A wide range of compounds has been investigated theoretically by extensive ab initio computations in order to study the vertical and horizontal periodic trends in element-ligand bond stabilities and the occurrence of low valencies in the chemistry of heavy elements. We have studied the following: (1) group 13 hydrides and fluorides, MX and MX3 (M = B, Al, Ga, In, Tl; X = H, F); (2) various hydrides and fluorides of the 6th period, viz., AuH, HgH2, TlX, TlX3 (D3h), PbX2 (C2-upsilon), PbX4 (T(d)), BiX (C(infinity-upsilon); 3-SIGMA-), BiX3 (C3-upsilon), and BiX5 (D3h and C4-upsilon) (X = H, F); (3) the thallium compounds TlX and TlX3 (X = H, F, Cl, Br, I). At the best level of treatment of electron correlation (quadratic configuration interaction), the calculated molecular properties are in good agreement with the available experimental results. From the ab initio calculated values of dissociation energies, D(e), of MX(n), we have derived the internal energies of X2 elimination, of disproportionation of MX(n-2), and of X2 exchange between different elements in the gas phase. Our conclusions are summarized here. (1) Low valencies in compounds of the heavy elements arise naturally as a consequence of the periodic trend toward lower M-X bond strength with increasing atomic number, in agreement with an earlier analysis of experimental thermochemical data by Drago. Relativistic effects in the 6th period significantly augment this trend, but are not dominant. (2) Within the group 13 compounds MX3, elimination Of X2 is strongly endoergic but is less so for the heavier elements M. For TlX3, DELTA-U0 for X2 elimination varies in the order X = F > Cl > Br > I > H, despite a diminishing relativistic reduction of DELTA-U0 along this series. (3) In the series TlX3, PbX4, BiX5, the tendency to eliminate X2 and achieve a lower coordination number increases sharply between lead and bismuth, i.e., Bi(V) is a markedly stronger oxidant than Pb(IV). Relativistic contributions are almost constant along the series, though they clearly assist the oxidizing ability of these elements in their highest valency. (4) For the 6th period MX(n-2) compounds, the fluorides all strongly resist disproportionation, whereas disproportionation of monomeric TlH and PbH2 is favored and is assisted by relativity. For the univalent group 13 hydrides and for BF, disproportionation is decisively assisted by the formation of M2 from atomic M. (5) Thallium(III)-X bonds are weaker than thallium(l)-X bonds by ca. 30% except in the case of X = H. The strong binding between highly electronegative fluorine and thallium(I) favors the formation of 3TlF and TlX3 from TlF3 and 3TlX (X = Cl, Br, I, H). (6) Mulliken population analysis reveals no discernible trend in ns population between AlX3 and TlX3 (X = H, F) and thus provides no computational evidence for an especially inert 6s pair. (7) Although relativistic contributions play an important role in the chemistry of the 6th period elements, there is no evidence that the 6s electrons are more inert than the s electrons of lighter elements. Thus, there is no special "inert pair effect" for the 6th period elements, and it is inappropriate to use this term to designate the low valencies of heavier elements.
