DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 141–146 141 Synthesis and structural characterization of indium compounds with bidentate amide ligands Jungsook Kim, Simon G. Bott and David M. HoVman* Department of Chemistry, University of Houston, Houston, Texas 77204, USA Received 11 September 1998, Accepted 11th November 1998 Indium trichloride reacts with 1 equivalent of MeN(SiMe2NMeLi)2 to give the dimer [ClIn(NMeSiMe2)2NMe]2 and with 4 equivalents of HNMeSiMe2NMeLi to give [Li{In(HNMeSiMe2NMe)2(MeNSiMe2NMe)}]2.In the structure of [ClIn(NMeSiMe2)2NMe]2 a chloride and one amide group of a [MeN(SiMe2NMe)2]22 ligand are bonded to each In atom in terminal positions and the other amide group of the chelating ligand is shared between two In atoms. The terminal chlorides have an anti-ClIn ? ? ? InCl arrangement. The amine group of the [MeN(SiMe2NMe)2]22 ligand does not interact with In. Variable temperature NMR spectra show [ClNn(NMeSiMe2)2NMe]2 undergoes a fluxional process, and a mechanism involving bridge–terminal amide exchange is proposed to account for the data.The molecule [Li{In(HNMeSiMe2NMe)2(MeNSiMe2NMe)}]2 has a Li2In2Si2N4 adamantane-like core and overall virtual D2 symmetry. We recently reported the synthesis of new indium amide complexes including the neutral triamide compounds [In{N(Si- HMe2)(t-Bu)}3] and [In{NR(SiMe3)}3] where R = Ph or But.1 The compounds were prepared for possible use in combination with ammonia as chemical vapor deposition (CVD) precursors to indium nitride films,2 a process that would be analogous to one used to prepare GaN films at low temperature.3,4 Because we were interested in using the compounds for CVD studies, we restricted our synthetic work to simple amide ligands in the expectation that their complexes would have the necessary volatility to be good precursor candidates.The diYculties and successes we had while carrying out the syntheses, as well as the realization that there are only a few reported well-characterized indium complexes with multiple amide ligands,5–8 prompted us to take a more general approach and examine the synthesis of other types of indium amide compounds.Herein we report the syntheses and structures of two new indium complexes that contain the chelating amide ligands [MeN(SiMe2NMe)2]22 and [Me2Si(NMe)2]22. Results and discussion Syntheses A summary of our synthetic results is presented in Scheme 1.Indium trichloride reacts with 1 equivalent of the potentially tridentate ligand MeN(SiMe2NMeLi)2 to give the dimer [ClIn- (NMeSiMe2)2NMe]2 1. The reaction between InCl3 and 4 equivalents of HNMeSiMe2NMeLi gives [Li{In(HNMeSi- Scheme 1 (i) 2MeN(SiMe2NMeLi)2, diethyl ether, 24LiCl; (ii) 8HNMeSiMe2NMeLi, diethyl ether, 26LiCl, 22Me2Si(NMeH)2. In N In N Cl MeN Me2Si N Me SiMe2 NMe SiMe2 Me N Me2Si Cl Me Me In MeN In MeN MeN NMe SiMe2 N NMe SiMe2 N Me2Si N Me2Si N MeN Si N Li Li Me H H Me H Me H Me N Si Me2 Me2 2InCl3 Me Me 1 2 (i) (ii) Me2NMe)2(MeNSiMe2NMe)}]2 2 in 96% yield.In the latter case, if a stoichiometry consisting of an approximately 2: 1 mixture of HNMeSiMe2NMeLi and Me2Si(NMeLi)2 or only 3 equivalents of HNMeSiMe2NMeLi are used the product yield is ª70%. Interestingly, the related reaction of 2 equivalents of Me2Si[N(SiMe3)Li]2 with InCl3 has been reported to give a monomeric species, [Li{In{[N(SiMe3)]2SiMe2}2}].9 Compound 2 is partially soluble in hexane, benzene, toluene, diethyl ether, THF and CH2Cl2 while 1 is very soluble in the same solvents except for hexane, in which it is only partially soluble.X-ray crystallographic studies X-Ray crystal structure determinations of 1 (Fig. 1) and 2 (Fig. 2) were carried