1782 J.C.S. DaltonStereochemical Role of Lone Pairs in Main-group Elements. Part 1.Structure and Bonding in Dichloro(l,4-dioxan)tin(ii)By Edward Hough and David G. Nicholson,* Department of Chemistry, University of Oslo, Blindern, Oslo 3,NorwayThe crystal structure of the title compound has been determined from X-ray diffractometer data by Patterson andFourier methods. Crystals are monoclinic, space-group C2/c with Z = 4 in a unit cell of dimensions a =7.825(1), b = 11.688(2), c = 9.172(1) A, p = 98.01 (I)'. The structure was refined by least-squares methodsto R 0.043 for 1 209 observed reflections. The structure consists of SnCI2 units linked by dioxan ligands, in thechair conformation, to form polymeric arrays. The stereochemical activity of the tin lone pair is clearly evident andthe metal-ligand bonding is discussed in terms of the tin 5s and 5p valence orbitals, without recourse to signifi-cant 5d contributions.THE stereochemical influence of lone pairs in $-blockelements in lower valence-states has attracted con-siderable attention.1.2 In many instances it has beenestablished that the lone pair plays a stereochemicalrole, although a number of examples are now knownwhere it is stereochemically ina~tive.~ In the case oftin@) chemistry much of the experimental informationis consistent with the lone pair being stereochemicallyactive.4 However, although this is a dominant featureexceptions have recently been reported : the high-temperature modifications of caesium trihalogeno-stannate(I1) 536 and one of the tin sites in tin(rr) iodide.'In addition to this aspect of tin(@ chemistry certainother features require structural elucidation.A numberof tin(r1) compounds act as acceptors towards suitabledonor molecules. Many behave as monofunctionalacceptors, but some complexes are known which containadditional ligand~.~ 1.r. and l19Sn Mossbauer spectro-scopic studies of multiligand tin(I1) adducts have estab-R. J. Gillespie, J . Chem. Educ., 1970, 47, 18.L. S. Bartell, J . Chem. Educ., 1968, 45, 754.D. S. Urch, J . Cltem. SOC., 1964, 5775.J. D. Donaldson, Progr. Inorg. Chem., 1967, 8, 287.J. Barret, S. R. A. Bird, J. D. Donaldson, and J. Silver,D. E. Scaife, P. F. Weller, and W. G. Fisher, J . Solid-StateR. A. Howie, W. &loser, and I.C. Trevena, Acta Cvyst., 1972,J . Chem. SOL. ( A ) , 1971, 3105.Chew., 1974, 9, 308.B28, 2965.lished that the additional donor molecules may be eitherbonded to tin or present only for crystal packing pur-p0ses.8~~ Previously, two single-crystal structure deter-minations had been performed on two tin(rr) adducts inwhich the donors are neutral ligands. The structure ofone of these, tin(I1) chloride dihydrate, has been shownby two studiesl09l1 actually to correspond to aqua-dichlorotin(r1) hydrate, SnCl,(OH,)*H,O. The resultsof both investigations are consistent with only one watermolecule being bonded to tin with the second held in thelattice by hydrogen bonds. The second complex issulphatobis(thiourea)tin(1I),l2 Sn(tu),SO,, (tu = thio-urea), which differs froni SnC1,-2H,O in that both ligands,although not equivalent, are bonded to tin.In boththese structures the influence of the lone pair is evident.The ligand molecules in the 1 : 1 complexes of tin(rr)halides with 1,4-dioxan, 1,4-thioxan, and 1,4-dithiancould conceivably exist in either the boat or the chairconformation. Thus, they could act as either uni- orJ. E. Cassidy, W. Moser, J. D. Donaldson, ,A. Jelen, andJ. D. Donaldson, D. G. Nicholson, and B. J. Senior, J . C h i n .D. G. Nicholson, J . Chem. Soc. ( A ) , 1970, 173.SOC. ( A ) , 1968, 2928.lo B. Kamenar and