Group I. The Alkali and Coinage Metals

BY B. C. CROSSE

Structural and Bonding Studies. — Whereas lithium alkyls are highly associated in hydrocarbon solvents, the addition of tetrahydrofuran leads to a large conductivity increase, indicating ionization. The nature of the ions formed has been established by electrodialysis of labelled n-butyl-lithium solutions in tetrahydrofuran. An equilibrium is established between the solvated dimer, monomer, and free ions:

Bu- + Li+, THF [??] BuLi,THF [??] (BuLi)2, THF [??] Bu- + Li2Bu+, THF

The addition of lithium 2-methoxyethoxide or 2-dimethylaminoethoxide greatly enhances the ionic nature of lithium alkyls. The resulting electronic spectra are those expected for solvent-separated ion pairs. The presence of species of the type RLi,2MeOC2H4OLi is indicated, and the term 'co-ordination-agent-separated' ion pair is suggested, to distinguish them from solvent-separated species.

The structures of some aromatic ion pairs, including cyclopentadienyl-lithium and 1-phenylallyl-lithium, have been studied by n.m.r. spectroscopy. The 7Li chemical shift is related to the distance between the plane of the aromatic anion and the lithium cation located above its π-electron cloud. N.m.r. has also been used to study the nature of the bonding in benzyllithium and its derivative n-C5H11CPh2Li. The variation of the n.m.r. data with solvent points to an appreciable sp3 character for the α-carbons, but the variation is less marked in the diphenylhexyl compound, due to the larger π-system, which favours sp2 character. A similar study of n-C5H11CPh2Li and n-C3H7CPh2Li was interpreted in terms of a dimeric structure in benzene solution and solvent-separated ion pairs in tetrahydrofuran.

The sole X-ray structural determination was of the polynuclear compound LiBMe4. In this, each lithium and each boron atom is four-co-ordinate and linked to a terminal methyl group, two bridging methyls, and an almost linear bridging methyl group in which the lithium-carbon bond length is the shortest known. Measurements of 7Li spin–lattice relaxation times suggested that (Me3SiCH2Li)4 is distorted from tetrahedral symmetry, and so even is the methyl-lithium tetramer above 0 °C. A Raman study of the t-butyl-lithium tetramer indicated that the extent of metal–metal bonding in this compound is insignificant.

n-Pentyl- and n-octyl-lithium appear to exist in solution as hexamers; this is deduced from the extent of their lowering of the vapour pressure of benzene. In contrast, phenyl- and p-tolyl-lithium in ether solution appear not to be associated. A variable-temperature 7Li n.m.r. study of the exchange reaction between them showed that only two species exist, and these are in rapid equilibrium. If there were any association, one would expect also mixed species such as p-MexC6H4Li2Ph.

The exchange reaction between phenyl-lithium and bromobenzene has also been studied. The addition of lithium bromide had an inhibiting effect on the exchange of phenyl groups, and this was explained by the formation of a phenyl-lithium–lithium bromide complex, for which a 1:1 stoicheiometry was established. The previous irreproducibility of results from phenyl-lithium exchange reactions was therefore attributed to different methods of preparation, and the resulting presence or otherwise of this complex. In the halogen–metal exchange reaction of 1- and 2-fluoronaphthalenes with lithium, the presence of the fluoronaphthalene radical anions has been detected below -30 °C by e.s.r. spectroscopy.

Polylithio-compounds. — Some polynuclear arenes undergo facile polylithiation by n-butyl-lithium and NNN'N'-tetramethylethylenediamine (TMED). Deuteriation experiments and mass spectral analysis show that biphenyl and indene will substitute up to six hydrogen atoms by lithium, while anthracene and fluorene form traces of the perlithio-derivatives as well as less highly metallated species. In each case the predominant species contained about three lithium atoms per molecule.

A perlithio-compound C5Li4 was obtained from the butyl-lithium–TMED treatment of penta-l,3-diyne. Hydrolysis led to a mixture of penta-1,4-diyne, penta-1,2-dien-4-yne, and some of the initial diacetylene. Traces of the perlithio-compound C9Li8 may have been formed when 1-phenylpropyne was heated with a fifty-fold excess of butyl-lithium, but the main products were C9H2Li6 and C9H3Li5. Several other polylithio-species were also detected. The presence of TMED reduced the degree of lithiation. Butyllithium in ether at 0 °C gave only the monolithio-compound. Derivatization by quenching with water, deuterium oxide, formaldehyde, acetone, or trimethylchlorosilane gave products derived from 1- and 3-phenylpropynes and phenylallene. Methylation of the di- and tri-lithio-compounds followed by hydrolysis or silylation proved that attack occurs first at the benzylic carbon atom:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Allenic structures have been suggested for the di-, tri- and tetra-lithio-species on the basis of their i.r. spectra. Attempts to polylithiate the aliphatic ring of 1,2,3-triphenylcyclopropane resulted in ring opening, giving a mixture of stilbenes after hydrolysis.

