Propylamine–borane

The title compound, C3H12BN, was solved using data collected from a multiple crystal (note incomplete data shell). The cell packing is dominated by bifurcated attractive N—Hδ+⋯δ−H—B interactions.

The title compound, C 3 H 12 BN, was solved using data collected from a multiple crystal (note incomplete data shell). The cell packing is dominated by bifurcated attractive N-H + Á Á Á À H-B interactions.

Comment
We have previously reported structures of ammonia borane (Bowden et al., 2007) (FUYVUQ03) and methylamine borane (Bowden et al., 2008) (EFAGEY) as part our studies of hydrogen storage materials. We were challenged to solve the title compound structure by the poorly crystalline platey crystals and to enhance our understanding of the solid state intermolecular interactions. Most other reported H 3 B-N containing boranes are present as solvent in clathrates (e.g. DATGAG, Alston et al., 1985); a recent exception is a calcium borylamine complex (VODWUH, Spielmann et al., 2008).

Experimental
Synthesis was carried out using a procedure analogous to that for methylamine borane (Bowden et al., 2008). An equimolar mixture of propylamine hydrochloride (dried at 110°C) and sodium borohydride were stirred in anhydrous tetrahydrofuran in a reaction flask at room temperature. Evolution of hydrogen gas was observed immediately, and overnight 1 mol of hydrogen was collected. The resulting suspension was filtered to remove sodium chloride, and tetrahydrofuran removed by rotary evaporation. A near quantitative yield of propylamine borane crystals remained.

Refinement
Diffraction data was extracted from the major of multiple intersecting lattices using RLATT (Bruker, 2004). The structure was solved by direct methods but refinement halted at R1 0.18 for the 684 unique data with I>2σ(I). Inspection of data showed a large number with F o >>F c indicating coincidental contributions from the other contributing lattice(s). A total of 172 reflections which met the two criteria [with q=1.3], (1) I(obs)/I(calc) > q and (2) (I(obs)-I(calc)) > qσ(I(obs)), were then excluded from the dataset. The conventional R1 for these rejected data was 0.44. The ratio criteria q was varied down to values of 0.9: although the R1 agreement factors converged at around a ratio of 1.0 (R1 0.051, for 491 I>2σ(I) data) no significant changes occurred in final su values or parameters compared with the slightly larger dataset. On the basis that another analysis of the data would be possible if the larger dataset was presented, the refinement was continued with the supplementary materials sup-2 (ratio 1.3) 512 independent remaining reflections (R1 0.0544). Nine further weak intensity reflections at high theta, outliers with I(obs)>>I(calc), were then omitted lowering R1 to 0.0508 for the final dataset (503 I>2σ(I)).
The X-H bond distances (where X = C1, C2, B1 & N1) were refined. All methyl and other H atoms were refined as riding on their parent atom with U iso 1.5 & 1.2 times respectively that of the U eq of their parent atom. Fig. 1. Molecular structure of the asymmetic unit (Farrugia, 1997); displacement ellipsoids are shown at the 30% probability level. Fig. 2. Cell contents view (Mercury; Macrae et al., 2006). For clarity only a limited set of atoms are labelled. Hydrogen bonds are shown as rippled lines (purple) with carbon, nitrogen & boron atoms gray, blue & pink respectively. Symmetry codes: (i) 1 -x, 1/2 + y,1/2 -z (ii) 1 -x,1 -y,1 -z (see Table 2).