ELECTRONIC PROPERTIES OF MONOCAPPEDMONOCAPPED PRISMANE AND BASKET IRON-SULPHURIRON-8ULPHUR CLUSTERS
B. S. SNYDER, M. S. REYNOLDS and R. H. HOLM* Department of Chemistry, Harvard University, Cambridge, MA 02138, U.S.A.
and
G. C. PAPAEFTHYMIOUPAPAEFfHYMIOU and R. B. FRANKELTFRANKELt Francis Bitter National Magnet Laboratory, Massachusetts Institute of Technology,Technology, Cambridge, MA 02139, U.S.A.U.S.A.
Abstract-The ground state electronic properties of the mixed-valence clusters Fe7S6Fe7S6 (PEq4Cl,(PEt&Cl? (1),(l), Fe6S6(PEt,)4L2Fe,&(PEt&L2 [L = Cl- (2).(2), Br- (4) (4), I-I~ (5) and PhSPhS- (6)], Fe&Se,Fe6Se6 (PEt3)4CI2(PEt3)&12 (3) and [Fe[Fe,S,(PEt),]‘-6S 6(PEt) 6] 1- (9), have been investigated by magnetic susceptibility, magnetization and MossbauerMiissbauer spectroscopic measurements. Cluster 1 has a (idealized C,,)C3v) monocapped prismane structure and clusters 2-6 and 9 adopt the CtiC 2v “basket”"basket" 1 configuration based on the [Fe[Fe6(~2-S)@3-S)4@4-S)]2+,1+6(J.lrS)(j.trS) 4(j.t4-S)] 2+, + core unit. From magnetic prop-prop erties, the ground states S = 1/2l/2 (1, 9) and S = 1 (2-5) were established. Unlike thethe other clusters, 6 did not show a Curie region of susceptibility; its ground state was not directly determined but is probably S = 1. MossbauerMiissbauer spectra were successfully analysed in terms of a 1:: 1:: 1 ironiron site population,population. Isomer shifts and quadrupole splittings were assigned to each site. Magnetically perturbed spectra obtained in applied fields of 60-80 kOe were analysed toto give the magnetic hyperfine parameters and magnetic hyperfine fields in clusters 1-51-5 and 9. The magnetic spectra demonstrate antiantiferromagneticferromagnetic spin coupling which affords thethe indicatedindicated ground states.
NearlyNearly allall iron-sulphuriron-sulphur clusters are characterized ments. For example, the ground spin states and byby corecore structuresstructures thatthat are built up entirely by verver- associated matters of the electronic structures of tex-sharingtex-sharing and/orand/or edge fusionfusion of Fe2S2 rhombs. II*2,2 cubane-type [Fe4S4(SR)4P-,2-,I-[Fe4S4(SR)4]3-.2-“- clusters have TheThe prototype singlesingle planar rhomb isis found in the been eluciated in some detail by these methods.6-8methods.6-8 binuclearbinuclear speciesspecies [Fe2S2L4]2-.[Fe2S2L412-. Clusters proceed in All of these clusters have the common property of complexitycomplexity fromfrom thisthis typetype toto thosethose with nuc1earitiesnuclearities molecular antiferromagnetism. 3,3, 4,4, 6,6, 7,7, 88 andand 18.18.22*3,3 Excluding organometallics As part of our work on higher-nuclearity Fe-S andand with thethe exceptionexception Of[of [Fe&(PEt,),]“~”Fe6S8(PEt3)6F+,I+ 4,54V5 in clusters, we have recently prepared and described thethe preceding set,set, thesethese clusters have as terminal the structures and certain reactivity aspects of a ligands,ligands, halide, RS-RS or RO-. The electronic propprop- new series of hexa- and heptanuclear clusters. 3,9-12 ertieserties ofof manymany ofof thesethese materials have been probed These differ from the above in having a mixed byby applicationapplication ofof MossbauerMiissbauer spectroscopyspectroscopy and by ligandligand set that includes tertiary phosphines and core magnetizationmagnetization andand magneticmagnetic susceptibilitysusceptibility measure- structures not yet found in the absence of phosphine. With reference to Fig. 1, the core geometry of Fe7S6(PEt3)4CI/Fe7S6(PEt3)4C139 (1) approaches that ...*AuthorAuthor toto whomwhom correspondencecorrespondence shouldshould be addressed. ofa monocapped hexagonal prism, while that of Fe,Fe6 3 tt PresentPresent address:address : DepartmentDepartment of Physics, California S6(PR3)4L2,Sb(PR&L2, L = halide I‘O*” 0,1 1 (2-5, 7) and thiolate3thiolate PolytechnicPolytechnic StateState University,University, SanSan Luis Obispo, CA (6, 8), and [Fe6S6(PEt3)6]1+[Fe,S,(PEt1)h]‘+12 12 (9) resembles a 94307,94307, U.S.A.U.S.A. basket, with the lower Fe-S-Fe fragment, which 1 2-5,7 6396,8
9 110 0 Fig. 1.1. Schematic structures of clusters 1-10l-10 illustrating the monocapped prismane (1) and basket (2-10)(2-10) topologies.topologies. Iron atoms are labelled so as to correspond to the site assignments in Table 2.