The reaction of dilithiomethane with various boron and silicon halides gave compounds of the type R2BCH2BR2 and (Me3Si)2CH2. The former are not stable and dismute into BR3 and (RBCH2)n.

Functionally substituted organic compounds, such as nitriles and sulphones, easily form gem-dilithio-derivatives on treatment with excess n-butyl-lithium. Subsequent reactions occur only at the α-carbon atom, despite resonance stabilization:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Acetonitrile similarly gives the dilithio-compound Li2C2HN with t-butyl-lithium in ether, but use of n-butyl-lithium–TMED leads to trilithioaceto-nitrile. The monolithio-derivatives of acetonitrile and benzonitrile have also been reported, obtained from the reactions of the nitriles with n-butyl-lithium in ether at room temperature.

Halogenocarbon Derivatives. — A special case of the gem-dilithiation of sulphones mentioned above is the treatment of N-(α-chloromethanesulphonyl)-morpholine with butyl-lithium at -75 °C. The remarkable compound (1) which results is the first α-chlorodicarbanionic structure ever reported; it is readily dialkylated. Other new halogenomethyl-lithium reagents reported include several of the type LiCCl2CR1(OR2)2 as well as LiCCl2CMe2OSiMe3. These were stable below -100 °C, above which temperature they decomposed due to solvent reduction or elimination of lithium alkoxide (not chloride). This preferred β-elimination may arise from intermolecular oxygen–lithium co-ordination. Neighbouring oxygen atoms have been shown to stabilize certain carbenoids, notably the α-halogenocyclopropyl-lithiums (2) and (3), toward α-elimination.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Trichloromethyl-lithium decomposes above -65 °C to release the carbene. Reaction with p-halogenonitrobenzenes only occurred above this temperature, so presumably the formation of 2-dichloromethyl derivatives (4) arises from carbene insertion into a carbon-hydrogen bond. The combined interaction of cyclopentadienyl- and methyl-lithiums with dichloromethane proceeded via a lithium halogenomethide to give, along with benzene, its rare valence isomer benzvalene (5) in good yield (up to 29%). Large-scale preparations are liable to detonate.

Serious detonations have also been experienced during preparations of the reagents perfluorophenyl-lithium and m-bromophenyl-lithium, which should in future be treated with caution. 4,4'-Dilithioperhalogenobiphenyls have been obtained by butyl-lithium treatment of perchlorobiphenyl and 4,4-dibromoperfluorobiphenyl. Monolithio-derivatives were also detected. The thermal decomposition of these compounds was studied. In the reaction of polychlorobenzenes with organolithium reagents, careful selection of the latter can lead almost exclusively either to metal-halogen exchange or to metal-hydrogen exchange.

Some Organic Reactions. — A wealth of literature on reactions of lithium alkyls is published every year; most of it may loosely be described as about organic chemistry. It is only possible here to consider a few of the more unusual items in this category. Selective lithiation of certain aromatic compounds can be achieved by the presence of a substituent group, usually an amino-group, which directs the lithium atom to a specific ring position through co-ordination. Thus, NN-dimethylaminoethylferrocene is metallated by butyl-lithium in the 2-position. 3-Dimethylaminomethyl-substituted thiophens lithiate at the 2-position (6), while the 2-substituted analogues lithiate at the 3-position (7), unless the 5-position is available. Lithium alkyls will also add to simple olefinic bonds when there is an amino- or sulphido-group suitably placed for intramolecular complexation in the new organolithium compound (8) that is formed. Several NN-disubstituted aminomethyl-lithiums R1R2NCH2Li have been prepared by metathesis of the tributyltin derivatives with butyl-lithium. The amino-substituent had a stabilizing effect on the carbon-lithium bond. Surprisingly, PhN(Me)CH2Li did not rearrange to the stable o-Li·C6H4NMe2, despite the convenient geometry for this rearrangement.

The degree of lithiation of di-2-thienylmethane was found to be critically temperature dependent, irrespective of the amount of butyl-lithium present. The inability to introduce a second lithium atom at lower temperatures has been attributed to the formation of an intermolecular complex (9). Above -10 °C the solvated monomer predominates, allowing dimetallation.

Unfortunately, lithium alkyls will react with some of the more important polar solvents. The kinetics of the well-known cleavage of tetrahydrofuran by n-butyl-lithium have been studied in detail, but the two new studies were in disagreement about reaction orders. The decomposition of deuteriated and alkyl-substituted tetrahydrofurans was also studied in the course of elucidating the mechanism of decomposition. Perdeuteriated tetrahydrofuran and diethyl ether were found to cleave much less readily than the 1H compounds. The secondary reactions occurring during the decomposition of tetrahydrofuran and diethyl ether by butyl-lithium were also examined. Ethylene formed in the decomposition was found to insert up to twelve times into the lithium–carbon bond. The reaction of butyl-lithium with NN-dimethylformamide has also been examined.

