UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Synthesis, characterization and magnetic properties of some transition metal pyrazolate complexes Ehlert, Martin Kurt 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1992_fall_ehlert_martin_kurt.pdf [ 9.86MB ]
JSON: 831-1.0061748.json
JSON-LD: 831-1.0061748-ld.json
RDF/XML (Pretty): 831-1.0061748-rdf.xml
RDF/JSON: 831-1.0061748-rdf.json
Turtle: 831-1.0061748-turtle.txt
N-Triples: 831-1.0061748-rdf-ntriples.txt
Original Record: 831-1.0061748-source.json
Full Text

Full Text

SYNTHESIS, CHARACTERIZATION AN]) MAGNETICPROPERTIES OF SOME TRANSITION METALPYRAZOLATE COMPLEXESbyMARTIN KURT EHLERTB.Sc., McMaster University, 1987A THESIS SUBMITIED IN PARTIAL FULFiLLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDepartment of ChemistryWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAMay 1992© Martin Kurt Ehiert, 1992In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission._________________________________Department of //4fr.1The University of British ColumbiaVancouver, CanadaDate z- //4g’ zDE-6 (2/88)ABSTRACTA number of binary, divalent transition metal (substituted)pyrazolate compounds havebeen prepared of the general formula, [M(pz*),Jx (M = Cu (eleven compounds), Co (sixcompounds), Ni (one compound), Zn (one compound); pz = pyrazolate and C-substitutedpyrazolates). [Cu(4-C1pz)2j(4-Clpz = 4-chioropyrazolate) was prepared in green and brownforms. Three of the copper(ll) pyrazolates prepared, namely, the pyrazolate,4-methylpyrazolate, and 4-chioropyrazolate (green) derivatives, have been characterized bysingle crystal X-ray diffraction which revealed them to consist of linear chains containingcopper(II) ions doubly bridged by pyrazolate ligands. By analogy with the structures of thesecompounds and using physical evidence from all of the [M(pz*)2jspecies, it has beenconcluded that all such compounds considered in this work possess the same polymeric linearchain motif. In addition to these compounds, a material with the formulaCu(4-Ipz)2.1(4-IpzH (4-Ipz = 4-iodopyrazolate; 4-IpzH = 4-iodopyrazole) was synthesizedand characterized.The binary copper(II) and cobalt(ll) (substhuted)pyrazolates are paramagnetic andtheir magnetic susceptibilities were measured over the 2-300 K temperature range. All ofthese compounds exhibit antiferromagnetic exchange coupling (J, the exchange couplingconstant, ranges from -58 to -105 cm1). The [Cu(pz*)2] (pz*= pyrazolate,4-methylpyrazolate, 4-chioropyrazolate (green), 4-bromopyrazolate) species arethermochromic; their colours change from green to blue upon cooling from 300 K to 77 K.Low temperature diffraction studies have shown the thermochromism to be concomitant withsubtle structural changes in the compounds. These changes may be responsible for anomaliesobserved in the magnetic properties of the 4-Cl (green) and 4-Br derivatives. The analogouscobalt(ll) complexes exhibit anisotropic magnetic susceptibilities and quantitative analysis oftheir magnetic behaviour indicates that the compounds undergo Ising-like, 1-dimensional,11antiferromagnetic exchange. The Ising-like behaviour is likely due to single-ion, zero-fieldsplitting effects.In addition to polymeric, binary transition metal pyrazolates, several oligometallic,pyrazolate complexes were prepared and characterized. Most of these compounds weresubjected to single crystal X-ray diffraction studies. Seven trimetallic copper(I) orcopper(IIlI) species and a dimeric copper(ll) complex were prepared. Two isostructuralzinc(ll) and cobalt(ll) dimers were prepared and studies have indicated that these compoundscan undergo incipient polymerization reactions to form oligometallic chain compounds.Trimetallic and tetrametallic cobalt(II) complexes were also synthesized. Finally, two novel,octametallic molybdenum clusters were prepared and subjected to preliminarycharacterization.ifiTABLE OF CONTENTSAbstract iiList of Tables xivList of Figures xviList of Abbreviations and Symbols xxivAcknowledgements xxviiiChapter 1 Introduction 11.1 Low-dimensional materials 11.2 Magnetism 31.2.1 Magnetic exchange 101.3 Pyrazole 181.4 Pyrazolyl containing transition metal compounds 201.4.1 Binary transition metal pyrazolates 201.5 Scope of this study and organization of the dissertation 251.6 Physical methods of characterization 271.6.1 Magnetic property measurement 271.6.2 X-Ray diffraction methods 311.6.2.1 Single crystal X-ray diffraction 311.6.2.2 Powder X-ray diffraction 321.6.3 Spectroscopic methods 321.6.3.1 Infrared spectroscopy 321.6.3.2 Electronic spectroscopy 331.6.3.3 Electron paramagnetic resonance spectroscopy 331.6.3.4 Nuclear magnetic resonance spectroscopy 341.6.4 Miscellaneous methods 34iv1.6.4.1 Elemental analysis .341.6.4.2 Thermal analysis 341.6.4.3 Mass spectrometry 351.6.4.4 Scanning electron microscopy 35Chapter 2 Poly(copper(ll) pyrazolates) 372.1 Introduction 372.2 Results and discussion 372.2.1 Copper(H) 4-X-pyrazolates (X = H, Me, Cl, Br) 372.2.1.1 Synthesis, physical and thermal properties 372.2.1.2 X-Ray diffraction studies 422. Room temperature single crystal diffraction studies 432. Room temperature powder diffraction studies 452. Low temperature diffraction studies 482.2.1.3 Spectroscopic behaviour 512. Infrared spectroscopy 512. Electronic spectroscopy 512. Electron paramagnetic resonance spectroscopy 522.2.1.4 Magnetic properties 552.2.2 Copper(II) 4-X-3,5-dimethylpyrazolates (X = H, Me, Cl, Br) 672.2.2.1 Syntheses, physical and thermal properties 672.2.2.2 X-Ray diffraction studies 692.2.2.3 Spectroscopic behaviour 692. Infrared spectroscopy 692. Electronic spectroscopy 732. Electron paramagnetic resonance spectroscopy 742.2.2.4 Magnetic properties 752.2.3 Miscellaneous poly(copper(II) pyrazolates) 84V2.2.3.1 Syntheses, physical and thermal properties .842.2.3.2 X-ray diffraction studies 842.2.3.3 Spectroscopic behaviour 852. Infrared spectroscopy 852. Electronic spectroscopy 872. Electron paramagnetic resonance spectroscopy 882.2.3.4 Magnetic properties 882.3 Summary and conclusions 94Chapter 3 Oligometallic copper pyrazolates 963.1 Introduction 963.2 Results and discussion 963.2.1 Copper(I) pyrazolates 963.2.1.1 Syntheses, physical and thermal properties 963.2.1.2 X-Ray diffraction studies 983.2.1.3 Spectroscopic studies 1033. Infrared spectroscopy 1033. Mass spectrometry 1043.2.2 Copper(I) and copper(II) carboxylpyrazolate complexes 1043.2.2.1 Syntheses, physical and thermal properties 1043.2.2.2 X-Ray diffraction studies 1073.2.2.3 Electron paramagnetic resonance spectroscopy 1113.2.2.4 Magnetic properties 1133.3 Summary and conclusions 121Chapter 4 Zinc(II) and cobalt(II) pyrazolates 1234.1 Introduction 1234.2 Results and discussion 1234.2.1 Zinc(II) 3,5-diniethylpyrazolates 123vi4.2.1.1 Syntheses, physical and thermal properties .1234.2.1.2 X-Ray diffraction studies 1264. Single crystal X-ray diffraction 1264. Powder X-ray diffraction studies 1274.2.1.3 Infrared spectroscopy 1274.2.2 Oligometallic cobalt(ll) 3,5-dimethylpyrazolates 1324.2.2.1 Syntheses, physical and thermal properties 1324.2.2.2 X-Ray diffraction studies 1354.2.2.3 Spectroscopic behaviour 1424. Infrared spectroscopy 1424. Electronic spectroscopy 1444.2.2.4 Magnetic properties 1494.2.3 Poly(cobalt(II) pyrazolates) 1624.2.3.1 Syntheses, physical and thermal properties 1624.2.3.2 X-Ray diffraction studies 1644.2.3.3 Spectroscopic behaviour 1694. Infrared spectroscopy 1694. Electronic spectroscopy 1704.2.3.4 Proposed structures and magnetic properties 1754.3 Summary and conclusions 199Chapter 5 Miscellaneous compounds 2015.1 Introduction 2015.2 Results and discussion 2015.2.1 Nickel(II) pyrazolate 2015.2.1.1 Synthesis and physical properties 2015.2.1.2 Characterization and proposed structure 2035.2.2 A copper(II) 4-iodopyrazolyl compound 205vii5.2.2.1 Synthesis and physical properties .2055.2.2.2 Magnetic properties 2065.2.3 Octametallic molybdenum oxo-pyrazolate clusters 2085.2.3.1 Syntheses, physical and thermal properties 2085.2.3.2 X-ray diffraction studies 2125.2.3.3 Infrared spectroscopy 2175.2.4 Further comment on the octametallic molybdenum clusters 2185.3 Summary and conclusions 219Chapter 6 General summary and suggestions for future work 2216.1 General summary 2216.2 Suggestions for future work 224Chapter 7 Experimental 2287.1 Introduction 2287.2 Syntheses 2287.2.1 Pyrazole derivatives 2297.2.1.1 4-Chioropyrazole4-ClpzH 2297.2.1.2 4-Bromopyrazole4-BrpzH 2297.2.1.3 4-lodopyrazole4-IpzH 2307.2.1.4 4-Nitropyrazole4-NO,pzH 2317.2.1.5 3,4,5-Trimethylpyrazole4-MedmpzH 2317.2.1.6 4-Chloro-3,5-diinethylpyrazole4-CldmpzH 232vii’ 4-Bromo-3,5-dimethylpyrazole4-BrdmpzH 2337.2.1.8 4-Iodo-3,5-dimethylpyrazole4-IdrnpzH 2337.2.1.9 4-Nitro-3,5-dimethylpyrazole4-NO2dmpzH 2347.2.2 Copper(1) pyrazolates 2357.2.2.1 Tris(ji-3 ,5-dirnethylpyrazolato-N,N’)tricopper(I),[Cu(4-Hdrnpz)]3 2357.2.2.2 Tris(j.t-3 ,4,5-trirnethylpyrazolato-N,N’)tricopper(I),[Cu(4-Medrnpz)]3 2367.2.2.3 Tris(p.