The Typical Elements PART111 Groups IV and V By J.D. Smith 1 Reactions of Hydrides The pyrolysis of silanes is proving as complicated and difficult to sort out as the pyrolysis of hydrocarbons but interest in the subject is maintained by the differences between the chemistry of carbon and that of the other Group IV elements. For example the initial step in the thermal decomposition of hydrocarbons is the breaking of a C-H or C-C bond with formation of radical intermediates R,C*. However it has been shown e.g. by deuterium substitution studies that the initial step in the decomposition of polysilanes results in elimination of silenes :SiH or :SiHR:'" RSiH,SiH -+ :SiHR + SiH or :SiH + RSiH3 Activation energies for this process (ca. 210 kJ mol-') are less than Si-Si bond energies (ca.340 kJ mol-') and it seems that the reaction must involve a 1,2-hydrogen shift with a transition state (1) A similar transition state for hydrocarbon decompositions would involve quinque- valent carbon and this is apparently energetically unfavourable. The detailed mechanism of the decomposition of monosilane SiH is still not clear but evidence from pyrolysis in the presence of acetylenelb shows that silyl radicals -SiH, rather than silene species :SiH, are the predominant intermediates produced in the first step. It has been suggested that formation of ground-state singlet :SiH from SiH is forbidden by orbital symmetry unless the reaction path is such as to add considerable strain energy to the activation energy; .SiH formation is thus favoured.Ground- state :SiH could be generated by an orbitally allowed process from disilane and ethynylsilane HCECSiH, from :SiH2 +HCGCH is indeed found as a major product in the pyrolysis of disilane-acetylene mixtures. After further studies on the pyrolysis of hexamethyldisilane a mechanism has been proposed'' which accounts for some of the divergent results obtained under a variety of conditions in earlier work. Initial steps in the decomposition give trimethylsilyl radicals Me,Si* or dimethylsilene Me,Si both of which react further to give a complicated mixture. 2Me,Si-+ Me,SiSiMe + Me,Si + Me,Si A new value of 337 kJ mol-' for the bond dissociation energy D(Me,Si-SiMe,) has been deduced. The thermal decomposition of hexamethyldilead in toluene can be (u)A.J.Vanderwielen M. A. Ring and H. E. O'Neal J. Amer. Chem. Soc. 1975,97,993; (b) C. H. Haas and M. A. Ring Inorg. Chem. 1975,14,2253; (c)I. M. T. Davidson and A. V. Howard J.C.S. Furuduy I 1975,71,69 and references therein; (d)D. P. Arnold and P. R.Wells J.C.S. Chem. Comm.,1975,642. A.J. Carty,R.H. Cragg and J. D.Smith explained in terms of the following reaction sequence Me,PbPbMe + Me,Pb + Me,Pb Me,Pb + Me,Pb + Me8Pb + 2Me,Pb + Pb In this case no evidence for methyl or trimethyl-lead radicals Me,Pb* was obtained.ld Information about reactive species may also be obtained from reactions of metal atoms obtained by evaporation at high temperatures (>lOOO"C) and in high vacuum (<1Torr). The atoms and molecules of volatile compounds are brought together at a cold surface on which products are condensed (see also p.179).'" Particular attention has been given to the chemistry of carbon silicon and transition- metal atoms but the first reports2b9c of reactions of thermally generated germanium atoms have now appeared. The halides of carbon and silicon react with germanium atoms to give trihalogenogermyl derivatives and trimethylsilane reacts to give the new compound Me,SiGeH,SiMe, Ge + 3MX,+ MX,- ,GeX + (MX,-There is interest in whether the initially produced atoms are in the ground ('P) or excited (IS or 'D) states and whether the electrons in intermediates such as HGeSiMe or ClGeCCl remain unpaired or relax to a paired configuration before further reaction.Silyl compounds react with tin tetrachloride in two distinct ways.,= With SiH, SiH3F SiH,Cl (SiH,),N [SiH,Mn(CO),] or Si21& chlorine is substituted for hydrogen and SiH2Cl derivatives are isolated. With SiH,Br [SiH,Co(CO),] (SiH3),S or (SiH3)3P the predominant reaction is exchange SiH3X + SnC1 + SiH,Cl +XSnCl The compounds XSnCl sometimes undergo further reactions. With (SiH3)20 there is evidence for both substitution and exchange. In general the exchange reaction occurs when the silyl starting materials SiH,X have heavier less electronegative groups x. The germyl derivatives H,GeM (M =Li Na K Rb or Cs) have been made from germane and the alkali metal in dimethoxyethane and the potassium rubidium and caesium compounds have been isolated as crystalline solids.3b Solvent-free H,GeLi and H,GeNa could not be obtained.A single-crystal X-ray investigation has shown that germyl-potassium and -rubidium like silyl-potassium -rubidium and -caesium have the sodium chloride structure in which germyl anions show free rotation at room temperature. Germylcaesium has the rare thallium iodide structure in which 2 (a) P. L. Timms Adu. Inorg. Chem. Radiochem. 1972,14 121; (6)M. J. McGlinchey and T.-S. Tan Inorg. Chem. 1975,14 1209; (c)R. T. Conlin S. H. Lockhart and P. P. Gaspar J.C.S. Chem. Comm. 1975,825. (a)S. Craddock E. A V. Ebsworth and N. Hosmane J.C.S. Dalton 1975 1624; (b)G. Thirase E. Weiss H. J. Hennig and H. Lechert Z. anorg. &m. 1975,417,221; (c)P. C. Angus and S. R. Stobart J.C.S.Dalton 1975,2342; (d)A. R. Dahl C. A. Heil and A. D. Norman Inorg. Chem. 1975,14,1095. (e) A. R. Dahl A. D. Norman H. Shenav and R. Schaeffer J. Amer. Chem. SOC.,1975,97,6364; cr)A. R. Dahl and A. D. Norman Znorg. Chem. 1975,14,1093; (g)A. R. Dahl C. A. Heil and A. D. Norman Inorg. Chem. 1975,14,2562; (h)J. E. Drake and C. Riddle J. Chem. SOC.