D. GrdeniC, J . Chevn. Soc., 1961, 3951.l1 H. Kiriyama, K. Kitahama, 0. Nakamura, and R. Kiriyama,l3 J. D. Donaldson, D. G. Nicholson, D.C. Pusley, and R. -4.Bull. Chem. SOC. Japan, 1973, 48, 1389.Howie, J.C.S. Dalton, 1973, 18101976 1783bi-dentate donors. From i.r. spectra it has been sug-gested13 that the ligand molecules in these adductsadopt the chair conformation and that accordingly theyare non-chelating, since this would require the boatform. As there is no spectral evidence for two distinctC-0-C or C-S-C ring stretching vibrations for the 1,4-dioxan and 1,4-dithian adducts, it seems likely that thestructures are polymeric. In order to clarify this situ-ation we have determined the crystal structure ofdichloro( 1 ,4-dioxan)tin(11) ,SnCl,( 1 ,4-C4H,O,).EXPERIMENTALPre$aration.-The complex was prepared (a) by adding1,4-dioxan (0.03 mol) to tin(I1) chloride (0.01 mol) in hotethanol ( 5 cm3) and cooling to room temperature,13 or (b) bycooling a hot solution of tin(I1) chloride in 1,4-dioxan.Cvystal Data.-C,H,Cl,O,Sn, M = 277.9, Monoclinic,a = 7.825(1), b = 11.688(2), c = 9.172(1) A, p = 98.01(1)O,U = 830.6 Hi3, D, = 2.1 g cm-3 (flotation), 2 = 4, D, =2.22 g ~ m - ~ , F(000) = 528.Mo-K, radiation, p(Mo-Ka) =36.7 cm-l. Absent spectra, consistent with space-groupsC2/c and Cc. The density of the complex implies a unit-cell occupancy of four and one molecule in the asymmetricunit if the space-group is Cc whereas, in the alternative,C2/c, the molecule is required to be centrosymmetric ortwo-fold symmetric.Intensity Data Collection and Stvucture Re$nemepLt.-Since the complex loses dioxan within a relatively shorttime, crystals were sealed in glass capillaries.Crystalsobtained from preparations (a) and (b), although consistentin cell dimensions, showed significant differences in behav-iour during data collection. Those prepared by method (a)exhibited extremely rapid and large random intensityfluctuations which appeared to arise from movements of theintensity maxima by as much as 0.04" in o. Since the effectwas most prominent for reflections in the 1601 zone i t mayimply laminar twinning or dislocation. After examinationof some twenty specimens from preparation (a) i t was foundthat only crystals obtained from preparation (b) were freefrom this effect and one was selected for the collection ofintensity data on an automatic Syntex PI four-circlediffractometer out to 8 30" by use of nionochromatisedMo-K, radiation.The crystal, dimensions 0.37 x 0.23 x0.12 mm, was stable during data collection. It wasmounted in a general orientation in order to minimisemultiple reflections. Of the 1243 reflections measured1 209 were classified as observed, i.e. Inet > 2.580(1). Datawere corrected for Lorentz, polarisation, and absorptioneffects. Scattering factors for light atoms were from ref.14a and for tin from ref. 14b; solution and refinement werecarried out by use of the ' X-Ray '70 ' system of program-m e ~ . ~ ~The solution of a Patterson synthesis in the space-groupC2/c places the tin atom either at the origin, thereby requir-ing i t to be in a centrosymmetric environment, or on thetwo-fold axis at O,y,l/4.The former is highly unlikely onchemical grounds and can therefore only be accommodatedif pseudosymmetry is introduced at the tin site by disorderin the crystal lattice. In space-group Cc the tin atom is* See Notice t o Authors No. 7, in J.C.S. Dultort, 1975, Indexissue.l3 J. D. Donaldson and D. G. Nicholson, J . Chewz. SOC. ( A ) ,1970, 145.