The Observation of a Pb-Li Bond ; Synthesis , Structure and Model Molecular Orbital ( MO ) Calculations on the Monomeric Ph 3 Pb-Li-( pmdeta ) Complex [ pmdeta = ( Me 2 NCH 2 CH 2 ) 2 NMe ]

a Department of Pure and Applied Chemistry, University of Strathclyde, Thomas Graham Building, 295 Cathedral Street, Glasgow G I IXL, UK b University Chemical Laboratory, Lensfield Road, Cambridge CB2 IEW, UK c Supercomputer Computations Research Institute B186, Florida State University, Tallahassee, Florida 32306-4052, USA d lnstitut fur Anorganische Chemie, Tammannstrasse 4, W-3400 Gottingen, Germany

The title compound has been synthesised by the cleavage reaction of Ph3Pb-PbPh3 with BunLi and has been shown to have a monomeric Pb-Li bonded structure in the solid state; ab initiocalculations have been used to probe the nature of the early main group metaI/heavy p block metal bonding involved.
Complexes containing bonds between transition metals and heavy p block metals are well known and have been characterised in the solid state.'Recently, we have been interested in the occurrence and the nature of bonds between early main group metals [M = Li, Na, etc. (group 1); Mg, Ca, etc. (group 2)] and heavy p block metals [E = Ga-TI (group 3); Sn, Pb (group 4); Sb, Bi (group S ) ] . 2 .3 Previous studies have centred on species which contain non-metallic p blockearly main group metal contacts, e.g.organometallics (containing C-M), metal amides (N-M), and metal alkoxides (O-M).jCompounds which contain metallic p block elements and alkali or alkaline earth metals, and in which there is a potential for bonding between these, have rarely been studied in their own right though such materials have been known for many years.5More commonly these species have been employed as in sitii reagents in organic or inorganic syntheses.
A case in point is the early main group metal triorganostannates (commonly formulated as 'R3SnM' or 'R3Sn-*Mf') which have been used extensively to promote the formation of Sn-C bonds by their reactions with electrophilic organic substrates such as ketones, epoxides and acid chlorides.