-4-chloro-3,5-dimethylpyrazolato-N,N’)tricopper(I),[Cu(4-Cldrnpz)]3 2377.2.2.4 Tris(jt-4-brorno-3 ,5-dimethylpyrazolato-N,N’)tricopper(I),[Cu(4-Brdrnpz)]3 2377.2.2.5 Tris(j.t-4-iodo-3 ,5-dimethylpyrazolato-N,N’)tricopper(I),[Cu(4-Idrnpz)]3 2387.2.2.6 Tris(j.t-4-iodopyrazolato-N,N’)tricopper(I),[Cu(4-Ipz)]3 2387.2.2.7 Tris(ji-indazolato-N,N’)thcopper(I),[Cu(indz)3] 2397.2.3 Copper(ll) pyrazolates 2397.2.3.1 Poly-bis(ji-pyrazolato-N,N’)copper(II),[Cu(4-Hpz)21 2397.2.3.2 Po1y-bis(i-4-rnethy1pyrazo1ato-N,N’)copper(ll),[Cu(4-Mepz),j 240ix7.2.3.3 Poly-bis(j.t-4-chloropyrazolato-N,N’)copper(ll),[Cu(4-C1pz)2] 2417.2.3.4 Poly-bis(j.t-4-bromopyrazolato-N,N’)copper(II),[Cu(4-Brpz)2J 2427.2.3.5 Bis(ji-4-iodopyrazolato-N,N’)hemi(4-iodopyrazole)copper(ll),Cu(4-Ipz)2.1(4-IpzH 2427.2.3.6 Po1y-bis(.t-3,5-dimethy1pyrazo1ato-N,N’)copper(II),[Cu(4-Hdrnpz),J 2437.2.3.7 Poly-bis(ji-3 ,4,5-trimethylpyrazolato-N,N’)copper(II),[Cu(4-Medrnpz)21 2447.2.3.8 Po1y-bis(.t-4-ch1oro-3,5-dimethy1pyrazo1ato-N,N’)copper(II)[Cu(4-C1drnpz),] 2447.2.3.9 Poly-bis(j.t-4-bromo-3 ,5-dimethylpyrazolato-N,N’)copper(II)[Cu(4-Brdrnpz)21 2457.2.3.10 Po1y-bis(.t-3-methy1pyrazo1ato-N,N’)copper(I1)[Cu(3-Mepz),1 2457.2.3.11 Po1y-bis(i-indazo1ato-N,N’)copper(ll)[Cu(indz),] 2467.2.4 Copper carboxylpyrazolates 2467.2.4.1 Bis(.t-4-brorno-3-carboxy-5-methy1pyrazo1ato-N,N’,O)tetrakis(4-brorno-3 ,5-dirnethylpyrazole)dicopper(ll),[Cu(4-Br-3.-CO,Mepz)(4-BrdmpzH)}, 2467.2.4.2 Bis[Q.t-3-carboxy-4,5-dimethylpyrazolato-N,N’,O)(3,4,5-trimethylpyrazole)cuprate(I)]copper(ll),[Cu(3-COdrnpz)(4-MedmpzH)]u,andxTris(jt-3,4,5-trimethylpyrazolato-N,N’)tricopper(I),[Cu(4-Medrnpz)J3 2477.2.5 Cobalt pyrazolates 2487.2.5.1 Poly-bis(j.t-pyrazolato-N,N’)cobalt(II),[Co(4-Hpz)2] 2487.2.5.2 Poly-bis(.t-3-rnethy1pyrazolato-N,N’)coba1t(II),[Co(3-Mepz)2J 2497.2.5.3 Poly-bis(j.t-3,5-dimethylpyrazolato-N,N’)cobalt(II),-[Co(4-Hdrnpz),j 2497.2.5.4 Pol y-bis(-3,4,5-thrnethy1pyrazo1ato-N,N’)cobalt(H),[Co(4-Medmpz),] 2507.2.5.5 Poly-bis(j.t-4-chloro-3 ,5-dimethylpyrazolato-N,N’)cobalt(II),[Co(4-C1drnpz)2] 2507.2.5.6 Poly-bis(j.t-4-brorno-3 ,5-dimethylpyrazolato-N,N’)cobalt(JI),[Co(4-Brdrnpz)-,] 2517.2.5.7 Bis[di-ji-3,5-dirnethylpyrazolato-N,N’-chloro(3,5-dirnethylpyrazole)cobaltate(II)jcobalt(I1)[Co(4-Hdrnpz)C1(4-HdmpzH)jo 2517.2.5.8 Dimeric/oligornetallic cobalt(II) 3,5-dimethylpyrazolate-3,5-dirnethylpyrazoleCo(4-Hdrnpz),.0.344(4-HdrnpzH) 2527.2.5.9 ji-Oxo-hexakis(i-3,5-dimethylpyrazo1ato-N,N’)tetracobalt(II)[Co4(4-Hdrnpz)60j 2537.2.6 Nickel pyrazolate 2547.2.6.1 Poly-bis(j.t-pyrazolato-N,N’)nickel(II),[Ni(4-Hpz),] 2547.2.7 Zinc pyrazolates 254xi7.2.7.1 Bis(j.t-3,5-dimethylpyrazolato-N,N’-3,5-dimethylpyrazolato-3,5-dimethylpyrazole)dizinc(II)[Zn(4-Hdmpz)2(4-Hdmpz)1 2547.2.7.2 Dimeric/oligometallic zinc(II) 3,5-dimethylpyrazolate-3,5-dimethylpyrazoleZn(4-Hdmpz).0.323(4-Hd zfl) 2557.2.7.3 Poly-bis(j.t-3 ,5-dimethylpyrazolato-N,N’)zinc(II),[Zn(4-Hdmpz)21 2567.2.8 Molybdenum oxo-pyrazolates 2567.2.8.1 [Mo(4-Hpz)6018-HpzH}.2(4-H zH) 2567.2.8.2 [Mo(4-Hpz),-HpzH].3(4-HpzH).H2O 2577.3 Unsuccessful preparations 2577.3.1 Mixed cobalt(II)/zinc(II) 3,5-dimethylpyrazolate 2587.3.2 Copper(II) 4-nitropyrazolates 2597.3.3 Copper(II) and cobah(I1) 4-iodo-3,5-dimethylpyrazolates 2597.3.4 Cobalt(1ll) basic pyrazolate 2597.3.5 Nickel(II) substituted pyrazolates 2607.3.6 Iron(II) pyrazolate 2617.3.7 Chromium(III) pyrazolate 2617.4 Physical methods 2617.4.1 Chemical analysis 2617.4.2 Spectroscopic methods 2627.4.2.1 Electron paramagnetic resonance 2627.4.2.2 Electronic spectroscopy 2627.4.2.3 Infrared spectroscopy 2637.4.2.4 Nuclear magnetic resonance spectroscopy 2637.4.3 Magnetic susceptibility measurements 263xli7.4.3.1 Gouy balance .2647.4.3.2 Vibrating sample magnetometer 2657.4.3.3 SQUID magnetometer 2667.4.4 Diffraction methods 2677.4.4.1 Single crystal X-Ray diffraction 2677.4.4.2 Powder X-ray diffraction 2677.4.5 Miscellaneous methods 2687.4.5.1 Melting point determination 2687.4.5.2 Differential scanning calorimetry 2687.4.5.3 Scanning electron microscopy 2687.4.5.4 Mass spectrometry 2697.4.5.5 Solubility studies 269References 270Appendix I Single crystal X-ray diffraction data 280Appendix II Powder X-ray diffraction data 321Appendix III Magnetic data 324Appendix IV Infrared data 343Appendix V Mass spectra 348Appendix VI Special apparatus diagrams 354xl”LIST OF TABLESTable Page1.1 Abbreviations for pyrazole and pyrazole derived moieties 191.2 Metal pyrazolates prepared by Vos and Groeneveld (37-41) 232.1 X-ray powder diffraction data for [Cu(4-Brpz)2] 472.2 Crystallographic parameters for [Cu(4-Hpz)2]at 294 K and 116 K 492.3 Room temperature and low temperature lattice constants for the greencopper(ll) 4-X-pyrazolates with estimated standard deviations in the lastdigit in parentheses 502.4 Measured g values for [Cu(4-Xpz),] compounds 552.5 Derived magnetic parameters for the [Cu(4-Xpz)2]compounds 592.6 Derived magnetic parameters for the [Cu(4-Xpz)2jcompounds in the 2-300 K range with %P modelled as a Curie-Weiss law paramagnet 602.7 [Cu(4-Xpz),] structural parameters (see Figure 2.11(c)) 622.8 Positions of the bands in the [Cu(4-Xdmpz)2]complexes and the4-XdmpzH molecules 732.9 Derived magnetic parameters for the [Cu(4-Xdmpz),] compounds withestimated standard deviation in the last digit in parentheses 802.10 Derived magnetic parameters for [Cu(3-Mepz)2jand [Cu(indz)2]withestimated standard deviation in the last digit in parentheses 923.1 Selected structural parameters for [M(4-X-3,5-R2pz)]3species (distances in A) 1024.1 Electronic spectral data for [Co(4-Hdmpz)Cl(4-HdmpzH)],Co andCo(4-Hdmpz),.0.344(4-HdrnpzH) 146xiv4.2 Ligand field parameters for [Co(4-Hdmpz)2C1(4-HdmpzH)JoandCo(4-Hdmpz).0.34 (4-HdrnpzH) (estimated uncertainties in parentheses) 1484.3 Derived magnetic parameters for Reedijk and co-worker’s trimetallictriazolyl Co(I1) complexes 1554.4 Derived magnetic parameters for [Co(4-Hdmpz)2C1(4-HdmpzH)]owithuncertainties in the last digit in parentheses 1584.5 Electronic spectral data for [Co(4-Hpz)2]and [Co(3-Mepz)2J 1724.6 Electronic spectral data for the [Co(4-Xdmpz) (X = H, Me, Cl, Br) complexes.1734.7 Ligand field parameters for the [Co(pz*)2]complexes (pz* = 4-Hpz, 3-Mepz, 4-Hdmpz,4-Medmpz, 4-Cldrnpz, 4-Brdmpz) (numbers in parenthesesrepresent estimated uncertainties) 1744.8 Derived magnetic parameters for the [Co(pz*)2]complexes (pz* = 4-Hpz,3-Mepz, 4-Hdmpz, 4-Medmpz, 4-Cldmpz, 4-Brdmpz) using theHeisenberg, S = 3/2, antiferromagnetically coupled chain model withestimated standard deviation in the last digit in parentheses 1884.9 Derived magnetic parameters for low temperature data fits of the S = 1/2,Ising x11 model coupled with a Curie-Weiss law paramagnetic impuritycomponent for the [Co(pz*),L pz* = 4-Hpz, 3-Mepz, 4-Hdmpz, 4-Medmpz, 4-Cldmpz, 4-Brdrnpz) complexes. Estimated standard deviationin the last digit in parentheses 1954.10 Derived magnetic parameters for low temperature data fits of the S = 1/2,Ising x11 and S = 3/2, Heisenberg models to the data for the [Co(pz*),j (pz*= 4-Hpz, 3-Mepz, 4-Hdrnpz, 4-Medmpz) complexes. Estimated standarddeviation in last digit in parentheses 1967.1 Diamagnetic contributions of ligands and metal ions 264xvLIST OF FIGURESFigure Page1.1 Structure of KCuF3.The solid lines indicate the pseudo-perovskite latticeand the dotted lines indicate the actual unit cell. CuF chains run along thec-axis direction. After Okazaki and Suemune (5) 31.2 The components of the spin Hamiltonian 91.3 Spin orientations in a planar net array; (a) paramagnetism,(b) ferromagnetism, (c) antiferromagnetism 111.4 The approximate