isis not part ofthethe aforementioned rhomb, serving as Their unique core structure suggests that thethe thethe handle. We designate these as "basket"“basket” clusters, basket clusters may have unusual electronic features. asas a means of distinguishing their [Fe6S6][Fe&] core toptop- Ground state properties, previously uninvestigated, ology fromfrom thethe idealizedidealized hexagonal prisms present have been examined by magnetic and MiissbauerMossbauer inin thethe "prismane"“prismane” clusters, [[Fe6S6L6]2-~3-.‘3-‘6Fe6S6L6F-,3-. 13-16 The techniques, and the leading results are reported here. structuresStlllCtUlW ofOf Fe6S6(PBu3)4CI2,FC&j(Ph~)~C12,‘1 II Fe6S6(PEt3)4(S-PFe6S6(PEt3)&F CC6H4Br)236H 4Br)z3 and [Fe[Fe6S6(PEt)6]‘+6S 6(PEt) 6] 1+ 12l2 define the basket EXPERIMENTAL topology,topology, which isis manifested in virtually isometric [Fe(1l2-S)(llrS)4(1l4-SW+,I+[Fe(~2-S)(~3-S),(~~-S)]2+~‘+ core units. This unique Preparation ofcompound9compounds topologytopology isis supported,supported, inin part, by the unusual coorcoor- dination at thethe FeS3PFeS,P sites. Whereas FeS3LFeS,L sites in The compounds Fe7S6(PEt 3)4C13 (1),9 Fe6S6 thesethese and numerous other clusters have a distorted (pEt3) 4Cl2 (2), II Fe6Se6(PEt3)4CI2 (3), II Fe6S6 tetrahedraltetrahedral configuration, the phosphine-ligated (PEt3)4Br2 (4), II Fe6S6(pEt3)412 (5),11 Fe6S6 sitessites exhibit a roughly trigonal pyramidal geometry, (PEt3MSPh)2 (6),3 Fe6S6(PEt3MS-p-C6H4Br)z with thethe Fe atom closely approaching the S3 plane. (8)3 and [Fe6S6(PEt3)6](BF4) (9)12 were prepared as These speciesspecies are also among the most reduced previously described. Fe-S clusters, with Fe mean oxidation states of ++2.142.14 inin Fe7S6(PEt3)4Cl3,Fe&(PEt3)&13, ++2.172.17 in [Fe6S6[Fe& Physical measurements (PEt(PEt&]‘+3) 6] I+ and ++2.332.33 in Fe6S6(PEt3)4L2.Fe6S6(PEt&L2. As a resultresult of theirtheir high nuclearities, unusual ligand sets All measurements were performed under strict and previously unencountered core structures, these anaerobic conditions. Magnetic susceptibility and molecules are of current interest as the newest magnetization measurements at applied fields of 10 additionsadditions toto thethe extensive family of Fe-S clusters, kOe and up to 50 kOe, respectively, were carried which now encompasses seven nuclearities and 10 out on a SHE 905 SQUID magnetometer operating corecore structuralstructural types.types. 1-3I-3 Most of these clusters exist between 1.8 and 300 K. Solid state measurements inin twotwo ormore oxidation levels, some ofwhich have were made on finely ground polycrystalline samples been isolated.isolated. (15-35(1535 mg) loaded into precalibrated containers and sealed with epoxy resin under a di~~og~-he~~dinitrogen-helium Fe”’Felli sites, ifif present, cannot bebe recognizedrecognized unam-unam atmosphere. Diamagnetic susceptibility correctionsicorrections 17 biguously fromfrom crystallographic data.data. AsAs willwill bebe were were applied. MlissbauerMossbauer spectra were determined seen, *HIH NMR spectra areare consistentconsistentwithwithretentionretention with a constant-accelerationconsent-~lemtion spectrometer equipped of thethe basket structure of halidehalideclustersclustersininsolution.solution. with a 57COs7Co source in a Rh matrix. Zero-field measure-measure Fe&(PEt&Cl,Fe7S6(PEt3)4Ch (1);(1); Fe6S6(PEt3).&12Fe6S6(PEt3)4Clz (2);(2); Fe6Se6Fe6Se6 ments were made between 4.2 and 300 K, with thethe fPEt&&(PEt3)40Z (3)(3); ; Fe~S~~Et~~~r~Fe6S6(PEt3)J)rZ (4)(4); ; Fe~S~~Et~~~I~Fe6S6(PEh)4I2 spectrometer operating in the time mode and the (5); Fe~S~(PEt~)~(SPh)~Fe6S6(pEt3MSPhh (6);(6); k&(PBu~MAFe6S6(PBu3)4CIz (7)(7); ; source maintained at room temperature. Magnetically Fe6S6(PEt3)XS-p-CgH4Br)2Fe6S6(PEt3)4(S-p-C6H4Br)2 tS>(8);; V%S6(PEt3)611+[Fe6S6(PEt3)6] 1+ perturbed spectra were obtained in lon~tudinallylongitudinally (9.(9). applied fields up to 80 kOe, with the source and K. absorber at 4.2 K. PolycrystallinePolycrystaIline samples were dis-dis Electronic properties persed in boron nitride powder and sealed with epoxy resin in plastic sample holders. ‘isomerIsomer shifts are Ground spin states, other magnetic aspects andand reported relative to Fe metal at 4.2 K. EPR electronic distributions inin clusters l-71-7 havehave beenbeen spectra were recorded at X-band frequencies at ca examined using thethe methods of magnetic sus-sus 19 K on a Varian E-I09E-109 spectrometer. Solution ceptibility, magneti~tionmagnetization and 57Fe MijssbauerMossbauer susceptibilities were determined by a NMR spectroscopy. Related results for cluster 9 have been method. 18’ * Solvent susceptibilities were taken reported”reported 12 and certain data are includedincluded here forfor from the literatureliterature” 19 and corrections for solution comparison purposes. Before each measurement, density changes with temperature were applied. the purity of a given compound was established from its distinctive, isotropi~llyisotropically shifted ‘HIH NMR spectrum (vide infra) previously recorded on an RESULTS AND DISCUSSION spectrum (vi& infia) previously recorded on an analytical sample. 3*11,123.11.12 Magnetic and MiissbauerMossbauer The following clusters, 1-9,l-9, are of primary inter-inter results are contained in Tables 1 and 2 respectively,respectively, est inin this investigation. The monocapped prismane and selected data are displayed inin Figs 2-7. The structure of 1 has been crystallographically estabestab- numerical designation of Fe sites inin Table 2 and lished.lished.’9 The basket configurations are assigned to Fig. 1correspond. The Curie constant in the Curie-Curie 2--62-6 on thethe basis of the crystal structures of 7 and Weiss law, xMXM = C/(T-Cf(T- @),8), has the values 8, and thethe correspondence of spectroscopic propprop- C = 0.375, 1.0001.OOO and 1.875 emu K mol- ’I for pure erties toto thosethose of the structurally defined clusters. S = 1/2,l/2, S = 1 and S = 3123/2 states, respectively, All clusters are mixed-valence but individual FellFe” or whengwhen gee = 2.
