28 J.C.S. DaltonCrystal and Molecular Structure of Deca-p-acetato-d ioxo bis( pyrid ine) -heptazinc(i1) and the Electron Paramagnetic Resonance Spectrum of itsCopper -doped CrystalsBy Donato Attanasio,' Giulia Dessy, and Vincenzo Fares, Laboratorio di Teoria e Struttura Elettronica eComportamento Spettrochimico dei Composti di Coordinazione del C.N.R., Via Montorio Romano 36,001 31Roma, ItalyCrystals of the title complex [{Zn3.50(0,CMe)5(py)},] (py = pyridine) are orthorhombic, space group Pbca, withunit-cell dimensions a = 16.598(3), b = 14.682(4), c = 19.1 11 (9) A, and Z = 4. The crystal and molecularstructure has been determined from diffractometric data by the heavy-atom method and refined by least squares toR 0.076. Seven zinc atoms, bridged by acetate groups, form a heptameric centrosymmetrical unit.In each of thetwo semi-units a central oxygen is surrounded tetrahedrally by three zinc atoms which are in a tetrahedral arrange-ment, and by one zinc atom in a tetragonally compressed octahedral environment. The low-temperature e.s.r.spectrum of 83Cu-doped crystals of the complex is reported. The paramagnetic ion is substituted a t the com-pressed tetragonal zinc site. The g, metal hyperfine-coupling, and quadrupole-coupling tensors have beenobtained from an analysis of the angular variation of the spectra. In spite of the low site symmetry indicated by theX-ray results, an almost pure d,* ground state is found to be present. A small 4s admixture is assumed to beresponsible for the reduced value of the isotropic part of the copper hyperfine tensor.As a continuation of our studies of copper(I1) complexesof a-nitroketones we have recently investigated the e.s.r.spectra of some of these complexes diluted in the dia-magnetic, isomorphous, zinc( 11) analogues.'.During the preparation of single crystals of copper(I1)-doped bis(nitroacetonat0) bis( pyridine)zinc (11) some solu-tions yielded, on standing, crystals of a different com-plex, which was later found to be an oligomeric basiczinc acetate derivative [hereafter referred to as (l)].I thas long been known that, upon alkaline hydrolysis,a-nitroketones readily undergo cleavage of the carbon-carbon bond in the a position with respect to the nitro-group. Thus the formation of the acetate anion startingfrom a solution containing nitroacetone in the presenceof water and pyridine is not surprising.Later, complex(1) was also obtained by appropriate crystallization ofzinc acetate dihydrate.Since the system (1) showed some interesting struc-tural and spectroscopic features, further work has beencarried out and we now report the determination of itscrystal and molecular structure, together with an e.s.r.study of its copper-doped crystals.EXPERIMENTALPreparation of complex (1).-The complex [Zn(na),-(OH,),] * (na = nitroacetonate) or [Zn(O,CMe),(OH,),] (150nig) was dissolved in a mixture of acetone (10 cm3) andpyridine (0.3 cm3), in the presence of ca. 1% of 63CuC1,*2H,O (purchased from Oak Ridge Kational Laboratories inthe form of copper oxide).n-Heptane (10 (3111~) was addedand the container stoppered and set aside a t room temper-ature. Large pale yellow crystals were obtained after ca. 1week.Small pale green crystals of [(Cu,Zn) (na),(py),] (py =pyridine) could be obtained after 1 or 2 d when the nitro-acetonato-complex was used as starting material. Usuallythese dissolve again to give crystals of copper-doped ( 1 )after some days.Starting with [Zn(O,CMe),(OH,),], light blue crystals of[(Cu,Zn) (O,CMe),(py)] could be obtained, together withcrystals of copper-doped ( 1 ) . The yield of the niono-pyridine derivative could be decreased by using smalleramounts of copper ion (ca. 0.1-0.3y0).Crystal Data.