out. Selected bond distances and angles are presented in Tables 1 and 2. Compound 1 is situated about an inversion center and 2 lies on a two-fold axis. The amine hydrogens in 2 were located in a diVerence map and subsequently refined with distance constraints.Compound 1 is isomorphous with [ClAl(NMeSiMe2)2NMe]2.10 Three other closely related Fig. 1 View of [ClIn(NMeSiMe2)2NMe]2 1 showing the atomnumbering scheme (50% probability ellipsoids).142 J. Chem. Soc., Dalton Trans., 1999, 141–146 structures are [(py)Zn(NEtSiMe2)2NMe]2, [Be(NMeSiMe2)2- CH2]2 (see I), and [MeIn(NBut)2SiMe2]2.11–13 The In atom in 1 has a distorted tetrahedral geometry with the four coordination sites occupied by a Cl and three amide nitrogen atoms, one of which is terminal (N2 in Fig. 1) and the other two bridging (N1 and N19). The amine group of the [MeN(Me2SiNMe)2]22 ligand is not bonded to In. In the structure the indium atom is a member both of a four-membered In2N2 ring and a six-membered InN3Si2 ring. The molecule 2 has a Li2In2Si2N4 core with an adamantane structure II that incorporates the bridging [MeNSiMe2NMe]22 ligands. Exocyclic to the core are four six-membered rings that include the [HNMeSiMe2NMe]2 ligands.Overall, the molecule has virtual D2 symmetry, where, in addition to the crystallographic two-fold axis passing through Li1 and Li2, there are two virtual two-fold axes, one passing through the In atoms and the other through Si1 and Si19. Each indium atom is coordin- Fig. 2 View of [Li{In(HNMeSiMe2NMe)2(MeNSiMe2NMe)}]2 2 showing the atom-numbering scheme (40% probability ellipsoids). Zn N Zn N py EtN Me2Si N Me SiMe2 NEt SiMe2 Me N Me2Si py Et Et Be N Be N NMe SiMe2 CH2 Me2 Si MeN Me2Si H2C Si Me2 Me Me I Table 1 Selected bond distances (Å) and angles (8) for [ClIn- (NMeSiMe2)2NMe]2 1 In–Cl In–N1 In–N2 In–N19 Cl–In–N1 Cl–In–N2 Cl–In–N19 N1–In–N2 N1–In–N19 N2–In–N19 N1–Si1–N3 N2–Si2–N3 In–N1–Si1 In–N1–C1 2.350(3) 2.204(8) 2.036(7) 2.199(7) 110.1(2) 111.8(2) 116.3(2) 110.4(3) 89.5(3) 116.4(3) 107.5(4) 106.9(4) 107.7(3) 113.6(7) Si1–N1 Si1–N3 Si2–N2 Si2–N3 In–N1–In9 Si1–N1–C1 Si1–N1–In9 C1–N1–In9 In–N2–Si2 In–N2–C2 Si2–N2–C2 Si1–N3–Si2 Si1–N3–C3 Si2–N3–C3 1.768(6) 1.716(9) 1.708(8) 1.756(9) 90.5(1) 114.5(6) 120.1(3) 108.2(7) 121.9(4) 119.8(6) 118.3(6) 125.1(5) 118.4(8) 113.9(7) ated to the amide end of two [HNMeSiMe2NMe]2 ligands and also shares two bridging [MeNSiMe2NMe]22 ligands, resulting in an In atom surrounded by four amide groups (N1, N29, N3 and N5 in Fig. 2) in a tetrahedral arrangement.The amine ends of the [HNMeSiMe2NMe]2 ligands (N4 and N6) are coordinated to lithium, which also interacts with the bridging [MeNSi- Me2NMe]22 ligands.The cation has a distorted tetrahedral geometry with widely varying angles (101–1198). Charge separation in the complex can be described as [In(NRR9)4]2 and Li1. In 1 the angles around indium vary from 908 to 1168 with the smallest angle associated with the bridging amide groups, N1–- In–N19, but in 2 the N–In–N angles are in the narrow range 104–1128. The terminal amide nitrogens in both structures have essentially planar geometries. The angles about these nitrogens span a narrow range (118–1228) in 1 but vary more widely (112–- 1308) in 2 with the larger angles (129 and 1308) being associated with In–N–Si and the smaller with In–N–C (112 and 1138).In 2 the amine nitrogens, N4 and N6, and the amide nitrogens that interact with the Li1, N1 and N2, have distorted tetrahedral geometries. The terminal In–N amide distances, In–N3 [2.107(5) Å] and In–N5 [2.109(5) Å] in 2 and In–N2 [2.036(7) Å] in 1, are similar to those found in [In{Nt-Bu(SiHMe2)}3(p-Me2NC5H4N)] [average 