摘要:

Geometries of Transition Metal Complexes ContainingSimple AlkenesBY LJUBICA MANOJLOVI&MUIR, KENNETH W. MUIR, AND JAMES A. IBERSDept. of Chemistry, Northwestern University, Evanston, Illinois 60201, USA.Receiued 16th January, 1969The results of precise X-ray studies of the fumaronitrile complex of hydridocarbonylbis(tri-pheny1phosphine)iridium and of the tetracyanoethylene complex of bromocarbonylbis(tripheny1-phosphine) iridium are presented and detailed comparisons are made with the results on other simpleolefin-metal structures available in the literature.In summarizing and evaluating what is known from diffraction studies aboutthe structures of transition metal complexes containing acyclic alkenes, we haveseveral objectives in mind. We wish to provide this kind of experimental evidenceas a basis for a consideration of the various theoretical models within this Discussion,although we realize that in doing so we may place some theoreticians at a dis-advantage.We wish also to emphasize how little detailed, accurate structuralinformation is available at present, even though the attachment of olefins to metalshas been of intensive experimental and theoretical interest over the last fifteen years.Among the structural results summarized are some, usually older, studies for whichthe limits of error are so large as to preclude a meaningful discussion of whether themolecule bound to the metal differs geometrically from the free molecule, and evenwhat the nature of the central C-C bond is. There is no longer any reason forsuch imprecise studies.We have chosen to consider complexes containing acyclicalkenes for, at present, there is greater scope for detailed theoretical calculations onsimple molecules and potentially on these molecules bound to metals. The practicalimportance of alkene-transition metal complexes, both in homogeneous catalysisand as models for heterogeneous catalysis, provides us with an additional interestin these systems.EXPERIMENTAL(a) H Y D R ID 0 CAR B 0 N Y L ( F U MAR 0 N IT R I L E ) b i S ( T R I P H E N Y L P H 0 S P H I N E ) I R I D I U MPreliminary X-ray and optical examination of crystals of TrH(CO)FUMN(PPh3)2(FUMN = fumaronitrile = trans-1,2-dicyanoethene) indicated that they belong to themonoclinic system.The systematic absences were hkl : h+k = 2n, and h01 : 1 = 2rz.These absences are consistent with the space groups C2/c and Cc. No piezoelectric effectwas observed and carefuI optical goniometric measurements suggested the point group 2/m.The unit cell dimensions, obtained by a least-squares technique from diffractometer settings,2are n = 21.289(14), b = 9.383(3), c = 21-650(10) A, = 123.87(2)*.* The density calculatedfor four molecules per cell was 1-524 g CM-~, compared with an observed density of 1-53(1)g ~ m - ~ .* Here, and ekewhere in this paper, limits of error are estimated standard deviations. They arein units of the least significant digit of the quantity to which they refer. Also in this paper we use1 A = 1Wp.m.8L. M A N O J L O V I ~ M U I R , K.w. MUIR AND J . A . IBERSMolybdenum radiation was used.85Jntensity measurements were made with a Picker four circle diffractometer using pro-cedures described in detail elsewhere.2 A total of3024 independent intensities were measured in the range 20(MoKa,) 549". The intensitiesof 2274 reflections were greater than thrice their estimated standard deviations and onlythese reflections were used in the analysis. The intensities were corrected for Lorentz-polarization effects and for absorption. The transmission factors ranged from 0.65 to0.73.The structure was solved by Patterson and Fourier methods in the space group C2/c.The iridium was found to lie on the diad axis with the hydridic hydrogen and the carbonylatoms disordered across this axis. Least-squares refinement of this model led to an agree-ment index R = 5.6 %.to Dshsymmetry with C=C = 1.392A. Anisotropic temperature factors were refined for allbut the ring carbon atoms. With the exception of the disordered hydridic hydrogen, thecontributions of the hydrogen atoms were included in the structure factor calculations.If the space group were Cc the molecule would not be required to possess twofoldsymmetry and an ordered structure would be possible. Refinement of an ordered modelin the space group Cc led to a significantly lower agreement index (R = 5.2 %). However,the temperature factors of several atoms became physically unreasonable and so did anumber of bond distances. For these reasons we believe the model refined in C2/c to bepreferable.In this refinement the phenyl rings were constrained( b ) BROMOCARBONYL(TETRACYAN0ETHYLENE)biS-(TRIP HE N Y L P H 0s PHI N E) I R ID I u MA brief account of earlier work on this compound has a~peared.~ Preliminary examina-tion of the crystals of IrBr(CO)TCNE(PPh,), (TCNE = tetracyanoethylene) indicatedthat the compound crystallizes in space group P21/n of the monoclinic system with fourmolecules in a cell of dimensions a = 17.659(5), b = 18.623(7), c = 11-602(4) A, p =94-99(2)".Intensity measurements were made on a Picker four circle instrument with the use ofCuKa radiation.A total of 8009 reflections were collected within the sphere 20(CuKcr,)<106". These data were corrected for Lorentz-polarization and absorption effects and thenaveraged to yield 3585 independent reflections.The transmission factors varied between0.3 1 and 0.42. In subsequent calculations only the 2767 independent reflections whoseintensities were greater than five times their estimated standard deviations were employed.The structure was solved and refined by standard methods. The only complication wasa disorder problem, which is common in systems of this type.5 As we discuss below, theconfiguration about Ir is that of a trigonal bipyramid with an apical Br atom and an apicalCO group. On the average inthe asymmetric unit let there be a molecules with Br " up " and CO " down ". Then therewill be (1-a) molecules with Br' " down " and CO' " up ". In the final refinement thefollowing variables relevant to this disorder were varied : the value of a and the positionaland anisotropic thermal parameters of Br and Br'.The positions of CO and CO' wereidealized and were not refined. In addition, in this final refinement the positional andanisotropic thermal parameters of all non-phenyl group atoms were refined ; for each phenylgroup the group parameters and an overall isotropic thermal parameter were refined. Thecontributions of the phenyl hydrogen atoms were added as fixed contributions to the structurefactors. This refinement of 179 variables converged to an agremeent index on F of 3.7 %.The value of u is 0*714(4), indicating that this is not a 1 : 1 disorder. The method outlinedhere for treating this type of disorder problem is a very effective one, and leads to far morerealistic results than may be obtained from a consideration of peak heights on a Fourier map.There is some disorder between the Br and CO groups.RESULTSThe inner co-ordination geometry of the FUMN complex is shown in fig. 1 andthat of the TCNE complex is shown in fig.2. (The H/CO disorder in the FUMNcomplex and the Br/CO disorder in the TCNE complex are not illustrated for th86 ALKENE-TRANSITION METAL COMPLEXESPFIG. 1.-A perspective view ofthe inner co-ordination sphere ofIrH(CO)FUMN(PPh3)2. Hereand in fig. 2 the 50 % probabilityellipsoids for thermal motion aredisplayed.FIG. 2.-A perspective view ofthe inner co-ordination sphereof IrBr(CO)TCNE(PPh&L . MANOJLOVI&MUIR, K .w. MUIR AND J . A . IBERS 87sake of clarity.) In both complexes the Ir has a trigonal bipyramjdal configurationif one considers the olefin as a single ligand. In each, the olefin and the two tri-phenylphosphine groups lie in the basal plane, with H (or Br) and CO at the apicalpositions. As is evident from fig. 1 the fumaronitrile molecule maintains its transcyanide configuration on complexing to iridium, although the cyanide groups bendaway from iridium as they do in the TCNE complex.Table 1 presents important intramolecular bond distances and angles for theFUMN complex and table 2 presents similar results for the TCNE complex.TABLE 1 .-SELECTED BOND DISTANCES (A) AND ANGLES (deg.) FOR IrH(CO)FUMN(PPh)&distanceIr-PIr-C(l)Ir-CP-P’P-C(1)P-cC( 1 )-cC( 1)-C( 1)‘C(1 k-cmC(2)-N c-0A2-31 7(3)2.1 lO(9)1*98(2)3.888(6)3*469( 10)2-99(2)2.87(3)1.43 1 (20)1-416(16)1*145(15)1 * 165(24)angleP-Tr-P’P-Ir-C( 1)P-Ir-C( 1)’P-Ir-CC( 1 )-Ir-C( 1)’C( 1 )-Ir-CIr-C( 1 )-C( 1 )’Jr-C( 1)-C(2)C( 1)’-C( 1)-C(2)C( 1 )-C(2)-NIr-C-0deg.114.1 (1)103.1 (3)142.8( 3)87-9( 6)39-7(5)89*3(8)70-2(3)1 17-8(8)117-5(12)179q 14)178-7(24)TABLE 2.-sELECTED INTERATOMIC DISTANCES (A) AND ANGLES (deg.) FORIrBr(CO)TCNE(PPh&A2*397(3)2.402( 3)2-508(2)2-146(11)2.1 5 I (I 1)1.506( 15)1*455(18)1 4 6 ( I 7)1*455( 16)1*447( 17)1.1 45( 16)1*157(16)I 108( 14)1-1 54( 15)3*573(11)3.616(12)3.229( 12)3-278( 1 1)angleP( l)-Ir-P(2)P( 1 )-Ir-C( I)P(2)-Ir-C(2)P( 1 )-Ir--C( 2)P(2)-Jr-C( 1)C( 1 )-Ir-C(2)Ir-C( 1 )-C(3)Ir-C( 1 )-C(4)Ir-C(2)-C( 5)Ir--C(2)-C( 6)Ir-C( 1 k C ( 2 )Ir-C(2)-C( 1 )C(2)-C( 1 )-C( 3)C(2)-C( 1 )-C(4)C(1 )-C(2t-C(5)C(1 &C(2)-C(6)C(3)-C( 1)-C(4)C(5&C(2)--C(6)deg.1 10.4( 1)103*6(3)105-0(3)14444 3)146*1(2)4 I *0(4)11 5*8(8)120.1(8)1 1 8 *0(8)1 17-3(8)69-6( 6)69*3(6)116-9( 10)1 16*9( 10)120.3(9)1 1 6*5( 1 0)110*4(9)11 1*2(11)DISCUSSIONTable 3 presents comparative structural information on substituted alkenecomplexes of transition metals that we found from a literature survey or that weremade available to us prior to publication.With the exception of the neutrondiffraction study of Zeise’s salt and the present X-ray studies by counter techniques,all results in table 3 were derived from photographic X-ray studies88 ALKENE-TRANSITION METAL COMPLEXEScompoundCOIo c \ Ioc’ Fi ,-iCOZHI CO,HCOIrBr(CO)TCNE(PPh3 )2COIrH(CO)FUMN(PPh3)2COTABLE 3.-cOMPARLSON OF STRUCTURAL DATAsymmetry imposed see note 1 ZL,:~ C-C, A LM-C-co1 TB 2*09(1) 1.40(2) 70*4(7)2.10(1) 70.6(7)2 TB 2*040(25) 1.421(43) 69*6(8)2 TB 2.09(3) 1-40(5) 71 (1)2*06(3) 1.30(4) 72(1)2-03(3) 1*40(4) 70U)1 TB 2.146(11) 1.506(15) 69-6(6)2-151(11) 69*3(6)2 TB 2.1 lO(9) 1 a43 l(20) 70-2(3)(disordered)KCPtCL(C2H4)IHzO 1 SP 2.141(10) 1*354(15) 71.42.1 36(9) 71.7/ciCI- Pt -jH2CI / CH2PtClZ(NHMe2)(CzH4) m SP 2*21(15) 1*47(18) -Ni(C2H4)(PPh3)2 1 T 2-02(2) 1*46(2) -PhP, 2-00(2)P h,P /”q 1 T 1.93 1.41 E!1 -94 68PtTCNE(PPh3)z 1 T 2-10(3) 1-52(3) -C NPh,p Ph3P=pt<cN r CN ‘\I2.1 l(3)PtFUMN(PPh3)z 1 T 2.025(6) 1*525(8) 73(2)CN2*162(6)I TB trigonal bipyramid, SP square planar, T trigonal.The alkene is counted as a monodentate1 i gand.2 M is the metal atom, C an olefinic carbon atom directly bonded to M, L is any other atom bondedto M, and X is an atom directly bonded to C. In X-ray studies no values are given for X = HL . M A N O J L O V I C ~ M U I R , K. w. MUIR AND J . A . IBERS 89ON TRANSITION METAL-OLEFIN COMPLEXESLC-M-C"39*0(6)40.5(9)39(037(1)4W)4 1 -0(4)39*7(5)LC-c-X"116(1)121(2)1 15(4)121(4)121(3)116*9(10)11 6.9( 10)120.3(9)11 6.5(10)1 17*5( 12)L(ccx)(ccM>o102(1)see note 4108( 1)103(2)103(2)109(2)115*2(10)109-4(9)1 1 1 * 1 (10)1 11.0( 10)11 1-8(8)11(2)171700.7(5)0-2(9)comments ref.L (CC)(LLM)"8 2 ) 679(2) Racemic 7Optically active.72(2) Three crystallo- 872(3) graphically86(2) independent89.5(6) Present work--molecules89.8(9) present work9 36-9 127.5 104.2 ca. 90 ca. 0 Zeise's salt.11 6.9 99.1 preliminary1 19.7 101.9 neutron diffraction120.8 98.7 analysis(by symmetry) analysis-.__-- --- -- - - 90 0 Two-dimensional 10434342(1)43m 1 16(3)1 17(3)6 - - Preliminary111213143 L(CCX)(CCM) is the dihedral angle between the planes defined by these sets of atoms.L(CC)(LLM) is the angle the C-C vector makes with the normal to the plane LLM.4 Underlined quantities were not given by the original authors but have been calculated or inferredby us90 ALKENE-TRANSITION METAL COMPLEXES( a ) THE OLEFINIC C=C BONDIn those cases where the standard deviations are sufficiently smal! to allow mean-ingful comparisons to be made, there is a lengthening of the olefinic bond on attach-ment of the olefin to the metal, with the exception of the C=C bond in Zeise's salt,where no such lengthening is evident.The difference between the C=C bond lengthin the TCNE and FUMN complexes of Ir appears to be significant. There is also asignificant change in the Ir-P distances in these compounds (cf. tables 1 and 2).From the constancy of Ir-P distances in the related oxygen complexes ofIrX(CO)(PPh3)2 5 9 l5 (X = halogen) we believe that the change from H to Br inthe axial position is not responsible for the observed changes in C=C and Ir-Pdistances in the TCNE and FUMN complexes.Rather we believe that the degreeof electron withdrawal in the olefins is the decisive factor in determining the differencesbetween these two complexes. This view is consistent with the chemical evidence l6that the stability of such complexes increases smoothly as the number of cyanideson the olefin is increased. In the Chatt-Dewar pi-bonding scheme for such complexesincreased electron withdrawing power on the olefin will increase the C=C bondlength because the electrons go into antibonding orbitals.Thus the trend in the olefinic bond lengths in the d8 complexes-Zeise's salt,to the acetonitrile complex of Fe, to the FUMN complex of Ir, finally to the TCNEcomplex of Ir-seems reasonable, although certainly other factors besides theelectron withdrawing properties of the olefin may be active here.Yet this explana-tion does not extend to the d10 complexes of FUMN and TCNE of Pt, where theolefinic bond lengths are essentially the same. However, comparison of the C-Cbond length in Zeise's salt with that in bistriphenylphosphine nickel-ethylene suggeststhat the Ni system is a better electron donor. In view of the consistent trend notedabove for the d8 systems, one might expect the C=C distance in the FUMN complexof Pt to be longer than that in the Ir complex.If there is sufficient donation to theolefin from the metal system, then it may be that additional cyanide groups on theolefin bring about no appreciable increase in electron donation. This is a rationa-lization for the equivalent bond distances in the TCNE and FUMN complexes ofPt, although the latter results are preliminary.The trends in the metal-P distances in these complexes can be rationalized in asimilar manner. The Ir-P distances of 2-3 17(3) A in the FUMN complex of Irare significantly shorter than those of 2*400(3)A in the TCNE complex, perhapsbecause the TCNE group has successfully competed for the electrons that wouldhave been available for multiple metal-P bonding. The fact that the Pt-P distancesin the TCNE and FUMN complexes are equal is consistent with the equality of theC=C distances in these complexes.( b ) DISTORTIONS OF THE ALKENE FROM PLANARITYIt appears from the data of table 3 that the C-C-X angle is reasonably constantfrom compound to compound.Accordingly, the dihedral angle between the CCXplane and the CCM plane provides an indication of the degree of bending back of theX groups from the metal in those cases where a plane of X groups cannot be defined.Another measure of this distortion is to compute, where possible, the dihedral anglebetween the two CX, planes. In Zeise's salt this angle is 145"' while in the TCNEcomplex of Ir it is 109.6(13)". Although the bending back in all of these complexesis least when X = H, it appears as though factors other than non-bonded interactionsm y be active here.No immediately discernible trends are apparent to usL. MANOJLOVIC-MUIR, K . w. MUIR AND J . A. IBERS 91(C) METAL-CARBON BOND LENGTHSThe metal-carbon distances in table 3 are more remarkable for their constancythan for any obvious trend that can be justified within the confines of the statedstandard deviations. Certainly the metal-carbon distance is not a sensitive measureof the degree of back bonding from metal to olefin. Only in the FUMN complexof Pt is there an apparently significant difference in the two M-C bond distances,although the limits of error on the bond distances seem surprisingly small, comparedwith the errors on the bond angles. This difference may not be real, as the resultsare preliminary.( d ) DISPOSITION OF THE c=c BOND WITH RESPECT TO THE PRINCIPALEQUATORIAL MOLECULAR PLANEIn those compounds listed in table 3 in which the metal is in a trigonal or trigonalbipyramidal configuration the olefin lies in, or nearly in, the trigonal plane.In thesquare-planar Pt(i1) complexes, on the other hand, the C=C bond lies perpendicularto the equatorial plane.In the parent complexes IrX(CO)(PPh,), (X=H or halogen) and in all five-co-ordinate addition complexes of these substances that have been studied by djffrac-tion methods, the triphenylphosphine groups remain trans, with the exception ofthe TCNE and FUMN complexes reported here and of the C2F4 complex ofIrI(CO)(PPh,)2.17 If one takes the geometry of the olefin as found in these studies,one concludes from appropriate molecular models that it is not possible to placethe olefin in the basal plane of a trigonal bipyramid that has apical trans phosphines.Similarly, given the phosphine configuration shown in the figures it is impossible toorient the C=C bond normal to the equatorial plane.Moreover, from a molecularmodel and the geometry of the ethylene group in Zeise's salt it is sterically impossibleto orient the C=C bond near to the equatorial plane. These arguments suggestthat non-bonded interactions may play a dominant role in the disposition of theC=C bond with respect to the equatorial plane. Yet these arguments start withthe geometries of the bound olefins and these in turn may be the result of factorsother than non-bonded interactions.In the trigonal complexes there is no obvious argument based on steric groundsthat favours the C=C bond being in the trigonal plane.Perhaps electronic factorsare important here, even though Cramer l8 has shown that there is essentially freerotation about the M- 11 bond in solution.CCAlthough in this paper we have chosen to discuss some aspects of molecularstructure within the Chatt-Dewar scheme, this should in no way be interpreted asan unqualified endorsement of the scheme. We are confident that suitably ingeniousarguments can be made within the framework of sigma bonding or excited statesto account for the observed trends.We are indebted to Prof. W. H. Baddley and Dr. W. C. Hamilton for makinginformation available to us in advance of publication. This work was supportedby the U.S. National Science Foundation and the U.S. National Institutes of Health.We are grateful to Prof. W. H. Baddley for supplying the crystals of IrH(CO)FUMN(PPh&.P. W. R. Corfield, R. J. Doedens and J. A. Ibers, Znorg. Chem., 1967,6,197. ' S. J. La Placa and J. A. Ibers, Acta Cryst., 1965, 18, 51192 ALKENE-TRANSITION METAL COMPLEXESJ. A. McGinnety and J. A. Ibers, Chem. Comm., 1968,235.J. A. McGinnety, R. J. Doedens and J. A. Ibers, Znorg. Chem., 1967, 6,2243.A. R. Luxmoore and M. R. Truter, Acta Cryst., 1962, 15,1117.C. Pedone and A. Sirigu, Inorg. Chem., 1968,7,2614.R. Spratley, K. Klanderman and W. C. Hamilton, private communication.lo P. R. H. Alderman, P. G. Owston and J. M. Rowe, Acta Cryst., 1960, 13, 149.l1 C. D. Cook, C. H. Koo, S. C. Nyburg and M. T. Shiomi, Chem. Comm., 1967,426.l2 W. Dreissig and H. Dietrich, Acta Cryst. B, 1968, 24, 108.l3 C. Panattoni, G. Bombieri, U. Belluco and W. H. Baddley, J . Amer. Chem. SOC., 1968, 90,798.l4 C. Panattoni, R. Graziani, U. Belluco and W. H. Baddley, private communication.l6 L. Vaska, Accounts Chem. Res., 1968,1,335.l8 R. Cramer, J. Amer. Chem. SOC., 1964, 86,217.’ C. Pedone and A. Sirigu, Acta Cryst., 1967, 23,759.S. J. La Placa and J. A. Ibers, J. Arner. Chem. Suc., 1965, 87,2581.N. E. Kime, Doctoral Diss. (Northwestern University, 1968)