TableTabIe 1. Magneticparetic properties of clusters X-7”1-70
Cc lIefT ClusterCluster TT(K)" (K)b (emu K mol- I)‘) S (PB) e(K) Ue
Fe,S6(PEtFGWEt3Mh3) 4Cl3 &-456-45 0.581 1/2If2 2.16 -1.88 --- CDCI 3soln cd CDC1 3 soln 298 - 3.953.95c.d. - - - I Fe6S6(PEt3)4CI2Fe&(PW&% 6-150 0.993 I 2.82 -4.49 1.89 --17.3 17.3 0.18 CDCD&&2Cl 2 soinsoln 190-302190-302 - 2.74--2.812.74-2.81” c.e - - - - Fe6Se6(PEt3)4Fe~~~~Et~)~~l~Cl2 6-75 1.211 1 3.11 -1.28 2.06 -2.27- 2.27 0.23 CDCD&&2C1 2 solnsoln 190-302190-302 - 3.30-3.44 a“All All datadata referrefer toto thethe solidsolid statestate unless otherwise indicated.indicated. h‘Curie Curie region.region. M c‘Calculated Calculated fromfrom thethe CurieCurie law,law, Jleffg& = = 2.83(x2.83(~~7’)T) 1‘l’;12; moments increase monotonically with increasing tem~ratur~.temperature. d.ed’“Comparison Comparison solution/solidsolution/solid statestate moments at 300-302 K: d3.95/4.09; e2.81/3.04. ffpcg lIefT == 3.401183.40~~ inin solidsolid state,state, 200 K. K. gToluene9Toluene solution.solution. h*ReferenceReference 12.12. Table 2. Mossbauer spectroscopic resultsresults at 4.2 K AC AX’x A y Az #0(jo,b b Cluster Cluster Site AEpb!1EQ Wd(Hx)d (Hy) (Hz) 1'1' I (mm(mms- s- ‘)) (mm S-I) Fe,S6(PEt3)4CI 3 1 0.67 0.51 0.005 0.748 0.649 0.79 (-0.37)( -0.37) (-55.0)(-55.0) (-47.7)(-47.7) 2 0.36 0.870.87 -0.008 -0.161 -0.231-0.231 1.00 (0.59) (11.8)(11.8) (17.0)(17.0) 3 0.36 0.280.28 -0.007 -0.001 -0.966-0.966 0.24 (0.52) (0.07) (73.2)(73.2) Fe6S6(PEt3)4CI2 1 0.63 0.600.60 --U201.120 -0.397 -0.058-0.058 1.oo1.00 (88.7) (10.0) (7.7)(7.7) 2 0.34 0.87 0.440 0.694 0.939 0.22 (-34.9)(-34.9) (-17.4)( -17.4) (-125)(-125) 3 0.32 1.10UO 0.448 0.090 0.023 0.81 (-35.5)(-35.5) (-2.3) (-3.1)( -3.1) Fe6SeiPEt3)4CI2 1 0.64 0.65 -1.013 -0.614 -0.318 0.00 (135) (78.3)(78.3) (44.7) 2 0.35 1.08 0.882 0.567 0.616 0.40 (-117)(-1l7) (-72.3)(-72.3) (-86.7)(-86.7) 3 0.40 1.29 1.293 0.843 0.578 1.oo1.00 (-172)( -172) (-(-107)107) (-81.3)(-81.3) Fe6S6(PEt3)4Br2 1 0.61 0.56 -0.019 -0.017 -0.364 0.00 (0.77) (0.06)(0.06) (44.3) 2 2 0.33 0.79 0.426 0.269 0.121 1.00 (-17.4) (-0.91)(-0.91) (-( -14.7)14.7) 3 3 0.29 1.04 0.450 0.013 0.096 0.00 ((--18.3)18.3) (-0.04)(-0.04) (-11.7) Fe6S6(PEt3)412 1 0.60 0.53 -0.009- 0.009 --0.0070.007 --0.0370.037 0.03 (0.21) (0.00)(0.00) (3.05) 2 2 0.38 0.84 0.665 0.015 0.258 0.47 ((--15.4)15.4) (0.00) (-21.3)(-21.3) 3 3 0.23 0.95 0.175 0.008 0.120 0.47 ((-4.06)-4.06) (0.00) (-9.90)(-9.90) Fe6S6(PEt3MSPhh 1 0.43 0.61 - 2 0.39 0.86 - - - 3 0.37 U51.15 - - - [Fe6S6(PEt3)6](BF 4) 1 0.41 0.53 0.564 0.588 0.459 0.00 ((-41.5)-41.5) (-43.2) (-33.7) 2 0.41 0.88 0.440 0.359 0.418 0.00 ((-32.3) -32.3) (-26.4) (-26.4) (-30.7) 3 0.38 1.25 0.027 0.023 0.018 1.001.oo ((-1.98)-1.98) (-1.69)(- 1.69) (-1.32)(- 1.32) FeCIFeCl,(PEt,)/lPEt3){ - 0.65 2.90 - - oaRelative toto Fe metal at 4.2 K. K. bSpectral‘Spectral fitsfits made with lineline widths inin thethe range 0.2~.340.25-0.34 mm S-Is- ’ and fJq = O. 0. C‘MagneticMagnetic hyperfinehyperfine parameters forfor thethe 14.414.4 keYkeV excited state of 5'Fe.5’Fe. dMagneticdMagnetic hyperfinehypertine field,field, kOe;kOe ; negative hyperfine fields oppose applied field. C‘Asymmetry Asymmetry parameters,parameters, (V(V,,-V,,)/..-Vyy)/V V,.zz . 1200f200 K.K. - 150 . . . c: o 001 5050 100loo 150150 200200 250250 300?a( T(K) T(K) -a..... o 1.0 l/l .c« 08 0.606 M(lle)M(IQ) 0.40.4 0.2 HITH/T (kOe/K) -2.0 0 2.0 Velocity (mm/s)(mm/sJ Fig. 2. Magnetic properties of cluster 1. Upper: temtem- perature dependence of the reciprocal molar magnetic Fig. 3. MijssbauerMossbauer spectra in zero applied magnetic field susceptibility.susceptibility. Lower:Lower : magnetization at applied fields of of clusters 1, 2, 6 and 9 (top to bottom). Solid lines are 12.5,12.5, 25 and 50 kOe at T = 1.8-25 K. Solid lines are fits to the data using the parameters in Table 2. calculatedcalculated Curie behaviour for S = 1/2l/2 and theoretical fitsfits toto thethe magnetization data using the parameters in Table 1. tive deviation of the Curie constant fromfrom thethe S = 1/2l/2 value and the small shift of magnetization data in the lower H/T region from the best fit for (1) Fe7S6(PEt3)4CI3 data in the lower H/T region from the best fit for that state. The collective results support an S = l/21/2 (a) Magnetism. This odd-electron cluster for-for ground state. The negative deviation of thethe sus-sus ll III mally contains 6Fe6Fe”++ FeFe”‘.• As shown in Fig. 2, ceptibility from the Curie-Weiss lawlaw above 45 K cluster 1 obeys the Curie-Weiss law at 6-45 K, with indicates the population of excited levelslevels with C = 0.581 emu K mol-I.mol- ‘. The magnetization was S>S > 1/2.l/2. At 298 K, the essentially equal solution found to be field-independent and to follow a and solid state magnetic moments of ca 4~s4J1-B show S = 1/2l/2 Brillouin function with saturation asympasymp- that the antiferromagnetic coupling (vi&(vide injkz)infra) isis a totically approaching g&sgeSJ1-e = 1~s.IJ1-e. The EPR spec-spec molecular property. trum in frozen THF solution at 20 K (not (b) MiissbauerMossbauer spectra. The monocapped pris-pris shown) reveals a nearly isotropic signal atg,atg e = 2.02, mane structure of 1 under idealizedidealized CsvC 3v symmetrysymmetry consistent with a doublet state. Also present are contains threethree Fe sitessites: : Fe(Fe(1)(J1-rS)(J1-4-S2hCl, l)&-S)&&)$l, weaker resonances at ge = 5.2 and 5.5, which may Fe(2)&-S),(p,-S)PFe(2)(J1-rSh(J1-4-S)P and Fe(3)(p.,-S)3P.Fe(3)(J1-4-ShP. The Fe(2,3)Fe(2,3) be indicative of a small amount of the S = 3/2 sites have approximate trigonaltrigonal pyramidalpyramidal geo-geo state.state.