-C,,H,,N,O,,Zn,, M = 1 238.2, colourlessorthorhombic prisms, a = 16.598(3), b = 14.682(4), G =19.111(9) A, U = 4 657A3, D, = 1.77 & 0.01 (by flotation),2 = 4, D, = 1.77 g cm3, F(000) = 2 470.76, p(Mo-K,) =37.31 cm-', space group Pbca (Oil, no.61) from systematicabsences and structure determination.All the data were collected on a Syntex P2, automaticfour-circle diffractometer by the 8-28 scan technique.Only the intensities of the 1 133 independent reflectionshaving I > 3 4 1 ) were used throughout therefinement. Thedata were corrected for background and for Lorentz andpolarization effects. Given the low value of p and thedimensions of the crystal (ca. 0.1 x 0.2 x 0.2 mm), anabsorption correction was not applied. Unit-cell para-meters were obtained from a least-squares fit to the angularsetting of 15 reflections having 28 2 2 5 " .Calculations.-Calculations were made on the Univac1108 computer a t Rome University using the system ofprograms of the Laboratorio di Strutturistica Chimica delC.N.R.Neutral-atom scattering factors, and correctionsfor the anomalous dispersion of zinc, were taken from ref. 5.Determination and Refinement of the Structure .-Thesolution and refinement of the structure proceeded bystandard methods. The three zinc atoms in generalpositions were located from a three-dimensional Pattersonsynthesis coniputed using the full set of observed inde-pendent terms. A Fourier map phased on the positions offour zinc atoms, the fourth required to lie on from sym-metry considerations, was sufficient to locate all the remain-ing non-hydrogen atoms.Block-diagonal least-squaresrefinement was carried out on I;. The function minimizedwas Zw(lFol - klFc[)2. The weighting scheme w = (a +bF, + cF,2)-1, where a = 20.0, b = 1.0, and c = 0.000 01,was used. With individual atoms given isotropic thermalparameters, refinement converged a t R 0.103. Refinementwas continued using anisotropic thermal parameters untilfinal shifts in the atomic parameters were <0.2 CT a t whichstage the final R was 0.076. Final least-squares atomicparameters with standard deviations are in Table 1 .Calculated and observed structure factors, and therma1979parameters, are listed in Supplementary Publication No.SUP 22342 (9 pp.).* Interatomic distances and angleswithin the chemical unit are shown in Table 2.Spectra.-Single-crystal e.s.r.spectra were recorded a t theX-band frequency using a Varian E-9 spectrometer equippedwith a variable-temperature accessory. Magnetic-fieldC(1intensities werelinearity of theFIGURE 1used without corrections, after checking thefield with a manganese(I1)-doped sample ofTABLE 1Atom co-ordinates ( x lo4), with standard deviations inparenthesesxla0456(3)1353(3)1668(2)- 868( 12)- 707( 14)302( 14)1576(13)897 ( 1 5)1 565(16)786( 19)1 680(16)2 527(17)2 777(15)882(12)1207(15)-1 148(21)- 2 OlO(26)891 (19)79 1 (24)1391(22)1883(26)1290(23)1458(34)3 038(25)3 945(18)1739(20)1661(23)1088(23)552(23)645 (20)Y lb0-2 036(2)- 326(2)- 829(2) - 1 014(12) - 2 224( 12)244( 14)52(13)- 2 635( 14)- 1 494(16)-2 821(16)- 2 OOO( 13)- lOl(17)- 629(15)840(11)526(14)-1 758(20)- 2 041 (34)369(19)851(27)-2 310(21)-2 925(22)-2 745(19)-3 538(26)- 320(22)- 188(32)315(15)751 (20)1428(27)1639(24)1 181(15)ZIC0353(2)1058(2)- 542(2)- 363(8)- 347( 10)-1 112(9)- 1 297(10)1 171(13)1 635(10)- 41 7( 12)-1 025(8)874( 11)207 (8)- 191(10)1910(10)- 