ISSN:0366-9033    

DOI:10.1039/DF9694700084

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Organometallic Chalcogen ComplexesPart 18.-A Diluted Single Crystal E.S.R. Study of the Electronic Structure ofTricobalt Enneacarbonyl Sulphide : Antiaromaticity in a Transition Metal CarbonylCluster SystemB Y CHARLES E. STROUSE AND LAWRENCE F. DAHLDept. of Chemistry, University of Wisconsin, Madison, Wisconsin 53706Received 17th January, 1969An e m . study of CO~(CO)~S ( S = 3) in solution and doped in single crystals of the diamagnetichost FeCo2(CO)9S has shown from a detailed analysis of the hyperfine interaction that the unpairedelectron is in a non-degenerate molecular orbital of a2 symmetry comprised primarily of an anti-bonding combination of dorbitals essentially localized in the plane of the cobalt atoms. These defin-itive results firmly substantiate the previous conclusions based on X-ray measurements of CO~(CO)~Sand FeCo2(CO)9S that the unpaired electron in CO~(CO)~S has an appreciable conjugative destabiliza-tion effect (or antiaromatic character) in drastically altering the tricobalt geometry of Co3(CO)9S.The bonding in Co3(CO)9S and in the closely related Ni3(C5H5)3(C0)2 complex is discussed in lightof these observations.This e.s.r.investigation is an outgrowth of the recent discovery l* that theleast stable valence electrons in certain metal atom carbonyl cluster systems havesuch a pronounced effect on the molecular geometries that it can be experimentallydetermined that these electrons are in antibonding metal symmetry orbitals. X-Raystructural determinations 3 9 of the paramagnetic complex Co,(CO),S and itsisomorphous diamagnetic analogue FeCo,(CO),S revealed that the m.0.containingthe unpaired electron in CO,(CO)~S must be composed mainly of strongly antibondingmetal atomic orbitals. This observation is based on the fact that the formal removalof the unpaired electron by substitution of an iron for a cobalt atom has resultedin a remarkable shortening of the average metal-metal distance from 2.637(7) Ain Co,(CO),S to 2.554(7) A in FeCo,(CO),S. These stereochemical implicationsnot only have been supported by qualitative relative energy-level correlation diagramsbased on molecular orbital symmetry arguments l * but also have been subsequentlysubstantiated on a general basis by the successful preparation and structural-bondingcharacterization of other metal atom cluster systems 5-8 which show this samegeometrical sensitivity with respect to the nature of bonding of their valence electrons.In order to obtain more definitive information concerning the electronic structuresof these complexes, a single crystal e.s.r.study of CO~(CO)~S diluted in the diamag-netic host FeCo,(CO),S has been carried out. Few e.s.r. investigations of transitionmetal atom cluster systems have been previously reported ' 9 lo mainly due to thesmall number of known paramagnetic metal cluster compounds which have beenisolated. From the observed 49-line e.s.r. pattern of the polynuclear metal halide[Nb6C1,2]3+ ion, MacKay and Schneider concluded that the unpaired electronwhich has the same hyperfine interaction with all six niobium nuclei ( I = 9/2 for93Nb : 100 %) is delocalized in a non-degenerate bonding metal d-orbital combinationover the entire octahedron of niobium atoms.Longuet-Higgins and Stone loproposed that the g values obtained from e.s.r. measurements of single crystals of994 E.S.R. STUDY OF ELECTRONIC STRUCTURE OF CO3(CO)9SNi3(C5H5)3(C0)2 are not inconsistent with their assignment from qualitative m.0.considerations of the unpaired electron to a bonding combination of nickel dn-orbitals. Their assignment was later shown to be incorrect from a bond lengthcomparison of the molecular parameters in Ni3(C5H,),(CO), with those in thediamagnetic jsomorphous analogue CON~~(C~H,),(CO)~ (which corresponds to theformal abstraction of the unpaired electron in Ni3(C5 H5)3(C0)2 by the substitutionof a cobalt for a nickel atom).' This paper briefly comments on the re-interpretationof the electronic structure of Ni3(C5H5)3(C0)2 in terms of the placement of theunpaired electron in an antibonding non-degenerate metal symmetry orbital asrequired by the experimental results.EXPERIMENTALThe CO~(CO)~S used in this study was prepared by the reaction of CO,(CO)~ with H2Sin hexane at 150°C and 100 atm CO as well as by other The FeCo,(CO),Swas prepared by the procedure of Khattab, Marko', Bor, and Marko'.I3Single crystals of FeCo2(CO)9S containing CO~(CO)~S were grown from carbon-monoxide-saturated hexane solutions at about 0". Because of the instability of Co3(C0),Sin solution and the difficulty in the analyzing for CO~(CO)~S in the presence of FeCo,(CO),S,the concentration of the paramagnetic complex in the solution as well as in the crystals isuncertain. Several crystals grown from solutions estimated to contain from 3 to 10 partsof Co3(C0),S to 100 parts of FeCo2(COIgS gave spectra similar to that of pure Co3(CO)9S.Two crystals which were grown from solutions estimated to have between 0.5 and 1.0 partsof Co3(CO)9S to 100 parts of FeCo2(C0)9S gave strong, well-resolved e.s.r.spectra.The e.s.r. spectra were made on a conventional Varian E-3 spectrometer equipped witha Varian single crystal goniometer. The g values were measured with respect to a DPPHstandard, and the line separations were determined directly from the magnetic field calibra-tion of the spectrometer.For the spectra recorded with the crystallographic a axis perpendicular to the magneticfield, the crystal was mounted on a quartz rod and X-ray oscillation photographs were usedto align the a axis to within 0-5" of the axis of the rod.The rod attached to the goniometerwas then positioned in the spectrometer with its axis perpendicular to the magnetic fielddirection.RESULTS AND DISCUSSIONBONDING WITHIN THE CO3(CO)gS CLUSTERIn order to interpret the e.s.r. measurements, it is necessary to obtain some insightinto the nature of the metal-metal interactions within the Co,(CO)gS cluster.Although greatly oversimplified, the following LCAO-MO picture based primarilyon symmetry arguments involving the nodal character of wavefunctions provides adescription of the metal-metal bonding which reasonably rationalizes the experimentaldata for this metal cluster.Similar m.0. schemes applied to other related triangularmetal carbonyl cluster systems have given rise to experimental predictions whichsubsequently have been ~ e r i f i e d . ~ - ~The Co,(CO),S molecule possesses the idealized geometry of trigonal C3"-3rnsymmetry shown in fig. 1. The apical sulphur atom is connected by three metal-sulphur bonds to an M3(CO)g fragment containing three equilaterally positionedM(CO), groups co-ordinated to one another by metal-metal bonds. Since allthree metaI-metal distances in the isostructural F~CO,(CO)~S were found to be thesame within experimental error (viz., 2.554 A, indiv.e.s.d., 0.007 A), a crystal dis-ordered model which assumed a statistical distribution of the iron and two cobaltatoms over the three metal positions in each molecule was utilized in the least-squareC . E . STROUSE AND L . F . DAHL 95refinement. This crystal disordering of the metal atoms is not unexpected in viewof the nearly identical covalent radii of iron (1.165 .$) and cobalt (1.157 A).The qualitative m.0. model applied to Co,(CO),S assumes that four of the ninevalence orbitals of each cobalt atom are primarily involved in localized electron-paira-bonding with the sulphur atom and three attached carbonyl groups, and hence theFIG. 1 .-Idealized molecular configuration of Co3(CO)9S and FeCo2(C0),S corresponding toC3,-3m symmetry.perfect-pairing approximation allows their separability from the metal-metal inter-actions.Local right-handed Cartesian co-ordinate systems were chosen for eachcobalt atom with the z axis centripetally directed toward the centroid of the cobalttriangle, the x axis located in the plane of the three cobalt atoms, and withthe y axis perpendicular to this cobalt plane. Consideration of the transformationalproperties of the nine valence orbitals of each cobalt atom under C3v molecularsynimetry shows that the 4s orbital and three 4p orbitals have the proper symmetryto interact strongly in forming a-bonds with the sulphur atom and carbonyl groups ;hence, for the sake of convenience these orbitals are regarded to be completelyinvolved in the cobalt-ligand o-bonding.This leaves only the five 3d orbitals percobalt atom to be considered as basis functions for the metal-metal interactions.Of the 27 electrons possessed by the tricobalt system in CO~(CO)~S, two are requiredto complete the three electron-pair o-bonds with the sulphur atom (i.e., the sulphuratom is reasonably considered to have a tetrahedral-like valency of one non-bondingand three bonding electron pairs with the unshared electron pair localized alongthe threefold axis and with its other four valence electrons used in the o-bondingwith the three metal atoms). As a consequence, each cobalt atom may be formallyregarded as possessing a fractional oxidation number of +2/3. There are therefore25 electrons left in Co3(CO)9S to be distributed over a set of the metal symmetryorbitals. The atomic d orbitals on the three cobalt atoms can be combined underCJV symmetry to give seven.bonding metal symmetry orbitals (viz., dz2(a,), dxz(e),dyz(al), dx2-y2(aJ, and dxy(e)) and eight antibonding ones (viz., dx,,(at), dx2-y2(e*),dyz(e*), dxz(nr), and d22(e*)) relative to the isolated basis functions of each cobaltatom. If it is further assumed that the energy level splittings from the cobalt-cobaltinteractions are sufficiently large compared to the splittings due to the non-sphericalligand environment about each cobalt atom, then orbital overlap considerationspredict (with the neglect of any configurational interaction) that the highest twoenergy levels will be the dx2(a~) and dz2(e*) orbitals.The addition of the 25 electronsavailable to the tricobalt system of Co,(CO),S results in the unpaired electron beingin one of these two highest energy orbitals. If the odd electron were in the degeneratee" orbital, the molecule by the Jahn-Teller theorem should distort, whereas n96 E . S . R . STUDY OF ELECTRONIC STRUCTURE OF COj(CO)gSapparent molecular deformation js observed to occur in Co3(CO)9S. Since thedegree of mixing which can occur among the cobalt orbitals belonging to the sameirreducible representation is not known, the atomic orbital character of the occupiedenergy levels in general cannot be specified. Nevertheless, it can be argued thatthe highest non-degenerate level (a;) and degenerate level (e*), each of which isstrongly antibonding and therefore presumably separated energetically to a considerableextent from the other levels, essentially retain their identity as dxz(at) and d,z(e*)orbitals which are localized in the tricobalt plane. From a valence-bond formalism,Co,(CO),S is an exception to the “ noble gas ” rule in that with the assumption ofCo-Co electron-pair bonds the entire molecule contains one electron in excess ofthe closed-shell electronic configuration for each cobalt atom.The effect of thisone antibonding electron is to reduce the valence-bond metal-metal bond order toless than 1.0. If this unpaired electron is selectively removed to give a closed-shellelectronic configuration for each cobalt atom, then from the widely accepted premisethat the corresponding occupied bonding and antibonding levels effectively cancelone another’s bonding it follows from orbital overlap considerations that the cobalt-cobalt bonding would be essentially due to six strongly bonding electrons locatedin the lowest energy d,~(a,) and dxz(e) orbitals.Hybrid combinations of these twoorbitals on each cobalt atom result in two equivalent cobalt orbitals (viz., dz2+dxz)which can form localized electron-pair Co-Co a-bonds by overlap with the corres-ponding identical hybrid orbitals on the other two cobalt atoms. Hence, the successfultransformation from an 111.0. representation to a valence bond representation whichconforms to chemical intuition supports the ma. argument that the highest twoenergy levels in Co,(CO),S are primarily dxz(az) and dzz(e*) type orbitals which arealso localized in the tricobalt plane.While the latter degenerate orbital can mixwith the s(e*) cobalt combination and withp and d sulphur orbitals of e representation,the a; cobalt orbital by its symmetry classification not only cannot have any scharacter but also is orthogonal to all of the sulphur orbitals.Since the X-ray studies on CO,(CO)~S and FeCo,(CO),S also provide convincingevidence that the odd electron must be primarily in a strongly antibonding metalsymmetry orbital while the solution and solid-state e.s.r. measurements at roomtemperature of CO,(CO)~S require that this orbital be non-degenerate, it is the purposeof this paper to show that the e.s.r.results unequivocally demonstrate that theunpaired electron be essentially delocalized in a d orbital combination of a, repre-sentation in the tricobalt plane in accord with the placement of the unpaired electronin a dxz(a;) type metal symmetry orbital.The stability of these metal carbonyl cluster systems which by this m.0. modelaccommodates valence electrons in antibonding as well as bonding metal symmetryorbitals is no doubt due to a considerable degree to the back-bonding ability of thecarbonyl ligands to remove electronic charge from the metal symmetry orbitals(mostly from the higher energy antibonding ones which are nearer in energy to theempty carbonyl rc* orbital combinations). However, it will be shown from the e.s.r.data that the unpaired electron for all practical purposes is not extensively delocalizedover the ligands but instead effectively resides in the tricobalt part of the molecule.ANALYSIS OF RESULTSSOLUTION E A R .SPECTRUM.-The e.s.r. solution spectrum Of CO,(CO)9S (fig. Z),first observed in saturated hexane and ether solutions, is independent of solvent andtemperature and gives only the expected loss of intensity upon dilution. Further-more, it is completely consistent with that expected for a molecule in which thC. E. STROUSE AND L. P. DAHL 97unpaired electron has the same hyperfine interaction with all three cobalt nucleiwhere I(59C0 : 100 %) = 7/2. Although the predicted 22 hyperfine componentsare not all completely resolved, the resolution is sufficient to obtain a value of (a) =30.9f0.5 gauss for the isotropic coupling constant with (9) = 2*022r4.0.005.wFIG.2 . T h e e.s.r. spectrum of a solution of CO~(CO)~S in hexane.SINGLE CRYSTAL E.S.R. sPECTRA.-The isomorphous co,(co)9s and FeCo2(CO)Sgcompounds crystallize in the triclinic space group Pi with four molecules in theunit cell (fig. 3). Hence, crystals of F~CO,(CO)~S containing a small amount ofCo,(CO),S give an e.s.r. spectrum consisting of overlapping signals from the twocrystallographically independent molecules. The crystallographic a axis which i98 E.S.R. STUDY OF ELECTRONIC STRUCTURE OF CO3(CO)9San easily recognizable morphological feature of the crystals, makes an angle of89.5" with the idealized three fold axis of molecule l(i.e., the unprimed molecule infig.3). Thus, it was relatively simple to rotate the crystal about an axis nearlyperpendicular to a plane containing the idealized three fold axis of molecule I.Spectral data from a mixed crystal grown from a 0.5 % solution of CO,(CO)~SFIG. 4.-E.s.r. spectra of a single crystal of FeCo2(CO),S containing a small amount of C O ~ ( C O ) ~ S .Spectra a and b were recorded with the threefold axis of one molecule parallel to the magnetic fielddirection; a was obtained at room temperature, and b was obtained at 77°K. Spectrum c is atypical spectrum obtained when neither molecule has its threefold axis parallel to the magneticfield direction. (These three spectra were not reproduced on exactly the same scale.C . E . STROUSE AND L .F. D A H t 99were taken at 15" intervals as the crystal was rotated about the a axis. The orienta-tions of the two independent molecules relative to the magnetic field direction werecalculated for each angular setting from the determined position of the b* axis whosemorphological direction had been established from single crystal X-ray photographs.Fig. 4a shows the spectrum recorded at room temperature with the idealized three-rFIG. 5.-The low field portion ofa single crystal e.s.r. spectrumshowing the 1,3,5,7, . . . splitting.FIG. 6.-The observed angular depen-dence of the hypetfine coupling constantcompared with that calculated using themodel described in the text.I I I \ I5 c ~ 15 30 45 608 (degrees)fdd axis of molecule I essentially aligned with the magnetic field.In this orientationthe threefold axis of molecule I1 makes an angle of 68" with the field. The signaldue to molecule I is the broad resonance consisting of 22 hyperfine components,while the signal due to molecule I1 is the sharp resonance centered at slightly higherfield100 E.S.R. STUDY OF ELECTRONIC STRUCTURE OF CO~(CO)CJSAs shown in fig. 4b a great improvement in resolution is obtained by the recordingof the single crystal spectrum at 77°K. Again the resonance due to molecule Iconsists of 22 well-resolved hyperfine components, but the resonance due to moleculeI1 now consists of a large number of partially resolved lines. Fig. 4c shows a typicalcomplex spectrum obtained at 77°K if neither independent molecule in the crystalis oriented with its threefold axis along the magnetic field direction.Fig. 5 givesthe low field portion of a spectrum obtained at 77°K for a particular orientation ofthe crystal such that a 1,3,5,7, . . . splitting of the 22 principal hyperfine componentsis clearly indicated.The value of A measured from the spectrum in fig. 4b is 74.3 gauss. The angulardependence of A over the range 0 = 0 to 0 = 45" (where 8 is the angle between thefield and the threefold axis) is displayed in fig. 6. The g values in the parallel andperpendicular orientations, measured at room temperature with a DPPH standard,are911 = 2.04+0.01 andg, = 2-02+0.01.INTERPRETATIONThe isotropic contact term in the hyperfine interaction of the nuclei of transitionmetal complexes containing an unpaired electron in a d orbital has been attributedto the spin polarization of the inner s electron^.^^ It has also been found that suchan interaction for cobalt complexes gives rise to a negative isotropic coupling constant(a).14Since the molecular orbital in Co,(CO),S containing the unpaired electron mustbe a non-degenerate one comprised of a linear combination of metal atomic orbitals,the isotropic hyperfine coupling constant for an electron in the m.0.will be one-thirdthat for an electron localized in one of the atomic metal orbitals. Thus, for a validcomparison one must relate the value of 3(a) = -93 gauss for Co3(CO),S withthe value of ( a ) for molecules containing a single cobalt atom.This value of - 93 gauss is remarkably close to the value of - 94.7 gauss obtainedfor co+'(3d8) doped in a MgO host lattice ; this latter number based on the observedA and g values reported for Co+l in MgO by Orton, Auzins, and Wertz l5 wascomputed from the relationship l4 (a) = - K = A - (g -2.0023)P, where P =2*0023gNPe~N(r-3). A value of (r3> = 5.41 a.u.was obtained by an interpolationof the (r3) values reported by Freeman and Watson l6 for Coo, C O + ~ , C O + ~ , andC O + ~ . This 3(a) value for CO,(CO)~S also can be compared with the ( a ) value of- 101 gauss found by Maki et aZ.17 for the low-spin ( S = 3) cobalt(I1) planar anionic[COS,C,(CN),]~- complex.The observation of the e.s.r. spectrum of CO,(CO)~S at room temperature bothin solution and in solid state shows that the unpaired electron is in a non-degeneratem.0.of either an a, or a2 representation (under C3" symmetry). The isotropicnature of the g value and its closeness to the free-electron value not only confirmthe non-degenerate character of the partially filled orbital but also supports thepremise that there is not extensive spin-orbit coupling in this system. The fact thatthe isotropic hyperfine coupling constant in CO,(CO)~S is approximately 3 that inthe above and in other cobalt compounds l4 in which the unpaired electron is localizedin a dorbital on a single cobalt atom indicates that theunpairedelectron in Co,(CO)gSresides in a non-degenerate cobalt symmetry orbital comprised primarily of atomicd orbitals (with little if any s character). This observation definitely favours an a2rather than an al orbital assignment since an a, (but not an al) cobalt symmetryorbital cannot possess any cobalt s characterC .E. STROUSE AND L. F. DAHL 101These conclusions parallel those of MacKay and Schnejder for the metal halidecluster [Nb6Cll2I3+ ; the fact that the hyperfine interaction constant foi [Nb6Cl1213+( S = 4) is only slightly less than one-sixth of the values reported for several monomericniobium(1V) complexes ( S = +) was cited by these workers as strong evidence thatprimarily d-orbitals are involved in the metal-metal bonding and that the unpairedelectron is in a non-degenerate orbital which is delocalized over the entire Nb6-octahedron.If the half-filled non-degenerate cobalt symmetry orbital of Co,(CO),S consistsof a single type of cobalt atomic d orbital whose principal axis is parallel to thethreefold axis, the magnitude of the direct dipolar contribution to the hyperfinecoupling constant along the threefold axis is given by I Ail I = 1/3 x 4/7 x P.Forthe previous choice of axial directions, the three possible atomic orbitals with aprincipal axis along the y direction are the dxz, dX2-,2, and the dY2 a.0.; for metalsymmetry combinations based on the former two a.0. the sign of Ail is negative,while the sign of Ail is positive for a metal symmetry combination made up of dy2a.0. On the other hand, if the half-filled metal symmetry orbital is made up ofcobalt d,~, dX2+, dxy, or dyz a.o., then I Ail I = 1/3 x 2/7 x P.Here the sign of A\!is negative for the d , ~ combination but positive for any combination involving theother three a.o.The three a, cobalt symmetry orbitals mentioned previously consist of linearcombinations of cobalt dz2, d,~-~2, and dyz a.0. It was shown that these orbitalsare all bonding orbitals and hence are completely filled. The two a2 orbitals madeup of linear combinations of dxz and dxy orbitals were both found to be antibonding.In the limit of DJh symmetry the &.,(a;) metal symmetry orbital becomes a;*, whilethe dxy(az) transforms to a;*. Hence, consider the limit where one at cobalt symmetryorbital is made up entirely of d,, a.0. and the other at orbital is composed entirelyof dx,, orbitals (i.e., no orbital mixing is assumed to occur).Ail for the system withthe unpaired electron in the dx,(a;) orbital is given by Ail = - 1 / 3 x 4/7 x P, whereasif the &,(a;) is the half-filled orbital Ajl = + 1/3 x 2/7 x P. Since Hartree-Fockcalculations by Freeman and Watson l 6 give a value of P = 273 gauss for the " free "C 0 + ~ ( 3 d ~ ) ion, one can set as an upper limit for the direct dipolar contribution tothe hyperfine coupling constant value I A' I max=52 gauss.With the reasonable assumption that the indirect dipolar contribution to thehyperfine coupling can be neglected for CO~(CO)~S due to the closeness of the glland gl values to the free-electron value, the equation for the hyperfine term A , ,simplifies to All = Ajl + ( a ) ; it is readily apparent that All (as well as Ail) musthave the same sign as ( a ) and hence must be negative.Therefore, the half-filledcobalt symmetry orbital of a; representation must be primarily of atomic dxzcharacter. This conclusion is completely substantiated by the use of the observed1 All 1 to calculate P values for the limiting cases of an electron isolated in a puredxz(a;) and an electron localized in a pure dxy(az) combination. For the dx,(a4)representation the calculated value of P is 228 gauss, while for the d,,(a?) representa-tion the calculated P value is 456 gauss ke., this latter value was calculated with theassumption that ( a ) and All were both positive in that Ail must be positive for thed,,(a;) cobalt symmetry orbital ; if ( a ) had been taken as negative, then All wouldbe positive such that P = 1105 gauss.Both of these P values are unrealistic. TheseP values can be compared with the P values calculated from the Hartree-Fock <r3)values l 6 for the various electronic configurations of cobalt ; the computed P valuesare listed as follows : Co0(3d9), 217 gauss ; Co1+(3d8), 245 gauss (interpolated) ;Co2+(3d7), 273 gauss; Co3+(3d6), 303 gauss; Co4+(3d5), 336 gauss. Since inC O ~ ( C O ) ~ S the formal oxidation number is +2/3 for each cobalt atom, it is expecteI 02 E.S.R. STUDY OF ELECTRONIC STRUCTURE OF CO3(CO)9Sthat the P value be near to the P value of 245 gauss for Col+. The closeness of theP value of 228 gauss for the d,,(az) representation with that of 245 gauss for Colfstrongly implies that the unpaired electron is in a nearly pure d,,(ai) cobalt symmetryorbital.In general, the expectation value of 1 / r 3 will be decreased with the formationof a covalent compound, but such a decrease would only further confirm the aboveconclusion.Quadrupole terms in the spin Hamiltonian have been neglected in the calculationof All. This omission can be justified by the observation that the 22 hyperfine linesin the e.s.r. spectrum in fig. 4b are equally spaced within experimental error. Indeed,as the half-filled cobalt symmetry orbital approaches pure dxz (or pure dxy) character,quadrupolar contributions to A 11 should vanish. *The angular dependence of the e.s.r. spectrum for various admixtures of d,,and dxy atomic orbitals to the a; orbital 4 containing the unpaired electron has beencalculated in which it was assumed that the hyperfine tensor of each cobalt nucleusis axially symmetric and that the principal axis of each tensor .makes an angle cowith the molecular threefold axis.Thus, $(a;) = d,, cos co+dxy sin co. Thiscomputer calculation utilized the ( a ) and All (with the appropriate signs) togetherwith the mixing parameter co. The principal values of the three hyperfine tensors wereobtained and then used with zero-order wavefunctions to calculate the positions ofthe three lines corresponding to MI = +19/2 for various field directions. Thesecalculations are outlined in the appendix. Over the range of 8 (defined as the anglebetween the magnetic field direction and the threefold axis) where it is possible todistinguish sets of lines with a given MI, the variation in the position of the centreline with 8 for the 4 orbital corresponding to co = 0" (i.e., possessing pure dxzcharacter) is in reasonable agreement with the experimental positions (see fig.6).In this 8 range, however, the introduction of a large amount of dxy character intothe # wavefunction (equivalent to an co value as large as 45") did not produce asignificant change in the angular dependence of the hyperfine constant. Whilethis first-order perturbation treatment correctly predicts the general form of thespectrum (viz., the correct number of lines and relative splittings), it is not adequateto calculate the magnitude of the splitting between states with the same M,.CONCLUSIONSThis e.s.r.study of CO~(CO)~S shows that the molecular orbital containing theunpaired electron is an antibonding a, orbital which consists primarily of cobaltdxz atomic orbitals localized in the tricobalt plane. This determination which iscompletely consistent with the X-ray results can be used to rationalize the structuraland electronic parameters of similar metal cluster systems.Since an a2 orbital in Co,(CO),S cannot interact directly with any atomic orbitalson the apical suphur atom, substitution of selenium for sulphur should not resultin a large change in the hyperfine coupling constant unless there is a recordering ofthe upper molecular orbital energy levels. One might expect, however, a less drasticchange in the metal-metal distance between Co,(CO),Se and FeCo2(CO),Se thanhas been observed in the sulphur analogues in that the presumed greater interactionof the a, and e orbitals of the metal system with the atomic orbitals on the seleniumatom would tend to make the geometry of the complex more dependent upon themetal-chalcogen bonding.These corresponding selenium compounds which havebeen recently prepared in our laboratories are being investigated both by e.s.r. andX-ray methods ; the results will be reported elsewhereC . E. STROUSE AND L. F. DAHL 103THE g-TENsoR.-The equations with notation developed by Longuet-Higgins andStone lo to calculate the components of the g tensor for the trinuclear transitionmetal complex Ni3(C5H5)3(C0)2 are also applicable for Co,(CO),S.For an electronin a cobalt symmetry orbital of a2 representation, g 11 = 2 +I4 j *< 4i I liy 1 Xi) 1 2,i 1 3 i = lwhere the first sum is taken over the three al molecular orbiials which are the odyones which contribute to the g component and the second sum is taken over thethree cobalt atoms. The qi which include both an energy difference denominatorand the spin-orbit coupling constant for a cobalt 3d electron are positive for thethree bonding a, orbitals. ]in the limit of D3,, symmetry there would be only twonon-zero terms corresponding to the interaction with orbitals of a; symmetry. Thethe non-vanishing terms occur only when the sum is taken over the five e molecularorbitals, while in the limit of D 3 h symmetry there would be only two non-zero terms.Of the 25 electrons to be distributed in the tricobdt part of Co,(CO),S among threebonding a , orbitals, two antibonding a2 orbitals, and five e orbitals, in order for theodd electron to occupy an a2 orbital there must be a single e orbital with higherenergy.Thus, one of the qt will be negative. Because of the large number of termsin the gil and gl equations and because of the lack of data on molecular orbitalenergy differences, no quantitative information was obtained from the measuredg values.ANTIAROMATICITY IN Ni3(C,H,),(CO),.-Longuet-Higgins and Stone lo reportedan e.s.r. investigation of the paramagnetic Ni3(C5H5)3(C0)2 complex from whichthey deduced that the measured g11 and gL values are consistent with the one unpairedelectron being in a bonding dn(ui) orbital.However, if the same kind of m.0. modelwhich successfully predicts the nature of the non-degenerate orbital containing theunpaired electron in CO,(CO)~S is applied in the analysis of the nickel-nickel inter-actions in Ni3(C5H5)3(C0)2, it is found that the unpaired electron instead is in anantibonding metal symmetry orbital of the same type dcharacter as the metal symmetryorbital containing the unpaired electron in Co,(CO),S. The m.0. model therebyprovides support for the postulation and resulting direct evidence * given by acomparison of the corresponding molecular parameters of Ni3(C5H5)3(C0)2 andCON~~(C,H~)~(CO)~ that the unpaired electron in Ni3(C5H5)3(C0)2 is in an anti-bonding metal symmetry orbital rather than in a bonding metal symmetry orbital.This LCAO-MO treatment presumes that each nickel atom utilizes five orbitals inlocalized co-ordination with a cyclopentadienyl ring and with the two carbonylgroups.In all, fifteen nickel orbitals are required for bonding to the cyclopentadjenyland carbonyl ligands ; under D 3 h molecular symmetry these orbitals must have thesymmetry 2a; +a; + 2ai + 3e'+ 28. If the 4s and three 4p a.0. on each nickel atomare considered to be completely involved in the nickel-ligand bonding (as also pre-sumed by Longuet-Higgins and Stone lo), then the resulting 4s and 4p nickel symmetryorbitals have the representations 2 4 + a; + 02 + 3e' + e". The only 3d nickel orbitalcombination possessing the required symmetry, a; + e", to complete the localizedbonding with the cyclopentadienyl and carbonyl groups involves the dyz atomicorbitals (Le., the local Cartesian co-ordinate system at each nickel atom was chosento be the same as that in Co,(CO),S rather than that arbitrarily picked by Longuet-Higgins and Stone lo).The remaining four 3d orbitals on each nickel atom arecombined under DJh symmetry to give six bonding nickel symmetry orbitals (viz.,dzz(ai), dxz(e'), dX~-,,~(ai), and dxy(e")) and six antibonding nickel symmetry orbital104 E . S . R . STUDY OF ELECTRONIC STRUCTURE OF CO3(CO)9S(viz., dxr(a;*), dX2-,,2(e'*), dx,(ai*), and dz2(e'*)). As with CO,(CO)~S the observedroom-temperature e.s.r. signal for Ni3(C5H5)3(C0)2 necessitates that the unpairedelectron be in a non-degenerate orbital; hence, with 19 electrons to be consideredfor the nickel-nickel interactions, the highest unfilled metal symmetry orbital musthave degenerate e symmetry.Orbital overlap considerations strongly indicate thatthe non-degenerate orbital containing the unpaired electron is the antibonding a;*orbital primarily involving dxz atomic orbitals localized in the trinickel plane. Theobserved g values of Longuet-Higgins and Stone lo can be rationalized equally wellfor an electron in this antibonding metal symmetry orbital.We acknowledge the National Science Foundation (GP-49 19) for their financialsupport of this work. The use of the CDC 3600 and 1604 computers at the Universityof Wisconsin Computing Center was made possible by the partial support of NSFand WARF through the University Research Committee.One of us, C. E. S.,is grateful to the National Science Foundation for a predoctoral NSF TraineeFellows hip.APPENDIXThe following treatment will be used to interpret the hyperfine splitting in the e.s.r.spectrum of C O ~ ( C O ) ~ S in a single crystal of FeCo2(C0)9S when the magnetic field is notaligned along the threefold molecular axis. If a right-handed orthogonal co-ordinatesystem is chosen with the origin at the centroid of the triangle of cobalt atoms, the z axisperpendicular to the tricobalt plane and the x axis passing through cobalt atom M(3) (seefig. l), then the symmetrical hyperfine tensor at the nucleus of co(3) is of the formRotation by the angle y about the z axis results inwhereAi = G:A;Gi,cosyi -sin?, 0cosy, 00 1and where y1 = yz+ 120" = y3+ 240".If the magnetic field is restricted to lie in the xzplane, the hyperfine Hamiltonian of cobalt nucleus i can be expressed l9 asHi = [sin28(Af),,+ 2 sin 8 cos 8 ( A f ) x Z + c o s 2 0 ( A 2 ) Z Z ] ~ ~ s ~ ~ ( ~ ) = CiMSm,(i),where, e.g., (A&z is the xz element of A? and where 8 is the angle between the magneticfield and the z axis. In order to calculate the first-order perturbation energies a set ofnuclear spin eigenfunctions 4j must be obtained such thatA basis set made up of nuclear spin product functions, represented by I m1(l)m1(2)m~(3)),was chosen. For MI = +21/2 there is only one product function, 1 3 $ z).ApplicatioC. E. STROUSE AND L. F . DAHL 105of the lowering operator to this state gives #s = 1/2/3[ I 3 4 z)+ I 3 3 $)+ I S 3 a)] whichis one of the three eigenfunctions for MI = + 19/2.In the absence of an external magnetic field, states with the same MI are degenerate andthus one can choose as eigenfunctions for the remaining two states any two mutually ortho-gonal functions which are also orthogonal to 4s. One such set is$ A = l/d2[I%%S)- I;%3)] and +*= 1/d6[1$$$)+ I22;)+21$3.g)].Application of a magnetic field breaks this threefold degeneracy. It can be seen, however,that 4~ and 4~ are symmetry functions of the system as long as the magnetic field is inthe vertical plane containing cobalt nucleus 3. If the component of the magnetic field inthe plane containing the three cobalt nuclei makes an angle of y with the vertical planecontaining cobalt nucleus 3, three symmetry functions &, # ~ t , and 4 ~ p can be constructedfrom #s, #A, and 4~ as follows : 4 ~ # = 4s, 4 ~ e = cos y 4 ~ + sin y 4 ~ , and 4 ~ # = -sin y 4 ~ +cosy 4 ~ .It is now possible to calculate the first-order energies corresponding to $s., AT,and 4 ~ ' . For 4s. the ES' = (4s. I H I 4s.) = (1/3) x (19/2) x [C, + C,+ C3],where H=2 Hi. The 4 ~ t symmetry function can be expanded as4 ~ # = (siny/d6+cosy/2/2) I ~33)+(siny/y/6-cosy/d2) 1 $$$)+(-2siny/d6)I $33).The correspondingj = 1= < 4 A # j H I4Ap) =4[(5X; + 7 Y; + 7Z:)C, M , + (7X2+ 5 Y i + 7 2 3 3 4 , + ( 7 x 2 + 7 Y i + 5Z;)C,Ms],where XA = (sin y/2/6+cosy/1/2), YA = (sin y/\/6-c0s y/2/2), and ZA = (-2 sin y/d6).The expression for the energy E B ~ corresponding to 4 ~ ' is that same as that for the energy E A ~except that the values of X, Y and Z becomeXB = (-siny/2/2+cosy/y/6), YB = (siny/y/2+cosy/2/6), and ZB = (-cosy/d6).This first-order perturbation treatment accounts for the threefold splitting of the principalhyperfine component corresponding to A41 = + 19/2.An extension of this treatment willaccount for the observed 1,3,5,7 . . . splitting of the 22 principal hyperfine components ofone CO~(CO)~S molecule when the magnetic field is not oriented along the threefold molecularaxis. From the above treatment an admixture of dxy orbital character to the half-filled a;orbital comprised primarily of d', atomic orbitals would result in a splitting of the typeobserved. However, any factor which would result in a non-equivalence of the three cobaltnuclei would also contribute to this splitting. This first-order perturbation treatment is validonly for small values of 0.L. F. Dahl. Abstr. Proc. 3rd Znt. Symp. OrganometaZZic Chemistry (Munchen, Germany, 1967),p. 92.L. F. Dahl, Proc. 1st Int. Symp. Znorg. Chim. Ada, (Venice, Italy, 1968), B 1.C. H. Wei and L. F. Dahl, Znorg. Chem., 1967, 6, 1229.D. L. Stevenson, C. H. Wei, and L. F. Dahl, submitted for publication.C. E. Strouse and L. F. Dahl, submitted for publication.H. Vahrenkamp, V. A. Uchtman and L. F. Dahl. J. Amer. Chem. Soc., 1968,90,3272.V. A. Uchtrnan and L. F. Dahl, submitted for publication.R. A. MacKay and R. F. Schneider, Inorg. Chem., 1967, 6,549.l o H. C. Longuet-Higgins and A. J. Stone, Mol. Phys., 1962, 5,417.l1 L. Markb, G. Bor and E. Klumpp, Chem. andZnd., 1961, 1491.l2 L. Mark6, G. Bor, E. Klumpp, B. Markb, and G. AlmAsy, Chem. Ber., 1963, 96,955.l 3 S. A. Khattab, L. Mark6, G . Bor, and B. Mark6, J. Organometal. Chem., 1964, 1,373.J4 B. R. McGarvey, J. Phys. Chem., 1967,71,51.' H. Vahrenkamp and L. F. Dahl, Angew. Chem., Znt. ed., 1969, 8, 144106 E.S.R. STUDY OF ELECTRONIC STRUCTURE OF co3(co)$l5 J. W. Orton, P. Auzins and J. E. Wertz, Phys. Rev., 1960, 119, 1691.l6 A. J. Freeman and R. E. Watson, Magnetism, vol. IIA, G . T. Rado and H. Suhl, ed., (Academicl7 A. H. Maki, N. Edelstein, A. Davidson and R. H. Holm, J.Amer. Chem. Soc., 1964, 86,4580.l8 W. Low, Paramagnetic Resonance in Solids, (Academic Press, Inc., New York, N.Y., 1960),l9 A. Carrington and A. D. McLachlan, Introduction to Magnetic Resonance, (Harper and Row,Press, Inc., New York, N.Y., 1965), p. 167.chap. 10.New York, N.Y., 1967), p. 104