* 8 Physical mixtures of these spin states have metry while thethe Fe(l)Fe(1) site isis more nearlynearly tetra-tetra been observed with Fe4S4Fe& cubane-type clusters.*clusters. 8 hedral. The Mossbauer spectrum inin zerozero field,field, This situation could account (in part) for the posi-posi shown in Fig. 3. was fitfit with a constrainedconstrained sitesite ...."'. 250 200 c: M o 1/X 150 e-O ~ 100 « 50 04.00 I.DO VelocityVelocity (mm/s)(mm/s) T(K) Fig.Fig. 4.4. MossbauerMiissbauer spectrumspectrum of cluster 11 inin an applied Fig. 5. Temperature dependence of the reciprocal molar fieldfield ofof 6060 kOe.kOe. TheThe solidsolid lineline isis a theoreticaltheoretical fitfit toto the susceptibility of cluster 2. The solid line is a Curie-Weiss datadata usingusing thethe parameters inin Table 2 with S = 1/2.l/2. fit to the data using the parameters in Table 1.I. population of 3:3 : 3 :: 1. This allowed immediate (2) Fe6S6(PEt3)4X2Fe6SS(PEt&X2 (X = Cl-, Br-,Be, I-,1-, PhS-) and identificationidentification of thethe Fe(3) site as that with isomer FeFe,Se6(PEt3),C126Se6(PEt 3)4Cl 2 shiftshift ~6 = 0.36 mm S-Is- ’ and the (unusually) small (a) Mugnetism.Magnetism. This set consists of thethe even-elec-even-elec quadrupole splitting AEAE,Q = = 0.28 mm s- 1.‘. The Fe(2) sitesite isis common to 1 and to basket clusters 2-8, 2-8, tron basket clusters 2-6 which formally contain II III where itit also approaches trigonal pyramidal stereostereo- 4Fe”4Fe ++2Fe 2Fe”‘.• As will be evident, certain electronic chemistry. On the basis of this commonality, the aspects of these clusters are unusual and cannot be Fe(2) site isis assigned the doublet with the same satisfactorily interpreted. Clusters 2-5 display Curie isomer shift as Fe(3), but a larger quadrupole split-split behaviour of magnetic susceptibilities over appreci-appreci ting. This leaves the Fe(1)Fe( 1) site, at which chloride is able temperature intervals but inin several cases with bound, as that with the largest isomer shift (0.67 large Weiss constants (Table 1). Curie constants areare mm s- I).‘). This assignment is consistent with data fairly close to the value for a S = 1 state. The results for cluster 2, which are fairly typical for for the pairs ((Me4N),[FeC14]20/Fe(PEt3)2C12Me4Nh[FeCI4]20/Fe(PEt3hCI2 ((:::::: z 0.1O. I compounds in this set, are shown in Fig. mm S-I)s- ‘) and (Me4N)[FeCI4]2°/Fe(PPh3)CI/I(Me4N)[FeC14]20/Fe(PPh,)Cl~2’ 5. Devi (::::::(Z 0.05 mm S-I),s- I), for which the indicated isomer ations from Curie behaviour are such as to indicate the population of higher spin states or a small TIP shift differences correspond to a decrease in 6~ upon replacing chloride with phosphine.*phosphine.* contribution. Magnetic M6ssbauerMijssbauer spectra (not shown) exhi-exhi Magnetization data for clusters 2-5 obtained at bited movement of inner and outer features inin 1.8-100 K and at three applied fields are presented opposite directions as the applied field was in Fig. 6. All show nested curves corresponding to increased from 60 to 80 kOe, consistent with anti-anti different values of the applied field and indicating parallel spin coupling. The spectrum at 60 kOe isis that the zero-field splitting parameter D # O. The shown in Fig. 4, together with a theoretical fit data were analysed under the simple spin Hamil obtained using the method described below forfor thethe tonian (1) basket clusters and the parameters in Table 2. In basket clusters and the parameters in Table 2. In H = D[S/-S(S+ 1)/3] +E(S/-Sy2) +gellBH ° S, this way, the S = l/21/2 ground state assignment was (1) corroborated. with S = 1 by calculating the spin projection along the magnetic field direction; E is the rhombic split ting parameter and the other symbols have their * Note, however, thatthat thethe structurestructure of Fe(PPhS)Cl,Fe(pPh 3)CI 3 hashas usual meanings. The experimental data were simu not been established. lated from the Curie to the saturation region by ~ 2.0 ... .. Fe&ee(PEt&CI2 : -. I .o T 0 I .o c: 0.5 o -...a. o ^m (/l .0 s O r 1.0 « Fe,See(PEt,),CI, 5 1010 15 20 1 -5.0 0a 5.0 HITH/T (kOe/K)(kOe/K) Velocity (mm/s) Fig. 6. Magnetization behaviour of clusters 3,3,2,4 2, 4 and 5 Fig. 7. Mossbauer spectra of clusters 5, 4, 2 and 3 (top (top(top toto bottom). For each cluster, curves in ascending to bottom) at 4.2 K in an applied field of 80 kOe. The order refer to applied fields of 12.5, 25 and 50 kOe. solid lines are fits to the data using thethe parameters inin The solid lines are theoretical fits to the data using the Table 2 with S = 1. parameters in Table 1. least-squares fits, using a simplex algorithm and the magnetization measurement of a 41 mM solutionsolution parameters in Table 1. In order to obtain satis-satis of 4 in dichloromethane yielded essentially thethe samesame factory fits, it was necessary to use D < 0 and result as thethe