susceptibility magnitude ellipsoid in an antiferromagneticHeisenberg chain 131.5 The approximate susceptibility magnitude ellipsoid in an antiferromagneticIsing chain 141.6 The approximate susceptibility magnitude ellipsoid in an antiferromagneticX-Y chain 141.7 Magnetic susceptibility versus temperature in reduced coordinates for S =1/2, 1-D Heisenberg, Ising, and X-Y antiferromagnetic coupling. Afterde Jongh and Miedema (1) 151.8 Orbital interactions in superexchange; (a) kinetic exchange, (b) potentialexchange 171.9 Sample configurations for (a) the Gouy method and (b) the Faraday method 281.10 Schematic diagram of a vibrating sample magnetometer 301.11 Schematic diagram of a SQUID magnetometer 312.1 SEM images of (a) [Cu(4-Hpz),j and (b) [Cu(4-Clpz)2](green). Thewhite line represents a length of 5 .tm 42xvi2.2 View of the crystal structure of [Cu(4-Hpz)2]down the crystallographicc-axis 432.3 Stereoview of the coordination about the Cu(II) centre in [Cu(4-Clpz)2](green); 50% probability thermal ellipsoids are shown for the non-hydrogenatoms 442.4 Atom labelling scheme for the crystallographic data on the copper(ll) 4-X-pyrazolates: X = H(l) for [Cu(4-Hpz)2],CH3 (C(3), H(2), H(3), and H(4))for [Cu(4-Mepz)2],and Cl for [Cu(4-Clpz)2](green) 452.5 Powder X-ray diffractograms of (a) [Cu(4-Hpz)2J(b) [Cu(4-Mepz)2],(c)[Cu(4-Clpz)2j(green), and (d) [Cu(4-Brpz)2j 462.6 Powder X-ray diffractogram of [Cu(4-Clpz)2J(brown) 482.7 Electronic spectra of the Cu(II) 4-Xpyrazolate polymers: (a) X = H, (b) X =Me, (c) X = Cl (green), (a X = Br, (e) X = Cl (brown) 522.8 Powder EPR spectrum of [Cu(4-Hpz),] at 77 K 532.9 Powder EPR spectrum of [Cu(4-Clpz),] (brown) at 77 K 542.10 Powder magnetic susceptibility plots for the Cu(II) 4-Xpyrazolatepolymers: (a) X = H, (b) X = Me, (c) X = Cl (green), (d) X = Br, (e) X = Cl(brown). The lines are the calculated curves for the best fit values of theparameters listed in Table 2.5. The solid upper curves represent fits to thefull data sets while the lower curves represent fits to only the highertemperature data. The dashed curves represent the best fit values calculatedfrom the parameters in Table 2.6 572.11 (a) HOMO for the pyrazolate ion. (b) and (c): two different views of theoverlap of the magnetic orbital on copper with the pyrazolate HOMO 632.12 Powder X-ray diffractograrns of the Cu(II) 4-X-3,5-dimethylpyrazolatecomplexes: (a) X = H, (b) X = Me, (c) X = Cl, and (d) X = Br 702.13 Methyl group labelling scheme for discussion ofI3c-ci bands 72xvii2.14 Electronic spectra of the Cu(ll) 4-X-3,5-dimethylpyrazolate complexes: (a)X=H,(b)X=Me,(c)X=Cl,and(d)X=Br 742.15 EPR spectrum of [Cu(4-Hdmpz)2]at —.90 K 752.16 Possible trirnetallic structure for [Cu(4-Xdmpz)2]compounds 762.17 SEM image of [Cu(4-Hdmpz)21r 772.18 Stepped linear chain model for the [Cu(4-Xdmpz)2jcompounds 782.19 Powder susceptibility versus temperature plots for the Cu(I1) 4-X-3,5-dimethylpyrazolate complexes: (a) X = H, (b) X = Me, (c) X = Cl, (d) X =Br. The lines are the calculated best fit curves using the parameters inTable 2.9: solid lines, fits to high temperature data only; dashed lines, fitsto whole temperature range; dotted lines, fit to whole temperature rangewith a Curie-Weiss law paramagnetic impurity term 792.20 Isothermal magnetization data for [Cu(4-Cldmpz)2J 822.21 SEM image of [Cu(indz),J. The white line represents a length of 5 im 852.22 Powder diffractograms of (a) [Cu(3-Mepz)2Jand (b) [Cu(indz)2} 862.23 Electronic spectra of (a) [Cu(indz),J and (b) [Cu(3-Mepz)2] 872.24 EPR spectra of (a) [Cu(indz),1 and (b) [Cu(3-Mepz)2]at 77 K 892.25 Powder susceptibility versus temperature plots for (a) [Cu(indz),] and (b)[Cu(3-Mepz)2].The lines represent the best fits to theory and arecalculated from the parameters in Table 2.10. The dotted lines are best fitsto the high temperature data only. The solid lines are fits to the full dataranges and the dashed line in (a) is the fit with a Curie-Weiss lawpararnagnetic impurity term 913.1 ORTEP diagram of the two crystallographically independent molecules of[Cu(4-Hdmpz)]3;50% probability thermal ellipsoids are shown for thenon-hydrogen atoms 99xviii3.2 ORTEP diagram of [Cu(4-Medmpz))3;5Q% probability thermal ellipsoidsshown for the non-hydrogen atoms 1003.3 Stereoscopic view of [Cu(4-Hdmpz)]3showing the trimer-trimer Cu... Cuinteractions 1003.4 DSC thermogram of [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)] 1073.5 Stereoview of [Cu(3-COdmpz)(4-MedmpzH)J.,Cu; 33% probabilityellipsoids are shown for the non-hydrogen atoms 1083.6 Stereoview of the packing arrangement for [Cu(3-CO,dmpz)(4-MedmpzH)]2u1093.7 Stereoview of [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)21;(a) monoclinic form,(b) triclinic form 1103.8 Stereoview of the packing arrangement in [Cu(4-Br-3-CO2Mepz)-(4-BrdmpzH),],; (a) monoclinic form, (b) triclinic form 1123.9 Room temperature EPR spectrum of [Cu(3-CO,dmpz)(4-MedmpzH)]u 1133.10 Plot of versus temperature for [Cu4-Br-3-CO,Mepz)(4-BrdmpzH),]2.Lines represent the best fit to the data of the Bleaney-Bowers equation witha Curie law paramagnetic impurity term (dotted) line) and a Curie-Weisslaw paramagnetic impurity term (solid line) 1143.11 Bis-.t-pyrazo1ato- (N( 1) ,N(2))-bis[N,N-dimethylethanolamino)-(1 -pyrazolyl)gallato(N(2),O,N)copper(II)] 1193.12 Magnetic susceptibility versus temperature plot for [Cu(4-Hpz)(Me2Ga(4-Hpz)(OCHCH,N(Me))],.The solid line represents the best fit to the dataof the Bleaney-Bowers equation including a Curie law paramagneticimpurity term 1204.1 DSC thermogram of [Zn(4-Hdmpz)-,(4-HdmpzH)12 1254.2 Stereoscopic view of the [Zn(4-Hdrnpz)(4-HdmpzH)], molecule; 50%probability thermal ellipsoids are shown for the non-hydrogen atoms 126xix4.3 X-Ray powder diffractogranis of (a) [Zn(4-Hdmpz)2(4-HdmpzH)],(b)[Zn(4-Hdmpz)2],(c) the dlineric/oligometallic mixture 1284.4 Infrared spectra of (a) [Zn(4-Hdmpz)2(4-HdmpzH)1,(b) [Zn(4-Hdmpz)2],(c) the dimeric/oligometailic mixture. Spectra obtained from solid stateNujolmulls 1294.5 Stereoscopic ORTEP diagram of the [Co(4-Hdmpz)2Cl(4-HdmpzH)]omolecule; 33% probability thermal ellipsoids are shown for thenon-hydrogen atoms 1364.6 Stereoscopic view of the unit cell packing and intermolecular hydrogenbonding in the ciystal structure of [Co(4-Hdmpz)2C1(4-HdmpzH)J 1374.7 Stereoscopic view of the unit cell packing in the ciystal structure of[Co(4-Hdmpz)2(4-HdmpzH)1 1384.8 Stereoscopic ORTEP diagram of the [Co4(4-Hdmpz)601molecule; 33%probability thermal effipsoids are shown for the non-hydrogen atoms. Co(4)lies directly behind the oxygen atom 1414.9 Electronic spectra of (a) Co(4-Hdmpz)2.0.34 (4-Hd pzH) and (b)[Co(4-Hdmpz)2C1(4-HdmpzH)] 1454.10 Plot of effective magnetic moment versus temperature for[Co(4-Hdmpz)C1(4-HdmpzH)]o.The solid line represents the best fit ofthe trimetallic Heisenberg model with J andJ13 variable. The dashed linerepresents the same model with .113 fixed at zero. The dotted line representsthe preceding model coupled with a molecular field term 1514.11 Polymetallic Co(II) complexes prepared by Reedijk and co-workers. Theblack dots represent Co(ll) ions and the circles represent various alkylated1,2,4-triazoles 154xx4.12 Spin state energy level diagram for [Co(4-Hdmpz)2C1(4-HdmpzH)]ocorresponding to the best fit parameters of the trimetallic Heisenberg modelwith (•) J andJ13 variable and (D) J fixed at zero. The number besidelevel is the sum of the spins of the terminal Co(U) ions 1604.13 SEM images of (a) [Co(4-Hpz)2],(b) [Co(3-Mepz)2],(c) [Co(4-Hdmpz)2],(ci) [Co(4-Medmpz),]. The white bar in each image representsa1engthof5jtm 1654.14 Powder diffractogram of [Co(4-Hpz)2]recorded using a Cu KaX-radiation diffractometer 1664.15 Powder diffractograms of [Co(4-Medrnpz)2jwith the sample packing force(a) parallel and (b) perpendicular to the sample substrate 1684.16 Electronic spectra of the Co(II) (substituted)pyrazolates; (a) the 4-Xdmpzderivatives and (b) [Co(4-Hpz),] and [Co(3-Mepz),J 1714.17 Proposed structure for the Co(II) (substituted)pyrazolate polymers (whereR is H or Me and X is H, Me, Cl, or Br) 1774.18 Powder susceptibility versus temperature plots for the CcITI)(substituted)pyrazolate polymers: (a) uncorrected and (b) correctedsusceptibility values in the cases of [Co(3-Mepz)21and[Co(4-Medmpz),] 1794.19 Isothermal magnetization plots for (a) [Co(4-Medmpz)2]and (b)[Co(3-Mepz),] 1814.20 Diagram showing the preferred orientation of crystallites in samples usedfor the magnetic studies of the Co(I1) (substituted)pyrazolates 183xxi4.21 Oriented pellet susceptibility versus temperature plots for the Co(ll)(substituted)pyrazolate polymers: (a) [Co(4-Hdn1pz)21,(b) [Co(4-Medmpz)2],(c) [Co(4-Cldmpz)2],(d) [Co(4-Brdmpz)],(e) [Co(4-Hpz)2], and (f) [Co(3-Mepz)].The open circles represent the randomorientation and the solid circles represent the perpendicular orientation 1864.22 Suceptibility versus temperature plots for the Co(ll) (substituted)pyrazolatepolymers: (a) [Co(4-Hdmpz),J, (b) [Co(4-Medmpz)2],(c) [Co(4-Cldmpz)2j,(d) [Co(4-Brdmpz)2j,(e) [Co(4-Hpz)2],and (I) [Co(3-Mepz)2}.The lines are the calculated curves for the best fit parameterslisted in Tables 4.7 through 4.9. The solid lines correspond to theHeisenberg model and the dotted lines correspond to the Ising model 1905.1 SEM image of [Ni(4-Hpz)2].The white line represents a length of 2 p.m 2025.2 Powder X-ray diffractogram of [Ni(4-Hpz)21 2035.3 Electronic spectrum of [Ni(4-Hpz)2j 2045.4 Magnetic susceptibility plot versus temperature for Cu(4-Ipz)2.(4-IpzH).The solid line represents the best fit of the Heisenberg, 1-D,antiferromagnetic coupling model to the data in the 100-300 K range. Thedotted and dashed curves are the best fits of the same model to the full datarange including Curie and Curie-Weiss law paramagnetic impurity terms,respectively 2075.5 DSC thermogram of[Mo8(4-Hpz)6O21-HpzH].3(4- zH).H 2105.6 Two stereoscopic ORTEP views of[Mo8(4-Hpz)6018-HpzH];32%probability thermal ellipsoids are shown for the non-hydrogen atoms 2135.7 Stereoscopic unit cell packing diagram for [Mo8(4-Hpz)6018(4-HpzH)6j.2(4-HpzH showing the disordered solvate pyrazole molecules 2155.8 Stereoscopic ORTEP view of[Mo8(4-Hpz)6021-HpzH1;33%probability thermal ellipsoids are shown for the non-hydrogen atoms 216XXII5.9 Stereoscopic ORTEP view of the disordered pyrazole and water moleculesin [Mo8(4-Hpz)60,1(4-HpzH)61.3(4-HpzH).H2O 2165.10 Framework for the mixed-valence and single-valence octametallicmolybdenum clusters 217xxiiiLIST OF ABBREVIATIONS AND SYMBOLSA electron-nuclear hyperfine coupling constantA Angstromacac acetylacetonateAnal. analysisborbr broadB magnetic inductiondegree(s) Celsiuscalcd. calculatedcm centimetrescm1 wavenumber(s)D dimensional or deuteriumD axial zero-field splitting parameterDSC differential scanning calorimetryE rhombic zero-field splitting parameterEPR electron paramagnetic resonanceF fit quality factorg gram(s)g Landé splitting factorg1 parallel g componentg1 perpendicular g componentG Gaussh hour(s)H or H magnetic fieldH HarniltonianxxivHAM hydrogenic atoms in moleculesHOMO highest occupied molecular orbitalJR infrared3 Joule(s)I magnetic exchange coupling constant“13 next-nearest neighbour exchange constantk Boltzmann’s constantk wave vectorK KelvinL orbital angular momentumlit, literaturem mediumM molarM magnetizationMe methylmm minute(s)mL millilitre(s)MLPM molten ligand precursor/metalmmol millimole(s)mol mole(s)MW molecular weightN Avogadro’s numberMR near infrarednm nanometre(s)NMR nuclear magnetic resonanceOAc acetateOe Oersted(s)xxvORTEP Oakridge thermal ellipsoid plotPh phenylpz pyrazolyl moietypercent paramagnetic impurityR.T. room temperatures strong or singletS or S total spinsh shoulderSEM scanning electron microscopySQUID superconducting quantum interference deviceT temperatureTHF tetrahydrofuranTIP temperature independent paramagnetismUV ultravioletVSM vibrating sample magnetometervw very weakZFS zero-field splittinga. angle or mix factor13 angle, mix factor, or in-plane bendXcalc calculated magnetic susceptibilityXg gram magnetic susceptibilityXM molar magnetic susceptibilitypowder magnetic susceptibilityparallel magnetic susceptibilityXi perpendicular magnetic susceptibilityout-of-plane bendchemical shiftxxviAll change in enthalpyhapticity or Brillioun function parametervolume magnetic susceptibilitymagnetic permeability or bridgingor B.M. Bohr magnetoneff effective magnetic momentv vibrational or electronic banda density or standard deviatione Weiss constant or diffraction angleapproximately> greater thanless than> very much greater thanvery much less thanproportional toelectronic transitionxxviiACKNOWLEDGEMENTSI would like to express my sincerest thanks to Drs. A. Storr and R. C. Thompson fortheir support, guidance, and patience during the course of this work. I would also like tothank the members of my guidance committee, Drs. F. Aubke, C. Fyfe, and 3. R. Sams for theconstructive criticism they provided during the final preparation of this dissertation.Of course, most of the work described here would not have been possible without theexpert assistance of the men in the electmnics, glassblowing, and mechanical shops.My gratitude is also extended to Mr. P. Borda for microanalytical services, Drs. S. 3.Rettig and 3. Trotter for crystal structure determinations in this department, Drs. R. Batchelorand F. W. B. Einstein at Simon Fraser University for a low temperature crystal structuredetermination, and Mr. N. Osbourne of the Physics Department at U. B. C. for assistance inobtaining powder diffraction data. I am also thankful to Drs. F. G. Herring and P. S. Phillipsfor their assistance in obtaining EPR spectra and to Drs. W. Hardy and 3. Carolan of thePhysics Department at U. B. C. for use of their SQUID magnetometer.I also thank my colleagues, both past and present, Mr. T. Otieno, and Dr. 3. Du, forassistance and many fruitful discussions and Mr. M. Olson and Mrs. T. Olson for theirthorough proof-reading of this work.Finally, I extend special thanks to my family and to Ms. S. Giles for her enduringsupport and encouragement during the course of this work.xxviiiCHAPTER 1IITRODUCTIONThe work which constitutes this dissertation lies in an area which crosses theboundaries of physics, metallurgy, chemistry, and, more recently, biology. The goal of thiswork was to synthesize new paramagnetic materials which would possess the molecularstructural motif of the linear chain and to examine the magnetic properties of those newcompounds with a view to interpreting the observed properties in terms of their structures andcompositions. Binary transition metal pyrazolates were chosen as materials appropriate forthis study. Historically, magnetism has long been the province of physicists. As testament tothis, the vast majority of Nobel prizes awarded for works which have made fundamentalcontributions to the understanding of magnetism have been awarded in physics. Thechemist’s contribution to studies of magnetism has become increasingly important in the pastfew decades as theories of magnetism are tested through the synthesis and characterization ofnew magnetic materials (1-4). This opening chapter is devoted to: an introduction to theconcept of dimensionality in matter and its relevance to magnetic properties, a description ofthe physical basis of magnetism with particular emphasis on those concepts and theorieswhich relate directly to the magnetic studies undertaken in this work, a review of binarytransition metal and post-transition metal pyrazolate compounds, and a discussion of thescope of this work and the methods of characterization employed.1.1 LOW-DIMENSIONAL MATERIALSDimensionality, as it pertains to the study of materials, is a designation of theanisotropy in compounds. The chemist’s principal interest in dimensionality is its use in the1description of bonding anisotropy. Physical space permits the existence of four orders ofdimensionality. Zero-dimensional (O-D) compounds are, in number, the most common typeof compounds studied by chemists. Such species are more commonly referred to as molecularspecies; they possess what might be termed “local” connectivity. Since such compoundspossess no extended bonding in any of the spatial directions they may be representedgeometrically as points. l-D compounds possess extended connectivity in one spatialdirection, in other words they are chains. 