ISSN:0366-9033    

DOI:10.1039/DF9694700093

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

GENERAL DISCUSSIONMi-. A. Hamnett (Oxford University) said: I would make two observations onPettit's paper. The first concerns the geometry of the silver complexes. It is assumedin the paper that the initial silver-butene complex has the silver ion placed on topof the ethylenic linkage where steric effects are minimized. The silver-butadienecomplex is also considered symmetric, though it is known from other evidence tobe asymmetric, the Ag+ ion being bound more firmly to one olefin than the other.It is therefore conceivable that in the transition state, the silver ion moves and theplane of symmetry is not conserved. This would invalidate the rigorous symmetryarguments put forward, though the essential features of the explanation would beretained if the argument be recast in topological form.The second concerns the reaction for which the compound (VI) may open byscission at bonds 1, 2 (using a Ag catalyst) or 3, 4 using a Ni catalyst.O+U 2 4CVI 1The kinetics of this reaction will befast slowmetal -F organic compound +metal complex-scission.Now Ag+ may complex with benzene strongly' and with olefins, but since I, 2 isthe weaker bond (as is known thermochemically and follows from arguments onaromaticity), for Ag+, it is this bond which breaks. For Ni no aromatic complexesare known and we assume that Ni only binds to be olefhic linkage so that onlybond 3,4 may break.Dr. D, S . Urch (Queen Mary College) said: Pettit has described how complexformation by metals may make " allowed " a chemical reaction which, by orbitalsymmetry considerations, would be " forbidden " if initiated thermally.But suchprocesses can often proceed photochemically. I wonder if it might be possible toobserve the converse, the effect of metal ions on those reactions which are photo-chemicaIly forbidden but thermally allowed. Is it possible to initiate photochemicallyand with a metallic catalyst a reaction which would otherwise onZy take place at anelevated temperature ?Dr. R. Pettit (University of Texas) said : With respect to the nature of the butadiene-silver ion complex I do not know of any evidence which points to an unsymmetricalstructure for the complex in solution. Concerning the difference in the reactionpathways for the ring opening reactions of benzotricyclooctadiene with Agf and Nicatalysts we concur with the comments of Hammitt.In support of this argument isthe fact that biscyclooctadiene nickel does not catalyze the ring opening of dibenzo-tricyclooctadiene 011).Mr. C. H. Campbell (Oxford University) said: T wonder if Cotton would commenton the possible analogy in electronic structure between (C5H5)3M~N0 and nickel-ocene. In nickelocene it is usually supposed that each ring acts formally as a five-' M. L. H. Green, Organometallic Compounds, p. 28.10108 GENERAL DISCUSSIONelectron donor in the usual way, so that this is a twenty-electron compound with thetwo extra electrons in a degenerate anti-bonding orbital, probably the eIg. If theelectronic structure of (CSHs),MoNO were analogous to this the orbitals corre-sponding to the el, of nickelocene would not be degenerate, because of the lowermolecular symmetry, so that compound would be expected to be diamagnetic, asit is found to be.It seems that such a structure would explain the rather long metal-carbon distances in (C5H5),MoN0, where the average distance for the two n-bondedrings is 2.47& compared with (C5H&MoH2, an eighteen electron compound ofotherwise similar structure, where the average distance is only 2.29 This differ-ence is closely comparable to the apparent effect of the anti-bonding electrons innickelocene, where the metal-carbon distance is estimated to be 2.2 A,2 comparedwith the value of 2.05 A found in ferr~cene.~The asymmetric orientation of the rings relative to the metal in (C5H5),MoN0would then be due not to any peculiarity of the electronic structure of the molecule,but simply to its low symmetry.Bennett et aL4 cited a number of compounds(including (C5H5)2MoH2), all with the usual eighteen valence electrons, which havea pattern of two long metal-carbon distances, one medium and two short, similarto that which has now been found for (C5H5),MoN0. They pointed out thatasymmetric bonding may occur for any cyclopentadienyl ring bound to a group ofless than threefold symmetry, due to the unequal interaction of the two el orbitalswith the metal when the dxz and dyz of the latter are not degenerate.Prof. F. A. Cotton (M.I.T., Mass.) said: Although the metal-carbon distancesrn (C5H&MoH2 are unequal they were considered by Mason to correspond, withinthe limits of experimental accuracy, to a structure in which each ring retains a planeof symmetry, whereas in (C5H5),MoN0 the metal-ring bonding is completelyunsymmetrical.This is to be expected whatever the electronic structure of thecompound since the (C5H5)2M~H2 molecule as a whole has a plane of symmetryand (C5H5),MoN0 has no symmetry at all.I14Y T FIG. 1.Mr. K. F. Wagstaff (Grangemouth) (communicated): It is well known that the non-planarity of alkenes bonded to transition metal ions implies a mixing of the carbon2s orbitals with the n and n* orbitals of the alkene. To gain insight into the factorsaffecting the degree of non-planarity observed it is instructive to calculate the effectGerloch and Mason, J.Chem. Suc., 1965,296.in Dunitz and Orgel, J. Chem. Phys., 1955,23,954 ; reference to Berndt, Hamilton and Hedberg,private communication.Dunitz, Orgel and Rich, Acta Cryst., 1956, 9, 373.* Bennett, Churchill, Gerloch and Mason, Nature, 1964, 201, 1318GENERAL DISCUSSION 109of this mixing on the overlap integrals between the n and n* orbitals of the alkeneand the metal d orbitals.If we consider a symmetrical alkene X2C : CX2 symmetrically bound to a transi-tion metal ion and if we assume that the C-C and C-X cr bonds are formed fromorthogonal hybrid carbon a.0. directed exactly along the bond axes then it is possibleby use of the orthogonality conditions and the geometry of the molecule to find theform and direction of the fourth hybrid orbital t,bk at each olefinic carbon atom (seefig.1).Let t,bk be inclined to the C-C bond axis at an angle 4 and be of the formwhere 2pk is a carbon 2p orbital with its axis of symmetry in this direction. Aconvenient measure of the non-planarity of the alkene is 8, the angle between thebisector of LX-C-X and the C-C bond axis. (The dihedral angle between thetwo CX2 planes is (180" -20).) If we denote L. C-C-X by cc) and L X-C-X by x,~k = (1 + 3L:)-'(2s + l $ p k ) , (1)cos e = -cOS W/cos (x12).)3.k = -cos o/cos 4 J-cos x.We also findandtan 4 = cot 6 (I +sec2co)-4 cosec (28),(4)In the special case where u) is kept constant at 120" the variation of 4 and & with0 has been calculated and is given in the table.e L(CCX)(CCM) x = LX-€-X (b ak0" 90" 120" 90" 0010" 95" 51' 118" 58' 86" 34' 12.020" 102" 8' I 1 5" 42' 82" 26' 5.7730" 109" 28' 109" 28' 76" 6' 3.6140" 118" 59' 98" 30' 62" 12' 2.7945" 125" 16' 90" 45" 00FIG.2.-Overlap of c)k with 3d orbitals ofFe+.4 - 10 - - 20 - 3 0 = - 4 0 ' 0'' C . A. Coulson, Vdence, 2nd ed. (O.U.P. 1961), p. 203110 GENERAL DISCU$SIONIf the M-C distance is r and LC-M-C is 28 the overlap integrals of rc/k withthe metal dz2 and dyz orbitals, respectively, areS($,&) = (1 + A",>-+(( 1 - 3 sin2p))S(nda,2s) + A,( 1 - 3 sin2p) sin(# - p) S(nda,2pa) -9~~ sin (28) cos (4 - ~(ndn,2pn)), ( 5 )andS(t,bk,d,,J = (1 + A i ) - * { q sin (2j3)S(ndo,2s)+9Ak sin (28) sin (4-j3)S(nda,2pa)+;lk cos (2p) cos (4 -p) S(ndn,2pn)), (6)where S(nda,2s) etc.are the standard overlap integrals corresponding to the inter-nuclear distance r. Eqn. (5) and (6) were solved as a function of 8 for the specialcase of an iron complex where r = 2*05& p = 20" and co = 120". The wavefunctions used were those of Slater for neutral atom carbon and those of Richardsonfor Fe+. The results are plotted in fig. 2. The overlap with the metal dz2 is ratherinsensitive to bending of the co-ordinated alkene, there being a shallow maximum at6-20". Overlap with dyz, however, increases to a pronounced maximum at 81:37".One might therefore expect that as back donation (from dyz) becomes more importantrelative to forward donation (into dzz) the angle 0 would increase. If so, there maybe an electronic explanation of the larger value of 6 for the TCNE complex of iridiumas compared with Zeise's salt.Dr.H. A. 0. Hill (Oxford University) said: With regard to the paper by Iberset al., how do the central C-C bond lengths in the tetracyanoethylene, trans-1,2-dicyanoethylene, monocyanoethylene and fumaric acid complexes compare withthose of the free ligands?Prof. J. A. Ibers (Northwestern University, Ill.) said : As far as I am aware, onlythe structures of ethene and tetracyanoethene are known. Both of these moleculesare planar with a central C-C bond length of 1.34 A. It is reasonable to expect thesame geometrical features in mono- and dicyanoethene.Dr. D. S. Urch (Queen Mary College) said: In ligand with extended conjugated71 systems such as TCNE and FUMN the bond lengths in the terminal groups shouldprovide a diagnostic test as to whether back-bonding to the n system of the ligand asa whole is taking place (in which case the C-N bonds should be somewhatlengthened), or whether metal-carbon a bonds are being formed from the centralC-C atoms with a change of hybridization.I should therefore like to ask Ibers if hehas any data about C-N bond lengths in the complexes which he has studied?Prof. J. A. Ibers (Northwestern University, Ill.) said : In answer to Urch, the C-Nbond lengths in the complexes studied are given in the text. Within the estimatedstandard deviations on these distances, there are no discernible trends that would aidone in discussing the bonding. Of more importance, perhaps, are trends in the non-central C-C bond lengths. This point has been discussed earlier.2Dr.Michael Green (University of York) said : In view of the interest in the geometryof metal-olefin compounds, I would draw attention to evidence for an outer-spherecomplex between PtC12- and ethylene. Substitution reactions involving platinum(I1)J. W. Richardson, et al., J. Chem. Phys., 1962,36, 1057.J. A. McGinnety and J. A. Ibers, Chem. Cornm., 1968, 235GENERAL DISCUSSION 111compounds normally have rate laws of the form : rate = k,[Pt(II)]+k,,[Pt(II)][L],where [L] is the concentration of incoming ligand. However, Milburn and Venanzihave shown that when Pt(I1) is PtCl2- and L is a substituted ethylene,CH,=CHCH,NHz, CH2=CHCH2SO;, or CH2=CHCH,0H, (6)the k l term in water is too small to be detected. We (M. Green and C. J. Wilson)have shown that the same is true for ethylene itself. Moreover, if the k , path existsat all, k , has a value for these olefins which is less than that normally observed, e.g.,for PtCl$-/ethylene, kI <2 x sec-1.2Thus, the olefin actually suppresses the k, path. As the k, term corresponds to attackby the solvent on the starting material, it is reasonable to postulate the existence ofouter sphere complexes such as (I), in which steric and electronic factors preventattack by water.sec-I but for PtCli-/*Cl-, kI = 3.9 x01Cl / c1\ I /Pt/ \c1 C1(I), (01 = olefin)R. M. Milburn and L. M. Venanzi, Inorg. Chim. Ada, 1968,1,97.M. A. Rucker, C. B. Colvin and D. S. Martin, Inorg. Chem., 1964, 3, 1373