polycrystalline sample. E/D > 0 with inclusion of ge ##- 2. The algorithm While thethe foregoing treatmenttreatment of magnetization allowed for simultaneous fits of the data at the threethree isis obviously phenomenological, itit isis presentedpresented asas aa applied fields. means of comparing and contrastingcontrasting thethe behavioursbehaviours The behaviour of cluster 3 is the simplest, inin of clusters 2-5. The largelarge ZFSZFS areare notnotwithoutwithout pre-pre that the saturation magnetization of 2,~~2PB is nearly cedent and are seeminglyseemingly withinwithinphysicalphysicalreasonreasonforfor reached and the data are well fit with unexceptional 24.2-4. While no meaningful comparisoncomparison cancanbebemademade parameters for a spin-triplet, including a zero-field with other systems,systems, owingowing toto thethe uniquenessuniqueness ofofthethe splitting (ZFS) of - 2.27 cm-cm- I. However, thethe satu-satu basket clusterclustertopology,topology, wewe notenotethatthatapparentapparentZFSZFS ration magnetization of the remaining clusters (2,(2, values of upup toto caca 2020 em-I’ havehave beenbeen observedobserved forfor 4,s)4, 5) becomes increasingly and markedly suppressed certain S = 5/25/2 Fe”’FeIII porphyrins,22.23 and that for in the order X = Cl- > Br- > II,1-, reachingreaching a mini-mini two S = 1 planarplanar Fell complexes values near 7070 mum with cluster 5 (0.29~~(0.29PB at 50 kOe). Fits requiredrequired cn-em-I’ havehavebeenbeen obtained.24,25obtained.24,25 AllAllofofthese these areare obvi-obvi increasing large negative D values, which areare con-con ouslyouslyapparentapparent values, dependentonon thetheoreticaltheoretical sistent with thethe respective Weiss constants. The sup-sup formalism.formalism. (For(For a discussiondiscussion ofofthisthis point, cf.cf. ref.ref. pression effect isis intramolecularintramolecular inin origin becausebecause aa 22.) InTn the presentpresent case.case, wewe areare applyingapplying a spinspin Hamiltonian formalism to a general case where itit 2 and 3 show a near-Curie dependencedependence ofof magneticmagnetic has not been otherwise testedtested; ; viz. to polynuclear susceptibilities with values consistentconsistentwithwith anan SS == 11 spin-coupled systems with incompletely quenched ground state. Ranges ininmagneticmagneticmomentsmoments overover thethe orbital angular momentum and probable spin-orbit temperaturetemperature rangerange are given inin TableTable 1. TheThe largerlarger interactions at local metal sites. magnetic moments of selenideselenide clustercluster 33 vsvs sulphidesulphide Thiolate cluster 6 is the most anomalous member cluster 2 inin solution andand inin thethe solidsolid statestate (Table(Table 1)1) of the set. In the solid state it does not exhibit a are consistent with an earlier observation of smallersmaller Curie region (nor any indication of a Curie para-para antiferromagneticantiferromagnetic coupling inin Fe&Se,Fe4Se4 thanthan inin Fe&,Fe4S4 magnetic impurity), is very weakly paramagnetic at clusters. 26 The variable-temperature ‘HIH NMRNMR 6 K and shows nesting ofmagnetization curves with spectra of clusters 2 and 3 are presented inin FigsFigs 88 a nearly zero saturation magnetization (0.1~~(O.I,uB at 50 and 9, respectively. The occurrence of twotwo equallyequally kOe; data not shown). It is conceivable that this intense methylene and methyl resonancesresonances forfor 2,2, species represents an extension of the behaviour from the inequivalent sites Fe(2,3), over thethe 19&190 of 2-5, but from the information available we are 300 K range, demonstrates retentionretention of thethe solidsolid unable to demonstrate the ground state spin. More-More state structure inin solution. The spectraspectra of 33 areare over, in toluene solution, ,ueffpeL,ff = 2.72pB,2.72,uB' essentially similar but thethe methyl resonancesresonances areare notnot wellwell the same as the solution moment for 2, for which a resolved. Line broadening at thethe lowerlower temperaturestemperatures triplet ground state has been established. On this is ascribed to the effect of increasing cluster para-para limitedlimited evidence, the S = 1 ground state designation magnetism rather than any dynamic process. The has been entered in Table 1. In the solid state the isotropic shifts of 3 are considerably largerlarger thanthan core structures of 8, the magnetic properties of those of 2, the difference being 32% at 300 K. The which are quite similar to those of 6, and also of 7, effect is qualitatively consistent with thethe magnetic which bears a similar relationship to 2, are virtually susceptibilities and suggests thatthat thethe isotropicisotropic shifts congruent. 3,3,”II It is unlikely, in either the solid or have a significant contact component. Other evi-evi solution states, that the magnetic differences dence for a contact contribution isis thethe alternating between 6 and other basket clusters arise from large phenyl proton shifts of cluster 6.36. 3 However, thethe core structuralstructural differences. shifts of 2 and 3 do not scale linearlylinearly with thethe (b)(b) Solution properties. In dichloromethane solusolu- susceptibilities, as would be the case for pure con-con tionstions at 190-30219&302 K, magnetic moments of clusters tact shifts. This indicates the presence of dipolar Jl--~~-~·_-_~300K 0.14 -0.15 -6.59 -7.72 ppm Jl~_~JLJ-.