2-D and 3-D compounds possess extendedconnectivity in two and three spatial directions, thus forming sheets and solid volumes,respectively. By way of example, carbon forms substances which exist in all four orders ofdimensionality: methane is O-D, polyethylene is l-D, graphite is 2-D, and diamond is 3-D.What has been described above is idealized structural low-dimensionality. Condensedmatter does not exist as true 0-, 1-, or 2-D substances. The delicate and microscopic nature ofa single molecule, chain, or sheet would thwart any attempt to prepare or examine in thelaboratory such a specimen. Thus, 0-, 1-, and 2-D compounds occur and are examined only asthree-dimensional aggregates such as crystals, films, etc. It is still resonable to describe suchmaterials as low-dimensional because, although the compounds occur as three-dimensionalaggregates, the individual molecules, chains, or sheets are held together primarily via van derWaals interactions.The dimensionality of bonding in a solid is often reflected in the dimensionality of itsphysical properties, however, this need not always be the case. KCuF3 crystallizes in thestructure, depicted in Figure 1.1, which is a slightly tetragonally distorted version of theperovskite structure (5). It is apparent that this compound has a 3-D [CuF3j bonding network,surprisingly however, the material behaves magnetically as a 1-I), or linear chain, system (6).This behaviour has been ascribed to a unique orienting of the magnetic copper(II) d-orbitalsin such a way as to prevent interaction of the magnetic orbitals in the plane perpendicular tothe c-axis of the crystal (6).2a0 FFIG. 1.1. Structure of KCuF3.The solid lines indicate the pseudo-perovskite lattice andthe dotted lines indicate the actual unit cell. CuF chains run along the c-axis direction. AfterOkazaki and Suemune (5).The study of low dimensional materials has blossomed in the last three decadesrevealing interesting conductivity, optical, intercalation, and magnetic properties (7,8). 1-Dand 2-D compounds are of particular interest in magnetochemistry because, from the point ofview of theoretical analysis, they are simpler systems to treat quantitatively than 3-Dsystems, yet they still possess extended connectivity. A fair amount of effort has beendirected towards developing quantitative models to describe the behaviour of such systems,particularly in the case of 1-D systems (some of these models are discussed below).Experimental and theoretical studies of low-dimensional materials and their magneticproperties interact synergically, furthering the understanding of magnetic phenomena.1.2 MAGNETISMMankind has been aware of magnetism for thousands of years. The ancients hadobserved that lodestone, Fe304 could align itself with respect to certain directions on theEarth and that separate pieces of lodestone could attract or repel one another depending on3how they were arranged (9). This long held empirical knowledge did not expand into anunderstanding of the fundamental origins of magnetism until the nineteenth, and especially,the twentieth centuries. In a broad sense, everything is magnetic because all substancespossess diamagnetism which is the response of moving paired electrons in an atom to theapplication of a magnetic field. The magnetism which is of primary concern here, and indeedthe magnetism which is the source of the behaviour of lodestone, is paramagnetism.Paramagnetism derives from the angular momentum of unpaired electrons. This electronangular momentum comes from two sources: orbital angular momentum, L, and a property ofthe electron known as spin, S.Early in this century, during the birth of quantum mechanics, experiments in atomicspectroscopy suggested that the electron had intrinsic angular momentum which is called spin(10). Electron spin has a magnitude JIh and possesses the quantum number . The name“spin” provides the connotation that spin angular momentum derives from rotatioral motionof the electron about its own axis. This metaphor for spin is sadly inadequate because theeigenfunctions for angular momentum due to rotation require integer quantum numbers. Theelectron spin has a half-integer quantum number so representation of spin by a classicalnotion such as electron rotation can be misleading. Spin is a purely quantum mechanicalphenomenon. Dirac’s effort to combine quantum mechanics with the laws of relativitydemonstrated that spin was a necessary consequence of the integration of those theories(10,11).It was stated above that paramagnetism is a consequence of the angular momentum ofunpaired electrons in compounds. Paramagnetic transition metal complexes often possesselectronic configurations or low-symmetry geometries such that the orbital angularmomentum in these systems is partially or wholly “quenched”, thus the paramagnetismobserved is primarily due to spin angular momentum. This being the case, models used toaccount for paramagnetism are frequently based only on spin properties and interactions inorder to simplify development of the models. This having been stated, one must not be4cavalier in disregarding orbital angular momentum when interpreting the magnetic behaviourof systems. Some complexes will possess non-zero L and this will complicate interpretationof the magnetic properties of these compounds.Before going on to describe how spin gives rise to observable magnetic properties it isuseful to consider what is observed when examining magnetic materials. When a substance isplaced in a magnetic field, H, it becomes magnetized, that is, magnetic dipoles in thesubstance, be they induced or permanent dipoles, align along a common direction and a netmagnetic moment forms in the substance. The intensity of magnetization possesses both amagnitude and a direction so, like H, it is a vector quantity and is represented as M. H and Mneed not be collinear, but for the moment it is assumed that the substance being magnetizedis magnetically isotropic, in which case M and H are always collinear. The magnetization ofthe substance results in an induced magnetic field in the substance, referred to as themagnetic induction and denoted as B. The relationship between these quantities can beexpressed asB=H+4itM, [1.1]which may be recast asB=i.LH [l.2aj[l.2b]where .t is the magnetic permeability. From the expression for ii. and [1.1] it is apparent that-. [1.3]ic is the volume magnetic susceptibility and it is a dimensionless quantity. A moreconvenient quantity from the magnetochemist’s standpoint is the molar magneticsusceptibility defined asXM= [1.4]where a is the density of the substance and MW is its molecular weight, thus XM has the unitscm3mol’ in the CGS system.5At this point it is useful to consider the limitations of the expressions for magneticproperties described above. Empirically it has been observed that XM oc M/H for diamagnetsand paramagnets so long as H is not too large and the temperature of the substance is not toolow (for typical laboratory fields this means temperatures >1 K). However, for ferromagneticsystems M is generally not linearly dependent on H except in the limit as H tends to zero.This study is concerned with substances that are occasionally diamagnetic, but generallyparamagnetic so it will be assumed that the linear relationship between M and H holds forthese compounds. It was stated earlier that the equations above apply to magneticallyisotropic substances. Most crystalline magnetic materials are not completely isotropic; insuch cases M and H need not be collinear. Because XM relates H, a vector of action, to M, avector of effect, and these two vectors are not required to share the same direction, XM maybe represented as a tensor. In the case of “moderate” temperatures and applied fields thesubstances of concern in this study show a linear relationship between M and H, so, morespecifically, XM may be represented as a tensor of second-rank,Xii X12 X13X21 X22 X23 [1.5]X31 X32 X33The susceptibility tensor, XM is symmetric about the diagonal, i.e.