Borane-Stabilized Isomeric Dimers of the Phosphaethynolate Anion

A self-consistent system of additive covalent radii, R(AB)=r(A) + r(B), is set up for the entire periodic table, Groups 1-18, Z=1-118. The primary bond lengths, R, are taken from experimental or theoretical data corresponding to chosen group valencies. All r(E) values are obtained from the same fit. Both E-E, E-H, and E-CH(3) data are incorporated for most elements, E. Many E-E' data inside the same group are included. For the late main groups, the system is close to that of Pauling. For other elements it is close to the methyl-based one of Suresh and Koga [J. Phys. Chem. A 2001, 105, 5940] and its predecessors. For the diatomic alkalis MM' and halides XX', separate fits give a very high accuracy. These primary data are then absorbed with the rest. The most notable exclusion are the transition-metal halides and chalcogenides which are regarded as partial multiple bonds. Other anomalies include H(2) and F(2). The standard deviation for the 410 included data points is 2.8 pm.

We demonstrate that the synthesis of new N-functionalized phosphinecarboxamides is possible by reaction of primary and secondary amines with PCO− in the presence of a proton source. These reactions proceed with varying degrees of success, and although primary amines generally afford the corresponding phosphinecarboxamides in good yields, secondary amines react more sluggishly and often give rise to significant decomposition of the 2-phosphaethynolate precursor. Of the new N-derivatized phosphinecarboxamides available, PH2C(O)NHCy (Cy=cyclohexyl) can be obtained in sufficiently high yields to allow for the exploration of its Brønsted acidity. Thus, deprotonating PH2C(O)NHCy with one equivalent of potassium bis(trimethylsilyl)amide (KHMDS) gave the new phosphide [PHC(O)NHCy]−. In contrast, deprotonation with half of an equivalent gives rise to [P{C(O)NHCy}2]− and PH3. These phosphides can be employed to give new phosphines by reactions with electrophiles, thus demonstrating their enormous potential as chemical building blocks.

The recent fits of additive covalent radii R(AB) = r(A) + r(B) for the title systems are reviewed and compared with alternative systems of radii by other authors or with further experimental data. The agreement of the predicted R with experiment is good, provided that the A-B bond is not too ionic, or the coordination numbers of the two atoms too different from the original input data, used in the fit. Bonds between transition metals and halides are not included in the single-bond set, because of their partial multiple-bond character.