(Items less than 10 pp. are supplied as full-size copies).in the general position x,O,z with no special symmetryrequirements. Attempted refinement of a disordered formof the structure in C2/c was unsuccessful. However,refinement was possible in both the other cases and Bconverged to 0.042 (Cyc) and 0.043 (CB/c),with anisotropictemperature factors for all atoms.We prefer the solution based on C2/c for the followingreasons. ( a ) Although the R factors are very similar, onlyhalf of the parameters required in CG are involved in C2/c.(b) The standard deviations in CG are some four times larger.(c) The bond lengths and angles for the dioxan unit are inconsiderably better agreement with those reported forTABLE 1(u) Final positional parameters ( x lo5 for Sn, X lo4 for otheratoms) with estimated standard deviations in parenthesesSn 0 0000 0 907(6) 25 000c1 0 240(3) 1580(2) 4 433(2)0 3 233(6) 03 07(4) 2 659(6)Atom X Y z4 073(10) 1302(7) 2 198(15)4 058(10) -0 691[7) 2 180[12)C(1)C(2)(b) .Anisotropic temperature factors,* with standard deviationsin parefithesesSn 39.3(3) 35.4(33 28.2(3) 0.0 6.5(2) 0.00 Z3(2) 40(2) 61(3) 0.9(2) lO(2) l(2)C(1) 29(4) 38(4) 151(10) l(3) 13(5) 24(5)Ul1 u,, u33 U l , u13 u23c1 tiO(1) 44(1) 33.4(7) -4.2(8) 5.7(7) - 7 .2 ( 6 )C(2) 27(3) 38(4) 105(7) 2(3) 4(4) -11(4)* The values listed are 103Cij, where the temperature factorhas the form: expi-22-2(Ul,a*2h2 + Uz2b*'k2 + U , , C * ~ ~ ~ +U,,a*b*izfi $- U,,b*c*kl -+ U,,a*c*hZ)].TABLE 2Bond distances (A) and angles (O), with estimatedstandard deviations in parenthesesSn-Cl 2.474( 2) Sn-0 2.52 7 (5)C1-Sn-Cl' 90.52( 7) 0-Sn-0' 168.5(2)0-Sn-CI' 86.9(1) C1'-Sn-0' 85.0( 1)C(1)-0 1.43( 1) C( 1)-C(1') 1.48( 1)C(2)-0 1.43( 1) C(2)-C(2') 1.51(1)C(2J-C (2')-0 109.8 (7) C(l)-O-Sn 124.3(4)C(2)-O-Sn 113.0(4)(a) Tin co-ordination(b) Diosan groupsC (I)-- (2) I 09.1 [ 7 ) C(1)-C(1')-0 110.9(8)other dioxan complexes (see below).It should be notedthat the CG solution places the tin atom in a very similarenvironment, since it is only slightly distorted from two-fold symmetry, and for the present level of discussion thisdifference is not significant.The atomic positions and thermal parameters are given inTable 1, and Table 2 contains the interatomic distances andvalence angles. Observed and calculated structure ampli-tudes are listed in Supplementary Publication No. SUP21759 (3 pp., 1 microfiche).*DISCUSSIONThe crystal structure (Figure 1) consists of SnCl, unitsseparated by a unit-cell translation in the a direction andl4 (a) ' International Tables for X-Ray Crystallography,' vol.111, Kynoch Press, Birmingham, 1962; (b) D.T. Cromer andJ. B.,Mann, Acta Cryst., 1968, A24, 321.l5 X-Ray ' program system, eds. J. M. Stewart, F. A. Kundell,and J. C . Baldwin, University of Maryland Technical Report 6758, 1967, revised version of 19701784 J.C.S. Daltonlinked by dioxan molecules to form linear polymericarrays. The tin atoms in these arrays are connected totwo c-glide related tin atoms by long-range Sn * * C1interactions, thereby forming sheets of SnCl, units. Asecond, interposing, set of polymeric arrays is generatedVbL!lCFIGURE 1 Crystal structure of dichloro( 1,4-dioxan) tin(11)FIGURE 2 Part of the polymeric array showing the Sn environ-ment (two-fold axes through the tin atom and the dioxanmolecule omitted)by the face-centring translation.The dioxan moleculesare in the chair conformation and their bond parametersare similar to those found in other dioxan complexe~.~~J~16 M. Davies and 0. Hassel, Acta Chem. Scand., 1963, 17, 1181.17 0. Hassel and J. Hvoslef, Acta Chem. Scand., 1954, 8, 1953.18 P. F. R. Ewings, P. G. Harrison, and T. J. King, J.C.S.10 J. K. Stalick, P. W. R. Corfield, and D. W. Meek, Inorg.Dalton, 1975, 1455.Chem., 1973, 12, 1668.Tin Co-odination.