~ ~260K __ 26OK -0.' -~6 -9.b7-9.b -10!74d7.l --A -=22=.c220O --'-'-KK I I I -0.57-0.157 -1.61 -1.61 -12.67-12.L7 -15~31-1dl ISOK lJ-J\'--- ~~_-----:-__ I I I I -0.87 -2.37 -16.45 -20.28 HHoo -- Fig.Fig. 8.8. TemperatureTemperature dependence of the the 'H‘H NMR spectra of cluster 2 in CDCDzClz2Cl 2 solution; chemical shiftsshifts are indicated.indicated. i bWPEbhC12 3OOK300K JJ ppm '10~71 0.B1 '9.\9-9.9I -lo!71 ~-.------.-,--,-, JA _28_0_K )10 -10~56 -1J.46 -~-""~ _~ 1Io0o ...... '...... • ...."'l'l:'MlI .. JA---.1.3& ·1.1I2 _...... __,, ...__... ....,...... -, -12.40 ·14.82 ...,...... ,...... '-'...... ,...... "...... _.... 240K ., I pi, ...... ~__....._ ....,...... ,,.,.'.....,,"'...._• ...... ,;,.;,..t 'I .. • I" • lk.. I I .1."I" -2.24 ·14.83 -17.99 HHoo - Fig. 9. Temperature dependence of the IH‘H NMR spectra of cluster 3 in CD2C12CD2Cl 2 solution; chemical shifts are indicated. shifts, a feature not unexpected in molecules with assignments, they are internally consistent within significant ZFS and magnetic anisotropies. the cluster set 1--6l-6 and we proceed on thatthat basis. (c)(c) MossbauerMiissbauer spectra. Basket clusters contain Substitution of thiolate for chloride in tetrahedraltetrahedral three types of Fe sites (Fig. 1): Fe(l)(Jl3-ShCu4-S)L, Fe&XFeS3X units of prismane and cubane-type clusters Fe(2)(Jl3-Sh(Jl4-S)P and Fe(3)(Jl2"S)(Jl3-ShP. The has been found to cause a decrease of ca 0.05-O.0.05-0.11 11 first of these is roughly tetrahedral, while the other IllInSmm s--1 ' in isomer shift. ’13.27,28 3*27*28 In the basket clusters two have distorted trigonal pyramidal geometry. this results in the much larger change of 0.20 mm The zero-field Mossbauer spectra of clusters 2 and s-1,s- ‘, as just noted. The smaller change inin isomerisomer 6 are displayed in Fig. 3;; the spectrum of 2 closely shift in the cubane-type clusters has been previously resembles those of 3-535 (not shown), in containing considered in terms of a decrease in covalency of threethree resolved features. All spectra were analysed the Fe-LFe--L bond. In the case of the present clusters, with a 1:: 1: I1 Fe site population. The assignment however, the much larger change in 6(j values sug-sug of thethe Fe(2) site has been discussed;discussed ; that of the gests that there is a detectable change in core charge Fe(3) site inin 6 is based on correlation of the largest delocalization upon ligand substitution. Earlier, itit quadrupole splitting (1.15 mm S-I)s- ‘) with those in had been shown that in clusters containing well-well 2-525 (0.95-1.29(0.95-1.29 mm S-I).s-l). This leaves the Fe(l) defined Fe (mean) oxidation states, s, isomer shifts sites inin 2-5 and in 6 as those with (j6 = 0.60-0.64 of tetrahedral Fe&FeS4 sites follow the empirical linearlinear and 0.43 mm S-I,s- ‘, respectively. The relatively large relationship of eq. (2) :2g:29 isomerisomer shift change (0.20 mm s-SK’)1) upon replacing 6(j = 1.44-0.43s. (2) chloride (2) with thiolate (6), while the isomer (2) shifts of the other sites change by no more than Its application to 6 yields an oxidation state of 0.05 mm s-1,s-l, is another indication that the 2.35+ at the Fe(l) sites, very close toto thethe cluster Fe(l)Fe( 1) sites contain halide. Also, the small variation mean oxidation state of 2.33 + and, therefore,therefore, con-con inin isomerisomer shift with halide is consistent with sistent with a highly delocalized cluster. Based previous behaviour of FeS3XFe&X sites in prismane on structural results for [Fe4S4(SR)4]‘-*2-,3-[Fe4S4(SR)4] 1-,2-,3- ‘*7,8,3o1,7,8,30 I clusters.clusters.‘+-16 3-16 While we cannot insist on these site the mean Fe-SFe--S terminal bond length of 2.255 wAinin 88 isis somewhatsomewhat moremore consistentconsistent withwith 2.50+2.50 + thanthan widthwidthnarrows,narrows,parallelingparallelingthethemonotonicmonotonicdecreasedecrease 2.33+,2.33 +, butbut notnot withwith 2.67+.2.67 +. NoNo suchsuch 6/s(j/s relation-relation inin thethe saturationsaturationmagnetizationmagnetization forfor 2-5.2-5.InasmuchInasmuch shipship hashas beenbeen establishedestablished forfor tetrahedraltetrahedral FeS,ClFeS3CI asas thesethese clustersclusters allall appearappear toto havehave SS== 11groundground sites.sites. WeWe surmisesurmise thatthat thethe chloride-boundchloride-bound sitessites states,states, theythey wouldwould bebe expectedexpected toto havehave aboutabout thethe Fe(Fe(l), 1), on chargecharge neutralizationneutralization grounds,grounds, areare biasedbiased samesame AiAi valuesvalues atat aa givengiventypetype ofof Fe Fe site.site.ThatThattheythey towardtoward iron(II1)iron(III) character,character, inin whichwhich casecase thethe largelarge dodo not,not, reflectsreflects differencesdifferences inin orbitalorbital statesstates ofof thethe isomerisomer shifts resultresult fromfrom thethe removalremoval ofof electronelectron individualindividual FeFe sites.sites. TheThe spectrumspectrum ofof 6 6 (not(notshown)shown) density fromfrom orbitals with 4s4s charactercharacter ratherrather thanthan evidencedevidenced aa smallsmallperturbationperturbationininan an 8080kOekOeappliedapplied