(A),1968,2709; Quart. Rev. 1970,24,263. The Typical Elements each germyl ion has seven caesium neighbours and each caesium seven germyl neighbours. At low temperatures (below -100°C for H,GeK -120°C for H,GeRb and -170 "C for H,GeCs) rotation of the germyl anion is frozen as shown by broad-line n.m.r. experiments. The GeH3- ionic radius (assuming values of K+ 1.33A; Rb+ 1.48A) is 2.29A only a little larger than that of the H3Si- ion (2.26 A).Bond angles (from broad-line n.m.r.) are for SiH3- 94 * 4" (cf.,PH3 94") and for GeH3-93*4" (cf. ASH, 92"). The conversion of bromogermane into a variety of germyl derivatives by use of lead salts has been e~plored.~' In general the conversions are cleaner and the yields are higher than with the more reactive silver salts. Good yields of germyl formate and acetate have been obtained and digermyl ether has been made from bromogermane and lead oxide. Reactions with lead cyanate or thiocyanate gave germyl isocyanate or isothiocyanate (Scheme 1). Germyl trifluoroacetate was made from bromogermane and silver trifluoroacetate. (H,Ge),O ,PbO H,GeBr 5H,GeX (X = HCO or MeC0,) 1PQNCSh H,GeSCN H,GeY [Y = C1 SMe or Mn(CO),] Scheme 1 The chemistry of phosphinogermanes R,Ge(PH,),- (R = alkyl or H) made from the chlorogermanes and LiAl(PH2)4 in glyme solvents has been investigated in some For example redistribution reactions have been studied by n.m.r.spectros~~py.~~ The products from dimethyl(phosphino)germane Me,Ge(PH,)H at equilibrium showed roughly statistical distribution of phosphino- and hydrido- substituents at germanium Me,Ge(PH,)H Me,GeH + Me,Ge(PH,) Redistribution reactions of methyldiphosphinogermane were also clearly detected 2MeGe(PH,),H + MeGe(PH,)H + MeGe(PH,) 2MeGe(PH,)H -+ MeGeH + MeGe(PH,),H From phosphinogermane H,GePH, the products indicated redistributions both at germanium and more slowly at phosphorus. New ternary hydrides H,Ge(PH,) and HGe(PH,), as well as Geh were formed rapidly and the products (H,Ge),PH (H3Ge)3P and PH3 after long reaction times.Pyrolysis of the more thermally stable dialkyldiphosphinogermanesR,Ge(PH,) (R=Me or Et) resulted in elimination of phosphine and formation of the cage compound hexa(dialky1germa)tetraphosphide (2) characterized by spectral data by reaction with hydrogen chloride (which confirms the presence of Ge-P and the absence of Ge-Ge and P-P bonds) and by an X-ray study3' (see next section). The Ge-P distance is 2.317 A. By careful separation of products twointermediates (3) and (4) were characterized but the detailed mechanism of the condensation process was not completely established. Me,GeCI LiAI(PH2), HCI Me,Ge(PH,) 12% 140"C +Me,GeCl + PH, -(Me,Ge),P (2) A.J.Carty,R.H. Cragg,and J.D.Smith H I P Me,Ge Me2Ge/\GeMe Me,Ge /’\\GeMe2 I I &I -1 bH2 PH2 PH2 PH2 PI4 (3) (4) Since oxidation products appeared to be catalysts in the formation of the cage compound (2) the reactions between trimethyl(phosphino)germane,Me,GePH, or dimethylbis(phosphino)germane Me,Ge(PH,), and oxygen in chloroform solution were studied in more detail. The products were intractable solids phosphines or phosphine oxidation products but traces of new germyl phosphonates (5) and (6) were obtained. One of these (6)has been made in 70% yield from the dimethylger- manium oxides (Me,GeO) and a ten-fold excess of anhydrous phosphorous acid.3f /-\ /O O\ Me,Ge GeMe Hydroly~is~~ of the methyl(phosphino)germanes Me,GePH and Me,Ge(PH,) appears to be quantitative e.g.2Me,GePH2 + H20 -+ (Me,Ge)20 + 2PH No evidence has been obtained for redistribution reactions at phosphorus such as those observed earlier for unsubstituted germylph~sphine:~” H,GePH (H,Ge),P + PH Hydrolysis of dimethyl(phosphino)germane,Me,Ge(PH,)H was more complex as the Ge-H bond was labile under the reaction conditions. With short reaction times (20-100 min) and a deficiency of water the compound (Me2GeH),0 was formed but this was thermally unstable at 32°C and readily converted by water into (Me,GeO),. With reaction times greater than 2 h the hydrolysis could be described by the following equation 1 2Me2Ge(PH2)H+ H20 + -(Me,GeO) + 2PH + Me2GeH2 n No Me2GeH2 was detected until after (Me,GeH),O had begun to disappear.3g 2 Cage Compounds A large number of molecules X,Y (7) where X =RE’ or E2,Y =R2E1 RE2 or E3 and E’ E2,and E3are elements from Groups IV V and VI respectively are known.Most of these have the adamantane structure with symmetry Tdor close to Td. The Typical Elements Full details of the crystal structure of (MeSi),S have been p~blished:~" the molecule is of a well established series and the germanium and tin analogues are isostructural. Hexa(dimethy1germana)tetraphosphide (2) has a similar structure (7; X = P Y = Me2Ge).3" The phosphorus atom of the cage compound P406(7; X = P Y = 0)is a donor towards some transition metals. Under a carbon monoxide atmosphere the carbonyl [Fe,(CO),] reacted with the oxide P406 in THF to give a series of compounds [{(C0)4Fe},(P40,)] (n= 1-4) which was characterized by n.m.r.Reactions with pentacarbonyl iron were more complicated and the products included the oxides P,O and P40,. Another new cage compound synthesized during 1975 is the selenide P4Se4 made by fusing together the elements at 300-350 0C.4cIts i.r. spectrum was consistent with the structure (9) similar to those of the known selenides P4Se (8) and P4Se 3 Polyatomic Anions Alloys of the post-transition metals Sn Pb Sb or Bi with alkali metals are remarkably soluble in liquid ammonia and a series of potentiometric and prepara- tive studies more than 40 years agoSa suggested that the coloured solutions contained the cluster anions Sng4- Pb7,94- Sb3,5,73- and Bi3,53-.Confirmation of these clusters by X-ray crystallography has so far proved impossible because on evaporation of the ammonia the Na' and M,"-ions revert to the metallic alloys. The isolation last yea? of the crystalline compound Na(crypt)'Na-[crypt = N(CH2CH20CH2CH20CH2CH2),N] has suggested a new approach to a long- standing problem. Small alloy samples have been treated with the crown ether in ethylenediamine. Crystalline compounds [Na(crypt)'],M,"- have been obtained (a)J. C.J. Bart and J. J. Daly J.C.S. Dalton 1975,2063; (b)M. L. Walker and J. L. Mills Inorg. Chem. 1975,14,2438;(c)Y. Monteil and H. Vincent Z. anorg. Chern.,1975,416,181; (d)G. J. Penney and G. M. Sheldrick,J. Chern. SOC.