ISSN:0366-9033    

DOI:10.1039/DF9694700107

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Study of the Bonding in Pentacarbonylmanganese Derivativesby Photoelectron SpectroscopyBY S . EVANS, J. C. GREEN, M. L. H. GREEN, A. F. ORCHARD AND D. W. TURNERInorg. Chem. Lab., and Phys. Chem. Lab., South Parks Rd., OxfordReceived 13th February, 1969The photoelectron spectra excited by helium 21.2 eV radiation in the complexes Mn(CO)5X,where X = H, C1, Br, I, CH3, CF3, COCF3 and Mn(C0)5, have been determined. The ionizationpotential data are discussed in terms of a simple molecular orbital description of the electronicstructures.The value of photoelectron (PE) spectroscopy as an electronic structural tool inorganometallic chemistry is virtually unknown. We have studied the PE spectraof a series of compounds Mn(CO),X, where X = H, C1, Br, I, CH3, CF3, COCF,and Mn(CO),, in order to explore the utility of this new technique for molecules ofconsiderable electronic complexity.EXPERIMENTALThe photoelectron spectrometer has been described.l The vapour of the compoundunder investigation was fed continuously into the ionization chamber at a pressure below2 0 0 ~ .Except for the hydride, Mn(CO)SH, the most volatile of the complexes studied, itwas necessary to place the sample as close as possible to the ionization chamber so as toallow free flow of its low pressure vapour. The spectra were calibrated by introducingxenon into the ionization chamber at the same time, as previously described.'12The compounds Mn(CO)5X were prepared and purified essentially by previousIy reportedand their purity ascertained from infra-red and mass spectra.8'12RESULTSThe photoelectron spectra are given in fig.1 and the ionization potential(1P)data are recorded in table 1. The accuracy of the LP data varies, but for the mostpart, the figures quoted in table 1 are correct within at least 0.25 eV (the maxima forrelatively sharp PE bands may be defined more accurately). In each of the PE spectrawe may distinguish three main regions of ionization, labelled A , B and C in table 1.The PE bands in region A have the relative intensities given in square brackets. Theseare rough estimates, and may be in error by as much as 10 %.Some of the spectra show a PE band centred at 12.6 eV due to the presence ofwater as a trace impurity. The spectrum of the chloride, Mn(CO),Cl, also containsa sharp band at 14.0 eV.This we attribute to trace amounts of carbon monoxide l3(see table 2), presumably arising from in situ decomposition of the sample. Boththe chloride and its bromine analogue are relatively involatile, so that small amountsof volatile impurities are readily detectable. There is apparently such tracecontamination in the spectra of each of these compounds.11TBLE IO IONIZATION POTENTIAL DATA a FOR THE Mn(C0)5XH C1 Br I CF 3 COCF38-80[2-0] 8.76[2*0] 8*35[2.5Ic8.6510.43 [2*0] 10*04[2*0] 9.57[2*0] 9.20[3] 9.011*00[1*0] 10-80[1.0] 10-37[1-0] 10.30[1](1 3-5-1 6.9) (14.0-1 7.0) (1 3.8-17.4) (1 3.7-17.6) (1 3517.6) (1 2-2-17.6)14.1 14.0 13.3 &[ 13*80 14.6 15.2 15.4 15.0 14.5 14.5C 17.97 18.7 18.6 18.6 18.5 18.514.4 14.315.6 15-4 15.016-419.9a Gross regions of ionization are in parentheses : otherwise the figures relate to PE band maxima.For the iodide, brackets. b The assignment, in terms of the m.0.ionized, does not refer to Mn2(CO)lo.This value gives the sum of the relative band areas for the two IP centred at 8.35 and 8.65 eV. d No114 PHOTOELECTRON SPECTRA OF Mn(CO)SXI =' IL . . . . . . . . . . . . . I21 20 19 10 I7 6 6 14 1) 12 I1 10 9 8Fig. 1 (a-d).I20 13 I6 I7 16 15 14 13 11 I 1 10 9 8I m' I11 20 I9 I0 17 I6 15 I 4 13 I2 I 1 I0 9 S 7Fi.2. 1 (e-h).FIG. 1 .-The photoelectron spectra of (a) Mn(CO)gH, (b) Mn(CO)5C1, (c) Mn(CO)sBr, ( d ) Mn(C0)5'1,( e ) Mn(CO)5CF3, cf) Mn(CO)SCOCF3, (9) Mn(CO)&H3, (A), Mn2(CO)I o.Horizontal scale givenin electron-volts (eV); the derived ionization potential data are given in table 1 .DISCUSSIONASSIGNMENT PROBLEMWe seek to explain the ionization potential (IP) data in terms of Koopman'sappro~imation,'~ according to which the individual vertical IP of closed-shellmolecules may be identified with the SCF energy eigenvalues of the occupied molecularorbitals (m.0.) from which the electrons are ejected. The molecular IP from whose rela-tion ships are strictly those of ionic energy levels, then receive a simple interpretationas one-electron orbital energies. The Koopmans approximation, however, sufferscertain deficiencies : in particular, errors arise from orbital rescaling, correlation andrelativistic effects.15 A substantial cancellation of the different errors does in facEVANS, GREEN, GREEN, ORCHARD AND TURNER 115occur, and is such that a Koopmans ab initio SCF estimate of a molecular IP isusually too large, e.g., see ref.(15), but not in a uniform and predictable fashion.Thus, if we use molecular IP empirically to diagnose the relative energies of theoptimum one-electron orbitals, we may not obtain the correct ordering of the SCFeigenvalues, see ref. (1 6). Nevertheless, an empirically established m.0. energylevel scheme, based on Koopmans' approximation, will have considerable allegoricalvalue : moreover, it should relate to semi-empirical m.0. theories which make realisticuse of atomic (or molecular) spectral and IP data.The empirical analysis of 1P data hinges upon the assignment question.Tofacilitate assignment we shall use group theory to establish the symmetry-dependentfeatures of the m.0. diagram tie., to construct the symmetry orbitals correspondingto a realistic choice of valence orbitals). We shall also invoke gross electronegativityarguments to limit further the possible form of the m.0. energy level scheme. Also,there js available a number of criteria for assignment of ionization bands, in thelight of conceivable m.0. schemes. In particular, we employ the following criteriarelating to orbital degeneracies. (1) The area under PE band, which reflects theionization cross-section of the occupied orbital concerned, should be simply relatedto orbital degeneracy.l3. l 8 The cross-sections should be similar for orbitals ofcomparable localization properties and energy.(2) Spin-orbit interactions mayresult in multiplet splitting of the ion term which, if resolved, will indicate the spatialdegeneracy of the ionized level. '-l If the effective spin-orbit coupling variessubstantially with position in the molecule the magnitude of the multiplet splittingmay indicate the localization tendencies of the ionized m.0. concerned : it may alsoenable one to distinguish between ionic terms having the same degeneracy butdifferent symmetries .Circumstantial evidence derived from the study of a series of related compounds,and related model systems, may also assist assignment. In addition, the analysis ofvibrational 3* (or vibronic) fine structure, relative IP band widths or the correlationwith electronic spectra, may often prove helpful.However, criteria concerning thelatter are of limited utility for the PE spectra of the Mn(CO),X speciesinvestigated here.Mn(CO),X PE SPECTRATABLE 2.-A GENERAL COMPARISON OF THE PHOTOELECTRON DATA * FOR CARBON MONOXIDE,Mn2(CO)lo AND Cr(CO),MndCO) 10 cr(c0) 6c CobA (7.9-9-3) 8.4 14.0 50B (12.8-16'4) (12.6-16.3) 16.9 17~C (17.3-18-5) (17.2-18.2) 19.7 40Unless in parentheses the data relates to ionization peaks.b We include the generally accepted one-electron assignment of the carbon monoxide IP.zoC The Cr(C0)6 PE spectrum shows vertical IP at 13.3, 14.1 and 17.5 eV.l8From the IP data in table 2, the PE spectra of Mn,(CO)lo and Cr(CO)6,18 eachshowing three main regions of ionization, have a strong resemblance. Unless thereis an inconceivably high effective charge on the metal atoms, the correspondencewith the PE spectrum of carbon monoxide l3 must be that indicated in this table.We assign the central region B in the Mn2(CO)lo and Cr(C0)6 spectra to ionizationsfrom m.0.of mainly carbonyl 50 and In character: the high energy region C weassociate with m.0. largely localized as the 40 112.0. of carbon monoxide. Weattribute the remaining structure (region A ) to ionization from m.0. of predominantlymetal 3d a.0. character. These general conclusions are reinforced by an extendedHiickel calculation on Cr(CO)6.2116 PHOTOELECTRON SPECTRA OF Mn(CO),XThe essential features of PE spectra of the Mn(CO),X compounds are alsoanalogous to those of Cr(C0)6.It therefore seems reasonable to suggest similar+ 3xe I EMnFIG. 2(a).-Av qualitative molecularorbital scheme for the Mn(C0); cationswith its 46 valence electrons. Thisbelongs to the C4" point group, and thecoordinate system is chosen centred atthe metal atom such that there are fourCO groups on the x and y axes and oneon the z axis (cf. table 3). The occupiedlevels are ringed, and orbitals of un-certain relative energy are bracketed.The Mn(C0); anion presumably has itstwo additional electrons in the 6alorbital.FIG. 2(b).-A qualitative view of the main features of the molecular orbital structures of Mn(C0)5Hand the halides, Mn(CO)5X, and their relationship with the Mn(C0); ion (cf.fig. ah))EVANS, GREEN, GREEN, ORCHARD AND TURNER 117assignments (see table 1). To be more explicit we refer our discussion to the m.0.diagrams, fig. 2(a) and fig. 2(b), for the Mn(C0)5 fragment, and the Mn(CO),Xspecies with monatomic X, respectively. Each molecule has C4" symmetry. WeTABLE 3The transformation properties and correlation of the metal valence orbitals in M(C0)6(Oh point group), square-pyramidal M(CO), (CaU) and M2(C0)10 with staggered (Dad)or with eclipsed (D4h) configurationametal a.o oh c4v D4d D4hdz2 eg" a1 a1+b2 a1g+a2udx2 - y 2 4 61 e; + eg bl,+bdUdXY 6 s b2 eg + e; b2,+blU4, t 4 g ea ed; + e: e,b+ e:4 2 t:9 eb ef + eg e:+ e,PS a19 a1 Q+b2 a 1 g - t a2uPz a1 a1 + b2 el g + a2uPxPYa The usual ccnventions are observed.Z2choose a valence orbital basis set restricted to metal 3 4 4s and 4p (transformationproperties given in table 3), together with the 4a, 5 0 , ln and 2n m.0.of carbonmonoxide. For halogen we employ valence-shell p (transforming as e+a,) ands ( a , ) a.o., and for hydrogen just the 1s (a,) a.0. Five carbonyl orbitals of csymmetry transform as 2a, + b , + e in the C,, point group : on the other hand, thecarbonyl n orbitals generate symmetry orbitals a , + a2 + bl + b2 + 3e. The detailedstructures of the m.0. diagrams reflect our conclusions, but are also based onqualitative arguments invoking overlap criteria, etc. In fig. 2(a), for example, the 6em.0. i s placed above 2b, on the grounds that the metal dyz and dzx orbitals interact lessstrongly with the carbonyl n* orbitals than does the dxy orbital.Those levels whichwe believe are of uncertain relative energy are bracketed in the diagrams.Certain details of the B and C regions of the PE spectra require special comment.Except for the methyl and perfluoroacetyl compounds, the Mn(CO),X spectraeach show a sharp edge at about 13 eV (see table 1). On overlap grounds this isprobably the adiabatic ionization from the 5e, m.0. mainly localized as carbonyl50. We also observe, for X = CH3, CF3 and COCF3, more structure in region Bthan appears in the other PE spectra. In particular, the methyl compound gives abroad band centred at about 12.6 eV which, by comparison with the methyl halidePE spectra,18 we assign to ionization of electrons essentially localized in the CH3a-orbitals.Similarly, we refer to the PE spectra of simple perfluoromethylderivatives,18 such as CHF3, and attribute the additional structure in region B ofMn(CO),CF3 to photo-ionization from orbitals that are largely fluorine 2p incharacter. The perfluoroacetyl compound, Mn(CO),COCF,, also exhibits additionalstructure in the region C of its PE spectrum : this may be associated either with theacyl CO group or with CF, itself.I8Apart from these observations, it appears that the general appearance and positionof the B and C regions of the PE spectra are relatively insensitive to the nature ofthe group X. In contrast, region A is markedly dependent on the nature of X, andis sufficiently resolved to permit the application of more detailed assignment criteria118 PHOTOELECTRON SPECTRA OF Mn(CO),XThe hydride, and also the methyl and perfluoro-methyl compounds, show twoPE bands in region A.The relative intensities of these bands are close to 2 : 1 inMn(CO),H, and about 3 : 1 in Mn(CO),CF3 : in the methyl compound, the intensityratio is inverted but approximately 1 : 2. Invoking our intensity criterion, (1)above, we therefore assign these two LP to the mainly metal e(dx,,dy,) and b2(dxy) m.o.,as shown in table 1. The hydride case is illustrated in fig. 2(b). The acyl compound,Mn(CO),COCF,, also apparently has two IP in region A , but these are incompletelyresolved. We presume they have the same origin as the IP of Mn(CO),H, etc.,assigned previously.The PE spectra of the chloride and bromide exhibit three low-energy bands,while Mn(CO),I has four bands. We assume that the first two bands of the lattercompound are the components, split under spin-orbit interactions, of a band whichcorresponds to the first band in the PE spectrum of either Mn(CO),Br or Mn(CO),CI.Now the only terms of the Mn(CO),X+ ion which are subject to multiplet splittingare the 2E species : and these should split symmetrically into states of equal degeneracy(the doublets E' and E"), which conforms with our observations.Moreover, sincespin-orbit effects must be much greater in the region of the iodine nucleus thanelsewhere in the molecule, we may also infer that the first IP of the halides relatesto ionization from an m.0.of largely halogen character. The observed doubletsplitting in Mn(CO),I is about 0.3 eV: if the orbital concerned were completelylocalized on the iodine atom, we would expect a splitting of some 0.6 eV (as calculatedfrom the free iodine spin-orbit coupling constant csp = 5070 ~ m - l ) ~ ~ , whereas for apure manganese 3d a.0. we expect a mere 0.03 eV (cJd = 240 cm-' for the Mn atom).The spin-orbit coupling constants being much smaller for chlorine and bromine (c3p = 590cm-', c4p = 2460cm-'), it is not surprising that we observe doubletsplitting only in the Mn(CO),I case. Thus, we assign the first IP in each of thehalide PE spectra to a 7e level, as shown in fig. 2(b). By analogy with the hydride,the remaining PE bands in region A, with relative intensities close to 2 : 1, are assignedto the 6e and 2b2 levels.Having assigned the essential features of the Mn(CO),X PE spectra, furtherdiscussion is warranted.With regard to the m.0. diagrams, fig. 2(u) and 2(b)on the acquisition of additional electron density in Mn(CO),X the energy levels of theparent Mn(C0): fragment will become more positive: and, since X interacts onlyindirectly with the CO groups, we expect the largely metal levels 2b2, fie, etc. torise further than those more localized on carbonyl. It is thus reasonable that theselatter m.0. are reiatively insensitive to the nature of X, as indicated by the PE spectra.In view of the relative Mulliken electronegativities of the monatomic X, weexpect the self-consistent charges on the manganese atom to diminish in the orderCl>Br>I>H. Therefore, the average of the 2b2 and 6e IP should be least inthe hydrides, and increase in the same sense.This is in accord with our observations.Also (table 1) the 6e-2b2 energy separation is less in the halides than in Mn(CO),H :for the mixing of halogen p , a.0. and the largely metal 6e level of Mn(C0)f mustreduce relatively the energy of the 6e m.0. of Mn(CO),X.The halide PE spectra also provide some evidence for metal-halogen d,-p,bonding on the bask of the following arguments. The inferred 7e (mainly halogenpn) m.0. energies, whilst in the expected order CI<Br<I, are distinctly higher andless differentiated than the energies of essentially halogen p n orbitals in simple halides :in the methyl halides, e.g., the relevant vertical IP are CH3Cl 11.30 eV, CH,Br10.54 eV and CH31 9.55 eV.18 The noticeably lower first IP of each of the Mn(CO),halides (table 1) may in general be associated with a higher electron density on halogen,but in particular with a substantial mixing of the halogen p x and the metal dxz,dyEVANS, GREEN, GREEN, ORCHARD AND TURNER 119(e) a.o., which leads to a relative lowering of the 6e level and a raising of the 7e levelas shown in fig.2(b). This metal-halogen n-interaction is greatest in the chloride,so that more uniform values of the Mn(CO)5X first IP (from 7e) are observed. Ourinterpretation is also consistent with the trend observed in the 6e-2b2 energy separation(table I), since the stabilization of the 6e level should diminish in the sense C1> Br > I.As a final comment on the Mn(CO), halides, we recall our earlier observationthat the PE spectra are consistent with the electron density on the metal decreasingin the order I>Br>Cl.This correlates with the relative thermal stabilities of thehalides. The non-existence of metal carbonyl fluorides presumably relates to theexcessively high positive charge that would appear on the metal atoms, leading tosubstantial weakening of the metal-carbonyl bond.In terms of our assignment, the methyl compound Mn(CO),CH, is distinguishedfrom its hydride and halide analogues in that the predominantly metal b2 m.0. isof higher energy than the similarly localized e m.0. Also, with the single exceptionof the carbonyl Mn,(CO),, itself, the average IP in region A of the Mn(CO),X PEspectrum is the least of the series Mn(CO),X: this reflects the strong a-donatingpower of the CH3 group compared with that of H and halogen.The markedstabilization of the 6e level relative to 2b2 may in turn be understood in terms ofthe differing interactions of the metal dxz,dyz a.0. and the dxy a.0. with the available x*acceptor orbitals, together with effects due to the mixing of the metal p x , p c andd,,, dyz a.0. The strong a-donor property of the CH3 group should markedlyreduce the electro-negativities of all the metal valence a.0. leading, in particular, toincreased mixing of the 4p and n' orbitals (both a , and e). This effect should lowerthe 6e and 6a, levels relative to 2b2.Thus, even if the direct interaction of the metal4px and 3dx a.0. is not particularly strong in Mn(CO)Q13, a substantial 4px-x*mixing rray be responsible for the reversal of the relative energies of the 6e and 2bzlevels. But the CH3 group also possesses vacant acceptor orbitals which, in termsof a local CfV microsymmetry, are of a , and e symmetry. The latter CH3 orbitalhas n-symmetry with respect to the Mn-C bond, and may mix strongly with metaldxz,drz. Thus, the pronounced stabilization of the mainly metal 6e level may bedue also to back-donation of metal electron density into the n* orbitals of the methylgroup.Turning to the IP data for perfluoro-methyl compound, Mn(CO),CF,, it appearsthat the CF3 group is roughly comparable to iodide in its a-donor strength.ThePE spectrum of Mn(C0),COCF3 suggests, on the other hand, that the perff uoro-acetylgroup is intermediate in behaviour to the CH3 and CF3 groups, both with respect toa-donor and also n-acceptor properties. These conclusions are reasonable on simpleelec tr onegat ivit y grounds.Mn2(CO),o PE SPECTRUMThe average IP in region A of the Mn2(CO)10PE spectrum is lower than thecorresponding average IP for any of the other Mn(CO)SX molecules studied here.This is as expected, since the effective charge on the manganese atom should be leastwhen X = Mn(CO)5. Also, the PE spectrum of Mn,(CO),, in region A differsfrom that of Mn(CO),H in apparently having a third band. This is only partlyresolved from the adjacent PE band so that its relative intensity is difficult to gauge :but the overall intensity ratios in region A are not far from 4 : 2 : 1 (ignoring theshoulder, the intensity ratio would be about 4 : 2.9).A tentative explanation ispossible in the following terms.The molecule Mn2(CO)lo is and therefore of Dqd symmetry, in thesolid phase. The mainly metal 36 m.0. of the Mn(CO), monomer species the120 PHOTOELECTRON SPECTRA OF Mn(CO),Xcorrelate in the manner indicated in table 3. A possible ground electronic configura-tion for the dimer is thus . . . (5e2)4(6e1)4(6e,)4(6a,)2, where the order 6e3>6e,seems probable on overlap grounds, but where the energy of 5e2 relative to the othere-levels is uncertain. Four low energy LP with relative intensities 2 : 2 : 2 : 1, wouldthen be expected.Therefore we assign the shoulder in region A of the Mn,(CO)loPE spectrum to the 6a, level, and the peaks to the three filled e levels : the two lowerlevels are assumed to lie too close in energy to be resolved.We cannot be sure that Mn2(CO)lo is indeed staggered in the vapour phasethough, on simple steric grounds for example, it seems a more probable configurationthan the eclipsed one of D4h symmetry. In the latter case we expect (cf. table 3)a ground electronic configuration such as . . . (2b2,)2((zb,,)2(6e,)4(6e,)4(a,,)2, whenthere should be five low-energy IP with relative intensities 1 : 1 : 2 : 2 : 1. Arationalization of the Mn2(CO)lo PE data for region A would therefore involve theassumption that rather more m.0.were close in energy. This cannot be excludedas a possibility, but it seems less likely than an assignment based on the assumptionof a staggered geometry for Mn2(CO),o.We gratefully acknowledge Clive Baker's advice and assistance with the measure-ment of the photoelectron spectra. Also we thank Turner and Newall, Ltd., fora fellowship (to J. C. G.).Note added in proof; We have re-examined pentacarbonyl-manganese chlorideand have succeeded in obtaining a photo-electron spectrum free of the impuritybands in the low ionization energy region (region A) that were probably due to forma-tion of the dimer [Mn(CO),CI],. The revised figures for the first three photoelectronband maxima are 8*83, 10.46 and 11.08 eV.D. W.Turner, Proc. Roy. Soc. A , 1968, 307, 15.D. W. Turner and D. P. May, J. Chem. Phys., 1966,45,471.R. B. King, J. C. Stokes and T. F. Korenowski, J. Organometal. Chem., 1968, 11,641.W. Hieber and G. Wagner, Z. Naturforsch., 1958, 136,339.W. Hieber and G . Wagner, Annalen, 1958,24,618.W. Beck, W. Hieber and H. Tengler, Chem. Ber., 1961,94, 862.E. W. Abel and G. Wilkinson, J. Chem. SOC., 1959,2,1501.M. I. Bruce, Adv. Organometal. Chem., 1968, 283.G. A. Junk, H. J. Svec and R. J. Angelici, J. Amer. Chem. Soc., 1968,90,5758.lo J. C. Hileman, D. K. Huggins and H. D. Kaesz, Znorg. Chem., 1962, 1,933.l1 E. 0. Brimm, M. A. Lynch and W. J. Sesney, J. Amer. Chem. Soc., 1954,76,3831.l2 D. J. Parker and M. H. B. Stiddard, J. Chem. SOC. A , 1966, 1,695.l 3 D. W. Turner in Physical Methods in Advanced Inorganic Chemistry, ed. H. A. 0. Hill andl4 T. Koopmans, Phybica, 1933, 1, 104.l5 R. E. Watson, Tech. Report no. 12 Solid State and Molecular Theory Group, (Massachusettsl6 P. E. Cade, K. D. Sales and A. C. Wahl, J. Chem. Phys., 1966,44,1973.l7 D. C. Frost, C. A. McDowell and D. A. Vroom, J. Chem. Phys., 1967,46,4255.l8 D. W. Turner, A. D. Baker, C. Baker and C. R. Brundle, Molecular Photoelectron Spectroscopy,l9 H. J. Lempka, T. R. Passmore and W. C. Price, Proc. Roy. Soc. A , 1968,304,53.2o G. Herzberg, Spectra of Diatomic Molecules, (Van Nostrand, New York, 1950) and B. J.21 K. G. Caulton and R. F. Fenske, Znorg. Chem., 1968, 7, 1273.22 J. S. Griffith, The Theory of Transition MetaZ Ions, (Cambridge University Press, 1961), andF. A. Cotton, Chemical Applications of Group Theory, (Interscience, New York, 1963).23 C. K. Jargensen, Orbitals in Atoms and Molecules (Academic Press, London, 1962), chap. 10,and T. M. Dunn, Trans. Faraday Soc., 1961, 57, 1441.24 L. F. Dahl and R. E. Rundle, Acta Cryst., 1963, 16,419.P. Day, (Interscience, London, 1965).Institute of Technology, Cambridge, Massachusetts, 1959).Wiley, 1969 (in press).Ransil, Rev. Mod. Phys., 1960, 32,245