an increaseincrease inin thethe 3d-type orbitals. TheThe isomerisomer shiftshift fieldfield (ca(ca 2.82.8 mmmm s-S-1’ spectralspectralwidth)width)andandresembledresembled change of 0.05 mm s-lS-l at thethe Fe(2,3) sitessites isis con-con somewhatsomewhatthatthatofof5.5. OwingOwingtotothetheuncertaintyuncertaintyininthethe sistent with a marginal increaseincrease inin iron(I1)iron(II) charac-charac groundground state,state, thethe spectrumspectrumwaswas notnot analysed.analysed. ter, in apparent responseresponse toto thethe effect of chloridechloride atat InInthethebasketbasketclustersclusters2-5,2-5,phosphine-ligatedphosphine-ligatedsitessites Fe(l). The mean Fe--ClFe-CI distance inin 7 (2.193(2.193 A),A), Fe(2,3)Fe(2,3) showshow positivepositive hyperfinehyperfinecouplingcouplingconstantsconstants when compared with the value inin [Fe&C14]2-[Fe4S4CI4F and correspondinglycorrespondingly negativenegative hyperfinehyperfine fields.fields. SiteSite [2.216(2) A],3’A], 31 is compatible with a mean oxidation Fe(Fe(l), 1), thoughtthought toto bebe moremore ferric-like,ferric-like, hashasoppositelyoppositely statestate> > 2.50+. In this argument, thethe site with thethe signedsigned parametersparameters whilewhile thethe sitessites Fe(l-3),Fe(I-3), inin thethe all-all most ferric character has the largest, rather thanthan phosphine basket clustercluster 9 havehave positivepositive AiAi andand the smallest, isomer shift. Whatever the merits of negative Hi parameters. These resultsresults areareconsistentconsistent this interpretation it is certainly the case that, on thethe with thethe same signsign ofofthe the isotropicisotropic shiftsshifts ofof2-52-5andand basis ofstructuraF·8-1structura12,‘-” 0 and spectroscopic evidence, 9,9,12 I2 so longlong as thesethese areare mainlymainly contactcontact inin origin,origin, clusters 1--6,14, as all other mixed-valence synthetic as appears toto be thethe case fromfrom other evidence.evidence. TheThe and biological Fe-S clusters with a nuclearity of spectra in Figs 8 and 9 show positive (upfield)(upfield)shifts,shifts, 3 or larger, are, electronically, largely delocalized when referred toto Et3PEt3P [6[(j 1.02 (CH,),(CH3), 1.36 (CH,),(CH2), systems. Note that for cluster 9, which formally CD ,Cl2Cl 2]2] as the diamagnetic reference. contains 5Fe5Fe”+Fe”’ll + Fe\ll and has only phosphine ter-ter All known basket clusters are schematically minalligands,minal ligands, isomer shifts are essentially identical. depicted in Fig. 1. A recent addition toto thisthis set isis Magnetically perturbed MossbauerMiissbauer spectra of FesSS(PBu3)4(SPh)333Fe6Ss(PBu3)4(SPhh 33 (lo),(10), whose core structure isis clusters 2-5 are presented in Fig. 7. For 2 and 3, nearly identical to those of 7 and 8. This molecule where thethe effect isis the clearest, an increase in the differs from clusters 2-9 in the handle of thethe applied fieldfield fromfrom 60 to 80 kOe caused features in basket is thiolate- rather than sulphide-bridgedsulphide-bridged; ; itit isis thethe wings andand inin thethe central portions of the spectra isoelectronic with 9 and, therefore, has a probable toto move inin opposite directions, signifying antianti- S = = 1/2l/2 ground state. Cai et ~1.~~al. 33 consider that the ferromagneticferromagnetic coupling,coupling, which produces a S = 1 atoms Fe(2,4) are more reduced in 10 than in 6 groundground state.state. SolidSolid lineslines areare theoreticaltheoretical fitsfits using because of Fe-S distances in the handle which are thethe HamiltonianHamiltonian (3)(3) andand aa fittingfitting program which isis 0.07 A longer than in the latter cluster. As yet, aa modified modified versionversion ofof thethe Lang-Dale minimizer.minimizer.32 32 no M6ssbauerMijssbauer or other spectroscopic and magnetic data have been reported that would permit a com-com HH = = GeJ.lBH·Ge~CLBH.SS ++ 1:[A(i)·~[A(i) . Ii·Ii. SS-g,~“H. - gnJ.lnH· IiIi ++ HQ(i)]Ho(i)] parison of electronic features with those of clusters (3)(3) 2-9. HHoQ == (eQV(eQV,,/12)[31,2-zz/12)[3I/-15/4+/'/CI/+I/)]15/4+~(1,~+1,~)] (4)(4) Acknowledgements-This research was supported by FitsFits werewere performedperformed inin thethe fastfast relaxationrelaxation regimeregime NIH Grant GM 28856 at Harvard University and by the withwith anan imposedimposed 11 :: I1 : : 11 sitesite ratio,ratio, assumedassumed randomrandom OfficeOffice of Naval Research program on Cluster Science crystallitecrystallite orientation,orientation, andand otherwiseotherwise asas describeddescribed andand Dynamics underunder Contract No. NOOOI4-89-J-I779N00014-89-J-1779 at 7 8 forfor [Fe4S4(SR)4P-[Fe,S,(SR),]3- clusters.clusters.7,8 • TheThe parametersparameters forfor MIT.MIT. TheThe FrancisFrancis Bitter National Laboratory isis supsup- eacheach sitesite areare thethe magneticmagnetic hyperfinehyperfine couplingcoupling concon- portedported byby thethe National ScienceScience Foundation. stantsstants AX'A,, AA,y andand A"A,, thethe EFGEFG asymmetryasymmetry parameterparameter 17,q, andand thethe signsign ofof thethe principalprincipal componentcomponent ofof thethe EFGEFG (V(V,).zz)' ValuesValues usedused inin thethe spectralspectral fitsfits areare listedlisted REFERENCESREFERENCES inin TableTable 2.2. DeviationsDeviations ofof thethe theoreticaltheoretical fitsfits fromfrom 1. J.J. M.M. BergBerg andand R.R. H.H. Holm,Holm, inin Iron-SulfurIron-Sulfur Proteins experimentalexperimental pointspoints maymay arisearise fromfrom intermediateintermediate (Edited(Edited byby T.T. G.G. Spiro),Spiro), Ch.Ch. 1.1. Wiley-Interscience,Wiley-Interscience, relaxationrelaxation phenomena,phenomena, asas observedobserved inin relatedrelated clusclus- NewNew YorkYork (1982).