(A),1971 245; E.Keulen and A. Vos Acta Cryst 1959,12,323. (a)E. Zintl J. Goubeau and W. Dullenkopf Z. phys. Chem. l931,154A 1; (b) F. J. Tehan B. L. Barnett and J. L. Dye J. Amer. Chem. Soc. 1974,% 7203; (c) J. D. Corbett and P. A. Edwards J.C.S. chem. Cornm. 1975 984; (d) A. Hershaft and J. D. Corbett Znorg. Chem. 1963 2 979; (e) R. M. Friedman and J. D. Corbett Znorg. Chem. 1973,12,1134;v> J. D. Corbett Znorg. Chem. 1968,7,198; (g)J. D. Corbett D. G. Adolphson D. J. Merryman P. A. Edwards and F. J. Armatis J. Amer. chem. Soc. 1975,97,6267;(h)W. Dahlmann and H. G. von Schnering Natunviss. 1972,59,420; ibid. 1973 60 429; (i)U. Frank and W. Miiller Z. Natutforsch. 1975 Mb,313. A. J. Carty R. H. Cragg and J.D. Smith and the structures of three anions Sng- PbS2- and Sb,,- have been deter- mined.Although the anion Sn94-has the same valence electron structureSC as the trigonal-bipyramidal (D3h) Big5+ isolated in 'bismuth monochloride' [(Bi,5+)2(BiC152-)4(Bi2Clg2-)],'" the structure (1 1) is and in [Bi+(Bi95+)(Hfc162-),]5e an antiprism capped on one square face (C4,)(Sn-Sn = 2.93-3.31 A). That such a configuration is obtainable in a crystalline compound is consistent with the molecular orbital scheme proposed earlier.5f The green Pb94- established in liquid ammonia and isoelectronic with Big5+ has not yet been obtained in a crystalline derivative but Pb5,- also with D3h symmetry (12) and Pb-Pb 3.00-3.23 A has been obtained.'" This ion is isoelectronic with the well established BiS3+ ion. The Sb,,- ion found 5g in [Na(~rypt)+],Sb,~- has approximate C, symmetry (13) (Sb-Sb 2.69-2.88 A); the cluster is similar to the P$- cluster in Sr3PI4 and Ba3P14'h and the valence structure is presumably the same as in the well-established compounds P4S3 P4Se, and As& [cf.structures (13) and (S)]. Planar five- membered rings Ge have been characterized" in the new lithium germanide LillGe6 made by fusing the elements in a tantalum vessel but there are no metal clusters in Lil,Sn and a number of other solid phases in the lithium-tin and lithium-lead systems. 4 Alkylamido-derivatives and &oxides The chemistry of dialkylamido-compounds of tin(1v) has been extensively studied but simple dialkylamido-derivatives Sn(NR2) are rare. Most of those reported contain large R groups e.g.Sn[N(SiMe,),] or S~[N(S~M~,)BU'],.~" It has now been found,66 however that bis(dimethylamido)tin(II) is accessible from the reaction between tin(I1) chloride and LiNMe,. It is a white crystalline solid m.p. 91-93 "C subliming at 70 "C under Torr and very reactive towards air and moisture. Only peaks from fragmentation of monomer are observed in the mass spectrum but bis(dimethylamido)tin(II)appears to be dimeric in cyclohexane. The n.m.r. spectrum at 40 "Cshows only one line but at -40 "Cthere are two equal intense signals and at least three weaker peaks. The spectrum suggests that the dominant species is the trans-isomer (14) which has two non-equivalent sets of methyl groups. There may (14) (a) D. H. Harris and M. F. Lappert J.C.S.Chem.Comm. 1974,895; (6) P.Foley and M. Zeldin Inorg. Chem. 1975,14,2264; (c)R. Gsell and M. Zeldin J. Inorg. Nuclear Chem. 1975,37,1133; (d)P. F. R. Ewings and P. G. Harrison J.C.S. Dalton 1975 2015. The Typical Elements be smaller concentrations of the cis-isomer also. At 40 "C exchange of alkylamido-groups between isomers and between bridge and terminal positions must be rapid on the n.m.r. time-scale to account for the single peak. The reactions of bis(dimethylamido)tin(II)appear to be similar to those of the tin(1v) compounds. For example dimethylamine is rapidly eliminated in reactions with ethanol or N-methyldiethanolamine EtOH MeN(CH,CH,OH) 1 Sn(oEt)Z '-ZMe,NH Sn(NMez)z -2Me,NH Sn(OCH,CH,),NMe Spectroscopic properties of tin(@ alkoxides and phenoxides Sn(OR) (R =alkyl or aryl) have been reported; some of these compounds may be sublimed but most are insoluble in organic solvents and are thought to have polymeric solid-state structures.Sn(OBu), made by the well established procedure from tin(I1) chloride butanol and triethylamine is dimeric in dichloromethane.6c Phenoxides may be made in quantitative yield from bis(methylcyclopentadieny1)tin and phenols.6d 5 Lone-pair Stereochemistry The valence-shell electron-pair repulsion (VSEPR) theory provides one of the simplest interpretations of the shapes of many compounds of the main-group elements. It is usually illustrated by use of compounds with electronegative atoms such as oxygen fluorine or nitrogen and it successfully predicts the unsymmetrical environments in compounds of elements with lone pairs such as Sn" Pb" SbII' or Bi"'.In solid compounds with the heavier non-metals however post-transition metals in low oxidation states often have symmetrical environments in which lone pairs do not appear to occupy