2.74
The Ph3Pb-anion of 2 is closer to a pyramidal structure than the corresponding Ph3Sn-anion in 1 [av.C-Pb-C, 94.3(3)".cf.C-Sn-C 96.1(2)" in 13.These geometries suggest that both metals are using principally p orbitals to bond to their Ph groups and that the interaction with Li is fundamentally with the s orbitals on Sn and Pb.The more compressed (pyramidal) angles in 2 are consistent, in this respect, with the expected increase in energetic separation cf.av.2.817(7) 8, in 121.
$ Crysral data: C27HzxLiN3Pb.h! = 618.73,monoclinic.space group P21, a = 12.377 2), h = 17.543(2).c = 12.663(2) A. f3 = 101.801(11)",0.71073 A, T = 153 K, ~(Mo-Kcu) = 6.286 mm-1.Data were collected on a Siemans-Stoe AED using an oil-coated rapidiy-cooled crystal of dimensions 0.65 x 0.38 x 0.30 mm by the 2@/w method (So< 28 d 60").Of a total of 9673 collected reflections, 9233 were unique.The structure was solved by direct methods (SHELX 92) and refined by full-matrix least-squares o n Fwith all data to R1 and wR2 values of 0.037 and 0.086, respectively (SHELX 92); largest difference peak and hole 1.77 and -1.72 eff-'.A semi-empirical method from psi-scans was employed for the absorption correction.All hydrogens were located in the difference Fourier map and their positions were refined in a riding model with common refined U values for chemically equivalent atoms.The absolute structure was determined by refining the Flack parameter to 0.495 (11).implying racemic twinning.Atomic coordinates.bond distances and angles, and thermal parameters have been deposited at the Cambridge Crystallographic Data Centre.See Notice to Authors, Issue No. 1. , respectively] both reflect, qualitatively, the observed trends and geometries within the solid-state structures of 1 and 2. Thus, both calculated models predict the observed compression in the C-Sn-C and C-Pb-C angles (ca.102" in both) and the expansion in the Li-Sn-C and Li-Pb-C angles (ca.116" in both).The Ph3Sn-and Ph3Pbunits in 1' and 2' therefore have a geometry significantly distorted from pure tetrahedral (sp3).The major influence, dominating the electron density on the heavy p block metal available for interaction to Li, is the effectiveness of charge 5 M O calculations: 1', l'.NH3; the geometries were optimised using the GAUSSIAN 90 suite of programs and a 3-21G basis set (ref.12).The total energies (in a.u.) calculated for the optimised geometries are: 1', -6690.646487:1'.NH3, -6746.58022.
2', 3', 3'.NH3, 4, 4'a and 4'b; the geometries were optimised using the GAUSSIAN 90 suite of programs and a double zeta basis set employing pseudo-potentials for Pb (ref. 13).Calculations were performed on an IBM RS6000 system.The total energies (in a.u) calculated for the optimised geometries are: 2 -700.928846; 3 ' -129.533395;3'.NH3 -185.781473;4' -5.572736 dispersion from Sn or Pb to their three phenyl rings.The phenyl groups in I' are significantly more negatively charged (-0.52e) than those in 2' (-0.44e).It is interesting to note that, whereas in 2' there is a uniform charge distribution within its phenyl rings (ca.-0.2Se for each C), in 1' the bulk of the negative charge is carried by the a-C (-0.43e).The outcome of the less effective dispersion of electron density in 2', is to leave the Pb more electron rich (+ 1.03e) than Sn in 1' (+ 1.29e).There is therefore more electron density on Pb than on Sn to bind to Li+.Some measure of the effects of solvation of the Li is given by model calculations on solvated Ph3Sn-Li.NH3, 1' .NH3.Solvation has the effects of elongating and weakening the Sn-Li The most surprising result shown by these calculations is the manner in which the heavy p block metals use their orbitals.In Ph3Pb-Li 2' the 6p, orbital is directed towards and, together with the 6s orbital, interacts with Li.The Pb-Li bond is therefore a result of s and p, rather than of pure s, interaction.Both the p, and p,, orbitals are used equivalently in their interaction with the phenyl groups.However, a significant interaction between Pb and these groups is with the 6p, orbital the other lobe of which is directed towards their centre.Such is borne out by examination of the valence orbital populations on Pb (6s,1.395;6p,and 6p,,0.423;6p,,0.727).It is possible that the compression in C-Pb-C angles towards a more pyramidal geometry in Ph3Pb-Li can be rationalised in terms of the phenyl rings maximising their interaction with the 6p7 orbital on Pb.The phenyl rings are therefore pulled towards the z-axis.Model calculations on PbH4 4' (H-Pb-H, 109.0";valence orbital populations; 6s, 1.34e; 6p,, 6p,, and 6p7, 0.67e) and on H3Pb-Li 4'a 101.2';6s,1.46e;6p,and 6p,,0.67e;6p,,0.93e)illustrate this trend most dramatically.
We gratefully acknowledge the SERC (D. S. W, M. G. D), the Associated Octel Co, Ltd., Ellesmere Port, UK (D. S. W, M. G. D), the Nuffield Foundation (D. S. W.), and the DAAD (D. S.) for financial support.One of the authors (D.M.) acknowledges the support of the US Department of Energy through contact no.DE-FCQ)5-85ER2500000. , 27th May 1992;Corn. 2l02699F
bond (now 2.771 A, cf.2.747 A in I' and further distorting the Ph3Sn-anion towards pyramidal geometry (C-Sn-C, 101.2', cf.102.6' in 1'.For reasons of their calculational complexity, solvated Ph3Pb-Li models have not been investigated.However, model calculations on unsolvated Me3Pb-Li 3' and solvated Me3Pb-Li 3'.NH3, show that a similar elongation of the Pb-Li bond and compression of the Me3Pb-anion results from solvation of Li.Further investigations on H3Pb-Li 4'a and isolated PbH3-4'b suggest that the geometry about Pb will become even more pyramidal as the Pb and Li centres eventually ion-separate (H-Pb-H in 4'a 101.2', and 93.2' in 4'b).The latter implies that the Pb-Li bond in Ph3Pb-Li.(pmdeta) 2 [C-Pb-C, 94.3(3)'] is comparatively weak.