[1.61and the elements of the tensor lying along the diagonal, Xu’ are termed the principalsusceptibilities. In anisotropic magnetic materials the principal susceptibilities are uniquebecause it is only along these directions that both H and M are truly collinear. XM is thecrystal susceptibility tensor and, not surprisingly, total determination of this tensor requires asingle crystal of the material of interest. Unfortunately, it is often difficult, if not impossibleto obtain single crystals large enough for susceptibility measurements. Such crystals need tobe much larger than crystals suitable for single crystal X-ray diffraction and, as discussed6below, for many of the compounds prepared and studied in this work, even crystals suitablefor X-ray diffraction studies were elusive. In cases where single crystals are not available it isthe average, or powder susceptibility,x = 1, [X11X22X33] [1.7]which is determined for the compounds using a randomly oriented, microcrystalline, orpowdered sample (throughout this dissertation, will refer to measured powder magneticsusceptibilities and Xcalc will refer to theoretically modelled susceptibilities). All of thestudies conducted in this work were performed on powdered samples and as can be seen from[1.7], all of the anisotropy information of the tensor is lost. This does not mean, however, thatthe discussion of the tensorial nature of XM was entirely superfluous because, as will bediscussed in Chapter 4, partial resolution of the magnetic anisotropy in cobakII) pyrazolateswas achieved.In order to treat quantitatively the susceptibility data obtained from magnetic studies,some functional expression, or system of expressions is necessary. Because magnetism isquantum mechanical in origin such expressions are necessarily derived from quantummechanics. Van Vieck (12) developed an equation that, when solved for a particularinteraction Hamiltonian and eigenvector basis set, gives a closed form expression formagnetic susceptibility,NL{[Ej(’)/kT]2Ej(2)}e=ej()1kT[1.8]7where are the energies of the zero-field eigenvalues of the spin Hamiltonian,= <N13JHPIN1j> [1.9aJare the first-order Zeeman energies, and‘‘ I<NIkIHI.tjlwj>12= .i.d E(°)-Ek(°)[1.9b]kjare the second-order Zeeman energies. The are the eigenvectors corresponding to theenergy eigenvalues E(°). The Van Vieck equation is valid when gjiH ‘t kT and is meant toapply to magnetically isotropic materials (11,13).The spin Hamiltonian provides a means for mapping the energy levels of a magneticsystem in order to incorporate those energy levels in the Van Vieck equation and thusdetermine the temperature dependence of the susceptibility. Figure 1.2 shows a fairlycomprehensive, though by no means complete, spin Hamiltonian for magnetic insulatingcompounds (11,13,14). The figure illustrates that the components of the spin Hamiltonianmay be grouped into two categories: terms which affect single magnetic ions and termswhich account for the interactions between ions. A brief consideration will be given to thesingle-ion effects here.H7e , the Zeeman term of the Hamiltonian, accounts for the interaction of a spin (orspin plus orbit) system with a magnetic field. H1 ,the spin-orbit coupling component of theHamiltonian is especially important in spin systems with a T ground state as such systemspossess non-zero orbital angular momentum. Spin-orbit coupling is also the mechanism viawhich a ground state, whether it possesses orbital angular momentum or not, may couplewith excited states which possess orbital angular momentum.‘1zfs accounts for zero-field8Hspin__[ Spin HamiltonianHzee = J.t(kL+gS).H Zeeman effect+ llntraSpin-orbit coupling ioniceffects+= D[S2- ‘ S(S+1)] + E(Sx2Sy) Zero -field splitting+Hex = 2JE Exchange1J+InterHb1 Jjj E(S.S) Biquadratic exchange ionicIJeffects+HDM =E DJ(SAS) Dzyaloshinsky-Moriya exchangelJFIG. 1.2. The components of the spin Hamiltonian.splitting, due to lower than cubic symmetry ligand fields, of the ±m states for spin systemswithS 1.The second category, interion effects, which is of particular interest in this study willbe discussed at length in the following section so it is not considered here. What is importantto recognize is that if it were necessary to apply the spin Haniiltonian as it stands in Figure1.2 to a given spin system, it would be an extremely formidable, if not impossible task: withall terms present the Hamiltonian is simply too complicated to handle. Even if solution of thetotal spin Hamiltonian was tractable and all of its contributions present in a material, it isdoubtful that sufficient experimental information could be acquired to uniquely determine all9of the values of the parameters for those terms. Fortunately, in many real magneticsubstances only some of the terms in the spin Hamiltonian make non-negligiblecontributions, thus simplifying the situation. For example, Cu(II), which is an S = 1/2 ion, nomatter what its coordination geometry, never experiences zero-field splitting effects. As wasmentioned earlier, the orbital angular momentum of first row transition metal ions is oftenwholly or partially quenched thus significantly reducing the importance of spin-orbitcoupling. With regard to interion effects, biquadratic exchange is often very small inmagnitude and Dzyaloshinsky-Moriya exchange has symmetry restrictions which are notoften met thereby precluding its occurrence in compounds. This leaves us with the Zeemaneffect which is fairly easily handled and magnetic exchange, which was a feature sought inpreparation and study of the compounds discussed in this work.1.2.1 MAGNETIC EXCHANGEIn an ideal paramagnetic substance the orientations of the unpaired spins present aretotally uncorrelated. Such a situation is illustrated schematically in Figure 1.3(a) for a planarmagnetic net where the arrows represent the orientation of a given spin vector. The converseis that in a non-ideal paramagnetic substance the unpaired spins are correlated to some degreeand in some particular fashion. Two archetypal examples of spin correlation areferromagnetism and antiferromagnetism; these are illustrated in Figures 1.3(b) and (c)respectively. It is magnetic exchange which causes correlation of unpaired spins in amagnetic material and the examples given in Figures 1.3(b) and (c) are but two of the manytypes of spin correlation observed in nature and predicted by theory. It is important tounderstand what kinds of exchange are possible in order to appreciate the rich diversity ofmagnetic phenomena and the challenges posed to the magnetochemist in uncovering thosephenomena in nature and interpreting them.Consider two equivalent, isotropic, S = 1/2, ions suspended in close proximity in a10‘A/_17,7/17(a)/// / //,// //(b)_______ A/%A/(c)FIG. 1.3. Spin orientations in a planar net array; (a) paramagnetism,(b) ferromagnetism, and (c) antiferromagnetism.crystal lattice and that some, as yet unspecified, mechanism permits magnetic exchangebetween the centres. This exchange manifests itself as correlated spin orientation between thetwo magnetic centres, but what are the possible correlations? To help answer this question leta Cartesian coordinate frame be imposed on the pair of ions. This reminds one that the spinvector on each ion may be decomposed into its Cartesian components S, Sb,, and S. Theexchange term of the total spin Hamiltonian is presented again because of its importance to11this discussion.Hex = -2JE [1.10]lJIt is apparent that the Hamiltonian is presented with the spin components explicitly shown. aand f may be considered “mix factors” which can independently vary between the values ofzero and one. There are three limiting cases of the values of a and f3 which describeimportant models for magnetic exchange. When a = = 1 the Heisenberg model is obtained.In Heisenberg systems all three components of the spin vectors couple equally, in otherwords, the Heisenberg model is an isotropic model. When a = 1 and 3 = 0 the Ising modelobtains and in this case only the S components of the spin vectors couple. In the third case, a= 0 and 3 = 1 giving rise to the planar, or X-Y model, in which case it is the only the S, andS,, components of the spin vectors which couple. Consequently, these latter two models areanisotropic. In the exchange Hamiltonian, J is the exchange, or coupling, constant analogousto coupling constants in nuclear magnetic resonance spectroscopy. The magnitude of J is ameasure of the strength of the exchange interaction and its sign determines the nature of thecoupling. If J>0 or J<0 the coupling is ferromagnetic or antiferromagnetic, respectively.There is another important constituent of [1.10] which