2.125(3) Å], [In{NPh(SiMe3)}3(OEt2)] [average 2.095(2) Table 2 Selected bond distances (Å) and angles (8) for [Li{In(HNMeSiMe2NMe) 2(MeNSiMe2NMe)}]2 2 In–N(1) In–N(3) In–N(5) In–N(29) Si(1)–N(1) Si(1)–N(2) Si(2)–N(3) N(1)–In–N(3) N(1)–In–N(5) N(3)–In–N(5) N(1)–In–N(29) N(3)–In–N(29) N(5)–In–N(29) N(1)–Si(1)–N(2) N(3)–Si(2)–N(4) N(5)–Si(3)–N(6) In–N(1)–Si(1) In–N(1)–C(1) Si(1)–N(1)–C(1) In–N(1)–Li(1) Si(1)–N(1)–Li(1) C(1)–N(1)–Li(1) Si(1)–N(2)–C(4) Si(1)–N(2)–Li(2) C(4)–N(2)–Li(2) Si(1)–N(2)–In9 C(4)–N(2)–In9 Li(2)–N(2)–In9 In–N(3)–Si(2) In–N(3)–C(5) 2.165(5) 2.107(5) 2.109(5) 2.161(5) 1.719(5) 1.730(5) 1.697(6) 110.5(2) 111.9(2) 103.9(2) 111.7(2) 110.2(2) 108.4(2) 104.9(3) 106.9(3) 106.9(3) 112.5(2) 111.4(3) 112.9(4) 93.1(3) 124.5(3) 100.6(4) 113.3(4) 125.3(3) 100.0(4) 112.8(2) 110.4(4) 92.7(3) 128.8(3) 112.8(4) Si(2)–N(4) Si(3)–N(5) Si(3)–N(6) N(1)–Li(1) N(2)–Li(2) N(4)–Li(1) N(6)–Li(2) Si(2)–N(3)–C(5) Si(2)–N(4)–C(8) Si(2)–N(4)–Li(1) C(8)–N(4)–Li(1) In–N(5)–Si(3) In–N(5)–C(9) Si(3)–N(5)–C(9) Si(3)–N(6)–C(12) Si(3)–N(6)–Li(2) C(12)–N(6)–Li(2) N(1)–Li(1)–N(4) N(1)–Li(1)–N(19) N(4)–Li(1)–N(19) N(1)–Li(1)–N(49) N(4)–Li(1)–N(49) N(19)–Li(1)–N(49) N(2)–Li(2)–N(6) N(2)–Li(2)–N(29) N(6)–Li(2)–N(29) N(2)–Li(2)–N(69) N(6)–Li(2)–N(69) N(29)–Li(2)–N(69) 1.752(5) 1.696(6) 1.757(6) 2.073(8) 2.036(8) 2.232(8) 2.210(9) 118.3(5) 114.8(4) 125.4(4) 112.5(5) 130.3(3) 112.1(4) 117.4(5) 116.6(5) 122.7(3) 113.9(5) 110.4(2) 109.3(6) 103.8(2) 103.8(2) 119.0(6) 110.4(2) 100.7(2) 110.1(6) 114.5(2) 114.5(2) 116.9(6) 100.7(2)J.Chem. Soc., Dalton Trans., 1999, 141–146 143 Å], [In(NPh2)3(py)] [average 2.083(3) Å],1 [In{N(SiMe3)2}3] [2.049(1) Å], [(Me3C)2In{NSiPh3(2,6-i-Pr2Ph)}] [2.104(3) Å],6 [InL3] (HL = 2,2,6,6-tetramethylpiperidine)] [average 2.078(5) Å],8 [MeIn(Nt-Bu)2SiMe2]2 [2.107(3) Å] 13 and [Et2In(NC5H4)] [2.166(4) Å].14 The In–N1 and N19 distances in 1 [average 2.202(8) Å] are slightly shorter than the In–Nbridge distances in [MeIn(NBut)2SiMe2]2 [average 2.267(4)].13 In 2 the interaction of the amide nitrogens N1 and N29 with the lithium cations lengthens their In–N distances about 0.05 Å compared to In–- N3 and In–N5, and causes them to be almost as long as the In–- N1 [2.204(8) Å] and In–N19 [2.199(7) Å] bridging amide distances in 1. The Li–N1 and –N2 distances in 2, which involve the nitrogens associated with the long In–N distances, are significantly shorter than the Li–N distances involving the amine N4 and N6 atoms.Spectroscopic characterization In the 1H NMR spectrum of 2 there are five sharp singlets and a doublet, all with equal intensity, and a quartet with one-third relative intensity. The five singlets arise from three sets of four methyl groups attached to Si and two sets of four methyl groups attached to nitrogen, and the quartet (NH) and doublet (NMe) arise from the amine groups of the [HNMeSiMe2NMe]2 ligands.In the 13C-{1H} NMR spectrum there are six singlets. These data are consistent with the solid state structure (i.e., with the molecule having virtual D2 symmetry). A medium intensity band at 3310 cm21 is observed in the IR spectrum that can be assigned to the N–H stretch. At room temperature the 1H spectrum of 1 consists of three closely spaced singlets in the SiMe2 region in a 1:1:2 ratio and one sharp singlet and two slightly broad singlets in a 1:1:1 ratio in the NMe region (Fig. 3). The SiMe2 singlet of relative Fig. 3 The NMe (left) and SiMe (right) regions of the 1H NMR spectra for [ClIn(NMeSiMe2)2NMe]2 (toluene-d8) recorded at various temperatures. intensity 2 is composed of two accidently degenerate singlets of equal intensity. The 13C-{1H} spectrum has seven