62( 13)- 201 (24)- 1 471(14)-2 159(15)1602(16)2 069(16)-909( 16)-1 394(22)399( 19)494(20)2 431(10)3 084,18)3 156(12)2 593(17)1981(14)MgO.Diphenylpicrylhydrazyl (dpph) was used as a gmarker (g 2.003 6). The crystal was mounted on a quartz29rod by means of apiezon grease and measurements werecarried out around three arbitrary rotation axes.In eachplane, spectra were recorded at 10" intervals a t ca. 110 K.The theory and the computer programs used to deriveprincipal values and orientations of the various tensors havebeen described elsewhere.* The experimental spectra werefitted to the spin Hamiltonian (1) using second-orderTABLE 2Bond distances (A) and angles ( O ) , with standarddeviations in parentheses(a) DistancesZn( 1)-Zn(2)Zn(l)-Zn(3)Zn( 1)-Zn(4)Zn( 1)-O( 1)Zn(1)-O(3)Zn(1)-O(l1)Zn(2)-Zn(3)Zn( 2)-Zn(4)Zn (2)-0 (2)Zn (2)-O( 5)Zn (2)-0 (7)Zn(2)-O(11)Zn (3) -Zn (4)Zn (3)-0 (6)Zn (3)-0(9)Zn(3)-0( 11)Zn(3)-N(1)Zn(4)-0(4)Zn (4)-0 (8)Zn(4)-0( 10)Zn (4)-0 ( 1 1)(b) AnglesO( 3)-Zn( 1)-O( 11)0 ( 1)-Zn ( 1 )-0 (3)O(3)-Zn( 1)-O( 11)0(2)-Zn(2)-0(5)O( 2)-Zn (2)-0 ( 7)0(2)-Zn(2)-0(11)O( 5)-Zn (2)-O( 7)0(5)-Zn (2)-0 (1 1)0(7)-Zn(2)-0(11)0(6)-Zn( 3)-0(9)0(6)-Zn(3)-0( 11)O(6)-Zn (3)-N (1)0(9)-Zn(3)-0(11)O( 9)-Zn (3)-N( 1)O( 1 l)-Zn(3)-N(1)0 (4)-Zn (4)-0 (8)O( 4)-Zn (4)-0 (1 0)0(4)-Zn(4)-0(11)O( 8)-Zn (4)-O( 10)0(8)-Zn(4)-0(11)3.157 (3)3.060 (4)3.197(4)2.18(2)1.95(2)3.215(5)3.181(5)1.95 (2)1.94(3)1.95( 2)1.96( 2)3.190(4)2.07(2)2.0 1 (3)1.96I2)3.07(2)1.94 ( 4)1.95(2)1.98(2)1.94( 2)2.21 (2)97.4( 7)97.3(7)101.0(11)87.4(7)108.4 (9)119.7(8)103.6(10)11 3.3 (8)109.2 (9)93.7( 10)101.0(8)95.9 ( 8)107.8( 8)98.8(10)147.3(9)103.7(8)103.1(9)I 20.0( 8)106.3(10)110.4(8) o( iO)-zi (4)-O( 1 i) 1 12.i ( 8 jZn( 1)-O( l ) - C ( 1) 13 l(2)Zn(2)-0(2)-C( 1)Zn( 1 )-0 (3)-C (3)Zn (4)-0 (4)-C (3)Zn (2)-0 ( 5)-C (5)Zn (3)-0 (6)-C (5)Zn (2)-O( 7)-C (7)Zn (4)-0 (8) -C (7)Zn( 3)-O( 9)-C( 9)Zn (4)-0 (lo)< (9)Zn(l)-O(ll)-Zn(2)Zn( 1)-O( 11)-Zn(3)Zn(l)-O(ll)-Zn(4)Zn(2)-0( 11)-Zn(3)Zn(2)-0( 11)-Zn(4)Zn(3)-0( 11)-Zn (4)O( 3)-C( 3)-0( 4)0 (5)-C (5)-0 (6)O( 7)-C( 7)-0 (8)O( 3)-C( 1)-O(2)0(9)-C(9)-0(10)Zn (3)-N (1)-C (1 1)Zn( 3)-N (1)-C( 15)1.32(3)1.27( 4)1.21 (4)1.27 (4)1.26 (4)1.24(4)1.26(4)1.29(4)1.28(5)1.29(4)1.37(3)1.35 ( 4)1.5 1 (6)1.50( 4)1.51(5)1.52(5)1.53( 5)1.41(4)1.38(5)1.43 ( 5 )1.36(4)120(2)121(2)139(2)127(2)138(2)134(2)131(2)136(2)131(2)1 09.4 (9)103.0(8)110.5(8)1 12.3( 8)11 1.4(8)121(3)121(3)124(3)122(3)119(4)110.0(9)111(2)126(perturbation theory.No assumptions were made con-cerning the relative orientation of the various tensors.DISCUSSIONDescription of the Structure.-The most noteworthyaspect of the structure of (1) is shown in Figure 1. Sevenzinc atoms bridged by acetate groups form a heptamericcentrosymmetrical unit. In each one of the two semi-units a central oxygen is surrounded tetrahedrally byfour zinc atoms. Five of the six edges of the resultingtetrahedron are bridged by acetate groups. The sixthone, Zn(l)-Zn(3), is open. There are, therefore, three* For details see Notices t o Authors No.7, J.C.S. Dalton, 1978,Index issue30IJ.C.S. Daltontypes of non-equivalent zinc atoms: Zn(l), which lieson the centre of symmetry, has an octahedral arrange-ment; Zn(3) is tetrahedrally co-ordinated to threeoxygen atoms and to the nitrogen atom; Zn(2) andZn(4) are also tetrahedrally co-ordinated, but to fouroxygen atoms. The structure strongly resembles thatof [Zn,O(O,CMe),] where a closed tetrahedron of zincatoms surrounds an ' inner ' oxygen.'In the title complex the pyridine molecule opens outone edge of