ISSN:0366-9033    

DOI:10.1039/DF9694700112

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Photo-electron and Ultra-violet Spectroscopy of TransitionMetal Carbonyl DerivativesB Y P. S. BRATERMAN AND A. P. WALKERDept. of Chemistry, University of GlasgowReceived 3rd February, 1969Principles which have been used to explain the ultra-violet spectrum of metal carbonyls M(C0)6are applied to substituted derivatives of type M(CO)sL, M(COhL2. A model is developed which isconsistent with the spectra of substituted chromium carbonyls, but not with the photo-electronspectrum of CH3Mn(CO)5.The spectra of the unsubstituted metal hexacarbonyls M(CO), (M = e.g., Cr(0),1-4 v (-I),"'" Mn(I),39 Re(1)2* 4, are well understood; the vibronic nature of thet2g-+es transition^,^ and the variation in position and spacing of the charge transferbands with charge in isoelectronic series,'-'" confirm the original assignment despite~riticisrns.~ In this paper we discuss the spectra of substituted metal carbonyls oftype M(CO)5L and trans M(C0)4L2, using models related to that of the hexacarbonyldiscussion.A iiaive Huckel treatment of the hexacarbonyl system, neglecting overlap andalso the interaction of n* orbitals on different ligands, gives for the metal p (or rathero* (tl,)) and d(t2J orbitals, and for the ligand n* orbitals, secular equations of theform :( 1 ) 2P(d9n*) I = O,t l u 1 $ $ ) a(n*) - E (2) 2p(P,n*) I = O,C C ( ~ * ) - Et 2 g : I2P(d,n*) - E(3) tig, tzt, : CC(X*)- E = 0,whence second-order energies :t2g(l) W9), -40 ; tl,(l> (6n*9), a(n*)-4P; t l g , tzu(n*), +*) ;Here a is a Coulomb energy (relative to unperturbed d(t2&), p is the resonanceintegral between a single CO group and a metal orbital, and D, P are second-orderenergy parameters :D = p*(d,n*)/a(n*) ; P = p2(p,n*)/[~(p) - CC(~*)].( 5 )if we take the observed splitting between the t2g(l)-)~* bands as equal to thedifference in one-electron energy between the three groups of n* orbitals, then inCr(C0)6 we have4P = 8,700 cm-', 4 0 = 6,800 cm-l. (6)(1 eV = 8,066 cm-')12122 CARBONYL SPECTRAA similar formal treatment may readily be applied to the pentacarbonyls M(CO),L(ignoring differences in Coulomb energy between p(s) and p(x,y), between cl(xy)and d(xz,yz), and between n*(5) and n* (1-4) (in the notation of fig. 1)). Thisgives rise to zeroth order functions (normalization constants omitted) (point groupG V )t a-'54FIG.l.-Axis system for M(C0)5L.IIX4a, :a2 :b, :b2 :e(x) :e(Y> :whence secular equationsa, :e :(7)and second-order energies :b2(6d9), - 4 0 ; e(6d9), - 2 0 - D' ; u ~ ( % * ~ ) , a(x*)-4P; e(6z*9), a(n*)-4P;e(67r*(5)9), a'(n*) - P'+ D' ; a2,bl(n*), cc(n*) ;e(%~*~), tx(n*)+20; bi(6n*9), u(n*)+40;e(6p9), a(p)+2P+Pr ; a,(6p9), cr(p)+4P. (9)Here primes refer to CO(5); p(d,p) is assumed zero. Taking linear combinationsbetween ;n*(5) and the other e(n*) sets so as to diagonalize ligand-ligand interactionsdoes not greatly alter the picture given by eqn. (9).6Two separate models have been proposed to cope with the proliferation of para-meters in eqn.(9). The first model '* ' treats only the interaction between CO(5P. S. BRATERMAN AND A. P . WALKER 123and the metal as significant ; this could be rationalized by assuming a'(n*) 4 u(n*),'' *or by assuming P'> D'>>P,Da9 This gives an energy ordering.e(6d9) < b,(d) (full) < e(6n*(5)9) <other n*. (10)On this mode, the highest occupied orbital should be non-degenerate; the firstcharge transfer band (b2-+e(5)) should be absent in trans M(CO),L2, and the spectrumshould be relatively simple.the difference between primed and unprimed quantitiesis ignored. The resultant energy order isb2(6d9) <e(6d9)(full)<a,(6n*9) <e(n*(5))~a~,b~(n*)<e{~n*~) ~ b ~ ( ~ n * ~ ) <The highest occupied orbital is degenerate; the first change transfer band (c-a,)should be present in trans M(C0)4L2 and the spectrum may be expected to be ofconsiderable complexity.The observed spectra of the compounds Mo(CO),P(NMe,), and trans Mo(CO), [P(NMe2)3]2 are shown in fig. 2. Similar spectraare obtained for the chromium and tungsten analogues, and also lo for the compoundsMo(CO), PEt3, trans Mo(CO), (PEt,),.In the seconde(6P9) <a1VP9)* (1 1)4.000n'E 3.0000*'0 e g 2.000c. - wI,O 0 021FIG. 2.-Spectra of MO(CO)~P(NM~~)~ and trans MO(CO),[P(NM~~)~]~.These spectra accord with the second model; the lowest acceptor orbital mustbe located on CO groups 1-4 since the lowest charge transfer band is only slightlyaffected by the second substitution (moving to lower energy but with little change inintensity). The b2-+al charge transfer is forbidden, and would not be expected tobe as strong as the band under consideration ; the preferred assignment is thus e-+a,,as required by the second model.The energy of the first transition is listed for a variety of compounds in table 1.(The band in Cr(CO),P Ph3 which we correlate with the band of Cr(CO), differsfrom that chosen (for the Mo analogue) by other workers,ll but we would not resta case on this spectrum, wh:cli is ill-resolved at low energies, and which may becomplicated by configuration interaction.Whatever the preferred assignment fortriaryl phosphine derivatives, substitution of a trialkyl phosphine into a metalhexacarbonyl has a marked effect on the spectrum.) This energy may be expectedto depend on three sets of factors.(a) Factors affecting the energy of the accepto124 CARBONYL SPECRAorbital. A good a-donor would be expected to raise the energy of the cs,(n*) orbitalindirectly by raising that of p(z), and perhaps directly also through ligand-ligandinteractions. (b) The effect of a-donation on the donor orbitals. A good a-donorcan partially screen the 3d electrons from the metal nucleus, thus facilitating metal-ligand charge transfer. (c) Substituent n-effects; the energy of the 6dg(e) orbitalsmay be written as -20- D’- D (substituent), where D (substituent) depends on then-acceptor ability of the ligand (and is negative where the substituent is a n-donor).The order of table 1 suggests that this is the biggest single effect.TABLE FI FIRST CHARGE TRANSFER BOND IN Cr(CO)5Lv(cm-1)35,70030,40029,40029,40029,25027,85023,50022,800~(l.cm- ‘mole - 1)13,1001,8501,7701,8501,6801,3803,5001,030solvent ref.CHSCN a 4hexane (b)hexane (b)hexane (b)hexane (b)hexane (b)cyclohexane (c)CHSOH (4(a) Solvent effects on this band are small (see ref.(4) for comparative data. (6) this work(c) W. Strohmeier and K. Gerlach, 2. Phys. Chem. (Frankfurt), 1961, 27, 439. ( d ) P. S. Bratermanand R. A. N. McLean, unpublished results.The theory described here requires that the difference in ionization potentialbetween b2 and e levels be a direct measure of D(C0)- I) (substituent). Thishypothesis may be tested by photo-electron studies, but these are unfortunatelylimited by problems of volatility.A start has been made with the photo-electronspectrum of CH,Mn(CO),, part of which is shown in fig. 3 (this spectrum was kindly7 8 9 10 I I 12ionization potential (V)FIG. 3.-Photo-electron spectrum of CH3Mn(C0)5 (courtesy Dr. D. R. Lloyd).supplied by Dr. D. R. Lloyd). Integrated intensity measurements make it clear thatthe first ionization is from b2, rather than, as predicted, from e. The anomaly cannotbe explained by the electrostatic field of the CH3 group, nor by its n-bonding (beinP . S . BRATERMAN AND A. P . WALKER 125negative relative to Mn, it is expected to destabilize d(e) preferentially, and it is, ifanything, a n-donor). Two explanations remain :(a) In this case, D’% D (specifically D’-2D- 5,000 crn-l), as required by theJim model.(b) In this case, /3(p,d) # 0; the pronounced field that may be expected along thez-axis will cause mixing of p(e) and d(e) orbitals and lowering of the latter.It is now necessary to obtain photoionization spectra of compounds of typeCr(CO),L, and also to obtain the spectrum of CH,Mn(CO), at low temperature(the room temperature spectrum is ill-resolved) for comparison with the spectra ofc hr omi um derivatives .The authors thank Dr. D.R. Lloyd (for the photo-electron spectrum), R.Poilblanc, R. A. N. Maclean and D. W. Milne (for spectroscopic data) and theS.R.C. for an equipment grant.H. B. Gray and N. A. Beach, J. Amer. Chem. SOC., 1963, 85,2922.P. S . Braterman and A. P. Walker, Proc. 3rd Int. S’mp. Organometal. Chern., (Munich, 1967).N. A. Beach and H. B. Gray, J. Amer. Chem. SOC., 1968,90,5713.E. W . Abel, R. A. N. McLean, S. P. Tyfield, P. S. Braterman, A. P. Walker and P. J. Hendra,J . Mol. Spectr., 1969, 30, 29.A. F. Schreiner and T. L. Brown, J. Amer. Chem. SOC., 1968,90, 3366.P. S . Braterman, unpublished work. ’ H. B. Gray, I. Bernal and E. Billig, J. Amer. Chem. SOC., 1962, 84, 3404.H. B. Gray, E. Billig, A. Wojcicki and M. Farona, Cun. J. Chem., 1963, 41, 1281.H. B. Gray, private communication.lo R. Poilblanc, private communication.l1 D. J. Darensbourg and T. L. Brown, Inorg. Chem., 1968,7,959

ISSN:0366-9033    

DOI:10.1039/DF9694700121

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Mossbauer Spectra, Structure, and Bonding in Iron CarbonylDerivativesBY R. GREATREX AND N. N. GREENWOODDept. of Inorganic Chemistry, University of Newcastle upon Tyne NEl 7RUReceiued 16th January, 1969The Mossbauer spectra of the mono-, di-, tri-, and tetra-nuclear iron carbonyl anions and carbonylhydride anions are reported and compared with the known spectra of the neutral binary carbonyls.Increase in anionic charge in the sequence 0, - 1, -2 results in progressively lower values of chemicalisomer shift 6 for each series. In addition, there is a systematic increase in 6 as the co-ordinationnumber of iron increases from 4 to 7. These trends are interpreted in terms of changes in the bondingin these compounds. Systematic variations in the quadrupole splitting A are also discussed.The effects on 6 and A of substituting carbon monoxide by tertiary phosphines are investigatedand the structure and bonding in the unusual new compound [(OC)3FePMe2Ph]3 are discussed.The Mossbauer spectra of the tetrameric species [x-C5H5FeC0I4 and [x-C5HSFeCO]i are alsoinvestigated.The recent rapid expansion in the synthetic and structural aspects of metal carbonylchemistry presents the theoretician with the challenge of developing the existingtheories of chemical bonding to describe the many new molecular species which arebeing made.Moreover, the complexity and lack of symmetry in many of thesemolecules, coupled with the large mass of many of their constituent atoms poseformidable problems for ab initio treatments.There are, however, a number ofexperimental techniques which yield information about the bonding and one ofthe most direct of these is Mossbauer spectroscopy.l In the present context, thetwo main parameters of chemical interest are the chemical isomer shift 6, which is ameasure of the s electron density at the resonant nucleus, and the quadrupole splittingA, which stems from the electric field gradient at the nucleus as a consequence ofasymmetry in the electronic environment. In principle, therefore, Mossbauerspectroscopy gives experimental values for the wave functions at specific pointsin the molecule. For 57Fe, an increase in 6 corresponds to a decrease in the value ofC J $s(0) I 2, the total s electron density at the iron-57 nucleus. Changes in 6 thereforereflect (a) changes in s orbital population, (b) changes in the radial extension of sorbitals due to involvement in covalent bonding, and (c) changes in the shielding ofelectrons in s orbitals by other electrons, notably those in d orbitals whose populationand distribution are affected by the details of bonding.The quadrupole parameterA reflects any imbalance in p and d electron density at the iron atom and also incorpo-rates the effect of the geometric disposition of surrounding atoms on the electricfield gradient at the resonant nucleus.We have recently applied iron-57 Mossbauer spectroscopy to a variety of problemsinvolving iron carbonyl derivative^.^-^ The present paper refers briefly to this workwhere appropriate, but is chiefly concerned with the discussion of new results on thefollowing systems :12R .GREATREX A N D N . N. GREENWOOD 127(i) mono-, di-, tri-, and tetra-nuclear anions and hydride anions of the binaryiron carbonyls ;6 this work was undertaken jointly with Dr. M. Kilner and Dr. K.Farmery of the University of Durham-it involved a variety of preparative pro-cedures and physical techniques and will be reported in full elsewhere;'(ii) various phosphine substituted iron carbonyl derivatives ;(iii) the uncharged tetrameric cyclopentadienyliron carbonyl (n-C,H,FeCO),and the cation (n-C,H,FeCO)B obtained from it by oxidation.8The systems in (i) were chosen because they provided an opportunity of investi-gating iron in a wide variety of stereochemical environments and formal oxidationstates.In particular, these compounds contain iron in formally negative oxidationstates; chemical isomer shifts of iron in such compounds have only briefly beenalluded to previo~sly.~ The phosphine derivatives in (ii) were studied to definemore closely the differences between carbon monoxide and various phosphines asligands. The tetrameric cyclopentadienyliron carbonyl system (iii) provided a rareopportunity of studying the effects on bonding of the removal of a single electronfrom a polynuclear metal cluster to give a cationic species having the same basicstructure.EXPERIMENTALTwo Mossbauer spectrometers were used. The first has been described,2 and the secondwas a commercial instrument model AM-] from the Nuclear Science and EngineeringCorporation, used in conjunction with a 400 channel R.I.D.L.analyser, model 34-12B.The source was a nominal 10 mC of 57C0 diffused into palladium as supplied by the Radio-chemical Centre, Amersham.All results were obtained using absorbers at liquid nitrogen temperature and were fittedby computer programmes developed for the purpose;'O they are quoted relative to theroom temperature spectrum of hydrated sodium nitroprusside. A confidence limit of&0.01 mm/sec is placed on most of the parameters tabulated though occasionally, wherethe peaks severely overlap, this limit is increased slightly. In many cases, spectra werealso run at room temperature but, as the results were unchanged except for the expectedsmall second-order Doppler shift, they are not included here.The iron carbonyl anions and hydride anions were prepared by Dr.K. Farmery andDr. M. Kilner who also supplied the sample of (CO),Fe(SPh),Fe(CO),PPh,. All werewell characterized and gave good Mossbauer spectra except for the tetranuclear hydrideanion [Fe,(CO), 3H]-. Many attempts were made to obtain pure compounds containingthis anion but they invariably showed two additional peaks whose intensity varied fromspectrum to spectrum and hence could be assigned to some impurity (probably the neutraltetranuclear hydride [Fe4(CO) I 3H2]). The reproducible part of the spectrum consistedof two broad peaks, one more intense and narrower than the other. The parameters forthis species are therefore only approximate and were arrived at by considering the spectrumto be merely a quadrupole split doublet.The new compound [(CO),FePMe2Ph], was suplied by Dr.J. Moss and Dr. B. L. Shawof the University of Leeds. The tetrameric cyclopentadienyl compIex [n-CSH5FeC0l4and its bromine derivative [n-CSHSFeC0I4Br3 were prepared according to the literature.sThe other derivatives [n-CSHsFeC0]4CI and [ ~ E - C ~ H ~ F ~ C O ] ~ ~ ~ (n - 5 ) were obtained bydirect reaction of chlorine and iodine with the neutral complex in dichloromethane solution.RESULTS AND DISCUSSIONBINARY CARBONYLS, CARBONYL ANIONS AND HYDRIDE ANIONSThe compounds studied have the formulae [Fe,(CO)+x~ll-,,] where x = 1-3 ;and [Fe,(CO)+,~, 1-x)-lJ2- and [Fe,H(CO)+,(, l-x)-l]- where -Y = 1-4.The serie128 MGSSBAUER SPECTRA OF IRON CARBONYL DERIVATIVES[FexH2(CO),,,, l-x)-l] is the subject of a separate investigation.' Known (or. . suggested) structures are shown in fig. 1. The- structures ofthe binary carbonyls1. Fe(CO)5 2. Fe2(CO)9bS t cuc tureunknown8. Fe(CO),H- 9. Fe2(C0)*H' 10. Fe3(CO), ,H- 11. Fe4(CO)1sH-FIG. 1.4tructures of iron carbonyls, carbonyl anions, and carbonyl hydride anions.themselves (compounds 1-3) are now well established 9 9 11-13 ; their Mossbauerspectra have also been the subject of numerous investigations 2* 3* l4-I6 and aregiven again here merely for the purpose of comparison; Of the dianions (compounds4-7) only the tetranuclear species has received a complete X-ray diffraction study ; l7conclusions regarding the structures of the di- and tri-nuclear anions are onlypreliminary.18* l9 The suggestion that the mononuclear dianion is tetrahedral isconsistent with its infra-red and Raman spectra.20* 21 In general, much less isknown about the hydride anions (compounds 8-1 1), and only the tri-nuclear specieshas had its structure established by X-ray analysis.22 The mononuclear hydrideanion has been the subject of controversy and its structure is still not establishedunequivocally.20 The structure of [Fe,(CO),H]- was recently established by Moss-bauer and vibrational spectroscopy and no structural information is availablefor the tetra-nuclear hydride anion (compound 11).The Mossbauer data are given in table 1.The results on the first three compoundsagree well with previous data; in addition, as shown in fig. 2, the central peak forFe,(CO),, is now partially resolved for the first time. The data on [Fe,(CO),H]-support the structure given in fig.1 and the results on [Fe2(CO)8]2- and [Fe,(CO),support the preliminary X-ray proposals l9 also shown in fig. 1 ; in allcases only a single, unbroadened quadrupole doublet is obtained, implying theequivalence of the iron environments within each molecule. However, it shouldalso be noted that the spectrum of the tetranuclear anion [Fe,(CO),3]2-, whicR . GREATREX AND N. N. GREENWOOD 129was obtained from both its Et,N+ and [Fe(py)J2+ salts, gave a barely resolveddoublet though the X-ray structure (see fig. 1) might lead one to expect a pair ofdoublets, one three times as intense as the other; it is indeed possible to analysethe spectrum on this basis, thus illustrating the need for caution in interpretingunresolved spectra.TABLE 1 .-M~SSBAUER PARAMETERS FOR BINARY IRON CARBONYLS, CARBONYL ANIONS, ANDHYDRIDE ANIONSno.compound chemical isomer quadrupoleshift ( 6 mm/sec)* split (A mm/sec)*0.17 2-570.42 0.42(a) 0.37 1-13(b) 0.31 0-138 [Et4Nf][Fe( C0)4H]- 0.09 1-369 [ E t 4N+][ Fez( CO),H]- 0.33 0-5010 [ E ~ ~ N + I [ F ~ ~ ( C ~ ) I iw- (a) 0.30 1.41(b) 0.28 0-1611 [PYH+I[F~~(CO>I~HI- -0-33 -0-67* for s7F0 1 mm/sec = 4.638 mJ/mole = 11.63 Mclsec. t for anion only; cation has 8 = 1.38 mmlsec, A = 1.19 mmlsec.-1.0 -0.5 0 0 - 5 1.0 1-5 2.0velocity (mm/sec relative to nitroprusside)FIG. 2.-Mossbauer spectra of (Q) Fe3(CO)1 and (b) [(OC)3FePMe2Ph],.Two general trends are observable in the chemical isomer shift values in table 1 :(i) increase in the anionic charge from zero, through - 1 to -2 results in progressivelyThis trend is illustrated in fig.3(a). (ii) The chemical isomer shift increases as theco-ordination number increases-see fig. 3(b). Thus for the species with co-ordinationlower Values Of 6, e.g., Fe2(C0)9 0.42 ; [Fe,(CO)sH]- 0.33 ; [Fe,(co)~]~- 0.18 mm/Sec.130 MoSSBAUER SPECTRA OF IRON CARBONYL DERIVATIVESnumber four 6 is 0.08 mm/sec, for co-ordination number five it falls in the range0*09-0.18 mm/sec, for co-ordination number six in the range 0.27-0.33 mm/sec andfor co-ordination number 7 in the range 0.31-0.42mm/sec. In plotting fig.3(b),iron-iron interactions have been included but the triply bridging carbonyls in[Fe,(CO), ,I2-, [Fe,(CO), J2- and [Fe4(CO), ,HI- have been ignored ; if they areincluded the trends remain but the ranges are extended somewhat.040 *3 f - 0.2 m0.1- 2 -1 0total charge on speciesFIG. 3.Trends of chemical isomer s( b )0 45 fj E;.0 82 g 3aI I I I4 5 6 7co-ordination number* ignoring triply bridged carbonylsift 6 with charge and co-ordination number.Trends in the quadrupole splitting parameter A are governed predominantly bythe geometrical distribution of the iron nearest-neighbour atoms and by the asym-metric delocalization of electron density on to the ligand since, ultimately, the electricfield gradient must be described by the distribution of the electronic wave functionsabout the nucleus :eq = e(3 cos28- l)(r3).Thus, the data in table 1 show that tetrahedral co-ordination gives the smallestquadrupole splittings (A - 0 mmlsec) and that trigonal bi-pyramidal co-ordinationgives the largest splittings (A = 2.1-2.6 mm/sec).Six co-ordination results in smallsplittings which, depending on the degree of distortion from true octahedral symmetry(A = 0 rnm/sec), fall in the range 0-1-0.6 mrnlsec, whereas seven-co-ordinate ironin these compounds gives larger values in the range 0-4-1-4 mmlsec. The value ofA for [Fe(C0)4H]- (1.36 mmlsec) is particularly interesting since it is intermediatebetween the values for the tetrahedral ion [Fe(C0)J2- (A = 0) and the trigonalbipyramidal species Fe(CO)5 (A = 2-57mm/sec).This implies that the threeequatorial carbonyl ligands in the pentacarbonyl are displaced so as to give C3vsymmetry in the hydride,20 as shown in fig. 1.A more detailed, though still qualitative, interpretation of the trends in chemicalisomer shift values can be made in terms of changes in the population and radialextension of the 3d and 4s orbitals on the iron atoms (or changes in the molecularorbitals incorporating these atomic orbitals in their basis sets). Thus, there is aprogressive decrease in chemical isomer shift (increase in s electron density at theiron nucleus) in the mononuclear series [(CO),Fe+-(CO)](D,,), [(CO),Fe+ H-](C,J,and [(CO),Fe+ : 2-](Ta).The polynuclear series behave similarly (fig. 3). ThesR. GREATREX AND N . N . GREENWOOD 131is arise as the result of three simultaneous interactions which increase the sron density at the iron nucleus and two weaker effects which tend to diminishelectron density : (i) increase in forward a-donation into orbitals involvingron 4s orbital as a result of both the increase in anionic charge and the reduction)-ordination number; (ii) increase in n-back donation of electron density onie ligands, with concurrent diminution of shielding by 3d electrons and resultant:ase in s electron density at the iron nucleus ; (iii) radial expansion of the non-ling electrons as a result of the progressive increase in anionic charge, withurrent diminution of shielding and increase in s electron density at the ironx i s ; (iv) increase in forward o-donation into orbitals involving the iron 3d;ah, thereby increasing the shielding and diminishing the s electron density atiron nucleus; (v) radial expansion of the bonding orbitals (including thoselving iron 4s electron density).lome authors 23* 24 have favoured (i) as the predominant influence on chemicalLei shift in low-spin iron complexes, and others Is* 2 5 9 26 have emphasized (ji),gh there is in fact insufficient information to decide this point and most treat-ts have ignored one or more of the above effects.Theoretical treatments areEciently precise to determine the magnitude of individual contributions to the. s electron density at the iron nucleus in real systems because of the number andnetric complexity of the orbitals involved and the problems associated with theivistic correction of wave functions for atoms as massive as iron.A more.aniental difficulty is that most currently used computational methods involveition or perturbation methods which minimize the total energy of the system,gh this is notoriously insensitive to the detailed form of the orbitals usedicularly when shielding efficiencies and densities at a central nucleus are required.Zis context, it is worth recalling that the electron density which determines theiical isomer shift amounts to only 10-lo of the total electron density on the iron1. The present data on a systematic series of related compounds thereforeide the theoretician with more precise information against which to test hisdations.n the absence of more precise theoretical guidance on the relative magnitudesie five effects listed above, an experimental approach was adopted in which theonyl ligands were replaced by tertiary phosphines rather than by H-; thishates the influence of anionic charge and alters the relative a- and n-bondingtensities of the ligands.The results are presented and discussed in the nexton.PHOSPHINE-SUBSTITUTED IRON CARBONYL DERIVATIVES'able 2 summarizes the results obtained on certain phosphine-substituted irononyl derivatives. Typical spectra are shown in fig. 2 and 4 which illustrate theence of substituting terminal carbonyls by tertiary phosphines and the degreesolution obtained.There is no general trend in either 6 or A, an observation:h demonstrates the fhtility of attempting to generalize the assignment of partialier shifts 27 to substituent ligands in systems of this type. A closer inspectionie results reveals that (i) substitution of CO by PPh3 or PEt, in the mononuclears decreases both 6 and A slightly; Pi) substitution of three bridging CO groupsNO bridging PMe, or SPh groups in the binuclear series decreases S and increasesibstantially (the same pattern of change is found when terminal carbonyls aretituted by PMe,Ph in the trinuclear series); (iii) substitution of a terminal COL tertiary phosphine in (OC) Fe(PMe,), Fe(C0) and (OC) Fe(SPh),Fe(CO)132 MOSSBAUER SPECTRA OF IRON CARBONYL DERIVATIVESTABLE 2.-MOSSBAUER PARAMETERS FOR PHOSPHINE-SUBSTITUTED IRON CARBONYL DERIVATIVES"no.1234567891011chemical isomershift (6 mm/sec)t0.170.1350.14-0.1 70.16(OC)3Fe(CO), Fe(C0) 3 0.42(OC), Fe(PMe2)JWCO) 3 0-21(OC), Fe( PMe2)2 Fe( CO),PEt 0.22(OC)3Fe(SPh)2Fe(CO)3 0.32(OC)3Fe(SPh)2Fe(CO),PPh3 (a) 0.31(6) 0.39Fe3(co)12 (a) 0.37(6) 0.31[ (OC)3FePMe2Ph] (a) 0.35(b) 0.28quadrupolesplit (A mm/sec)t2.572.3 12.760.420.690-661 -071 -020.861.130-131.150-572.42-2.54* Data for the first four compounds were taken from the compilation by R. H.Herber, Int. Atomict for s7Fe 1 rnmlsec = 4.638 mJjmole = 11.63 Mc/sec.Energy Agency (Vienna) Tech. Rept. Series, 1966, 50, 121.37,00036,00035,00034,00034,0003 3,00032,000-1.0 - 0 - 5 0 0 .5 1.0 1.5velocity (mmlsec relative to nitroprusside)FIG. 4.-Mossbaue r specma of (a) (0 C) Fe( SPh)2 Fe( CO) and (b) (OC) Fe(SPh)z Fe(C0) ZPPh 3.increases 6 slightly but diminishes A. The only combination unobserved on carbonylsubstitution is the simultaneous increase of both 6 and A.The foregoing effects of phosphine substitution on the chemical isomer shiftcan be understood in terms of the chemical characteristics of the group trans to thecarbon monoxide molecule being replaced. When this group is also a carbonyl,6 decreases due to the greater c donating ability of the phosphine and the ability ofthe iron atom to delocdize any excess (shielding) d electron density on to the rernain-ing carbonyl ligands.The substantial diminution in 6 when bridging carbonylgroups are replaced by bridging phosphido- and thio-groups stems from the samR . GREATREX AND N. N. GREENWOOD 133cause. By contrast, when a terminal carbonyl in these latter compounds is thensubstituted by a tertiary phosphine (which is a weaker n-acceptor than carbonmonoxide) the ability to delocalize the shielding d electrons diminishes and 6 showsa slight increase. In this particular case it would appear that the diminished ndelocalization more than compensates for the increased 0 donor strength of thephosphine. A similar effect appears to be operating in the mononuclear series where6 initially decreases from 0.17 to about 0.14mm/sec, then increases again to 0.16mmlsec on substitution by the second phosphine.However, some of these effectsare sniall and approach the limits of experimental accuracy of the data. The overallimpression is one of insensitivity of 6 to the substituent in these low-spin neutralcomplexes. The quadrupole splitting is also rather insensitive to substitution of thecarbonyl groups except when the co-ordination number of the iron changes, as ingoing from compound 5 to compound 6 or 8.The filial compound in table 2 merits further discussion since it provides anexample of the success of Mossbauer spectroscopy in providing structural informationas yet unobtainable by other means. The spectrum is shown in fig. 2 together withthat of Fe3(CO),, from which it was synthesized ; 28 it comprises a pair of quadrupolesplit doublets with an intensity ratio 2 : 1 thus implying a trinuclear species ratherthan a di- or tetra-nuclear species (other evidence was contradictory on this point).The large change in the quadrupole splitting of the apical iron atom from 0.13 to0.57 nimlsec might be thought to suggest that substitution has occurred solely at thisiron atom as in fig.Sa. However, this is chemically improbable and further inspectionFIG. 5.-Possible structures for[(OC)3FePMe2Ph]3 : (a) rejectedstructure ; (b) suggested structure.of the data reveals that equal changes have occurred in the chemical isomer shifts of allthree iron atoms. Whilst this is not conclusive evidence that substitution has occurredsingly on each iron atom as in fig. 5(b) it is strongly suggestive and this structure istentatively adopted.The relatively large change in the quadrupole splitting of theapical iron atom reflects the changed geometry at this point whereas the much smallerchange at the other two iron atoms parallels the insensitivity to change alreadynoted for terminal substitution on iron atoms of bridged dimers.CYCLOPENTADIENYLIRON CARBONYL TETRAMER AND ITS UNIPOSITIVECATIONThe compound [;n-C,H,FeCO], was first prepared by King and has the structureshown in fig. 6 ; it consists of a tetrahedron of iron atoms carrying a cyclopentadienylring normal to the diagonal above each apex and a carbonyl group above eachtriangular face. A cationic species was formulated as the tribromide salt[n-C,H,FeCO],Br, and it was suggested that other intractable black product134 MOSSBAUER SPECTRA OF IRON CARBONYL DERIVATIVESobtained with certain other oxidants might contain the same cation.As thesematerials were too insoluble for study by most other techniques, the Mossbauerspectra of the bromide and similar oxidation products were investigated to obtainFIG. 6.-Structure of [x-C5H5FeC0J4.further information on the structure and bonding in these species. The results arein table 3 ; the spectra of all four compounds are simple quadrupole split doubletswith essentially the same value of the chemical isomer shift. This suggests thatoxidation of [n-C5H,FeC0J4 to [n-C,H,FeCO],+ involves a molecular orbitalaffecting all four iron atoms equally since removal of the electron from an orbitalbased predominantly on one iron atom would give a pair of quadrupole doubletsTABLE 3.-MOSSBAUER PARAMETERS FOR [n-CsH5FeCOl4 AND ITS CATIONIC DERIVATIVESchemical isomer quadrupoleshift (d mm/sec)* split (A mm/sec)* compund[z-CSH5FeC0I4 0-66 1.76[n-CSH ,FeC0I4Cl 0.67 1.38[n-C5H5FeC0I4Br3 0.67 1 *40[n-C5H5FeC0I4IN 0.67 1 *42* for 57Fe 1 mm/sec = 4638 mJlmole = 11.63 Mc/sec.with an intensity ratio 3 : 1.(A less attractive alternative is that the unpaired electronis located on a particular iron atom but that it hops from one iron atom to the nextwith a frequency which makes all four atoms appear identical on a time scale oflo-’ sec.) The near constancy of 6 can be taken to imply removal of a ligand-basedelectron or alternatively a compensating set of changes at the iron atoms whichleave the resultant s electron density at the iron nuclei essentially unchanged.Suchchange in 6 as is observed on oxidation is in the opposite direction to that obtainedfor the anionic species already discussed.The quadrupole splitting for the neutral complex [n-C,HSFeC0I4 is typical ofcompounds containing the cyclopentadienyl group.2- 27 Oxidation diminishes thisquadrupole splitting by a similar amount for each of the three halogen derivativesthus suggesting the presence of the same species in each case. Further evidence onthis point comes from the infra-red spectra-all three compounds were found tohave closely similar spectra with a single peak at 1695 cm-l for the chloride andtribromide and at 1700 cm-l for the polyiodide, characteristic of the triply bridgingcarbonyl groups.In the neutral tetramer this peak appears at a lower wave number(1620 cm-l) as expected. Closer inspection of the values of the quadrupole splittinR . GREATREX AND N. N. GREENWOOD 135shows that A for the chloride is slightly less than that for the polyhalide derivatives,suggesting increased polarization of the electron cloud of these latter anions by thecation.N. N. Greenwood, Chem. in Brit., 1967,3,56, and ref. therein.T. C. Gibb, R. Greatrex, N. N. Greenwood and D. T. Thompson, J . Chem. SOC. A , 1967,1663.T. C . Gibb, R. Greatrex and N. N. Greenwood, J . Chem. SOC. A , 1968,890.K. Farmery, M. Kilner, R. Greatrex and N. N. Greenwood, Chem.Comm., 1968, 593.R. Greatrex, N. N. Greenwood and P. L. Pauson, J. Organometul. Chem, 1968,13,533.for recent reviews see : (a) R. B. King, Adu. Orgummetal. Chem., 1964, 2,218 ; (b) W. Hieber,Angew. Chem 1960,72,795 ; (c) M. C. Baird, Pro.qr. Inorg. Chem., 1968,9, 1. ' K. Farmery, M. Kilner, R. Greatrex and N. N. Greenwood, to be published.R. B. King, Inorg. Chem., 1966,5,2227.N. E. Erickson and A. W. Fairhall, Inorg. Chem., 1965,4,1320.lo B. J. Duke and T. C. Gibb, J. Chem. SOC. A, 1967, 1478.l 1 A. W. Hanson, Acta Cryst., 1962, 15,930.x 2 H. M. Powell and R. V. G. Ewens, J. Chem. SOC., 1939,286.l3 C. H. Wei and L. F. Dahl, J. Amer. Chem. Sac., 1966,82,1821.l4 M. Kalvius, U. Zahn, P. Kienle and H. Eicher, Z. Naturforsch., 1962, 17,494.l5 R. H. Herber, W. R. Kingston and G. K. Wertheim, Inorg. Chem., 1963, 2, 153.l6 E. Fluck, W. Kerler and W. Neuwirth, Angew. Chem., (Eng. ed.), 1963,2,277.l7 R. J. Doedens and L. F. Dahl, J . Amer. Chem. SOC., 1966,88,4847.l8 0. S. Mills, quoted by R. S . Nyholm, Proc. Chem. SOC., 1961,273.l9 0. S. Mills, A. A. Hock, and G. Robinson, 17th Int. Congr. Pure Appl. Chem., (Munich, 1959),p. 143, abstract A.2o H. Stammreich, K. Kawai, Y . Tavares, P. Krumholz, J. Behmoiras and S . Bril, J. Chern.Phys.,1960, 32,1482.W. F. Edgell, J. Huff, J. Thomas, H. Lehmann, C. Angel1 and G. Astata, J . Amer. Chern. Soc.,1960, 82, 1254.22 L. F. Dahl and J. F. Blount, Inorg. Chem., 1965,4,1373.23 R. L. Collins and R. Pettit, J. Amer. Chem. SOC., 1963,85,2332; J. Chem. Phys., 1963,39,3433.24 R. L. Collins, R. Pettit and W. A. Baker, J. Inorg. Nuclear Chem., 1966. 28, 1001.2 5 J. Danon, Int. Atomic Energy Agency (Vienna) Tech. Rept. Series, 1966, 50,89.26 N. E. Erickson, Ado. Chem. Ser., 1967,68,80.27 R. H. Herber, R. B. King and G. K. Wertheim, Inorg. Chem., 1964,3,101.28 J. Moss and B. L. Shaw, private communication