-The tin environment (Figure 2)is two-fold symmetric by symmetry requirements, butis close to point-group CzV. The co-ordination poly-hedron is best described as distorted five-co-ordinatewith the fifth position evidently being occupied by astereochemically active lone pair.This arrangementis fully consistent with the distorted trigonal bipyramidpredicted by the VSEPR model 1 for discrete AX,Y,Especies (A = central atom; E = lone pair; X, Y =ligands). Whilst the two Cl-Sn bonds form a right-angle the O-Sn-0 bonding angle is only 11" short oflinearity. Repulsion forces between the Sn-0 bondingelectrons and the tin lone pair may be relieved either by areduction in the O-Sn-0 angle or by lengthening of theSn-0 bonds. The dominating effect of the tin lone-pairand its tendency to repel the oxygen atom is opposed bythe role of dioxan as a link between SnC1, units since areduction in the O-Sn-0 angle must result in distortionof the oxygen environments. The Sn-0 bonds (2.527 A)are some 0.2 A longer than that in tin(I1) chloride dihy-drate,ll and even longer (by 0.34 and 0.39 A) than thosein the p-di ket onate, bis ( 1 -phenylbut ane- 1,3-dionat 0 ) -tin(I1) .18 This compound is more appropriate for com-parison because it exhibits the same co-ordinationnumber and point-group for tin as the present structure.However, whereas the diketonate structure consists ofdiscrete molecules the dioxan adduct is polymeric.Moreover, although there is some constraint imposedon the bond angles subtended a t tin in the diketonate bythe oxygen atoms being part of two chelating units, theconsiderably larger deviation (30") of the axial O-Sn-0bonds from linearity and the much shorter Sn-0 bondsdo suggest an increase in stereochemical activity of thelone-pair.It therefore seems evident that the lone-pair-bond-pair repulsions in the dioxan adduct are re-duced by bond-lengthening. This clearly results inweaker Sn-0 bonds which is also consistent with theready loss of dioxan from the adduct at room temper-ature. An interesting feature is that, in contrast to thesituation in SnS0,-2tu,12 where there are major differ-ences between the two Sn-S bond lengths (2.62 and 2.86A), both oxygen atoms in the present compound aresimilarly bound to tin.The Sn-C1 interatomic distances are shorter than thosein the dihydrate (2.50 and 2.56 A),11 but similar to thosein the trichlorostannate(r1) anion (2.440, 2.445, and2.447 A).19 The two long-range Sn * - C1 13.451 A)interactions which connect the polymeric SnC1,-dioxanarrays complete a distorted octahedral co-ordinationabout tin.Interactions of this type are described assecondary bonds by A1cock,20 and defined as thoseinteratomic distances, often regarded as nonbonding,which are longer than normal bonds but shorter than vander Waals distances.There has been much discussion concerning the natureof bonding in compounds of main-group elements.This applies especially to the question of whether or not2o N. W. Alcock, Adv. Inorg. Chem. Radiochem., 1972, 15, 21976 1785d orbitals are significant to o-bond f ~ r m a t i o n . ~ l - ~ ~ Inthe case of tin(I1) chemistry 1 : 1 complexes betweentin(I1) acceptors and various unidentate donors are wellknown. This is consistent with the available 5p orbitalin tin(I1) acceptor compound^.^ Simple descriptions interms of hybridisation language place considerableweight on the importance of d-orbital participation whenthe co-ordination number exceeds four (includingstereochemically active lone pairs) and sp3 hybridisationbecomes inadequate.This model requires the use of dorbitals in order to accommodate the excess of electrons.Thus, in order to form a complex with a second donormolecule, tin would have to make use of its 5d orbitals.In view of the energy separation between the tin 59 and5d orbitals this is ~nlikely.