co-ordination positions. The compound CsSn"Br has now been by a single-crystal X-ray study to have the ideal perovskite structure at room temperature so each tin atom is surrounded by six bromine atoms at the corners of an octahedron as suggested earlier from '"Sn Mossbauer data. The compound is a black semiconductor and thus differs from the white compounds MSnBr (M =Na K Rb or NH,). It is suggested that the colour the electrical properties the low Mossbauer chemical shifts and the symmetrical environment of the tin(I1) can all be accounted for by postulating that the 5s2 electrons are delocalized into a band formed by the bromine orbitals.Low-temperature Mossbauer data indicate that the tin(@ environment becomes less symmetrical at low temperatures. Coloured high-temperature phases of other caesium bromostan- nates(r1) e.g. CsSn",Br or Cs4SnI1Br6 may also be obtained and their properties are accounted for in the same way. At 20 "C the colours fade and the crystal symmetry becomes non-cubic; it is suggested that this indicates depopulation of the conduction band. The white compound CS,S~'~B~~ has a cubic structure closely related to that of CsSn"Br, and a series of'solid solutions ~&'"Br6-CSSn"Br may be obtained.The intense colours of these substances which contain both SnIV and Sn" may also result from delocalization of the SnII lone pair into a conduction band formed by the bromine atoms. Electronic structures of mixed-valence compounds of antimony7' and hexahalogenotellurates(~v)~~ may be discussed in similar terms. (a)J. D. Donaldson J. Silver,S. Hadjiminolis and S. D. Ross,J.C.S. Dalton 1975,1500;(b) L.Atkinson and P. Day J. Chem. SOC.(A) 1969,2423 2432; (c) J. D. Donaldson S. D. Ross J. Silver and P. J. Watkiss J.C.S. Dalton 1975 1980. A.J. Carty,R. H. Cragg and J. D. Smith 6 Trends in Bond Order The value of structural information is considerably enhanced where data are available for a closely related series of molecules. This point is illustrated by two examples.Reactions between the lithium salt of diphenylketimine and the appropriate Group IV tetrachlorides yield the derivatives E(NCPh,) (E = Si Ge or Sn) ECl + 4Ph,CNLi -+ E(NCPh,) + 4LiCl Surprisingly these compounds are not isomorphous and there is a systematic variation in the E-N distances and E-N=C angles in the series (Table 1).8a Table 1 E-N Bond lengths and E-N=C angles in the compounds E(NCPh,) Mean Mean E E-N/A E-N=C/O E-N/A (calc.) DiferencelA Si 1.7 17( 10) 137.0 1.879 0.162 Ge 1.872(5) 127.0 1.928 0.056 Sn 2.06(4) 121.3 2.108 0.048 Each molecule has a tetrahedral arrangement of nitrogen atoms about the Group IV element. The structural differences between the silicon and germanium compounds cannot be attributed simply to differences in size as the radii of silicon and germanium are quite similar.A possible explanation is that in the series from tin to silicon the hybridization at nitrogen changes from sp2 to sp as the lone pair is transferred to a p-orbital which can more effectively interact with silicon through (p-d)~ bonds. Single-bond lengths may be estimated (Table 1)by combining well established E-C C-C and C-N bond lengths and there is a systematic shorten- ing decreasing in the series Si > Ge > Sn in accord with the n-bonding hypothesis. The (p-d)n overlap may be achieved by a variety of combinations of the four E-N orbitals and it is noticeable that the E-N=C bond angles are extremely variable especially in the silicon and germanium compounds.The choice of angles in a given crystal thus seems to depend on packing considerations -which explains the variety of molecular conformations found. In the silicon compound there are two crystallo- graphically independent molecules with a range of E-N=C angles. Another example of a systematic study designed to detect trends in molecular parameters has involved the trimethyl-phosphine and -arsine oxides and sulphides. Table 2 Molecularparameters for Me,EY (E = Por As Y = 0 or S) Me3P0 Me3PS Me,AsO Me3AsS E-Y/A 1.476(2) 1.940(2) 1.631(3) 2.059(3) E-CIA 1.809(2) 1.8 18(2) 1.937(2) 1.940(3) LYEC/" 114.4(7) 114.1(2) 112.6(3) 113.4(4) Electron diffraction measurements (Table 2)8b show that the E-C distances decrease very slightly but systematically in the series X3E X,ES and X3E0 (a)N.W. Alcock M. Pierce-Butler G. R. Willey and K. Wade J.C.S. Chem. Comm. 1975 183; N. W. Alcock and M. Pierce-Butler J.C.S. Dalron 1975 2469; (b) G. J. Wilkins K. Hagen L. Hedberg Q. Shen and K. Hedberg J. Amer. Chem. SOC.,1975,97,6352. The Typical Elements 127 (X = Me) as has been found for X = F or CI. There is also a systematic decrease in P-0 and P-S distances in the series Me3PY Cl,PY and F3PY. Trends are attributed to changes in bond polarity. By making reasonable assumptions about the lengths of E-Y single bonds it is found that P=S and As=S bonds have orders close to two but P=O and As=O bonds appear to have higher orders. The rotational freedom of the methyl group increases in the series Me3P0 <Me3PS-Me3As0<Me3AsS.7 Gas-phase Basicities One of the major problems in the interpretation of results of preparative experi- ments in terms of molecular properties is in the allowance which should be made for solvation effects. The availability in recent years of measurements made on molecules in the gas phase e.g. by ion cyclotron resonance spectro~copy,~~ has enabled much progress to be made. The gas-phase basicity of trimethylarsine has now been measured and this may be placed in the context of various amines and phosphines (Table 3)?' Table 3 Proton affinities A ionization potentials I and bond dissociation