ISSN:0366-9033    

DOI:10.1039/DF9694700126

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Bonding and Structure in Fe(I1) Low-spin Compounds usingthe Mossbauer EffectBY G. M. BANCROFT, M. J. MAYS AND B. E. PRATERUniversity Chemical Laboratories, CambridgeReceived 10th February, 1969Mossbauer centre shifts (C.S.) and quadrupole splittings (Q.S.) are presented for a wide range ofsix-coordinate low-spin compounds, many of which have not been prepared previously. The centreshift is an additive property of the number and type of ligands in a particular compound, and partialcentre shifts (P.C.S.) values have been derived for twenty-two ligands. The P.C.S. values can berelated to the bonding properties of the various ligands, and the bonding properties of the novelligand N2 have been examined and compared with those of CO. There is an inverse relationshipbetween the P.C.S.values and the spectrochemical ranking of ligands. The trans : cis 2 : 1 quad-rupole splitting ratio holds for ligands such as C1-, SnCl; as well as for CN-, and the variation ofQ.S. with change in neutral Iigand L for trans-FeL,Cl, is discussed in terms of the bonding propertiesof the neutral ligands.The Mossbauer centre shift (C.S.) and quadrupole splitting (Q.S.) are potentiallyuseful for elucidating bonding and structural changes in solids. In several Fe(I1)low-spin compounds, the changes in C.S. have been attributed to n-bonding effects,land a correlation has been noted between C.S. and the spectrochemical A for a smallrange of 1igands.l Quadrupole splittings in the trans and cis isomers of octahedralFe(1I) compounds have been rationalized using the point-charge model outlinedby Berrett and Fitzsimmons,2 while Dale et aL3 have used molecular orbital theoryto account for their results.of a large number of Fe(I1) six-coordinate low-spin compounds, many of which have not been prepared previously.A partialcentre shift (P.C.S.) value can be assigned to each ligand and in the majority ofcases this value is essentially independent of the other ligands bonded to the ironatom. There is a good correlation between the P.C.S. values and the spectrochemicalA values over the whole range of the spectrochemical series, and this correlation isrationalized in terms of the 0 donor and z acceptor properties of the ligands. TheP.C.S. values have been used to examine the bonding properties of novel ligandssuch as molecular nitrogen, and these values should also be helpful in preparingnew Fe(I1) compounds showing intermediate spin properties or high spindow spinequilibria.The geometry of several of the complexes is assigned using the quadrupolesplitting values and the quadrupole splittings are rationalized in terms of a latticecontribution from the '' ionic " ligands such as C1, Br, I, and a valence contributionfrom the neutral ligands such as RNC and phosphines.We report Mossbauer spectraEXPERIMENTAL AND RESULTSThe preparation, analyses, infra-red spectra and conductivities of the new com-pounds will be presented in a later p~blication.~ The Mossbauer techniques and13G .M. BANCROFT, M. J . MAYS AND B .E . PRATER 137spectrometer have been published previously.6 A 57C0 in Pd source was used forall spectra. All absorbers contained about 5 mg/cm2 of natural iron, and all C.S.values are quoted relative to stainless steel (add 0.16 mm/sec to convert to sodiumnitroprusside).The Mossbauer centre shifts and quadrupole splittings from a selection of ourcompounds and others reported in the literature are listed in tables 1 and 2. TypicalMossbauer spectra of trans-FeCl,(ArNC), and cis-FeH,(CO), are shown in fig. 1.Many of the spectra in this study were distinctly asymmetric (fig. la). A discussionof the potential value of this asymmetry in determining the sign of the field gradientwill be publi~hed.~,I I I i I I I-3 -2 -1 b 1 2 3velocity (mmlsed(a)-r----f---I-- ' I 1-3 -2 -1 0 1 2 ?velocity (mmlsec)(6)FIG.1 .-Mossbauer spectra of (a), tran~-FeCl~(ArNC)~ at 295°K ; (b) cis-FeH,(CO), at 80°K.The small peaks in (b) are due to Fe(CO)+Looking at table 1, a number of general observations can be made. The iso-cyanide compounds give much lower C.S. values than the compounds containingphosphine ligands. The SnClT; ligand gives rise to lower centre shifts than thehalides, while hydride and silyl ligands in cis-FeH,(CO), and cis-Fe(SiH,),(CO),give rise to centre shifts which are nearly as negative as that in Na2Fe(CN),NO*2H,0.The compounds trans-[FeH(N2)(depe),]+[B(C6H5)4]- and trans-[FeH(CO)(depe),]+[B(C,H,),]- have very different centre shifts. In the depe compounds, the bromideand iodide have significantly higher C.S.values than chloride (compounds 18, 19,20).From table 2, it is evident that the 2 : 1 trans-cis quadrupole splitting ratioholds for ligands such as C1-, SnCl;, CN- and RNC. Also the Q.S. is additive forcompounds containing two different ionic ligands. Thus the value forcis-FeCl(SnCl,)(ArNC), is intermediate between those of cis-FeCl,(ArNC), an138ci~-Fe(SnCl,),(ArNe)~. In the chlorides, the Q.S. value varies from 1.54 mmlsecfor trans-FeCl,(ArNC), to 1.29 mm/sec for trans-FeCl,(depe), to 1 - 13 mm/secfor trans-Feel2( dep b) 2.BONDING IN Fe(I1) LOW-SPIN COMPOUNDS123456789101112TABLE l.-Roo~ TEMPERATURE CENTRE SHIFTS *t rans-FeCl ( ArNC), tcis-FeCI,(ArNC),tran~-Fe(SnCI~)~ (ArNC)4~is-Fe(SnCl~)~(kNC)~[Fe(SnCl3)(ArNC),]C1O4cis-FeH, (C0)4cis-Fe(SiH3), (CO),Fe( MeNC) (HSO4)2Fe(EtNC) 6(C104)2K4FdCN) 60.200120.080.1 1&0.030.090-06f 0.020.02-0.15-0.10- 0.020.000.051314151617181920212223Na3[Fe(CN)SNH3]. H20trans-FeC12(depe), ttrans-FeBr,(depe),trans-FeI (depe)t ram-FeCl,(depb) ttrans-[ FeH(N2)(depe) 2]+trans-[FeH(CO)(depe),]+[B(C 6H5) 41-[B(C,H5)43--0.16- 0.010.060.100.1 00.430.500-490.430.16- 0.04* error in results = f0-01 mmlsec, except where quoted (see ref.(4)).t ArNC = p-methoxyphenyl isocyanide ; depe = bis(diethy1phosphino)ethane ; depb = o-phenylenebis(diethy1phosphine).12345678910111213TABLE 2.-ROOM TEMPERATURE Q.S.*tran~-FeCl,(ArNC)~ 1-54ci~-Fecl,(ArNC)~ 0.78tran~-Fe(SnCl~)~(ArNC)~ 1 -05cis-Fe(SnC1 3)2(ArNC)4 0-52[FeCI(ArNC)S]C104 0.73c ~ s - F ~ C ~ S ~ C I ~ ( A ~ N C ) ~ 0.61[Fe(SnC13)(ArNC)S]C104 0.32trans-FeC12(depe), 1 *29trans-FeBr ,(depe), 1-37trans-FeT,(depe), 1.33t ram-FeCI (dep b) 1.13trans-Fe( EtNO4(CN); 0.60cis-Fe(E tNC)4(CN)i 0.29* error = f0.01 mm/sec; the sign of the Q.S.is taken to be negative for all trans compounds.st ref. (2).DISCUSSIONCENTRE SHIFTSWe postulate that the C.S. is a simple algebraic sum of partial centre shift (P.C.S.)values for each ligand6i = 1 C.S. = 2 (P.C.S.)i. (1)We calculate our C.S. values relative to stainless steel, and assume that the P.C.S.value for our isocyanide ligand, p-methoxyphenyl isocvanide, is 0.00 mm/sec, aG .M. BANCROFT, M. J . MAYS AND B . E . PRATER 139has been found for other isocyanide ligands which have been studied (table 1). Wealso assume that any change in C.S. is due solely changes in isomer shift (I.S.).*For the above postulate to be valid for a large range of Fe(I1) compounds, theC.S. must be insensitive to any small changes in bond angles and Fe-ligand bonddistances which occur from compound to compound.Also the P.C.S. values for a particular ligand must be comparatively insensitiveto the bonding properties of the other ligands in the complexes. Thus, ideally, thechange in P.C.S. value from one ljgand to another should be large in comparisonwith the variation in the P.C.S. value of one ligand in different complexes.We have calculated P.C.S. values (table 3) at 295°K mainly because more dataare available at this temperature.These values have been calculated from 32 com-pounds and have been used to predict C.S. values for a further 32.49 Only sixout of the 32 predicted values do not give results in good agreement (-0.05 mm/sec)with the observed values. These discrepancies are undoubtedly due to a change inP.C.S. value of ligands, such as CO and RNC, from compound to compound.Despite these discrepancies, the derived values are widely applicable and the generallygood agreement lends support to our initial assumptions.NO+H-SiH 5ArNCMeNCEtNCcoCN-SnCI ;niox/21.so;-TABLE PA PARTIAL CENTRE SHIITS* AND PARTIAL A VALUES;P.C.S.d (cm-1) P.C.S.- 0.20 15,400 NO, 0.05- 0.08 7,700 NH3 0.05- 0.05 Pd4t 0-050.00 depb/2 0.050.00 depe/2 0.060.00 diPY 12 t 0.060.00 8,200 P h d 2 t 0.070.01 7,900 PY 0.070.01 5,600 c1- 0.100-04 Br- 0-130.04 I- 0.1 36 (cm-1)6,7005,4006,1006,1005,7005,7004,6002,7002,4001,700error usually f0-01 mm/sec ; $ error estimated at f250 cm-l.f niox = cyclohexane-1,2-dione dioxime ; py = pyridine ; pc- = phthalocyanine ; dipy = 2,2’-dipyridyl ; phen = 0-phenanthroline.The range of P.C.S. values can be attributed to differences in c donor and xacceptor properties of the ligands studied. In these compounds, the most importantbonding interactions are Q bonding from the ligand to a hybrid d2sp3 iron orbital,and n back-bonding from the filled iron dxy, d’z, dxz orbitals to empty n* orbitalsof the neutral ligands.Both interactions lead to an increase in s electron densityat the nucleus, and a decrease in C.S. For Q bonding (L+M), donation of electronsinto the 4s orbital will influence the s electron density at the nucleus more stronglythan donation into 3d or 4p, (which shield the 3s and 4s electrons), and thus the selectron density at the nucleus will increase as the c donor power of the ligandincreases. With increased n backbonding into empty n* orbitals of the ligands,the dxy, dyz, dxz orbitals become more delocalized and shield the 3s+4s electronsless, increasing the s electron density at the nucleus.* C.S. = 1.S.S-S.O.D. where S.O.D. = second order Doppler shift and includes Z.P.M.=zero-point motion shift. % 140 BONDING IN Fe(1I) LOW-SPIN COMPOUNDSThus, as we would expect from the above considerations, the ligands giving themost positive P.C.S. values (I-, Br-, C1-) are the most ionic in character with com-paratively little covalent bonding, whereas hydride (strong B donor) and NO+ (strongn acceptor) have the most negative P.C.S. values. The C.S. values for trans-[FeH(N2)(depe)2]+[B(C6H5)4]- and tran~-[FeH(Co)(depe)]+[B(C~H~)~]- (table 1)suggest that CO is an appreciably better 0 donor and/or TC acceptor than N,.The P.C.S. values for CO (calculated from cis-FeH,(CO),), CN- (calculatedfrom K,Fe(CN),), and RNC (calculated from Fe(EtNC) 6(C104)2) are similar andreflect the similar (Q+ n) bonding characteristics of each ligand in the above com-pounds.However, there are several pieces of evidence which suggest that the P.C.S.values of RNC and CO and thus their bonding properties vary from compound tocompound. For example, cis-FeCI,(ArNC), has an appreciably lower C.S. thantrans-FeCl,(ArNC),, while cis- and trans-Fe(SnCI,),(ArNC), have similar shifts.The difference in the values for the chlorides suggests that the P.C.S. value of ArNCdecreases, (becomes more negative) by 0.02 mm/sec on going from the trans to thecis complex. In the cis compound, two of the ArNC groups should become moreeffective n-acceptors because they will no longer be competing for d electron densitywith trans ArNC groups. The c donor power of the isocyanide ligands may alsoincrease in the cis compound, and this may make an additional contribution to theobserved change in C.S.values. If we tentatively attribute differences in the C.S.values of cis-trans pairs to differences in 71: bonding, then we can separate B and ninteractions using the P.C.S. values as an estimate of ~ + n , and C.S.,,,,,-C.S.,i,as an estimate of the difference in n bonding for the two ligands. For example, onthe above basis, we would suggest that the SnClJ and ArNC groups have similar7r accepting properties, but that ArNC is an appreciably better 0 donor. Thisprediction is consistent with the conclusion reached by Parshall from n.mr. datafor the SnCl; ligand.atQ .- - 1 1II I I-0.20 -0.15 -0.10 -0.05 0 0.05 0.10 0.15 0.20P.C.S. (mmlsecfFIG.2.-Plots of P.C.S. values against partial A@) values (see table 3).The order of P.C.S. values is made more meaningful by the correlation with thespectrochemical ranking of ligands (fig. 2). Unfortunately, most partial A (6) valuesfor Fe(I1) compounds are inaccurate because the d-d bands are often masked bG . M. BANCROFT, M. J . MAYS AND B . E . PRATER 141charge transfer bands. The values in this figure have been calculated using a methodsuggested by McClure and results mainly given by Stephens.1° The values haveoften been estimated from the spectra of Co(1II) complexes containing the appropriateligand. However, despite the uncertainties in the partial A values, a good generalcorrelation is evident.This correlation is not surprising because just as the C.S.should decrease withan increase in cr and 7c bonding, so partial A values should increase. Thus,Jorgenson l1 has expressed A asA = A(Voct)+a(L+M)+n(LcM) (2)neglecting 7c bonding from the ligand to the metal. A(Voct), the crystal field term,is usually considered to be small, and A should then increase with increasing Qdonation and 7c back-donation. Thus NO+, the strongest field ligand gives riseto the lowest (most negative) P.C.S., while bromide and iodide, the weakest fieldligands, give rise to the most positive P.C.S.Since there is an inverse relationship between P.C.S. and partial A, we wouldexpect to find a limiting value of the C.S., beyond which there will be a transitionfrom a low spin to a high spin or intermediate spin complex.From our data, andfrom a study of Fe(I1) hydrotris (1 pyrazolyl) borate compounds,12 we suggest thatthis upper limit should be about 0.5 mmlsec. This relationship should be useful inpreparing compounds which exhibit intermediate spin or high spin-low spin equilibria.QUADRUPOLE SPLITTINGSThe Q.S. values are valuable in elucidating the geometry of these complexes,and are often useful in examining the Q donor and 7c acceptor properties of the ligands-especially when used in conjunction with P.C.S. values.This ratio can be predictedusing a simple point charge or the McClureappr~ach.~ The latter two approaches suggest that, although the 2 : 1 ratio willhold for a given combination of ligands, if one of the neutral ligands is changed, asignificant change in Q .S . will result. However, the 2 : 1 ratio is still extremelyuseful in assigning the structures of these compounds. Thus, the trans and cisstructures can be assigned to all the isocyanide complexes. The depe dihalides giveQ.S. values closer to those of trans-FeCl,(ArNC), than cis-FeCl,(ArNC), suggestingthat they are trans. FeCl,(depb),, however, gives a Q.S. intermediate between cis-FeCl,(ArNC), and trans-FeCl,(ArNC),, and on the Mossbauer evidence alone, itwould be difficult to assign this structure. N.m.r. and dipole moment data suggestthat these phosphine complexes are trans.I4 In order to assign the geometry ofany Fe(I1) octahedral structure containing the ligands in our larger s t ~ d y , ~ partialquadrupole splittings (P.Q.S.) values are calculated.The approach used for cal-culating these values, and for evaluating bonding properties of the neutral ligandswill now be outlined.The quadrupole splitting AEQ for a nucleus of quadrupole moment Q in anelectric field gradient q is generally expresssed asThe 2 : 1 trans : cis ratio is widely appli~able.~.a general ligand-field model,In both cis and trans isomers, q = 0 an142where q = field gradient due to the valence electrons (qvalence) or due to externalcharges (q,attice) ; Q = N 0-3 barns.' R and yco = Sternheimer antishielding factors,(1 - y m ) = 10.1;BONDING I N Fe(I1) LOW-SPIN COMPOUNDSCiZi(3 C O S ~ Oi - 1)qlattice = Y r:where Bi = angle from the 2 axis, ri = Fe-ligand bond distance, Zi = charge onligand ;qvalence = - C (3 COS' O i - 1 ) <r-3)i,d electronswhere (r3), is the mean value of r3 averaged over a particular d orbital, and(3 cos2 O1- 1) for the various d orbitals are +4/7 for d,~, -4/7 for dxz-y2 and dxyand +2/7 for dxz and dyz.The total qvalence is considered to be the sum of the contri-butions made by individual electrons. Thus, for the Fe(I1) free-ion ( f 2 g ) , 6 qvalence =O.Depending on the electron donating or withdrawing power of the ligands, the valueof (r3> will vary for different orbitals, thus changing the " effective " populationof these orbitals, and give rise to a qvalence.In the following treatment, we assume that (i) the Fe-L bond distances (for agiven ligand L) are constant from compound to compound, and all complexes havethe highest possible symmetry.(ii) The contribution to the Q.S. from parts of thelattice other than the nearest neighbours is neglected. Generally this will not bea good assumption, but there are two pieces of evidence which strongly support itfor the compounds in this study: (a) the frozen solution spectrum of trans-FeC1,(depe), is identical to that of the solid; (b) the trans : cis 2 : 1 ratio is based on theassumption that the contribution from other than nearest neighbours is neglected,*and the 2 : 1 ratio holds well.To obtain a " reference value " on which to base our P.Q.S. vaiues, we calculatea hypothetical Q.S. for the trans C1-Fe-C1 group assuming that YFe-Cl = 2.3&4and that the bonds are ionic.AEQlattice from this calculation equals - 1.5 mmlsec.The negative value of the Q.S. is consistent with the area ratio data.5 Taking intoaccount covalency and polarization which will tend to decrease the value of AEQlattice,we assume the Q.S. contribution from the FeCl, linkage to be about - 1.2 mmlsec.We then take the total quadrupole splitting in the chloride compounds aswhere the valence contribution arises from the bonding properties of the neutralligands and can be expressed for the trans compounds FeCl,L, as( 5 )where the n are the " effective " populations of the various d orbitals and K = (4/7)0 donation by the neutral ligand in the trans compounds will be effectively intoc / , ~ ~ ~ J r which will create a positive q in opposition to the " chloride only " value.However, 7c back-donation in the trans chloride compounds can have a dual effect.If L withdraws electrons mostly of dxy type, then n,, becomes smaller than +(n,,+nyz)and qvalence will be negative, enhancing the field gradient from the chlorides. If, onthe other hand, the dx, orbital is essentially non-bonding, as has been suggested forcyanide complexes,16 then 7c bonding via dxz, dyz will decrease the magnitude of theQ.S.* It seems highly unlikely that contributions from other than nearest neighbours would be 2 : 1.7 We neglect donation to the 4p orbitals which will give a smaller contribution to qvdence.AEQtotal = - 1.2 mm/sec+AEQvalence,qvalence = K [ n x y - H n x y + n y z ) ] + K C n x ~ - y 2 - 3 n z 2 3 ,+--3)(1- R)G.M. BANCROFT, M. J . MAYS AND B. E. PRATER 143For any compound of the type FeA,B,, AEQ can be expressed asAEQtrans = 4(P.Q.S.), -4(P.Q.S.),,andUsing the P.Q.S. value for C1- = -0.30, the P.Q.S. value for some other ligandshave been derived (table 4). The P.Q.S. values are useful in assigning the stereo-chemistry of a compound containing any of the ligands we have studied. Theyappear to be much more sensitive than the P.C.S. values to the other ligands present ;yet, there is reasonable agreement between predicted and observed values, and thegeometry can be usually assigned with confidence.AEQcis = -2(P.Q.S.), + 2(P.Q.S.),.TABLE 4.-PARTIAL QUADRUPOLE SPLITTINGSBr- - 0.32 depb/2 -0-02I- -0.31 depe/2 +0.02c1- - 0.30 RNC +0*09SnCl; -0.17CN- - 0.06Using eqn.(5) and the P.C.S. values for ArNC, depe and depb, we now commenton the trend in Q.S. of the dichlorides trans-FeCl,(ArNC),, trans-FeCl,(depe),and trans-FeCl,(depb),. The P.C.S. values indicate that ArNC is an appreciablybetter CT donor and/or 7c acceptor then depe or depb. The larger Q.S. value fortrans-FeCl,(ArNC), indicates that either the d electron density withdrawn by ArNChas more dxy character than for the phosphorus ligands and/or that ArNC is apoorer CT donor. The former explanation is more consistent with the P.C.S. values.The two phosphine ligands have similar P.C.S. values, yet different P.Q.S. values.This suggests that depb is a poorer acceptor of dxy electrons than depe and/or thatit is a better Q donor, while the total Q donor + 7c accepting capacity is similar for both.(a) N. E. Erickson, Ph.D. Thesis, (Columbia, U.S.A. 1964).(6) N. E. Erickson in The Mossbauer Eflect and its Applications in Chemistry (Amer. Chem.SOC. Publ., 1967).R. R. Berrett and B. W. Fitzsimmons, J . Chem. SOC. A , 1967, 525.B. W. Dale, R. J. P. Williams, P. R. Edwards and C . E. Johnson, Trans. Faraday Soc., 1968,64, 620.G. M. Bancroft, M. J. Mays and B. E. Prater, Chem. Comm., 1968, 1374; 1969, 39.G. M. Bancroft, M. J. Mays and B. E. Prater, J. Chem. Soc., in press.Y. Hazony, J . Chem. Phys., 1966, 45, 2664.D. S. McClure in Advances in the Chemistry of Coordination Compounds, (MacMillan, N.Y.,1961), p. 498.C. K. Jorgenson, Absorption Spectra and Chemical Bonding in Complexes (Pergamon Press, 1962)' G. M. Bancroft, A. G. Maddock and J. Ward, Chem. Ind. 1965,423.13 R. V. Lindsay, Jr., G. W. Parshall and V. G. Stolberg, J. Amer. Chem. SOC., 1965,87, 658.lo P. J. Stephens, D. Phil. diss., (Oxford, 1964).l2 J. P. Jesson, S. Trofimenko and J. F. Wecker, J. Chem. Phys., 1968, 48, 2058.l3 M. G. Clark, private communication, to be published.l4 J. Chatt and R. G . Hayter, J. Chem. Soc., 1961, 5507.l5 J. D. Artman, A. H. Muir and H. Wiedersich, Phys. Rev., 1968. 173, 337.l6 H. B. Gray, I. Bernal and E. Billig, J. Amer. Chem. SOC., 1962,84,3404