Addition of n-butyl-lithium to isoprene gave a mixture of methyloctenyllithiums, from which varying proportions of 2-methyloct-l-ene and 3-methyloct-2-enes were obtained by changing the solvent. It seems that methyloctenyl-lithiums undergo a reversible solvent-dependent rearrangement. The reaction of 1,1-diphenyl-n-butyl-lithium with isoprene has been studied by n.m.r. spectroscopy, and similar results were found. Products of different structure were obtained with different solvents, whether a given solvent was introduced before or after the addition reaction.

In the preparation of neophyl-lithium (PhCMe2CH2Li), ca. 6% of the rearranged product PhCH2CMe2Li was also formed, probably by a free-radical mechanism. Attempts to increase the proportion of rearrangement failed.

Triazolyl- and tetrazolyl-lithium compounds, though stable at low temperatures, undergo fragmentation at room temperature with release of gaseous nitrogen.

Phenyl and cyclohexyl isocyanides give α,α-adducts with butyl-lithium, though for p-tolyl isocyanide hydrogen-abstraction competes with the addition reaction:

BuLi + p-Me·C6H4NC -> p-LiCH2·C6H4NC + p-Me·C6H4N=C(Li)Bu

Isocyanomethyl-lithium, prepared from methyl isocyanide and butyl-lithium, was found to be a convenient reagent for introducing aminomethyl groups into carbonyl compounds (Scheme l). It thus provides the first step in the homologation of cyclic ketones.

Finally, carbon monoxide will insert into lithium t-butylamide to give the stable t-butylcarbamoyl-Iithium, from which derivatives such as ButNHC-(:O)SiMe3 are isolable. Its stability is attributed to steric factors since derivatives of the corresponding dimethyl compound cannot be isolated.

ButNHLi + CO -> ButNHC(:O)Li [??] ButNH-C-OLi

2 Sodium and Potassium

Any contribution under this heading must always comprise a highly personal selection of literature, it being so difficult to define what truly belongs hereunder. Pride of place must surely go to the synthesis of alkylpotassium compounds from the reaction of lithium alkyls with potassium (—)-(1R)-menthoxide. The similar reaction of butyl-lithium in the presence of benzene led to phenylpotassium, but toluene and cumene underwent α-metallation, giving benzylpotassium derivatives.

Remarkably stable α-halogenomethyl derivatives of sodium and potassium were reported from the reactions of polyhalogenomethanes with alkali-metal bis(trimethylsilyl)amides:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

The intermediacy of a similar carbenoid was also established in the ring-expansion reaction of fluorenone that occurs on treatment with Ph2CCl2 and sodium.

An important tripartite study of the sodium–acetylene system has been published. Firstly, dissociation pressures were measured for the equilibrium:

NaHC2(s) [??] Na2C2(s) + C2H2(g)

Then, the reactions of propyne and acetylene at the surface of liquid sodium were studied. The influence of dissolved barium upon the latter reaction was also examined; it was found to have a marked retarding effect. The solid product formed in this case was probably BaC2.

When sodium atoms were trapped in a solid benzene matrix, the absorption spectra and thermal stability of the resulting system indicated the formation of a complex; this had a symmetry not higher than C6v.

The nature and reactions of the anion radicals of aromatic hydrocarbons have been widely investigated. This consideration will be restricted mainly to alkali metal–naphthalene complexes, which have been studied in some detail. From molecular orbital calculations, the spin density at the metal nucleus of sodium naphthalenide has been derived as a function of the metal-hydrocarbon distance. Kinetic measurements made on the atom-transfer reaction between sodium naphthalenide and naphthalene enabled an estimate to be made of the metal–ring distance in the transition state. The effect of perdeuterio-substitution on this exchange reaction was studied by e.s.r. spectroscopy. The difference in the rate constants for:

K+(C10H8)- + C10D8 -> C10H8 + K+(C10D8)-

and the reverse reaction was surprisingly large for a secondary isotope effect. Electron-transfer equilibria in the system sodium biphenylide, sodium naphthalenide, and the parent hydrocarbons, were investigated in a range of ethereal solvents to determine the role of tight and loose ion-pairs. The effect of cation solvation on the reactivity of alkali-metal adducts of naphthalene and biphenyl was examined by following the metallation of 9-phenylxanthene and triphenylmethane spectrophotometrically in polyethers. Thermodynamic constants and e.s.r. spectra were determined for alkali metal–naphthalene ion-pairs in several ethereal solvents and solvent mixtures; a variety of different species were identified. A reaction occurred when potassium naphthalenide solution in 1,2-dimethoxyethane was stored over a potassium mirror; it was shown that the decomposition product must be potassium-substituted 1,4-dihydronaphthalene, arising from reduction of the radical anion to the dianion and subsequent protonation by the solvent.A kinetic study has been made of the reaction of sodium naphthalenide with water, using a stopped-flow device. The intermediately formed dihydronaphthyl anion reacts much faster with water than does the radical anion. Stopped-flow techniques were also used to study the protonation of alkalimetal adducts of anthracene and terphenyl by water and alcohols.