singlets, four in the SiMe2 region and three in the NMe region. These data are consistent with the solid state structure.Variable temperature NMR (Fig. 3) was used to determine why two of the NMe resonances in the room temperature spectrum of 1 are broad. As the temperature of the NMR sample is raised, the two broad NMe resonances broaden further, collapse into the baseline at ª60 8C (DG‡ = 16 kcal mol21 at 60 8C),15 and re-emerge at 70–80 8C as a broad hump. At 90 8C, the highest temperature examined, the coalesced resonances are beginning to sharpen back into a singlet. The sharp NMe resonance observed in the room temperature spectrum remains sharp in the entire temperature range examined.In the SiMe2 region, the two separated singlets merge into one peak as the temperature is raised while the accidentally degenerate peaks never separate and presumably merge, thereby resulting in two singlets being observed in the region at high temperatures. Conversely, as the NMR sample is cooled to below room temperature, the two broad NMe resonances sharpen and the resonances in the SiMe2 region sharpen and shift slightly; thus, at 210 8C there are three sharp singlets in the NMe region and four sharp equal intensity singlets in the SiMe2 region, as is consistent with the solid state structure.The variable temperature data indicate that the two amide methyl groups and, separately, two sets of two methyl groups attached to Si of the [MeN(SiMe2NMe)2]22 ligands are made equivalent by a dynamic process. Possible mechanisms to account for the NMR data include a concerted bridge–terminal amide exchange mechanism (Scheme 2, reading bottom to top) or, more likely, a mechanism involving In–N bond opening and rotation that passes through an intermediate with C2 symmetry (Scheme 2, reading top to bottom). A dimer–monomer equilibrium does not account for the data if the reasonable assumptions are made that the monomer would have a C2v trigonal planar ClInN2 core and the amine nitrogen undergoes rapid inversion in the temperature range examined.In contrast to the solution dynamic behavior of 1 the aluminium analog [ClAl(NMeSiMe2)2NMe]2 is reported to be Scheme 2 b a In N In N Cl N Si N Si N Si N Si Cl g' g a 'b b 'a d c' d' c e f e' 'f In N In N Cl N Si N Si N Si N Si Cl e' e c c' d d' b b' a' a g f g' f' In N In N Cl N Si N Si N Si N Si Cl e' e c d' d c' a' b' g f g' 'f In N In N Cl N Si N Si N Si N Si Cl g' e c b' d a' b c' d' a g f e' f' In N In N Cl N Si N Si N Si N Si Cl g' g a b' b a' d c' d' c e f e' f' B A a,a' b,b' e,e' f,f' A A B c',c d',d g',g f',f144 J. Chem.Soc., Dalton Trans., 1999, 141–146 stereochemically rigid on the NMR time scale near room temperature. 10 The related compounds [(py)Zn(NEtSiMe2)2NMe]2, [Be(NMeSiMe2)2CH2]2 (see I) and [MeIn(NBut)2SiMe2]2, however, all exhibit fluxional NMR behavior,11–13 which was attributed, respectively, to a monomer–dimer interconversion, an interconversion among oligomers, and an intramolecular dynamic process.In the room temperature 1H NMR spectrum of a CD2Cl2 solution of 1 there are in addition to the primary resonances discussed above four equal intensity sharp singlets in the SiMe2 region and three equal intensity sharp singlets in the NMe region. The relative intensities of these resonances are approximately 10% of the primary resonances. The resonances are present with about the same intensities in samples prepared from diVerent batches of crystals and from crystals grown from disparate solvent systems as well as when toluene-d8 or benzene-d6 is used as the NMR solvent instead of CD2Cl2, although there is more overlap of the resonances with the