the tetrahedron and, for stoicheiometricreasons, one of the two zinc atoms of the opened edgehas to change its co-ordination geometry, by ' duplicat-ing ' its three co-ordination directions.The resultingslightly distorted octahedron of oxygen atoms is there-fore tetragonally compressed, the bond distance Zn( 1)-O( 11) being that relative to a tetragonally co-ordinatedoxygen. The Zn-0 distances [2.20(2) A] in the d,, planecontaining Zn(l), 0(1), and O(3) are actually significantlylonger than in the d,l direction [Zn(l)-O(11) 1.95(2) A],confirming the description above and the e.s.r. data.From a comparison with data available in the liter-ature, we note that, about the tetrahedral zinc atoms,Zn(2)-0 and Zn(4)-0 distances are in the narrow range1.94-1.96 A ( 0 0.02 A) comparable with those found in[Zn,O (O,CMe),] and [Zn{ SC( NH2),),( O,CMe),] whileZn(3)-0 distances are spread over the range 1.96-2.07by the great distortion of the co-ordination tetrahedronowing to the steric hindrance of the pyridine molecule,the six tetrahedral angles around Zn(3) being greatlydifferent from the ideal value of 109.5" [range 96-147"(see Table 2)].Only two intermolecular contacts shorter than 3.5 Aare present: C(12) O(4) and C(12) 0(10), 3.37and 3.43 A(o 0.04 A) respectively.E.S.R.Results.-According to the X-ray resultsgiven above, for a general orientation of the externalmagnetic field, the e.s.r. spectrum of a crystal of copper-doped (1) is expected to be a superimposition of several0FIGURE 2 First-derivative X-band spectrum of a single crystalof copper-63-doped (1) for an arbitrary orientation of the ex-ternal magnetic field. Absorptions from the four inequivalentsites are indicatedcopper signals. Apart from possible copper-copperinteractions, giving rise to triplet spectra, 28 partlychemically, partly magnetically, inequivalent sites areavailable for copper substitution.Despite this, the* *FIGURE 3 First-derivative X-band spectrum of a single crystalThe external magnetic field lies in aSo-called ' forbidden ' lines are indi-of copper-63-doped ( 1 ) .crystallographic plane.cated by arrows in the low-field part of the spectrumexperimental spectra depict a much simpler situation.The major features of the spectrum (Figure 2) can beanalyzed by assuming four groups of four signals,coalescing to two sets in the crystallographic planes(Figure 3) and to only one set along the crystallographicaxes.On this basis, we assume that copper substitutionmainly occurs for only one of the seven possible zincatoms. The behaviour of the e.s.r. spectra, suggestingan almost pure d,a ground state, indicates that theparamagnetic ion enters the compressed tetragonal Zn( 1)site, which is the only one consistent with such a ground-state orbital.In addition, other weak signals, due to a chemicallydifferent copper ion, can be observed in the spectra.They show clearly resolved hyperfine interaction withone nitrogen nucleus (14N, loo%, I = +) and are attri-buted to substitution of some of the copper ions at theZn(3) tetrahedral site. Since these signals have ratherlow intensity and have been observed only in few of theseveral crystals examined, no detailed analysis has beenpossible and they will not be mentioned further.Finally, AMz > 0 transitions have been detected.Their intensity results from mixing of spin states withdifferent Mz values by nuclear Zeeman and quadrupoleinteractions.Keeping in mind that our measurementswere carried out at the X-band frequency, where nuclearZeeman effects are small, the high intensity