influences the nature of spininteraction in a magnetic material, that is the summation. Generally, exchange interactionsare strongly dependent on the distance between magnetic centres and vary as --r10, where “r”is the interspin distance (1). It is apparent then that the most important interactions betweencentres in a magnetic lattice will be those between nearest neighbours. Furthermore, thesummation is dependent on the dimensionality of the system under consideration; magneticchains have distinctly different properties from magnetic sheets or 3-D lattices.If all permutations of the components of [1.10] discussed above are considered onerecognizes that a very large number of magnetic exchange models is potentially available tothe magnetochemist. The availability of models is limited by the fact that in many cases no12—/7 ‘..F—T=OKdecreasing TFIG. 1.4. The approximate susceptiblity magnitude ellipsoid in an antiferromagneticHeisenberg chain.analytic solution to a particular Hamiltonian exists and often even the calculation ofnumerical solutions is too difficult a task to be undertaken. Fortunately for this work, areasonable number of expressions for magnetic models have been developed for 1-D systems.Figures 1.4 to 1.6 represent the qualitative temperature variation of the susceptibilitymagnitude ellipsoid (analogous to ORTEP thermal motion representations) for Heisenberg,Ising, and X-Y, 1-D antiferromagnetically exchange coupled systems. In these figures asegment of the chain is shown in the background where the black and white spheres representthe alternating orientation of the spin vectors, the hallmark of antiferromagnetic interaction.Attention is focused on a magnetic centre by illustrating the temperature dependence in theforeground, but it is important to realize that the ellipsoids represent the bulk susceptibilitiesof the chain compounds, not merely a single magnetic centre. Figure 1.7 illustrates themagnetic susceptibility versus temperature behaviours for the systems illustrated in thepreceding three figures. It is apparent from Figure 1.7 that the dimensionality of the spinexchange can have profound effects on the temperature dependence of the susceptibility. One13-../ 1J%/ .4I %/.4 S.—/ I ../ I 4.—.../ I .4/ .4 ——.4 __4_I 5—___T=OKdecreasing TFIG. 1.5. The approximate susceptiblity magnitude ellipsoid in an antiferromagneticIsing chain.•o •V•Q’._/4-—..4 —.4.I ——/ I .4% 4..—I I 4-/ I 4-4%I .4.T=OKdecreasingTFIG. 1.6. The approximate susceptiblity magnitude ellipsoid in an antiferromagnetic X-Ychain.140.30.2x!1Ng20.10.0FIG. 1.7. Magnetic susceptibility versus temperature in reduced coordinatesfor S = 1/2, 1-D Heisenberg, Ising, and X-Y antiferromagnetic coupling. Afterde Jongh and Miedema (1).can see that the anisotropic Ising and X-Y models yield susceptibilities which are themselvesanisotropic; and Xj refer to the principal susceptibilities of the susceptibility tensor paralleland perpendicular to the coupled spin component direction, respectively (it should be pointedout that the parallel and perpendicular directions in these models need not correspond to thechain axis as the principal direction). The temperature dependence of x11 for the X-Y modelhas not yet been determined. Another striking feature illustrated by these diagrams andsomething which is characteristic of many low-dimensional systems is that long range orderis not achieved at finite temperatures in these models. In other words, the magneticsusceptibility does not go to zero as the temperature goes to zero. The Heisenberg, Ising, andX-Y exchange models in 1-D systems are characterized by short range order only, that is,0 1 2 3 4 5J15only finite lengths of the chains have their spins correlated.Up to this point, the possible mechanisms which facilitate magnetic exchange havenot been considered. The spin Hamiltonian is to magnetism as rate laws are to chemicalkinetics. In other words, the spin Hamiltonian allows quantitative treatment, within aquantum mechanical framework, of the magnetic behaviour of a given system, but it does notyield insight into the underlying mechanism(s) governing that behaviour. There are severalpossible sources of spin interaction in magnetic materials and the most important of theseare: dipolar coupling, direct exchange, and superexchange.Dipolar coupling is a concept of classical origins in which paramagnetic centres arecorrelated because their respective spins are dipoles and have associated local magnetic fieldswhich influence the orientations of neighbouring dipoles. Dipolar effects are proportional tothe square of the magnetic moment of the magnetic centre and inversely proportional to thecube of the inter-magnetic centre separation. The upshot of these behaviours is that dipolarcoupling effects become important only in materials in which the magnetic centres have largemagnetic moments (e.g., Mn2 Fe3, some lanthanides and actinides) and/or the magneticcentres are in close proximity (e.g., metals) (1).Direct exchange involves the direct overlap of unpaired electron containing orbitalson neigbouring magnetic centres. This phenomenon is important in itinerant magnets such asiron metal, Alnico alloys etc. As with dipolar exchange, direct exchange requires smallseparations, generally less than 3 A, between the magnetic centres.The magnetic compounds studied in this work are insulating transition metalcoordination complexes in which metal ions are separated by intervening diamagnetic ligandsand the interion separations in these materials are generally quite large, certainly greater than3 A. Thus the two spin coupling mechanisms described above are unlikely to makesubstantial contributions to the magnetic properties of these compounds. It is the thirdmechanism listed above, superexchange, which is dominant in mediating the interactionbetween magnetic ions in such insulating materials. Superexchange, the concept that16interposing non-magnetic moieties, be they atoms or molecules, can act as agents for spincoupling between magnetic centres, was first proposed by Kramers in 1934 (15). In 1950,Anderson presented a powerful theory, based on Kramers’ ideas, to explain the mechanism ofsuperexchange (16). Later on in that decade, Goodenough (17) and Kanamori (18) developedrules to predict the sign of the exchange from knowledge of the electron orbitals and theirsymmetries in the magnetic system under consideration. A basic notion which stems fromAnderson’s theory is that there are two basic types of exchange interactions present insuperexchange: kinetic exchange and potential exchange. Kinetic exchange occurs whenthere is a nonorthogonal orbital pathway via a bridging ligand linking magnetic centres andthis gives rise to antiferromagnetic coupling between the spins. This process is illustrated inFigure 1.8(a). Potential exchange occurs when an orthogonality exists in the pathwayFIG. 1.8. Orbital irteractions in superexchange; (a) kineticexchange, (b) potential exchange.between the spin centres and this gives rise to a ferromagnetic interaction between the spins.Potential exchange is illustrated in Figure 1.8(b). It is apparent that the strength and nature ofexchange coupling in systems where superexchange is the source of the coupling, is stronglydependent on the symmetry and geometry of orbital overlaps between bridging ligands andmagnetic ions. At the risk of being overly pedantic, this means that the magnetic properties of(a)(b)17such materials are heavily dependent upon their specific compositions and structures. Thisstatement lies at the heart of magnetochemistry.Magnetochemists endeavour to find correlations between the structures andcompositions of paramagnetic materials and their resultant magnetic properties. To this end,vast numbers of compounds, primarily transition metal complexes, have been synthesizedarid subjected to study. Comprehensive reviews of such studies have been published(1,3,4,19,20).It is hoped that the preceding treatment of magnetic properties and theoreticaldescriptions of those properties, in spite of its very superficial nature, provides someperspective for magnetochemical studies. Many of the references cited in this section provideexcellent, detailed discussions of the subjects mentioned here and there