primary resonances in the hydrocarbon solvents, e.g., see Fig. 3. The intensities of the resonances do not change nor do the resonances change shape as a function of temperature (e.g., they do not broaden at high temperatures). From the observations it can not be determined whether the resonances are due to an isomer of 1 or a persistent impurity, but an isomer that would plausibly account for the data is III.A referee suggested that one might expect III to be fluxional, which is not observed. Conclusion Indium trichloride reacts with 1 equivalent of MeN(SiMe2NMeLi) 2 to give the dimer 1 and with 4 equivalents of HNMeSiMe2NMeLi to give 2. In the structure of 1, which is isomorphous with the known Al derivative, a chloride and one amide group of a [MeN(SiMe2NMe)2]22 ligand are bonded to each In atom in terminal positions and the other amide group of the chelating ligand is shared between two In atoms.The terminal chlorides have an anti-ClIn ? ? ? InCl arrangement. The amine group of the [MeN(SiMe2NMe)2]22 ligand does not interact with In. Variable temperature NMR spectra show 1 undergoes a fluxional process that makes the bridging and terminal amide groups and, separately, two sets of two methyl resonances of the [MeN(SiMe2NMe)2]22 ligand equivalent at high temperatures.A mechanism involving bridge–terminal amide exchange is proposed to account for the data. The molecule 2 has an adamantane-like Li2In2Si2N4 core with four InNSiNLiN rings fused to the core in such a way as to give the molecule virtual D2 symmetry. Experimental General techniques and reagents All manipulations were carried out in a glove box or by using Schlenk techniques. Solvents were purified by using standard techniques after which they were stored in the glove box over 4-Å molecular sieves until needed.H2NMe was purchased from Matheson and Me2SiCl2 from Aldrich. The former was used as received and the latter was degassed with an argon stream before it was used. The lithium salts of the amines were prepared by reacting the amines in hexanes with the appropriate amount of n-BuLi, washing the resulting solid with hexanes, and then In N In N MeN Me2Si N Me Si Me N Me2Si Cl Me MeN Si Cl Me Me2 Me2 III drying in vacuo.NMR spectra were collected on a 300 MHz instrument. Syntheses The amines MeN(SiMe2NHMe)2 and Me2Si(NHMe)2. These compounds were prepared by using a slight modification of the literature procedure.16 Methylamine was added via a syringe needle over the surface of a stirred solution of Me2SiCl2 (30 cm3, 0.25 mol) in cold (5–10 8C) diethyl ether (300 cm3). The amine addition continued for 4 h during which time a white solid formed. After the amine addition was stopped, the reaction mixture was refluxed for 1 h.The mixture was then cold-filtered (0 8C). The solvent was removed in vacuo from the filtrate, and the residue was fractionally distilled at atmospheric pressure, giving Me2Si(NHMe)2 as a colorless liquid (bp 107 8C at 760 mmHg). Yield, 12 g (40%). Low pressure distillation of the residue gave MeN(SiMe2NHMe)2 as a colorless liquid (bp 70 8C at 0.01 mmHg). Yield, 3.8 g (7.4%). 1H NMR (C6D6) for MeN(SiMe2NHMe)2: d 0.085 (s, 12, SiMe2), 0.28 (broad, 2, NH), 2.37 (slightly br s, 6, NMe), 2.43 (s, 3, NMe). 