of the for-bidden signals points to the presence of a rather largequadrupole coupling tensor. The principal values of thetensors, the angles that their axes make with the crystalaxes, and their relative orientations are reported inTable 3. The final values have been obtained byaveraging the experimental results for three of the fourmagnetically inequivalent sites.Extensive overlappingof the signals led to large inaccuracies in the results forthe fourth site.The g and hyperfine-coupling tensors are not diagonalin the same axis system; they have one common axis1979i.e. g, and AICu, while the other components, in theperpendicular plane, are rotated by ca. 28" with respectto one another. The common axis of the two tensorsTABLE 3Principal values and principal axes of the g tensor, the metalhyperfine-coupling, and quadrupole-coupling tensorscm-l) for copper-doped (1) at 110 KPrincipal Principal Estimated Direction angles bg1 2.293 0 0.000 2 73 70.5 16g2 2.267 4 0.000 2 54.5 50.8 79.51.996 1 0.000 2 40.5 51 79.8 gsgav." 2.185 4164.55 0.3 89.7 89.5 179.532.45 0.5 62.5 27.5 89.5axis value error --A ,A2A3 18.5 0.5 27.5 117.5 90A13".e 37.93.35 0.05 70 23 10297 100 168P Ip, - 0.80 0.1p, - 2.55 0.1 21 110 93Zn-0 (3) 77 81 164Zn-O( 1 ) 131 133 108Zn-O( 1 1 ) 41 129 78Defined as the largest difference between the correspondingprincipal values of the magnetically inequivalent sites in theunit cell.* The eigenvectors of the g tensor and of the metal-oxygen bond distances are given with respect to the crystal-axis system. The direction angles of the A and P tensors arereferred to the g-axis system. e Average values; Aav. wascalculated assuming that A , is opposite in sign to A , and A , (seetext).was located along the tetragonal axis of the molecularframe, i.e. along the Zn-O(l1) bonding direction. Thepresence of large distortions in the bonding angles, asshown by the X-ray results [O(l)-Zn-O(l1) and O(3)-Zn-O(l1) are both ca.97"], implies that the othermagnetic axes cannot coincide with any simple mole-cular direction. However, g, and g, roughly point alongthe in-plane metal-oxygen bonds.The largest ACu component is associated with thesmallest g value, the latter being very close to the free-electron g value. These observations indicate that themolecular orbital of the unpaired electron is an essenti-ally dZl type.s Indeed the g, value is even smaller thang,, suggesting negligible mixing with other d orbitals.Considering the low symmetry of the metal-ion environ-ment, as revealed by the X-ray measurements, this is arather surprising result.For true d,a ground states,g, (or g3) is expected to be G2.002 3, if higher-orderterms of the type ( h / A ) 2 are taken into account[equation (2)]. Small admixtures of other d orbitals,mainly d5*-y*, may raise g, well above this value.More detailed information can be achieved followingthe approach of Swalen et aZ.1° The ground-stateorbital wavefunction is written as a linear combinationof all the five d orbitals. The g and A C u values are thenused to derive the five coefficients and the dipolar andisotropic terms. The results of Table 4 show that theexperimental g values can be reproduced, with reason-able accuracy, by using only three instead of five co-efficients. Admixture of the dza-ys and d,, orbitalsappears to be unimportant. A small difference in thecoefficients and/or energy of the two out-of-planeTABLE 4&Orbital mixing coefficient and calculated magneticparameters for copper-doped (1)Parameter Axis g ACUC,'+' 0 ~, 0 x 2.2940 -21.4CY 2 