energies D(B'-H) (all in kJ mol-l) Base B NH3 MeNH2 A 841 878 I 983 865 D(B+-H) 512 431 Me2NH 904 795 387 Me3N 921 754 364 PH3 MePH2 783 841 961 879 432 408 Me2PH 889 817 394 Me3P 926 773 387 ASH^ 756 954 397 Me3As 876 761 326 For all three groups of compounds amines phosphines and arsines the gas-phase basicities increase as hydrogen atoms are replaced by methyl groups.The effect is greater in phosphines and arsines than in amines in which it is suggested rehybridi- zation energy in going from unprotonated to protonated amine opposes the effect of methyl substitution. Ion-molecule reactions of trimethylarsine are very like those of trimethylphosphine . Another approach which leads to detailed information about bonding in simple molecules is exemplified by studies" on the phosphines R,PX3- (R =Me or But; X= H C1 or F; n = 1-3) and on related compounds such as (Me2N),PC13- (n = 1-3) and R2NPF (R = Me or Et).The measurement of the He(1) photoelec- tron spectra of a complete series of similar compounds enables many of the (a)J. L. Beauchamp Ann. Rev. Phys. Chem. 1971,22 527; (b)R. V. Hodges and J. L. Beauchamp Inorg. Chem. 1975,14 2887; (c)M. F. Lappert J. B. Pedley B. T. Wilkins 0.Stelzer and E. Unger J.C.S. Dalton 1975,1207;(d)0.Stelzerand E. Unger Chem. Ber. 1975,108,1246;(e) D. C. Mente and J. L. Mills Inorg. Chem. 1975 14 1862; (f) L. J. V. Griend and J. G. Verkade J. Amer. Gem. Soc. 1975,97,5960;(g) K. 0.Christe C. J. Schack and R. D. Wilson Inorg. Chem. 1975,14,2224;(h) R. Savoie and P. A. Gigukre J. Chem. Phys. 1964,40,2698; (i) K. 0.Christe Inorg. Chem. 1975,14 2230 2821; 6)S. P. Mishra M.C. R. Symons K. 0.Christe R. D. Wilson and R. I. Wagner Znorg. Chem. 1975,14 1103. A. J.Carty,R. H. Cragg and J.D.Smith complexities which result from band overlap to be resolved so that a detailed picture of molecular orbital energies may be built up. The first band in the p.e. spectrum is assigned to the phosphorus lone pair and by plotting the values of the corresponding ionization potential against parameters such as J(PB) in phosphine-borane com-plexes or the A carbonyl-stretching frequencies in complexes ci~-[Mo~(CO),l,~~ the conclusion is reached that the lone-pair ionization potential provides a good measure of relative basicities within a related series. Correlations with Hammett constants XuPhor are less clear (except for alkyl phosphines R,-,,pH,,).One of the attractions in using ionization potential data from p.e. spectra as a measure of basicity is that information may be obtained about a very wide range of compounds. The basicity of the trimethyl derivatives of the Group V elements towards a series of boron Lewis acids has been studied by classical methods. The expected trends in complex stability have been ~onfirmed.'~ One of the manifestations of amine and phosphine basicity is in the ready formation of onium salts in acidic media and interest in these continues. For example in solutions usually in liquid sulphur dioxide containing HS0,F-SbF and phosphorus halides the species PHF3+ PHF2CI' PHC13+ PHC12Br+ PHClBr,' and PHBr,' have been identified by their 31P n.m.r. spectra from which one-bond P-H coupling constants may be clearly mea~ured.'~ The spectrum of PHF3+ for example shows a doublet of quartets with 'J(PH) = 1190.6 Hz.The new salts OH,'[EF,]-(E =As or Sb) obtained as well-defined crystalline solids from the H,O-HF-EF system," appear to be the most stable oxonium salts known and the most suitable for detailed study of the cation. Thus OH,+[SbF,]- decomposes only above 350 "C. X-Ray powder data suggest that OH,'[AsF,]- has a structure similar to that of Ag+[AsF,]- and that OH,+[SbF,]- is similar to KMF,(M = Re W,or Mo). The i.r. spectra of the cations are assigned by comparison with the isoelectronic ammonia; the bands are somewhat affected by cation-anion interactions but much less so than those of mineral acid hydrates such as H30+c104-.9h The sulphonium salt SH,'[SbF,]- was made similarly by condensing hydrogen sulphide on to a frozen solution of SbF in HF but attempts to make SH3+[AsF6]- were not successful hydrogen sulphide reacted quantitatively with arsenic(v) fluoride to give arsenic&) sulphide." By careful experiments it was possible to obtain both i.r.and Raman spectra of the cation SH3+ and to assign peaks by comparison with the isoelectronic PH,. The peaks are much better defined than those of OH3+. Protonation of HCI was also almost certainly achieved in HF-SbF but the white solid adduct decom- posed below room temperature and so full characterization of the cation was not possible. Attempts to make the analogous NHF3+ salts from NF and SbF,-HF were apparently unsuccessful and starting material was recovered from the reaction mixture at -78 'C." The difluoroammonium cation NH2F2+ was however isolated in hexafluoro-antimonate(v)or -arsenate(v) salts and in spite of frequent explo- sions n.m.r.i.r. and Raman data were recorded. The n.m.r. parameters for the NF2H2+ ion are in good agreement with those already formed for ions in the series Nb' NH3F' and NF,+ and the vibrational spectra (except for solid-state effects) were assigned by comparison with CH,F2. The dangerous instability of the difluoroammonium salts at roem temperature was attributed to exothermic elimina- tion of hydrogen fluoride. Irradiation of NF4+[AsF6]- and NF,'[SbF,]- with 6oCo y-rays gave samples with e.s.r. signals assigned to the NF3+ radical cation.9j The Typical Elements 8 Diphosphines Conditions have been described"" for the synthesis of diphosphine H2PPH2 in ca. 30%yield by passing phosphine through an electric discharge. This seems to be an improvement on the more usual method from the hydrolysis of calcium phosphide which gives unpredictable yields. Methylphosphine under similar conditions gives the new diphosphines MePHPH and MePHPHMe. Methyldiphosphine was ther- mally unstable and could not be isolated pure but its formation was clearly charac- terized by n.m.r. spectroscopy. 