1H NMR (C6D6) for Me2Si(NHMe)2: d 20.056 (s, 6, SiMe2), 0.21 (broad, 2, NH), 2.39 (d, JHH = 6.6 Hz, 6, NMe). The dimer [ClIn(NMeSiMe2)2NMe]2. A diethyl ether solution (5 cm3) of MeN(SiMe2NMeLi)2 (0.22 g, 1.0 mmol) was added dropwise to a slurry of InCl3 (0.22 g, 1.0 mmol) in cold (278 8C) ether (25 cm3). The mixture was stirred for 24 h while the temperature was allowed slowly to warm to room temperature. A white precipitate formed.The ether was removed by vacuum distillation and the residue was extracted with hexane (3 × 10 cm3). The extracts were combined and filtered through Celite. The hexane was removed in vacuo and the residue, a white solid, was held in vacuo for 24 h. This material is pure product by 1H NMR. Yield, 0.20 g (57%). Colorless cubic crystals can be grown from ether at low temperature (235 8C). A satisfactory nitrogen analysis was not obtained (Found: C, 23.76; H, 6.13; N, 10.83.C14H42N6Cl2In2Si4 requires C, 23.76; H, 6.00; N, 11.88). See the text for complete details regarding the 1H NMR spectra of this compound. 1H NMR (C6D6): d 0.16 (s, 12, SiMe2), 0.20 (s, 6, SiMe2), 0.22 (s, 6, SiMe2), 2.60 (s, 6, NMe), 2.66 (s, 6, NMe), 2.91 (s, 6, NMe). 13C-{1H} NMR (C6D6): d 22.19 (SiMe2), 20.74 (SiMe2), 0.04 (SiMe2), 1.36 (SiMe2), 32.5 (NMe), 33.1 (NMe), 34.3 (NMe). IR (Nujol, CsI, cm21): 1307w, 1257s, 1219w, 1170m, 1151w, 1130w, 1076m, 1045m, 997w, 893m, 856m, 819w, 792m, 760w, 680w, 669w, 642w, 542w, 488w, 453w and 407w.The dimer [Li{In(HNMeSiMe2NMe)2(MeNSiMe2NMe)}]2. A diethyl ether solution (5 cm3) of HNMeSi(Me2)NMeLi (0.50 g, 4.0 mmol) was added dropwise to a slurry of InCl3 (0.22 g, 1.0 mmol) in ether (25 cm3) at room temperature. The mixture was stirred for 24 h and then the ether was removed by vacuum distillation. The residue was dried for 24 h after which it was extracted with hexane (10 × 10 cm3). The extracts were combined and filtered through Celite, and the hexane was removed in vacuo from the filtrate.The residue, a white solid, is pure product by 1H NMR. Yield, 0.45 g (96%). If the reaction is carried out by using a 2 : 1 mixture of HNMeSiMe2NMeLi and Me2Si(NMeLi)2 or only 3 equivalents of HNMeSiMe2NMeLi the yield is about 70%. Colorless crystals of the product can be formed by dissolving the solid in a hexane–ether mixture (1 : 9) and cooling (235 8C for 24 h) (Found: C, 30.03; H, 8.15; N, 17.52.C24H76N12In2Li2Si6 requires C, 30.50; H, 8.12; N, 17.79). 1H NMR (C6D6): d 0.22 (s, 12, SiMe2), 0.36 (s, 12, SiMe2), 0.46 (s, 12, SiMe2), 0.51 (q, J HH = 6.6, 4, NH), 2.21 (d, J HH = 6.6 Hz, 12, NMe), 2.79 (s, 12, NMe), 3.06 (s, 12, NMe). 13C-{1H} NMR (C6D6): d 21.91 (SiMe2), 21.07 (SiMe2), 0.48 (SiMe2), 30.1 (NMe), 34.0 (NMe), 34.6 (NMe). IR (Nujol, CsI, cm21): 3310m, 1246m, 1168m, 1062m, 1026m, 1003m, 854m, 823m, 763m, 690m, 671m, 511w, 474w and 443w.J.Chem. Soc., Dalton Trans., 1999, 141–146 145 Crystal structure determination of [ClIn(NMeSiMe2)2NMe]2 Crystal data. C14H42Cl2In2N6Si4, M = 707.42, triclinic, a = 8.313(2), b = 9.550(2), c = 10.244(2) Å, a = 102.06(2), b = 97.98(2), g = 110.03(2)8, U = 727.4 Å3, T = 23 8C, space group P1� , Mo-Ka (l = 0.71073 Å), Z = 1, Dc = 1.62 g cm23, F(000) = 356. Colorless rods. Crystal dimensions: 0.08 × 0.11 × 0.41 mm, m = 19.2 cm21. Data collection. The crystal was mounted in a capillary under an argon atmosphere. Enraf-Nonius CAD-4F (k geometry) diVractometer.q–2q scan mode with scan width Dq = 0.8 1 0.35tanq, scan speed range 0.67–88 min21, graphite-monochromated Mo-Ka radiation; 1766 reflections measured (38 £ 2q £ 448, h, ±k, ±l), 1766 unique, 1437 observed with F > 6s(F). Lorentz and polarization corrections were applied. A semi-empirical absorption correction was applied based on y scans of 5 reflections having c angles between 70 and 908.Three standard reflections were measured every 3600, and these showed no significant variation. Structure solution and refinement. The Laue symmetry was determined to be 1� , and the space group was shown to be P1 or P1� . P1� was assumed to be the correct setting, which was con- firmed