ISSN:0366-9033    

DOI:10.1039/DF9694700136

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

GENERAL DISCUSSIONMr. A. F. Orchard (Oxford University) said: I have some additional commentson the interpretation of the Mn(CO)&H3 photoelectron spectrum. It may be thatKoopman's approximation is yielding misleading results here. However, one canidentify a number of factors that may contribute to an ordering b2 > e for the highestoccupied molecular orbitals. One possibility, given a particularly high electrondensity on the manganese atom, is an indirect stabilization of the mainly-3dn e levelas a result of an increased 4pn-n* interaction. Alternatively, there may be effectsarising from a more positive potential of the bonded methyl group. This wouldlead to greater degree of 4pn-3dn mixing, since the off-diagonal Fock integral connect-ing these orbitals becomes more negative.A direct lowering of the e level wouldthen result. In addition, direct interaction with a more positive potential shouldsomewhat stabilize 3dn relative to 3d6. None of these effects alone provides acompelling explanation of the problem : but, taken together, they present a crediblepicture.We have also tentatively suggested that the bound methyl group might, in thisparticular situation, possess significant n-acceptor properties. This notion is, froma general chemical point of view, somewhat heretical. But it does provide the moststraightforward explanation of the anomalous Mn(CO)&H3 photoelectron spectrum.Dr. M. G. Clark (University of Cambridge) said: I would like to say a few wordsabout the correlation of quadrupole splitting with stereochemistry, as exemplified bythe ratio,QS(trans-MA,B,) : QS(cis-MA2B4) = -2 : 1mentioned by Bancroft, Mays, and Prater.l My basic premise is that relations ofthis kind, which are observed in various Sn(1V) and low-spin Fe(I1) systems,2*reflect an underlying symmetry feature.Particular rationalizations, such as point-charge or molecular-orbital models, are then seen as manifestations of this symmetryproperty. Consider, for example, the electric field gradient (EFG) tensor at thenucleus of the central atom M in an octahedral system MA2B4, where A and B aremonatomic ligands. Trans-MA2B4 has point symmetry D4h, hence the asymmetryparameter q of the EFG tensor must vanish ( q r 0 ) . The C,, symmetry of cis-MA2B4 does not demand q r O , but theoretical models of the -2 : 1 quadrupolesplitting ratio give q = O .Thus the relationship is associated with a situation inwhich the system has symmetries higher than those strictly required by its pointgroup. This has been termed " intermediate symmetry 9y.5Suficient conditions for simple relationships between quadrupole splitting andstereochemistry are : (1) additivity : the total EFG tensor at the nucleus of the centrall Incidentally the minus sign, although demanded by theory, does not seem to have been verifiedexperimentally.Sn(1V) : B. W. Fitzsimmons, N. J. Seeley and A. W. Smith, Chem. Comm., 1968, 390 ; B. W.Fitzsimmons, N. J. Seeley and A. W. Smith, J. Chem. SOC. A , 1969, 143 ; B. W. Fitzsimmons,Chem. Comm., 1968,1485.Low-spin Fe(I1) : R.R. Berrett and B. W. Fitzsimmons, J. Chem. SOC. A , 1967, 525 ; G. M.Bancroft, M. J. Mays and B. E. Prater, Chem. Comm., 1968,1374 ; G. M. Bancroft, M. J. Mays,and B. E. Prater, 1969, this discussion.For the theory of EFG tensors see M. H. Cohen and F. Reif, Solid State Phys., 1957,5,321.J. S. GriBCith, Mof. Phys., 1964,8,217.14GENERAL DISCUSSION 145atom M is the sum of individual contributory EFG tensors, one for each ligand. Thisis also a necessary condition in any theory purporting to reasonable generality.The existence of a set of “partial quadrupole splittings ” implies additivity. Ingeneral, additivity will require that contributions to the field gradient from distantparts of the solid be negligible.(2) AxiaZly symmetric bonds: each individualcontributory EFG tensor is axially symmetric about the relevant ligand-central atomaxis, and thus may be characterized by a single parameter 4.’ This condition isnot necessary ; for example, consider MA2P2, where A is a monodentate ligand con-tributing an EFG tensor characterized by qA, qA, and /l is a bidentate chelate with eachpoint of ligation contributing an EFG tensor characterized by qb, qB. Denoting themonodentate ligand by a line to define its relative asymmetry, we consider isomers (I)I nand (11). Summing over individual contributions, the EFG tensors at M are, for (I) : :I [ o 0 -2[%(1 -rts)-q*lq/d - q/3> - qA( - VA) 0-qb> -qA(l +qA)and for (11) :O l 1- !dqb(l - qb) - q A ( l - YA)] 0-*[%(l -qB)-qA(l +?/A)] * 1 0l o 0 qfi(l -qf3) -qAFor a ++ Mossbauer transition, such as 57Fe or ‘19Sn, the ratio QS(1) : QS(I1) =-2 : 1.Note that although the point symmetries of (I) and (11) are so low (C2hand C1, respectively) that the EFG tensors are no longer axially symmetric, theEFG tensor for (11) still has more symmetry than is strictly required by the pointgroup.Dr. M. J. Mays (University of Cambridge) said: Fig. 3 of the paper of Greatrexand Greenwood correlates the change in chemical isomer shift with the total chargeon the species and with coordination number. It is not clear, however, whether thereare two separate trends, since total charge and coordination number are both allowedto vary in each of the figures shown.Are there in fact two separate trends or arethey interconnected ?Dr. M. G. Clark (University of Cambridge) said: One feature of Greatrex andGreenwood’s paper is that it gives yet another example of a positive correlationbetween centre shift and coordination number. Such correlations arise in widelydiffering systems2 The existence of a set of partial centre shift values at leastFor the theory of EFG tensors see M. H. Cohen and F. Reif, SoZid State Phys., 1957,5. 321.e.g., G. M. Bancroft, A. G . Maddock, and R. G. Burns, Geochim. Cosmochim Acta, 1967,31,2219, found such correlations in iron silicates146 GENERAL DISCUSSrONapproximately independent of coordination number would usually be a sufficientcondition for such a correlation, although the converse is not necessarily true.On the question of the range of validity of partial centre shift (PCS) and partialquadrupole splitting (PQS) values, such as those proposed by Bancroft, Mays, andPrater, it seems intuitively likely that, since the centre shift does not involve anyexplicitly angular-dependent factors, PCS values have the greater range of validity.This consideration points out a fundamental difficulty concerning theoretical investi-gation of the isomer shift, which is the main contributor to the centre shift.Becauseof the explicitly angular-dependent factors in the quadrupole coupling, symmetryarguments can play a major role in its study. This is not so with the isomer shift,and since the strongest points of any theory of bonding tend to be those which aresymmetry-based, one is in an unfavourable position for theoretical analysis.Finally,with regard to PQS values, my previous remark suggests that both simple and elaborateparameterizations may sometimes give similar experimental consequences.Prof. N. N. Greenwood (Newcastle upon Tyne) said: In answer to Mays andClark; the right-hand side of fig. 3 includes the effects of both coordination numberand charge, but the correlation of chemical isomer shift with coordination number is,in fact, independent of the effect of charge as can be seen from the vertical columnsin the following data (6 in mm sec-l relative to nitroprusside) :co-ord.no. neutral species mono-anions di-anions4 [Fe(CO)4]2- 0.085 Fe(CO)5 0.17 [Fe(CO)&II- 0.09 [Fe2(C0),l2- 0.18[Fe3(CO), 1]2-0.166 Fe3(CO)l,(b)0.31 [Fe2(CO),H]- 0.33 [Fe4(CO)13]2-0.28[Fe3(CO)l ,H]-(b)0-287 Fe3(CO)I 2(a)0.37 [Fe3(CO)1 1H]-(a)0.30Fe2(CO)9 0.42The fact that an increase in coordination number increases S and therefore impliesa diminution is s electron density at the nucleus is not surprising.The effect is wellestablished in both iron and tin Mossbauer spectroscopy.Dr. G. M. Bancroft (Oxford University) said: Are Greatrex and Greenwood notrelying more on chemical insight than Mossbauer spectra in distinguishing betweenstructures (a) and (b) of [(CO),FePMe,Ph], ? The only Mossbauer parameterwhich changes significantly is the Q.S. of the apical iron atom.Prof. N. N. Greenwood (Newcastle upon Tyne) said: In reply to Bancroft and Mays,more than chemical intuition is involved in deciding in favour of fig.5(b) as thestructure of the compound [(CO),FePMe,Ph],. The paper indicates the Mossbauerevidence in favour of a trimer, rather than a dimer or tetramer, and also suggestsreasons for placing the three phosphines on separate iron atoms rather than on one.There is also mass spectrometric evidence which indicates that each iron atom isassociated tith one phosphine ligand only (J. Moss and B. L. Shaw, unpublishedresults).Dr. A. J. Rest (Cambridge) said: I am particularly interested in (HFe(C0)4)-(structure 8 in fig. 1 of paper by Greatrex and Greenwood) because we have recentlyprepared HMn(CO), in an argon matrix at 15°K by photolyzing HMn(C0)5.1A. J.Rest and J. J. Turner, Chem. Corn., 1969,375GENERAL DISCUSSION 147The intermediate value of the quadrupole splitting for (HFe(CO),)- was taken byGreatrex and Greenwood as evidence for a heavily distorted five co-ordinated specieswith C3" symmetry. We find that HMn(CO), has an infra-red spectrum consistentwith C,, symmetry analogous to HCo(CO), with little distortion of the equatorialcarbonyl groups in the trigonal bipyramid towards the hydride ligand. The photolyticproduction of HMn(CO), is quantitative and can be reversed using a different photo-lytic source.Prof. N. N. Greenwood (Newcastle upon Tyne) said: In connection with Rest'sremarks, we agree that it is difficult to decide whether or not the three non-axialcarbonyl groups in [Fe(CO),H]- are planar or distorted towards the hydrogen atom(as shown in fig.1.8 in our paper). We postulated the distorted structure partlybecause of the low value of the quadrupole split and partly because virtually allknown carbonyl complexes for which accurate structural data are known are distortedin this way. The structure is also consistent with the vibrational spectroscopicevidence quoted in ref. (20) of our paper.Dr. M. G. Clark (University of Cambridge) said: In connection with the paper byGreatrex and Greenwood, I wonder if Greenwood would like to comment furtheron the correlation of isomer shift with total charge on species? Whereas the correla-tion of isomer shift with coordination number has been observed in a variety ofsystems, its correlation with total charge on the species seems to have been less wellexplored.Prof. N.N. Greenwood (Newcastle upon Tyne) said: In reply to Clark, there seemsto be no problem in understanding how an increase in charge on the carbonyl speciescan, in principle, influence the chemical isomer shift. The experimental evidenceon this point is unequivocal; the theoretical interpretation must consider the 5effects listed in our paper, though it is not possible to decide on a detailed allocationof the changes in chemical isomer shift to each individual mechanism separately.The results imply that an increase in anionic charge increases the s electron densityat the iron nucleus and this seems reasonable.On the question of the influence of varying charge whilst keeping the structureconstant, there is some evidence to show that the resultant changes in chemicalisomer shift depend entirely on the nature of the ligands surrounding the centralatom and on the spin-state of this atom.Thus, octahedral complexes of high-spiniron(II1) have chemical isomer shifts in the region of 0.7 mm sec-l with respect tohydrated sodium nitroprusside and addition of an electron to give the correspondingoctahedral complexes of high-spin iron(I1) increases this by nearly 1 mm sec-lto about 1.6 mm see-l because of increased screening of the 3s electrons from thenucleus by the added 3d electron. By contrast, reduction of the low-spin complex[Fe"'(CN),]3- to [Fe"(CN),I4- leaves the chemical isomer shift virtually unaltered(to within -0.02mm sec-l) because the shielding effect of the added 3d electron(which makes the complex diamagnetic) is almost exactly counterbalanced by thedelocalization of d electron density on to the ligands.Intermediate cases are alsoknown between these extremes of complete localization of the added electron andcomplete delocalization of equivalent electron density via d,-p, interactions on toligand-based orbitals. For example,l reduction of the trismaleonitriledithiolatecomplex [FeJV( S 2C 2(CN) 1 J2- to [FeIJ1(S ,C,(CN) ] 3] 3- increases the chemicalisomer shift from 0.50 to 0.65 mm sec-' indicating that the added electron has enteredT. Birchall and N. N. Greenwood, J. Chem. SOC. A , 1969,286148 GENERAL DISCUSSIONan orbital which has both metal and ligand character. Such examples illustratethe care which must be exercised in interpreting formal oxidation states of metal andin classifying ligands as stabilizing either high or low oxidation states-frequently thetotal electron density at the central atom remains unaltered.Dr.H. A. 0. Hill (Oxford University) said: With regard to the paper by Bancroftet aZ., how sensitive to the anion is the quadrupole splitting in those compoundscontaining a complex Fe(I1) cation? A dependence on the anion would obviouslymake difficult an estimation of the qlattice contribution to the electric field gradientand hence the partial QS in these complexes.Dr. G. M. Bancroft (Cambridge University) said : In reply to Hill, we have assumedthat the contribution to the QS from parts of the lattice other than the nearest neigh-bour ligands can be neglected.This assumption is most likely to break down for thecationic and anionic compounds. Indeed, the agreement between the predicted andobserved values is the poorest for [Fe(SnCl,)(ArNC),]ClO,.Prof. N. N. Greenwood (Newcastle upon Tyne) said: The concept of partial centreshifts is undoubtedly valuable in certain restricted ranges of compounds and it isimportant to try to establish the range of validity of the concept. The compoundsdiscussed by Bancroft are all 6-coordinate complexes of low-spin iron(I1) and assuch the range of centre shifts is small. In such cases the temperature of comparisonmay be important if the compounds have significantly different values for the tempera-ture coefficient of 6.The similarity in centre shifts also requires a restricted rangeof partial centre shifts and most of the ligands listed in table 3 fall in the range 0-0.1mm sec-1 ; agreement between observed and calculated values is therefore not toosurprising.The concept certainly does not work for complexes in which the ligand is nC5H5,nor does it work when the spin-state of Fe(I1) changes, or when the coordinationnumber of the iron atom alters. For example, it is not applicable to the carbonylcomplexes discussed in the preceding paper. Is it possible, therefore, to defineconditions when the concept is applicable and when it is not ?Dr. G. M. Bancroft (Cambridge University) said : With regard to Greenwood’squestion on the sensitivity of the C.S., the C.S., as seen in table 1, is surprisinglysensitive to the nature of the ligands. The range of C.S. values for these compoundsis almost 0-7 mm/sec-about 40 % of the total range for all iron compounds. Thetemperature coefficient of the C.S. for all these compounds is similar from 80 to295°K; perhaps more important are the possible variations in the Z.P.M. shift.However, these variations in the S.O.D. shift should lead to variations in P.C.S.values of one ligand from compound to compound of only about 0.01 mm/sec-and the total range of P.C.S. values is 0.29 mm/sec. The good agreement betweenpredicted and observed values for such a large number of compounds containingligands having such widely varying P.C.S. values, strongly supports the concept forsix-coordinate Fe(I1) low-spin compounds.The concept also applies for Fe(1l) six coordinate high spin compounds studiedby Hazony et aZ.l Although it is difficult to generalize at this stage, it seems likelythat the concept of P.C.S. values will usually work for a given coordination number,valency and spin state of iron or any,other Mossbauer isotope.Y. Hazony, R. C. Axtmann and J. W. Hurley, Jr. Chem. Phys. Lett., 1968,2,440