1,2-Dimethyldiphosphine was apparently formed as two diastereoisomers (15 and 16; R=H) with a gauche conformation. When a mixture of phosphine. and acetylene was passed through the discharge small amounts of buta- 1,3-diyne HCr C-CECH and the new compound ethynyl- phosphine HCrCPH, (yield 9Yo) were found.Isolation of these products is interesting since their formation involves the breaking of the H-C bonds in acetylene. .. R RBMe=fi R-0.R Me R Me R R (15) (17) The factors which determine the conformations of diphosphines are still not clear. Analysislob of the vibrational spectra of H,PPH in gas liquid and solid phases suggests that the only conformation is gauche; the microwave study described in 1974"' was ambiguous on this point since any trans-isomer (17; R = H) would have no dipole moment and so be undetected. The hydrogen atoms in diphosphine appear to be too far apart to have much effect on the conformation which must be dominated by lone-pair interactions.Further evidence for the predominantly trans conformation (17) in P2F4 and P,(CF,) has been obtained from electron diffraction studies:'Od the P-P distance in P2F4 (2.281 A) is the longest yet determined. There is clearly no evidence for n-delocalization across the P-P bond in contrast to the situation in molecules such as Me2NPF2 or H2NPF2 where the P-N bonds are short. It is strange that although the N-C bonds in N,(CF,) are comparable with those in other amines the P-C bonds in P,(CF,) are longer [1.914(4) A] than in most phosphines. Trends in bond lengths and angles in diphosphines (Table 4) cannot be predicted by VSEPR theory which suggests that angles adjacent to more electronegative substituents should be decreased.Table 4 Structural parameters for diphosphines P2X4 X P-P/A LPPXP LXPXI" Me 2.192(9) 101.1(7) 99.6( 10) H 2.218(4) 95.2(6) 91.3( 14) CF3 2.182(16) 106.7(7) 103.8(8) F 2.28 l(6) 95.4(3) 99.1(4) lo (a) J. P. Albrand S.P. Anderson H. Goldwhite and L. Huff Inorg. Gem. 1975 14 570; (b) J. D. Odom C. J. Wurrey L. A. Carreira and J. R. Durig Inorg. Chem. 1975,14,2849; (c) J. R. Durig L. A. Carreira and J. D. Odom J. Amer. Chem. Soc.,1974,96 2688; (d) H. L. Hodges L. S.Su and L. S. Bartell Inorg. Chem. 1975,14,599 and references therein; (e) H. C. E. McFarlane and W. McFarlane J.C.S. Chem. Comm. 1975 582; (f)R. K. Harris E. M. McVicker and M.Fild J.C.S. am. Comm. 1975,886; (g) G. R. Newkome J. D. Sauer and M. L. Erbland J.C.S. Chem.Comm. 1975,885. A. J. Carty R. H. Cragg and J.D. Smith Conformations in diphosphines have also been investigated by n.m.r. spectroscopy. An analysis of the spectra of Bu'MePPMeBu' indicated that only the gauche-isomer with the conformation (16; R = But) was present in appreci- able concentrations at the temperatures studied. Replacement of methyl by t-butyl groups gives a much more negative value for the coupling constant 'J(PP) and it has been suggested that this indicates a change in hybridization at phosphorus with an increase in the CPC angle and an increase in s-character of the P-P bond. It has also been shownlof that in diphosphines R1R2PPR1R2 the parameter ['J(PC) + 2J(PCC)] which is easily found from the 13C n.m.r. spectrum may be used to assign conformations.For the series (MeEtP), (MePr'P), and (MeBu'P), the proportion of the meso-diastereoisomer (15) (in which the larger groups cannot adopt a trans configuration) relative to the racemic (16) decreases showing the effect of steric interactions between the larger groups. A new preparation of diphosphorus tetraiodide in 75-80% yield from potassium iodide and phosphorus(II1) chloride has been described. log 9 Methylidynephosphine Modern instrumentation allows detailed data to be obtained on short-lived species. Methylidynephosphine HCP which may be obtained by passing phosphine at low pressure through a carbon arc rapidly polymerizes above -70 "C. If however the gaseous reaction products with or without vinyl chloride as solvent are condensed at -100"C the n.m.r.spectrum may be recorded at that temperature.'lU The value of 'J(13CH) (21 1Hz) is in the range expected for C(sp)-H bonds and *J(HCP) 43.9 Hz)is much larger than is normally found in phosphines or phosphonium salts. These results are consistent with the proposal that the molecule in solution is best described by the structure HCS-EPs+ as suggested by earlier microwave data.llb 10 Phosphinidene and Arsinidene Complexes A number of phosphine complexes may be metallated by butyl-lithium to give yellow crystalline lithio-derivatives as indicated in Scheme 2.12u These compounds may be kept as solids for several days but they are sensitive towards moisture and are pyrophoric and decompose even at -20 "C in THF or dioxan.The dilithiophos- phine complex (18) reacts12b with NN-dichlorocyclohexylamine to give the red [(PhPH,)Mn(CO),(CsH 511 1BuLi-pentane [(PhPHLi)Mn(CO),(C5Hs)]+ [(PhPLi,)Mn(CO),(C5H5)] 1D2O 1D*O (18) [(PhPDH)Mn(CO),(C 5H s)1 [(PhPD 2)Mn(CO),(C 5 Hs11 Scheme 2 (a) S. P. Anderson H. Goldwhite D. KO,A. Letsou and F. Esparza J.C.S. Chem. Comm. 1975,744; (b) J. K. Tyler J. Chem.Phys. 1964,40 1170. l2 (a) G. Huttner and H.-D. Muller 2. Narurforsch. 1975 30b 235; (b) G. Huttner H.