subsequently by successful refinement. The structure was solved by using the MolEN Patterson interpretation program, which revealed the position of the In atom. The remaining non-hydrogen atoms were located in subsequent diVerence Fourier syntheses.The usual sequence of isotropic and anisotropic refinement was followed. Hydrogen atoms attached to carbon were then entered in ideal calculated positions and constrained to a riding motion such that U(H) = 1.3U(attached C). After all shift/esd ratios were less than 0.01, convergence was reached with R, R9 = 0.040, 0.048 (goodness-of-fit = 1.27). The weighting scheme was w = [0.04F2 1 (s(F))2]21. No unusually high correlations were noted between any of the variables in the last cycle of full-matrix least squares refinement, and the final diVerence map showed a maximum peak of about 0.62 e Å23 located near In.All calculations were made using the MolEN package of programs.17 Crystal structure determination of [Li{In(HNMeSiMe2NMe)2- (MeNSiMe2NMe)}]2 Crystal data. 2Li1?C24H76N12Si6In2 22, M = 945.18, monoc = 17.736(6), b = 12.778(4), c = 21.201(7) Å, b = 103.64(2)8, U = 4669 Å3, T = 250 8C, space group I2/a, Mo-Ka (l = 0.71073 Å), Z = 4, Dc = 1.34 g cm23, F(000) = 492.Crystal dimensions: 0.20 × 0.25 × 0.35 mm, m = 11.5 cm1. Data collection. The crystals were handled under mineral oil. The crystal chosen for analysis was transferred to a cold nitrogen stream for data collection on a Nicolet R3m/V diVractometer equipped with an LT-1 low-temperature device, w mode with scan width Dq = 1.30 1 (Ka2 2 Ka1)8, scan speed range 1.5–15.08 min21, graphite-monochromated Mo-Ka radiation; 3329 reflections measured (48 £ 2q £ 458, ±h, k, l), 2327 independent with I > 3s(I).Lorentz and polarization corrections were applied; however, no correction for absorption was made due to the small absorption coeYcient. Two standard reflections were measured every 2 h or after every 100 data points collected, and these showed no significant variation. Structure solution and refinement. The Laue symmetry was determined to be 2/m, and the space group was shown to be Ia or I2/a.Because the unitary structure factors displayed centric statistics, I2/a was assumed to be the correct setting from the outset, which was confirmed subsequently by successful refinement. The structure was solved by using the SHELXTL Patterson interpretation program, which revealed the position of the In atom in the asymmetric unit, consisting of one-half molecule situated about a two-fold axis. The remaining nonhydrogen atoms were located in subsequent diVerence Fourier syntheses.The usual sequence of isotropic and anisotropic refinement was followed. Hydrogen atoms attached to carbon were then entered in ideal calculated positions and constrained to a riding motion with a single variable isotropic thermal parameter for the SiMe3 hydrogens and a separate variable for the NMe hydrogens. The two amino hydrogens were located in difference maps and allowed to refine with distance constraints. All non-Li atoms occupy general positions, and both the Li atoms lie in special positions on a two-fold axis.The isotropic thermal parameters of both Li atoms refined to unreasonably small values (average 0.002 Å2); therefore, in the final least squares refinement the Li isotropic thermal parameters were fixed. The possibility that some other cationic species occupies the Li positions was considered because of the irregularity in the Li refinement. Based on the way in which the compound was synthesized the only other species that can reasonably be considered to occupy the Li positions is Na.Whether the atoms are Li