈NC〉=145.5(2); ∠NBN=127.8(5)°, ∠BNCE=119.9(3)°, ∠BNCZ=127.3(2)°, θ(NBNCE)=−167.6(12)°, and ΔθZ=182.3(21)° giving θ(N–B–N–CZ)=ΔθZ+θ(NBNCE)=14.7(15)°. The location of the hydrogen atoms relied heavily on MO-constraints using only one distance and one valence-angle giving 〈CH〉=110.64(13) pm, and 〈NCH〉=111.4(2)°, and two torsion-angle variables ultimately also MO-constrained to θ(BNCEH)>=−3.1° and θ(BNCZH)=13.9°. This GED structure has a 〈2NC+BN〉 average of 144.50(15) and a difference Δ=〈NC〉−BN=3.1(5) pm. These two parameters and the <NC>, BN and <CH> distances translate into estimates of 144.0, 3.0, 145.0, 141.9, and 109.3pm, respectively. These parameters and the angular features of the GED structure are reproduced fairly well by the B3PW91 computations, but considering the experimental uncertainties one of the four computational methods cannot be unambiguously favored over the others.