-D. Muller A. Frank and H. Lorenz Angew. Chem.Inremar. Edn. 1975 14 572; (c) M. Baudler and M. Bock 2. anorg. Chem. 1973,395 37; (d) G. Huttner H.-D. Muller A. Frank and H. Lorenz Angew. Chern. Inremar. Edn. 1975.14,705; (e)G. Huttner and H.-G. Schmid Angew. Chem. Internat. Edn.1975,14 433; G. Huttner J. V. Seyerl M.Marsili and H.-G. Schmid ibid. p. 434. The Typical Elements compound [(C5H5)(CO)2MnPPh]3 (19) and this has been shown by an X-ray study to contain the ligand (PPh)3. Free triphenylcyclotriphosphine may be isolated'2c but it rearranges above -20 "C to the pentaphosphine (PhP),. The complex [(C,H,)- (C0)2MnPPh]3 is stable in air. It decomposes on heating to give a new compound PhP[Mn(CO),(C,H,)] (20); the X-ray structure shows that the co-ordination at phosphorus is planar and the Mn-P bonds (2.184 f0.002 A) are unusually short (cf. 2.26-2.40 8 in Mn-phosphine complexes). The compound thus appears to be the first example of planar phosphorus in which electrons from the manganese are used to complete the valence shell.A similar complex of arsenic has been isolated. The phenylarsine complex (21) may be metallated with n-butyl-lithium and the resulting dilithio-derivative (22) reacts with NN-dichlorocyclohexylamine to give the intensely coloured arsinidene complex PhAs[Cr(CO),12 (23). An X-ray study shows that the co-ordination at arsenic is planar and that the As-Cr bonds (2.38 A) are short compared with those in R,As complexes. The Mn-P-Mn groups in (20) and Cr-AsZCr group in (23) may be described as three-centre 4~-systems; the intense (E = 20 000) absorption involving charge transfer from metal to ligand is then ascribed to the 'A + 'B2 transition (Figure 1). Cr As Cr Figure 1 Molecular orbital diagram for the arsinidine complex PhAs[Cr(CO),12 A.J.Carty,R.H. Cragg and J. D. Smith 11 Phosphazenes After the silicones the phosphazenes constitute the most important group of polymers based on inorganic backbones. The P-CI bond in polydichlorophosphazene (24) is hydrolytically unstable but high molecular weight polymers (NPX,) (25; e.g. X = OPh or OCH,CF,) have important uses as elastomers or water repellants. These materials which cannot be made by direct polymerization of cyclic (NPX2)3-5 are accessible by nucleophilic substitution reactions on polydichlorophosphazene. Substitution however is more simply studied in cyclic derivatives; for example the reaction between (NPCI,) and sodium 2,2,2-trifluoroethoxide has yielded the first complete set of products N3P3Cl,(OR)6- (R = CH2CF3,n = 0-6).13' Successive substitution is non-geminal and trans and the sequence of reactions appears to be determined mainly by steric factors.Substitution reactions of ortho-diphenols however may lead to degrada- tion of P-N rings; it is thought that the introduction of a five-membered exocyclic ring as in the undetected (26) increases the susceptibility of the N3P3 ring to nucleophilic attack as the steric strain is relieved with formation of phosphoranes such as (27) or (28). Support for this suggestion comes from the isolation of the intermediate (27) from the reaction between N3P,Cl6 and o-aminophenol. 13' Reac-tions are complicated by scrambling of cyclic ligands during the degradation. Many derivatives N3P3Xn(NR2)6-n(X = halogen) have been made either by the reaction between hexachlorocyclotriphosphazene and amines or by treatment of the hexakisamido-derivativesN3P3(NR2)6with hydrogen halides 13d N3P,Cl NRzH N3P3CIn(NR2)6-n N3P3(NRZ)6 The importance of steric effects in shielding chlorine atoms from nucleophilic attack by amine has been illustrated by isolation of the compound N3P3CI[N(CH2Ph),]- NMe2)4 from the reaction between N3P3CI5N(CH2Ph) and dimethylamine in t~luene;'~' in many aminations the final chlorine is rapidly displaced with formation of the hexakisamido-derivative,but forcing conditions are required here Me NH N 3p3cls"(CH2 Ph)21 75-bo ' N3P3Cl"(CHzPh)z] (NMez) l3 (a)H.R.AUcock Chem.Rev.1972,72,315; R. Chem.inBritain 1974,10,118;(b)J.L.SchmutzandH. Allcock Inorg. Chem.1975,14,2433;(c)H. R. Allcock R. L. Kugel and G. Y. Moore Inorg. Chem. 1975,14 2831; (d) S.N.Nabi R. A. Shaw and C. Stratton J.C.S. Dalton 1975,588;(e)Masood-ul-Hasan R. A. Shaw and M. Woods J.C.S. Dalton 1975 2202; cr) T. S.heron K. Mannan S. S. Krishnamurthy,A. C. Sau A. R. V. Murthy R. A. Shaw and M. Woods J.C.S. Chem. Comm. 1975,975; (g) D.W.J. Cruickshank Acta Cryst. 1964,17,671;(h)H. T. Searle J. Dyson T. N. Ranganathan and N. L. Paddock,J.C.S. Dalton 1975,203;(i)H.P. Calhoun and J. Trotter,J.C.S. Dalton 1974,377,382; (j)H.P. Calhoun R. H. Lindstrom R. T. Oakley N. L. Paddock and S. M. Todd J.C.S. Chem. Comm. 1975,343;(k)H.P. Calhoun,R. T. Oakley and N.L. Paddock,J.C.S. Chem. Comm. 1975,454;(I) 0.J. Schemer and N. Kuhn Chem. Ber. 1974,107,2123;R.Appel and M.Halstenberg J. Orgunometalfic Chem.,1975,99,C25. The Typical Elements aNH2 OH \ N II /P HN aN”’ OH H I / oNH* OH (27) (28) The reaction be tween hexachlorobis(e thy1amino)cyclote trap hosp hazene (29) and an excess of dimethylamine yields among other products the unusual bicyclic phosphazene N,P,(NMe,),(NHEt)(NEt) (30).13’ The P-N bonds in the ring (mean 1.602A) are comparable with those in other phosphazene rings in which the bond order is considered to be greater than one. The P-N bonds at the bridgehead however appear to be inequivalent; the longer bond has a length (1.77 A) compara-ble with that in sodium phosphoramidate’3g (1.769 A) which is normally considered to be a P-N single bond.A further group of phosphazene derivatives may be obtained by replacement of the chlorine atoms in chlorophosphazenes by alkyl or