or Na, charge balance requires the cation-to-In atom ratio be one (i.e., the positions cannot be half occupied). The possibility that the salt contains Na rather than Li was excluded for the following reasons: (1) a search of the Cambridge Crystallographic database for Li–N and Na–N distances where nitrogen is attached to at least two carbon atoms revealed that the distances are in the range 1.89–2.56 and 2.34–3.44 Å, respectively. In the present case, the Li–N distances range from 2.04 to 2.23 Å. This suggests that Li at 100% occupancy is the more reasonable choice.(2) The compound was synthesized in high yield using the Li salt HNMeSi(Me2)NMeLi. A high yield would not be expected if the Na came from a contamination source, such as the Li amide salt, Celite filter aid or glassware. (3) When Na was refined in the Li positions the isotropic thermal parameters became unreasonably large (average 0.14 Å2).After all shift/esd ratios were less than 0.2, convergence was reached with R, R9 = 0.034, 0.036. The weighting scheme was w = [s(F)]22. No unusually high correlations were noted between any of the variables in the last cycle of full-matrix least squares refinement, and the final diVerence map showed a maximum peak of about 0.7 e Å23 located 0.47 Å away from Li(1). There was also a peak of about 0.5 e Å23 located 0.7 Å away from Li(2).Calculations were made using Nicolet’s SHELXTL PLUS (1987) package of programs.18 CCDC reference number 186/1244. Acknowledgements Acknowledgement for support is made to the Robert A. Welch Foundation (S.G.B. and D.M.H.), The Energy Laboratory at UH (D.M.H.), and The Institute for Space Systems Operations (D.M.H.). We thank Dr. James Korp for his technical assistance with the crystal structure determination of [Li{In(HNMe- SiMe2NMe)2(MeNSiMe2NMe)}]2. References 1 J. Kim, S. G. Bott and D. M. HoVman, Inorg. Chem., 1998, 37, 3835. 2 S. Strite and H. Morkoç, J. Vac. Sci. Technol., B, 1992, 10, 1237; S. Strite, M. E. Lin and H. Morkoç, Thin Solid Films, 1993, 231, 197. 3 R. G. Gordon, D. M. HoVman and U. Riaz, Mater. Res. Soc. Symp. Proc., 1991, 204, 95. 4 R. G. Gordon, D. M. HoVman and U. Riaz, Mater. Res. Soc. Symp. Proc., 1992, 242, 445. 5 H. Bürger, J. Cichon, U. Goetze, U. Wannagat and H. J. Wismar, J. Organomet. Chem., 1971, 33, 1. 6 M. A. Petrie, K. Ruhlandt-Senge, H. Hope and P. P. Power, Bull. Soc. Chim. Fr., 1993, 130, 851. 7 G. Rossetto, N. Brianese, A. Camporese, M. Porchia, P. Zanella and R. Bertoncello, Main Group Met. Chem., 1991, 14, 113.146 J. Chem. Soc., Dalton Trans., 1999, 141–146 8 R. Frey, V. D. Gupta and G. Linti, Z. Anorg. Allg. Chem., 1996, 622, 1060. 9 M. Veith, M. Zimmer and S. Müller-Becker, Angew. Chem., Int. Ed. Engl., 1993, 32, 1731; Angew. Chem., 1993, 105, 1771. 10 U. Wannagat, T. Blumenthal, D. J. Brauer and H. Bürger, J. Organomet. Chem., 1983, 249, 33. 11 A. J. Elias, H.-G. Schmidt, M. Noltemeyer and H. W. Roesky, Organometallics, 1992, 11, 462. 12 D. J. Brauer, H. Bürger, H. H. Moretto, U. Wannagat and K. Wiegel, J. Organomet. Chem., 1979, 170, 161. 13 M. Veith, H. Lange, O. Recktenwald and W. Frank, J. Organomet. Chem., 1985, 294, 273. 14 M. Porchia, F. Benetollo, N. Brianese, G. Rossetto, P. Zanella and G. Bombieri, J. Organomet. Chem., 1992, 424, 1. 15 H. Günther, NMR Spectroscopy—An Introduction, Wiley, Chichester, 1980, ch. VIII. 16 L. W. Breed and R. L. Elliott, Inorg. Chem., 1964, 3, 1622. 17 MolEN, An Interactive Structure Solution Program, Enraf-Nonius, Delft, 1990. 18 G. M. Sheldrick, SHELXTL PLUS, Release 3.4 for the Nicolet R3m/v Crystallographic System, Nicolet Instrument Corp., Madison, WI, 1987. Paper 8/07041E