ISSN:0366-9033    

DOI:10.1039/DF9694700144

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Structure of Molecules of Type (CH,),X-C = C-Y (X= Si,Ge, Y=H, Cl) determined by Electron DiffractionBY WERNER ZEIL, JOACHIM HAASE AND MARWAN DAKKOURILehrstuhl fur Physikalische Chemie, Universitat UlmReceived 30th December, 1968The structure of the molecules (CH3)3Si--C= C-H, (CH3)3Si-C= C-Cl and (CH3)3Ge-C=C-CI as gases have been determined using electron diffraction. The lengths of the C=Cbond show no significant differences from the corresponding bond length in carbon compounds.We conclude therefore that the length of the C r C bond is insensitive to x-electron delocalization.We have reported the structure of the molecule (CH3)3C-C=C-Cl as deter-mined by the gas electron-diffraction method and by microwave spectroscopy,land the values of nuclear quadrupole coupling constants of the following molecules :(CH3)3C-C = C-Cl, (CH3)3Si-C = C-C1, (CH3)3Ge-C- C-C1, (CH3)3Sn-C = C-Cl, measured by nuclear quadrupole resonance spectroscopy have beenpublished.2 From these results there is strong evidence of a bond type in the halo-genated acetylenes which will be described(CH3)3C-C C-Cl(CH3)3X-C= C-ClWith silicon and germanium compoundsby the following mesomeric formulas :- +(CH3) ,C-C==C=Cl- +(CH3)3X-C=C=Cl (X = Si, Ge, Sn) - +(CH~)3X=C=C=C1+ -(CH3) 3X=C=C-CIof the type described above we wouldtherefore expect an increase of the length of the C=C bond and a decrease of theforce constant of the C 3 C bond with increasing number of the mesomeric structures.The prediction concerning the dependence of the force constant have been f~lfilled.~TABLE 1(mm)(CH,),SI-C = C-H 250500(CH,),Si-C 3 G-Cl 250500lo00(CH,),Ge-C = C-C1 250lo00distance range of s-values(A-1)4-0-26.02.0- 14.34.0-26.01 -8- 12.00.75-7.54.0-25.01.0-7-0number ofplztes8646686In this work we report the determination of the structures of the following mole-cules : (CH3)3Si-C=C-H, (CH3)3Si-C=C-Cl, (CH&Ge-C=C-Cl.Themethod used was the electron diffraction by the gases. The compounds were pre-pared by methods described el~ewhere.~ The electron diffraction patterns of the14150 MOLECULES DETERMINED BY ELECTRON DIFFRACTIONmoIecules under investigation have been recorded on our KD-G2 apparatus,developed by us in collaboration with Balzers AG (former Trub-Tauber & Co.)under the sponsorship of the Deutsche Forschungsgemeinschaft.Exposures weremade with 2-3 nozzle-to-photographic-plate distances in each case. Table 1 givethe distances and the range of s-values. The s-value is given bys = (47r/A.) sin (912)16 I2 18 EL 301Si-CsC-H /FIG. 1.-Intensity curve of (CH3)3Si-C=C-H; expt., (a); theor., (b); (c).I I- h ~ n - h n- --Ad---. * . . I . L.-o * )I ' * DI 15b i - c E w i /FIG. 2.-Intensity curve of (CH3)3Si-C=G-Cl; expt., (a); theor., (b); (c)WERNER ZEIL, JOACHIM HAASE AND MARWAN DAKKOURI 15100n-- 4 - -!I, ~ 1 ~ ~ ~ ~ * . . . 1 1 6 12 I8 2.4LGe-C=GCI/Fro. 3.-Intensity curve of (CH3)3Ge-C=Ml; expt., (a); theor., (b); (a)-@), (c).0I 2 3 4 5 6bi-C=C!-H /FIG.4.--Radical distribution function of (CH3)3Si--C= C-H ; expt., (a) ; theor., (b) ; (a)-(&), (c)152- x1ICMOLECULES DETERMINED BY ELECTRON DIFFRACTIONI 2 I 5 6 J\ -Si--C=C-Cl/FIG. 5.-Radial distribution function of (CH&Si-CrC-CI ; expt., (a) ; theor., (b) ; (a)-@), (c).\-Ge-C=MlFIQ. 6.--Radial distribution function of (CH3)3Ge-C~C-Cl ; expt., (a) ; theor., (b) ; (a)-@), (c).WERNER ZEIL, JOACHIM HAASE AND MARWAN DAKKOURI 153The accelerating voltage was 45 to 46 kV. This voltage corresponds to a wavelength of 0.0615 to 0.0604A. The wave length was calculated from the diffractionpattern of ZnO and the distances between the nozzle and the plate in the differentpositions. The experimental data were corrected and processed in the usual way?The resulting molecular intensity curves are shown in fig. la, 2a and 3a.The radialdistribution curves obtained by Fourier inversion from the observed intensity curvesare shown in fig. 4a, 5a, 6a.Fig. 1-6 give also the theoretical curves (6) of the intensity functions and of theradial distribution functions together with the curves of the difference ( c ) betweenobserved and theoretical curves. The theoretical curves have been calculated fromthe molecular data given in tables 2, 3 and 4. In fig. 7,8 and 9 the results concerningTABLE 2.-INTERATOMIC DISTANCES ?'@, BOND ANGLES AND MEAN AMPLITUDES /" OF(CH3)3Si-C = C-HH-Cr-C-G-C-SiSi-C,,,Si ... HSi ... =CH . . . C =Si . . . HXetC,-Hc, .. - c,c, ... c-c, ... =cCm * * H,,t ----------4 C,SiC,4: HCmHr i j (A)1-05 f0.0081.20 f0.0081.825 jC0-0061-865 f0.0041.100 &0*0082-475 f0.0023.018 lf0.0083.02 50.013-04 fO.012.27 f0.014.075 &0*0064.08 f0.015-06 -fO.Ol107'59'108'25'I i j (A)0.055 f0-0080.055 f0.0080.065 f0.010.073 f0.010.063 f0.010-077 f0-0080-080 f0.020.079 f0.020.085 f0-020.100 f0-020.100 i-0-020.073 f0.020.183 f0.02TABLE 3.-hl"TRATOMIC DISTANCES Yg, BOND ANGLES AND MEAN AMPLITUDES !i/ OF(CH3)3SiC = C-CICI-c -c=c-= C-SiSi-CmSi ... Hc, . . . c,,Si ... =C c, . . . c=c1 ... ECc, ... =cSi . . . C1 c, . . . c1C,-Hr i j (A)1-630 f0.0051.210 k0-0081.825 &0*0081.855 &0-0061 *095 f0.0082-488 f0.0053-040 f0.0083.035 f0.012.993 f0-012.850 f0.014.040 f0.014-672 f0.0065.549 k0-006I i j (A)0.060 f0.010.055 f 0.0080.065 f0-0080.070 f0-0080.060 kO.010.077 f0-0080.080 f0.020.077 f00-20.080 f0.020.070 f0.0150.100 f0.020.078 f0.010-190f0~0154 MOLECULES DETERMINED BY ELECTRON DIFFRACTIONTABLE 4.-INTERATOMIC DISTANCES rq, BOND ANGLES AND MEAN AMPLITUDES li/ OF(CH,),GeC = C-CIa - C r-CEC-=C-GeGe-C,Ge ...HGe ... =Cc1 ... =cGe . . . C1C,-Hc, . . . c,c, ... CGc, ... r Cc, . . . CIr i j (A)1630 f0-0051.215 f0.0051.930 f0.0071.960 f0-0051 -095 f 0.0052-520 f 04053-180f0.013.150 f0.013.200 f0.012.840 f0.014.250 f0.0074.775 f0.0055780 f0.02It j (A)0.060 f O-010.055 40.010.065 &O-010.070 50.010.060 f0.0080-077 f0.0080.080 f0-020.077 f0.020.080 f0-020.070 f0.010.100 f0-010.078 f0*010.190 f0-02the geometry of the molecules are shown.The diagrams show all the measuredvalues. In table 2, 3, 4 the interatomic distances, the calculated bond angles and themean amplitudes are given. The most interesting distances concern the questionk - C = C-H /FIG. 7.71.- interatomic distances of (CH&3i--Cr G H .h-c= C-CIFIG. 8.-The interatomic distances of (CH&Si-C= C-Cl.TABLE 5.-BOND DISTANCES, FORCE CONSTANTS AND NUCLEAR QUADRUPOLE COUPLING CONSTANTSHjC-C f C-HHjC-CrC-Cl(CH3)jC-C C-H(CHj)jC-C z C-ClH3Si-C = C-H(CH3)&-C zi C-H(CH3)3Si-C = U-CIH3Ge-C=C-H lo(CHj),G+C C-CId(A)1.105 1 -459 1.206 1.0561.117 1.458 1 *207 1 *6371.529 1 0498 1.210 1.0561.525 1.468 1.210 1.6371.488 1 -826 1 0208 1.0581.865 1.825 1 *200 1.0501.855 1-825 1.210 1,6301.521 1.896 1.208 1 a0561.960 1.930 1.215 1-630z-x X-Ca C I C c-Yangle zxz z-x108'39' 5.06108'8' 4.84110'40' 4.501 lO"9' 4.231 10'12'107'59' 2.80110'05' 2-75109'54'108'18156 MOLECULES DETERMINED BY ELECTRON DIFFRACTIONwhether there are any correspondence between the physical behaviour and the meso-rneric structures proposed by the chemists ; these are the C = C bonds and the Si-C = ,Ge-C= and the =C-Cl bond distances.Table 5 gives our results relating tothe force constants and nuclear quadrupole coupling constants. The table also‘Ge-C=C-ClFIG.9.-The interatomic distances of (CH3),Ge-C = C-Cl.7contains values, reported by other authors for molecules of a similar t~pe.~-lO Thereare no significant differences, within the limit of error in the length of the C r C bond,between the molecules with two and four mesomeric structures. This is contraryto the behaviour of the force constants and the nuclear quadrupole coupling constantswhich show considerable differences. The Si-Cr and -C-C1 bond lengthsare in good agreement with the earlier measurements on similar molecules. Thelong distance of the Ge-Cr bond is remarkable.If the force constants and nuclear quadrupole coupling constants indicate theexistence of the pd-n-bond, then we must conclude that the geometry of moleculesof the type (CH&X-C= C-Y will not be affected within the error limits of electrondiffraction data, by this bond type.J. Haase, W. SteingroB and W. Zeil, 2. Naturforsch., 1967, 22a, 195; H.-K. Bodenseh,R. Gegenheimer, J. Mennicke and W. a i l , 2. Naturforsch., 1967, 22a, 523.W. Zeil and B. Haas, 2. Naturforsch., 1967, 22a, 2011.W. Huttner and W. Zeil, unpublished.W. Steingrof3 and W. a i l , J. Organometal. Chem., 1966, 6,109.W. Zeil, J. Haase and L. Wegmann, 2. Intrumentenkunde, 1966,74,84 ; L. Wegmann, J. Haaseand W. Zeil, 6th Int. Congr. Electron Microscopy, (Kyoto, 1966).J . Haase, H.-D. Kamphusmann and W. Zeil, 2. phys. Chem., N. F., 1967,55,865. ’ L. F. Thomas, E. T. Sherrard and J. Sheridan, Trans. Faraday SOC., 1955,51,619. * C. C. Costain, J. Chem. Phys., 1955, 23,2037.C. L. Gerry and T. M. Sugden, Trans. Faraday Soc., 1965,61,2091.lo E. C. Thomas and V. W. Laurie, J. Chem. Phys., 1966,44,2602

ISSN:0366-9033    

DOI:10.1039/DF9694700149

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Electron Impact Studies on Organo-Beryllium and -AluminiumCompoundsBY D. B. CHAMBERS, G. E. COATES AND F. GLOCKLINGDept. of Inorganic Chemistry, The University, DurhamReceived 14th January, 1969Electron impact studies on R2Be, RzAIH and R3Al compounds at low source temperaturesrevealed the presence of associated electron deficient ions. General decomposition modes arediscussed together with bond energy data on beryllium dialkyls.Most alkyl derivatives of beryllium and aluminium are associated through bridgingcarbon atoms to yield dimeric, or for dimethylberyllium polymeric electron deficientstructures.1 Although the energy of bridging metal-carbon bonds is low (-105ourca Temperature 45-W°C .c -_- l5I34I i"9 I0Source TernDerature 1 % " ~mleFro. 1.-Mass spectra of diethylberyllium (70 eV).15158 ELECTRON IMPACT STUDIESkcd per Be-C or A 1 2 bond) it has proved possible to detect associated metalions formed by electron impact, provided close attention is paid to the experi-mental conditions, especially source temperature.For example, the mass spectrumdiethylberyllium (fig. 1 and 2) illustrates how the abundance of monomeric ionsC,HbBe+ increase with source temperature, whereas trimeric ions C,HbBe$ and themuch more abundant dimeric ions C,H,Be; decrease with temperature. Thesechanges are ascribable to increased dissociation of associated molecules at the highersource temperatures prior to ionization. A similar effect is observed with di-n-propyl-'0 H; $ 2 0lo!40 60 80 100source temp "c.1 -----L" - ~ - ~ -40 6'0 8'0 160 120 140 160 7th 2bOsource temp., "CFIO.2.Variation of ion abundances with source temperature for diethylberyllium.and di-iso-propylberyllium although dimer ions are of lower abundance, and above140" the proportion of hydrocarbon ions with m/e<43 rises steeply due to thermaldecomposition of the parent monomer molecules before ionization. At low sourcetemperatures, dimeric ions C,H,Be; are more abundant for PriBe than for PriBe;this may be due to the higher inductive effect of isopropyl groups increasing theelectron density in the bridging bonds. Although di-isobutylberyllium is dimericin benzene only monomer ions have been detected in its mass spectrum, and at highsource temperatures the abundance of hydrocarbon ions with nz/e I 5 6 increasesrapidly.For di-tert-butylberyllium, which is always monomeric, hydrocarbon ionsconstitute 35 % of the total ion current at 60" and 65 % at 240". Solid dimethyl-be1 yllium contains infinite chains of beryllium atoms with bridging methyl groupswhereas its unsaturated vapour consists mainly of monomer units. When thevapour is in equilibrium with solid, dimer and trimer molecules are also present inthe gas phase.l Conditions in a mass spectrometer are similar to that of the unD . B . CHAMBERS, G . E . COATES AND F . GLOCKLING 159saturated vapour, and only very low abundances of dimer and more highly associatedions (with up to 8 Be atoms) are detectable.and Me,Alz) are of appreciable abundance only at low source temperatures. Bycontrast, dimethylaluminium hydride, which exists as a cyclic hydrogen- bridgedtrimer in inert solvents, produces the trimeric ion Me,Al,H; and dimeric ionsC,H,Al; are of high abundance.This is consistent with the higher heat of associa-tion (- 15 to -20 kcal mole-l) of bridging Al---H---A1 bonds. The variation of ionabundances with source temperature (fig. 3) can only be interpreted in terms of thermalrearrangement (disproporti onation) of (Me,AlH), species producing trimethylalu-minium. This is an example where mass spectrometry provides a more sensitivemeans of detecting a reaction than vapour density studies., DiethylaluminiumTrimethylaluminium resembles beryllium dialkyls in that dimer ions (Me,AlCaHbAl+ ions formed bydecomposition of Me3Al+IT .O{CaHbAl+ ion5 formed bydecomposition of methyl-aluminium hydride ions.I* Me3A12H:+Me4A12H+50 60 80 lb lio l;rO 160 lFj3 200source temp., "CFIG.3.-Rearrangement of dimethylaluminium hydride with source temperature.ethoxide is dimeric in solution and in the vapour phase. The greater strength of thebridging Al---0---A1 bonds is reflected in the mass spectrum where, even at asource temperature of 195" most of the ion current is carried by organo-Al,Oz andorgano-Al,O+ species.DECOMPOSITION MODESFor beryllium alkyls even-electron metal-containing ions are not strikingly moreabuiidant than odd-electron ions, whereas for the limited number cf organoaluminiumcompounds examined even-electron ions predominate.Illustrative examples ofmetastable-confirmed decompositions are shown in fig. 4 and 5. Generally theparent ion formed from associated molecules is of low abundance and first loses analkyl radical, producing an even-electron ion which maintains its even-electroncharacter by successive elimination of neutral molecules. With aluminium alkylsthe neutral molecule eliminated can be metal-containing :Et2A12Hi -+Et,Al+ + AIH160 ELECTRON IMPACT STUDIESDi-krt-butylberylliurn produces the following decompositions, although theyBu\Be+'+Me,C+ + Bu'Be'are not metastable-supported :C,H,Be+'-+C,H: +Be'T r i m e r i o n s-C H Et2Be3H3+ ----+ 2 4 EtBe3HqfD i m e r ionsE t B e +' 4 2EL: B e 3- 3 2c €1 Be+' c,H,B~+* __t -HZ 4 10 C 11 Be+ r , 9-C2:i4C 2 5 H I + B e I\, C ,H4Bet 2 -H-----c,H,B~+ 'FIG.4. -Metastable confirmed fragmentation of diethylberyllium at 70 eVD. 3. CHAMBERS, G . E. COATES AND F . GLOCKLING 161If the C3Hz ion has the structure [CH2 : C : CHI+ then the observed process iscompatible with the ionization potentials [I(CH2 : C : CH)' = 8.25 eV ; I(Be) =9-32 eVJ. In the mass spectrum of dimethylberyllium low-abundance polymericions are related by successive loss of C,H,Be (probably Me,Be)- CzH6Be - CzH6BeC12H,3Bei + CloH2,Be~ 4 C,H,,Bel.Since these are all even-electron ions they must be fragments from an unobservedparent ion. It seems unlikely that (Me2Be)i' is the parent ion since four carbonI 'I \FIG. fi.-Fragmentation of aluminium alkyls.*atoms and fifteen hydrogen atoms must then be lost, either as radicals or moleculesto give C12H33Bei.For monomeric dialkylberyllium ions a favourable decompo-sition mode is elimination of alkane producing an odd-electron ion of high abundance :(CnH2n+ 1)2Be+' +CnH2nBe+' + CnHZn+ 2The extent to which structures may be written for fragment ions is extremelylimited, and ions such as Et2Be2H+ derived from diethylberyllium may be formu-lated in a variety of ways. On the basis of what is known of alkylberyllium hydridemolecules the structure [EtBe-- - H--- BeEt]+ might be favoured. Similarly, thetrimeric ion Et,Bef could be formulated as [I] or [II]* dotted arrows indicate processes not supported by metastable peaks.1 62 ELECTRON IMPACT STUDIESEt---Be 1+\ /' z,'*' \EtBe E tEt----BiEt JDlr Et Et l + ,' ',,,' '*\ /' '\ %'\,, ,,." '\\, .,,"Be Be BeEtEt EtAPPEARANCE POTENTIALS OF BERYLLIUM ALKYLSAppearance potentials of R2Be+' ions, at high source temperature (200") aredue almost entirely to the ionization of monomeric R2Be molecules, and are probablya measure of the vertical ionization potential. The ionization efficiency curves aresimilar to those produced by the inert gases (fig. 6) and, using AV, the ionizationpotentials of beryllium dialkyls have been evaluated by the Warren method (table 1).The ionization potential of dimethylberyllium is higher than that of any other fullyalkylated organometallic compound so far examined and the decrease in going todiethylberyllium is similar to that observed in other groups of the periodic table.i'I Im 10electron energy, eV (uncorrected)Fm.6.Tonization efficiency curves for Me2BeD. B . CHAMBERS, G . E . COATES AND F . GLOCKLING 163TABLE 1 .-IONIZATION POTENTIALS OF BERYLLIUM DIALKYLSI.P. (ev)Mez& 10.67 f0-079.46 f0.05 Et,BePrzBe 8-71 f0.06PriBe 8.80 f0.02BuiBe 8-74 f0.05Table 2 lists the appearance potentials of some fragment ions but in most cases" tailing " of the ionization efficiency curves was found as in fig. 7a-d. If an ionRBe+ is formed by the process :R2Befe+RBe++R'+2e20001000500'0050f 103 uc = 0 - -1Ion current 6 8 10 12 14 l 8 - -electron energy, eV (uncorrected)FIG. 7.4onization e9i:ien:y curves for C4H9Be+ and C4HsBe+* derived from BuiBe.then D(RBe-R) = A(RBe)Lze -I(R,Be). The observed " tailing " suggests thatthe dissociative process occurs with excess kinetic energy; hence the derived bonddissociation energies (table 3) must be regarded as upper limits.The low value ofD(Bu'Be-Bu')+ is probably due to inaccuracy in measuring I(BuiBe) because ofthe low abundance of the molecular ion.In beryllium dialkyls it is highly probable that ionization removes a bondingelectron so that the beryllium-carbon bond energy in the molecule, D(RBe-R164 ELECTRON IMPACT STUDIESTABLE 2.-APPEARANCE POTENTIALS OF FRAGMENT IONS FROM BERYLLIUM DIALKYLScompound ion appearance potential (eV)Me2Be CHZBe+' 11.92 f0.052-67 f0-020.35 f0.031.51 fO.059-86 f0.050.81 fO.059 . 6 0 fO.010.65 40.019-14 f0.03C,H,Be+ 1040 A-0.05TABLE 3 .-BOND DISSOCIATION ENERGIES IN BERYLLIUM DIALKYLSdissociation energy(k3.2 kcal mole-')MeBe-Me)+3tBe-Et)+?rr"Be-Prn)+VBe-Pri)+3u'Be-Bu')f3eCHz-H)fkCZHa-H)+3er"C3H6--H)+k'C3H 6-H)'3eiC4H8-H)+46.147.342-748.729.186.771.570.776.077.7will be in excess of -45 kcal mole-l. The bond dissociation energies:D(BeC,H,,-H)+ + (BeC,H2,)+ + Heaxe all about 74 kcal mole-1 with the exception of D(BeCH2-H)+ for which thevalue is 86.7 3-4 kcal mole-1. This difference may be explained if some stabilizingrearrangement occurs when the product ions (BeC,H,,)+' contain more than onecarbon atom. For example, the structure of the BeC,HZo ion m y be representedasl G. E. Coates and K. Wade, Orgmometallic Compounds, vol. 1, (Methuen, 1968).T. H. Wartik and H. I. Schlesinger, J. Amer. Chem. SOC., 1953,75,835.J. W. Warren. Nature, 1950, 165,810

ISSN:0366-9033    

DOI:10.1039/DF9694700157

出版商:RSC    

年代:1969

数据来源: RSC