~Alternatively, the tin environment may be described interms of a qualitative MO picture without recourse tosignificant 5d involvement, although a small contri-bution from higher orbitals will be contained within thetotal wavefunction. The 5s and 5p valence orbitals inthe fragment SnC1,02 (idealised to point-group C2,)transform irreducibly as A,(s), B,(p,), B,(p,), and A,@).Following a similar treatment used in describing theelectronic structures of antimony(II1) halides,% theelectronic structure would contain four lowest-energy(la,, lb,, lb,, 2aJ bonding levels which possess majorligand character, and a higher-energy level which islargely associated with the tin lone pair.It seemsreasonable to infer that the latter can be identified as the3a, level, as the crystallographic environment of the tinatom is consistent with the lone pair being directed alongthe symmetry axis. Since the tin 5s and 5pz orbitalshave the same symmetry they can mix, so that the 3a,level represents a large degree of the overall 5s contri-bution to the lone pair. Ligand lone-pair al grouporbitals can also interact with the tin lone pair thereby21 K.A. R. Mitchell, Chertz. Rev., 1969, 69, 157.22 W. G. Salmond, Quart. Rev., 1968,22, 253.23 C. A. Coulson, J . Claem. Soc., 1964, 1442.24 R. S . Berry, M. Tamres. C. J . Ballhausen, and H. Johansen,Acta Chem. Scand., 1968, 231.resulting in stabilised (bonding) and destabilised (anti-bonding) higher levels.the lone pair on the central atom is explicitly describedin terms of contributions from a number of a, symmetrymolecular orbitals and nonplanarity is mainly attribut-able to the balance attained between bonding and anti-bonding a, orbitals.The weak Sn-0 bonds can beexplained if they are mainly associated with the a, levelsand the normal Sn-C1 bonds together with the C1-Sn-C1angle would then be predominantly associated with thel b , and l b , levels. The two more distant chlorine atomsare largely associated with higher levels. These relateto lone pairs essentially localised on these ligands.In any discussion on stereochemistry at a particularsite in a solid-state compound, constraints imposed bythe crystal lattice must be considered. In the presentstructure the polymeric nature of the complex mustconstrain the O-Sn-0 angle (see above). This means thatthe energy of the antibonding contributions to bondingcannot be further reduced and as a result the Sn-0bonds are rather long and weak.It is of special interest to compare this situation withthat prevailing in bis( l-phenylbutane-l,3-dionato) tin-(11) where the tin(I1) atom in the fragment SnO, againlies on a diad axis and has approximate Czv symmetry.The constraint on the O-Sn-0 angle present in the dioxanadduct is not present here and the axial O-Sn-0 bondangle attains 150.4". This considerable reduction inangle is matched by much shorter axial Sn-0 bondlengths (2.290 A). We feel that these differences areconsistent with the lower energies of the 3a, and highera, levels, which permit reduced antibonding contri-butions to the axial Sn-0 bonds (see Walsh 26).We thank Dr. Bernt Klewe for assisting with data col-lection and for his interest in the problem.[5/2421 Received, 12th December, 1975125 D. G. Nicholson and P. Rademacher, Acta Chem. Scand.,28 A. D. Walsh, J . Chem. Soc., 1953, 2260, and the five papersAs was previously pointed1974, A28, 11 36.following