0.0395 z 1.9954 164.6c 2 0.9983cxz 0.0430 Y 2.2690 - 31.1P 369K 0.089orbitals, d,, and d,,, is enough to account for the observedin-plane g anisotropy.As far as the ACU principal values are concerned,satisfactory agreement with experiment, together withreasonable values of P and K , could be obtained onlywith the choice of signs of Table 4.Different choicesFIGURE 4 X-Band powder spectrum of copper-63-doped (1)invariably led to poor calculated ACU values or unrealis-tic, or negative, values of P and K. However, since theisotropic spectrum could not be measured we do nothave experimental support for this result.The cal-culated P value shows almost no reduction with respectto the free-ion value (360 x cm-l), whereas the Kvalue is much smaller than that usually observedll(ca. 0.3). I t is well known that the isotropic part of thehyperfine splitting is the result of two main contri-butions of opposite sign, i.e. spin polarization of theinner-shell s electrons and direct participation of the4s orbital in the ground-state molecular orbital. Al-though the former is usually the dominating effect, asmall 4s participation can strongly reduce the observedvalue of Aim, or even change its sign. In C a notationthe ground state of copper-doped (1) is an A,-type orbitaland hence 4s mixing is symmetry-allowed.The effecton the isotropic part of the hyperfine interaction can bequite large, even for a small value of the mixing co-efficient. For instance, 57; admixture would raise thecalculated K value to 0.20.Strongly tetragonal systems, such as the one we aredealing with, are expected to give rise to large quad-rupole interactions. Compared to other results reportedin the literature 1 2 9 1 3 our P values appear to be rathersmall. We believe that this is a consequence of thetype of data analysis used. The results of Table 3 wer32 J.C.S. Daltonobtained by fitting the line positions of the AM1 = 0transitions with the aid of second-order perturbationexpressions. I t is well known that, in the presence oflarge quadrupole effects, this approach may be in-adequate to reproduce the positions of the ' allowed ', aswell as of the ' forbidden ', lines.13 Probably accurate Pvalues may be obtained only by exact diagonalization ofthe energy matrix. However, trial calculations, per-formed using both the strong-field approximation andthe complete spin Hamiltonian, have shown that theother magnetic parameters are rather insensitive to theexact value of the quadrupole-coupling tensor.[8/074 Received, 17th January, 19581REFERENCESD. Attanasio, J . Magnetic Resonance, 1977, 26, 81.D. Attanasio and M. Gardini, J . Magnetic Resonance, in thepress.R. G. Pearson, D. H. Anderson, and L. L. Alt, J . Amer.Chem. SOC., 1955, 77, 527.D. Attanasio, 1. Collamati, and C . Ercolani, J.C.S. Dalton,1974, 2442.5 ' International Tables for X-Ray Crystallography,' KynochPress, Birmingham, 1974, vol. 4.C. P. Keijzers, G. F. M. Paulussen, and E. de Boer, Mol.Phys., 1975, 29, 973.H. Koyama and Y. Saito, Bull. Chern. SOC. Japan, 1954, 27,112.8 L. Cavalca, G. Fava Gasparri, G. D. Andreetti, and P. Dom-iano, Acta Cryst., 1967, 22, 90. * A. Abragam and B. Bleaney, ' Electron ParamagneticResonance of Transition Ions,' Oxford University Press, London,1970, p. 464.lo J . D. Swalen, B. Johnson, and H. M. Gladney, J . Chem.Phys., 1970, 52, 4078.l1 B. R. McGarvey, J . Phys. Chem., 1967, 71, 51.l2 B. Bleaney, K. D. Bowers, and M. H. L. Pryce, Proc. Roy.SOC., 1955, A228, 166.l3 H. So and K. I,. Belford, J . Amev. Chem. SOC., 1969, 91,2392