(1982). ters,6ters,6 andand fromfrom thethe presencepresence ofof crystallites.crystallites. TheorTheor- 2. J.-F.J.-F. You,You, B.B. S.S. Snyder,Snyder, G.G. C.C. PapaefthyrrriouPapaefthymiou andand eticaletical fitsfits ofof thethe magneticmagnetic spectraspectra requiredrequired propro- R.R. H.H. Holm,Holm, J.J. Am.Am. Chern.Chem. Soc.Sot. 1990,1990,112,1067. 112, 1067. gressivelygressively smallersmaller valuesvalues ofof hyperfinehyperfine couplingcoupling 3. M.M. S.S. ReynoldsReynolds andand R.R. H.H. Holm,Holm, Inorg.Znorg. Chern.Chem. 1988,1988, constantsconstants alongalong thethe halidehalide seriesseries asas thethe spectralspectral 27,4494.27, 4494. 4.4. A.A. Agresti,Agresti, M.M. Bacci,Bacci, F.F. Cecconi,Cecconi, C.C. A.A. GhilardiGhilardi andand 19. 19. W.W. Gerger,Gerger, U.U. MayerMayer andand V.V. Gutmann,Gutmann, Mh. Chern.Chem. S.S. Midollini,Midollini, Inorg.Znorg. Chern.Chem. 1985,24,689.198524,689. 1977,1977,108,417. 108,417. 5.5. F.F. Cecconi,Cecconi, C.C. A.A. Ghilardi,Ghilardi, S.S. Midollini,Midollini, A.A. OrlandiniOrlandini 20. 20. N.N. N. N. GreenwoodGreenwood andand T. C. Gibb, MossbauerMiissbauer SpecSpec- andand P.P. Zanello,Zanello, J.J. Chern.Chem. Soc.,Sot., DaltonDalton Trans.Trans. 1987,1987, troscopy,troscopy, Ch.Ch. 6.6. ChapmanChapman andand Hall, London (1971).(1971). 831.831. 21. 21. T.T. Birchall,Birchall, Can.Can. J.J. Chern.Chem. 1969,47,1969,47, 1351.1351. 6.6. V.V. Papefthymiou,Papefthymiou, M.M. MillarMillar andand E.E. Miinck,Miinck, Inorg.Znorg. 22. 22. S.S. Mitra,Mitra, inin IronIron Porphyrins (Edited(Edited by A. B. P. Lever Chern.Chem. 1986,25,3010.1986,25,3010. andand H. B. Gray), Part 2, Ch. 1. Addison-Wesley, 7.7. M.M. J.J. Carney,Carney, G.G. C.C. Papaefthymiou,Papaefthymiou, M.M. A.A. Whitener,Whitener, Reading, U.S.A. (1983).(1983). K.K. Spartalian,Spartalian, R.R. B. B. FrankelFrankel andand R.R. H.H. Holm,Holm, Inorg.Znorg. 23. 23. V. R.R. Marathe and S. Mitra, J. Chern.Chem. Phys. 1983, Chern.Chem. 1988,27,346.1988,27,346. 78,915.78, 915. 8.8. M.M. J.J. Carney,Carney, G.G. C.C. Papaefthymiou,Papaefthymiou, K.K. Spartalian,Spartalian, 24. 24. C. G. Barraclough, R. L. Martin, S. Mitra and R. R.R. B.B. FrankelFrankel andand R.R. H.H. Holm,Holm, J.J. Arn.Am. Chern.Chem. Soc.Sot. C. Sherwood,Sherwood, J. Chern.Chem. Phys. 1970,53, 1643. 1988,1988,110,6084. 110, 6084. 25. 25. P. D. W. Boyd, D. A. Buckingham, R. F. McMeekMcMeek- 9.9. I.I. Noda,Noda, B.B. S.S. SnyderSnyder andand R.R. H.H. Holm,Helm, Inorg.Znorg. Chern.Chem. inging and S. Mitra, Inorg.Znorg. Chern.Chem. 1979, 18, 3585. 1986,1986,25,3851. 25, 3851. 26. 26. M. A. Bobrik, E. J. Laskowski, R. W. Johnson, W. 10. 10. B.B. S.S. Snyder,Snyder, M.M. S.S. Reynolds,Reynolds, I.I. Noda andand R. H. O.0. Gillum, J. M. Berg, K. O.0. Hodgson and R. H. Holm,Holm, Inorg.Znorg. Chern.Chem. 1988,27,595.1988, 27, 595. Holm, Inorg.Znorg. Chern.Chem. 1978,1978,17, 17, 1402. 11. 11. B.B. S.S. SnyderSnyder andand R.R. H.H. Holm,Holm, Inorg.Znorg. Chern.Chem. 1988,1988, 27. M. J. Carney, J. A. Kovacs, Y. Zhang, G. C. PapaPapa- 27,2339.27, 2339. efthymiou, K. Spartalian, R. B. Frankel and R. H. 12. 12. B.B. S.S. SnyderSnyder andand R. H. Holm,Helm, Inorg.Znorg. Chern.Chem. 1990, Holm, Inorg.Znorg. Chern.Chem. 1987,26, 719. 29,274.29, 274. 28. 28. R. B. Frankel, B. A. Averill and R. H. Holm, J. Phys. 13. 13. M.M. G.G. Kanatzidis,Kanatzidis, N. C. Baenziger, D. Coucouvanis, (Paris) 1974,35,1974, 35, C6-107.C&107. A.A. SimopoulosSimopoulos andand A. Kostikas, J. Arn.Am. Chern.Chem. Soc.Sot. 29. 29. G. Christou, P. K. Mascharak, W. H. Armstrong, 1984,106,4500.1984,106,4500. G. C. Papefthymiou, R. B. Frankel and R. H. Holm, 14. 14. M. G. Kanatzidis, W. R. Hagen, W. R. Dunham, R. J. Arn.Am. Chern.Chem. Soc.Sot. 1982,104,2820. K.K. Lester andand D. Coucouvanis, J. Arn.Am. Chern.Chem. Soc.Sot. 30. 30. T. O'SullivanO’Sullivan and M. M. Millar, J. Am. Chem.Chern. Sot.Soc. 1985,1985,107,953. 107, 953. 1985,1985,107,4096. 107,4096. 15. 15. M. G. Kanatzidis, A. Salifoglou and D. CoucouCoucou- 31. 31. M. A. Bobrik, K. O.0. Hodgson and R. H. Holm, vanis,vanis, Inorg.Znorg. Chern.Chem. 1986,25,2460.1986, 25, 2460. Inorg.Znorg. Chern.Chem. 1977,16,1977, 16, 1851. 16. 16. D. Coucouvanis, M. G. Kanatzidis, A. Salifoglou, 32. 32. G. Lang and B. W. Dale, Nuci.Nucl. Znstrum.Instrurn. Meth. 1986, W. R. Dunham, A. Simopoulos, J. R. Sams, V. PapaPapa- 108,108,3541. 3541. efthymiou, A. Kostikas and C. E. Strouse, J. Am.Arn. 33. 33. J.-H. Cai, C.-N. Chen, Q.-T. Liu, B.-T. Zhuang, Chern.Chem. Soc.Sot. 1987,1987,109,6863. 109,6863. B.-S. Kang and J.-X. Lu, Jiegou Huaxue (J. Struct. 17. 17. C. J. O'Connor,O’Connor, Prog. Inorg.Znorg. Chern.Chem. 1982,29,203. Chern.)Chem.) 1989,8,220. 18. 18. D. H. Live and S. I.I. Chan, Anal. Chern.Chem. 1970, 42, 791.