aryl groups. These phos- phazenes are characterized by longer P-N bonds and greater charge localization so that the nitrogen atoms become Thus the cyclic methylphosphazenes (NPMe2)3-5 form salts such as N,P,Me,,HCl N,P4Me,,2HC10, and N5P5Me10,H2CuC14,H20 or complexes such as N,P,Me8,2HgC12 N4P,Me8,4AgN0, and N,P,Me8,HC1,CuC1,. All three phosphazenes form quater- nary salts N,P,Me,,RI (n= 3-5; R =Me or Et) with iodoalkanes. The properties of these alkylphosphazenes are attributed to the less electronegative substituents at NHEt c1 NHEt NMe, \ / \ / CI-P-N=P-CI P-N= P-NMe II I II\ I N N N NEt N I I1 I \II CI -P=N-P-CI Me,N -P=N -P\ / \ CI NHEt Me,N / NMe 134 A.J.Carty R. H. Cragg and J. D. Smith phosphorus so that the d-orbitals are more diffuse than in the halogenophos- phazenes and less suited to overlap with nitrogen p-orbitals. Some of the structural inf~rmation’~~” on protonated or quaternized phosphazenes is shown in Figure 2. The high-energy of the allowed transition at ca. 190nm associated with ring electrons confirms that the phosphorus and nitrogen orbitals have appreciably different electronegativities. Me H H I 1.70 ’ I pON. 1.69 I .69 1.541 1.56/ 9 1.561 p,N. p\ N N N N 1.611 / I .60\ / I.!; N .-i.??N” Figure 2 Mean bond lengths (A) in (a) N4P4Me9+,(b) N4P4Me8H+,(c)N4P,Me8,HCuC13,and (4 NJ’sM~IoH~~+ The reaction between octamethylcyclophosphazene and methyl-lithium in dimethyl ether gives the new carbanion [N,P,Me,(CH,-),] which has been identified by its reactions with iodomethane and the halides Me3EX (E= Si Ge or Sn).13’ N4P4Me8 MeLi-Et,O b [N4P,Me4(CH2-),] 5N4P4Et8 JMe,ECI N4P4Me4(CH 2EMe3) Deprotonation occurs at P-methyl rather than at N-methyl groups and it is thought that the carbanion may be stabilized by conjugation within the r-system of the ring.Deprotonation of N-phosphazenium iodides is more complicated. 13k Thus the compound P,N,Me,I (31) reacts with sodium bis(trimethylsily1)amide to give the ylidic diazaphosphorin (32) confirmed by hydrolysis to the phosphine oxide (33) and protonation to the C-hydriodide (34). The Typical Elements The initial step in the reaction appears to be deprotonation at a P-methyl group as in the phosphazenes.With potassium t-butoxide however the phosphazenium iodide (31) is converted into a linear phosphine oxide (35) by a process which appears to involve nucleophilic attack of t-butoxide on phosphorus followed by elimination of isobutene. The reaction between lithium bis(trimethylsily1)amideand phosphorus(II1) halides gives 50-70% yields of the iminophosphine (Me,Si),NP=NSiMe (36).13' This compound which is very reactive towards air or moisture may be converted into a 1 :1 adduct (37) with trimethylsilyl azide and on distillation the bisiminophos- phorane (Me3Si),NP(=NSiMe3)2 (38) is obtained This may then be treated with more of the iminophosphine to give a four-membered ring compound (39) with both tervalent and quinquevalent phosphorus.2(Me,Si),NLi + PCl Et,O (Me,Si),NP=NSiMe (36) 4 ZMe,SiN, -N (Me,Si),NP(=NSiMe,) t--(Me,Si),NP( =NSiMe,) ,Me,SiN (38) J(Me,Si),NP=NSiMe (37) SiMe SiMe (39) 12 Reactions of Phosphorus Esters More experiments to identify intermediates in the conversion of phosphites into phosphonates by alkyl halides or halogens have been described. In earlier studies much information was obtained by use of optically active corn pound^.^^^ The X -R'X *(RIO),&OR1)+ (R10),PR2+X-(R10),P(0)R2 I R2 -R'X T (40) I (41) observed stereospecificity which depends on the groups R' and R2,is thought to be a function of the lifetime of the five-co-ordinate intermediate (40).If this is short and alkyl halide R'X is rapidly eliminated either directly or via a phosphonium salt (41) optical purity is maintained; if the five-co-ordinate intermediate persists long enough to allow pseudorotation optical purity is lost. Five-co-ordinate species may be stabilized in cyclic phosphites [e.g. (42)]. Intermediates in Arbuzov-type reactions have now been detected directly14' by 31Pn.m.r. spectra of samples at -85 "C in chloroethane solution. The peak due to the five-co-ordinate intermediate (43) l4 (a)C. L. Bodkin and P. Simpson J.C.S. Perkin ZZ 1972,2049; (6) A. Skowronska,J. Mikolajczak and J. Michalski,J.C.S. Chem. Comm. 1975,791,986;(c) W. J. Stec,T. Sudol and B. Uznanski J.C.S.Gem. Comm. 1975,467. A.J. Carty R. H. Cragg andJ.D. Smith c1 (-SCI,) T weakens when the sample is warmed to -40 "C and a new peak corresponding to the phosphorokhloridate product (44)appears. Both phosphonium and five-co-ordinate intermediates have been detected in chlorinations of phosphorus thioesters. The cis-cyclic thionate (45) is converted via the phosphonium intermediate (46) into the trans-sulphenyl chloride (47) with full retention of configuration. Chlorination of the five-membered cyclic phosphorothionate (48) proceeds uia five-co-ordinate intermediates (49) and (43). The easy rearrangement of the isocyano-derivative Me,C(CH,O),P(O)NC to the corresponding cyano-compound was described last year. The prediction that the corresponding isocyano-derivative of tervalent phosphorus (50) should rearrange even more easily has been confimed,'qC since attempts to isolate the compound from the deselenization of the isoselenocyanate compound (51) by reaction with methyl diphenylphosphinite Ph,POMe yielded only the cyano-compound (52).The crude compound obtained from Me2C(CH20),PCl and potassium isoselenocyanate was used in the deselenization reaction since attempts to distil the compound Me2C(CH20)2PNCSe gave only the rearranged product (53). >c>p<N >c4PNCSe +40"C vacuum (53) Ph,POMe (51) I