are other texts whichconsider these topics further and cover areas of magnetic behaviour not mentioned here(21,22). There are other aspects of magnetism which are relevant to this work, but in order toavoid nestling the reader in the arms of Morpheus, introduction of those concepts will bemade at the necessary points in the dissertation.1.3 PYRAZOLEPyrazole is an aromatic, five- membered, nitrogen containing heterocycle, 1, whichwas first synthesized by BUchner in 1889 (23). The pyrazole moiety is quite resistant tohydrolytic or oxidative degradation and the chemistry of pyrazole has been well developedsince Btichner’s original synthesis of the compound (24,25). Pyrazole and many pyrazolederived compounds were employed in the present work. Throughout this dissertation,systematic abbreviations are used to denote these pyrazolyl moieties and these abbreviationsare listed in Table 1.1.18H H1TABLE 1.1. Abbreviations for pyrazole and pyrazole derivedmoietiesCompound Abbreviation’pyrazole 4-HpzH3-methylpyrazole 3-MepzH4-methylpyrazole 4-MepzH4-chioropyrazole 4-ClpzH4-bromopyrazole 4-BrpzH4-iodopyrazole 4-IpzH4-nitropyrazole 4-NO2pzH3,5-dimethylpyrazole 4-HdmpzH3,4,5-trimethylpyrazole 4-MedmpzH4-chloro-3,5-dimethylpyrazole 4-CldmpzH4-bromo-3,5-dimethylpyrazole 4-BrdmpzH4-iodo-3,5-dimethylpyrazole 4-IdmpzH4-nitro-3,5-dimethylpyrazole 4-NO2dmpzHindazole indzHagimilar abbreviations are used for the corresponding pyrazolate moieties exceptthat the terminal’H” in the abbreviations is omitted.5 3N N2H191.4 PYRAZOLYL CONTAINING TRANSITION METAL COMPOUNDSThe pyrazolyl moiety has an extensive transition metal coordination chemistry andthe development of this chemistry, up to 1984, has been thoroughly discussed by Trofimenkoin two review papers (26,27). A more recent review of aromatic nitrogen heterocycles asbridging ligands covers some of the developments in pyrazolate transition metal chemistrysubsequent to Trofimenko’s reviews (28).The pyrazolyl moiety occurs in transition metal complexes of three general forms:1. Complexes involving pyridine-like coordination of pyrazole or substituted pyrazoleto the metal through N(2).2. Complexes involving N(2) coordination of the pyrazole to the metal centre, butwith the pyrazolyl moiety bonded as a functional group to a larger, polydentate ligand. Thepyrazolyl group is usually bound in the ligand via N(1) and the ligand may be neutral ornegatively charged.3. Complexes involving pyrazole or C-substituted pyrazole in its deprotonated form,the pyrazolate anion. In such cases the pyrazolate anion acts almost exclusively as a 1,2-exobidentate ligand.Binary transition metal pyrazolates, which fall into the last category, are the focus of thiswork and so it is appropriate to review the history of such compounds.1.4.1 BINARY TRANSITION METAL PYRAZOLATESThe first report of a binary transition metal pyrazolate appeared in the literature justover a century ago when Büchner prepared Ag(I) pyrazolate, [Ag(4-Hpz)] (23). To this day,the structure of this compound is not known. A more recent study of [Ag(4-Hpz)} suggestedthat the compound is at least trimeric, but that a polymeric structure cannot be excluded (29).20Büchner also prepared a series of [Ag(pz*)] compounds (pz* = 4-Brpz, 4-Ipz, and 4-NO2pz)(30). After Büchner’s initial foray into the field of transition metal pyrazolates the area laydormant for 35 years until Fischer noted that solutions of cobalt, iron, and zinc formedprecipitates upon addition of 3,5-dimethylpyrazole (31). In 1930, Heim employed 3,5-dimethylpyrazole in the determination of cobalt by precipitation from basic aqueous solution(32). This interest in 3,5-dimethylpyrazole as a reagent for the gravimetric determination ofcobalt was more thoroughly pursued in the late 1950’s by Pflaum and Dieter (33) whodetermined that the blue-violet precipitate formed from Co(II) salts and 3,5-dimethylpyrazolein basic aqueous solution had the formulation [Co(4-Hdrnpz)2]These workers rationalizedthe observed insolubility of the compound by proposing that it possessed a sandwichstructure akin to that of cobaltacene. There was no corroborating evidence to support theirproposed structure and although the structure of the compound is still not known withabsolute certainty, a sandwich-type structure for [Co(4-Hdmpz)2]is highly improbable. Thestudies of transition metal pyrazolates listed so far were concerned with the use ofpyrazolates as reagents in analytical determinations; little importance was placed ondiscovering the underlying structures, reactivities and properties of these compounds. Thepaucity of studies on those aspects of transition metal pyrazolates may have been due, in part,to the intractable nature of these compounds. The transition metal pyrazolates are generallyinfusible, involatile, and insoluble solids which makes probing the atomic nature of thesecompounds a daunting task.The next report of transition metal pyrazolates appeared in 1967 (34). Seel andSperber described the synthesis of Fe(II) pyrazolate, [Fe(4-Hpz)2],and Fe(llI) pyrazolate,[Fe(4-Hpz)3], from dicarbonylcyclopentadienyliron dimer anddicarbonylcyclopentadienyliron chloride, respectively, and pyrazole in benzene, toluene, ormesitylene. This report appeared as a communication, so neither analytical nor spectroscopicdata were provided for these compounds. No subsequent report on these compounds, by the21authors, has appeared in the literature.Cobalt pyrazolates again became a subject of study in 1970 when Bagley, Nicholls,and Warburton published a study of pyrazole, and pyrazolate complexes of Co(ll) (35). Theyprepared [Co(4-Hdmpz)2Jfrom Co(II) acetylacetonate, [Co(acac)2]4 and hydrazine inisopropanol. addition, Co(II) pyrazolate, [Co(4-Hpz)2],Co(II) 3-methylpyrazolate, [Co(3-Mepz)2], and [Co(4-Hdmpz)2]were prepared from Co(ll) salts and the appropriate pyrazolein slightly basic solution. Infrared and electronic spectroscopic studies were conducted on thecompounds and r•om temperature magnetic moments were measured from which the authorssuggested that the compounds possess polymeric structures in which the pyrazolate ions actas bridging ligands and the Coal) cations are tetrahedrally coordinated. In his 1972 reviewarticle, Trofimenko reported, either as unpublished results or private communications fromW. Mahier, on the synthesis of [M(4-Hpz)2](M = Cu, Ni) via pyrolysis of the correspondingM(4-HpzH)(OAc)2(OAc = acetate) complex or reaction of the appropriate metal salt withpyrazole in aqueous ammonia (26). Trofimenko also reported the preparation of similarchelates from 4-bromo-, 3,4-dibromo-, 3,4,5-tribromo-, 4-methyl-, 4-isopropyl-, and 4-cyanopyrazole and that specifically in the case of [Cu(3,4,5-Brpz)21the compound exists asa tetramer in benzene solution (26). In 1973, Singh, Satpathy, and Sahoo reported thesynthesis and characterization of [Co(4-Hdmpz)j,., Ni(ll) 3,5-dimethylpyrazolate dihydrate,[Ni(4-Hdmpz)2.2O], and Cu(l) 3,5-dimethylpyrazolate, [Cu(4-Hdmpz)] (36). Thecompounds were prepared by condensation of hydrazine with the appropriate divalent metalacetylacetonate in water (in the case of Cu(ll) reduction of the copper took place). Theauthors presented infrared and electronic spectroscopic data for the compounds, analyticaldata (it should be noted that their calculated and found analytical values for [Cu(4-Hdmpz)jare incorrect and actually correspond to a compound with the formulation [Cu(4-Hpz)]), andvariable temperature magnetic susceptibility data for the Coal) and Ni(II) complexes. In thesame year, Okkersen, Groeneveld, and Reedijk synthesized the series of complexes,[M(pz*) (where M = Cu, pz* = 4-Hpz, 3-Mepz, 4-Brpz, 4-Ipz; M = Ag, pz = 4-Hpz, 3-22Mepz, 4-Brpz, 4-Ipz, 4-NO2pz, 4-Etpz) (29). They provided analytical, JR and massspectrontry data on the complexes.Beginning in 1977, a series of papers was published by Vos and Groeneveld ondivalent and trivalent transition metal and post-transition metal pyrazolates (37-41). Theauthors reported the synthesis of 54 metal pyrazolates of which 37 were strictly binary metalpyrazolates. Vos and Groeneveld prepared compounds using metal ions of Mn, Co, Ni, Cu, Zn,and Cd and pyrazolate anions of 4-HpzH, 3-MepzH, 4-HdmpzH, 4-ClpzH, 4-BrpzH, 4-IpzH,and 4-NO2pzH. The metal pyrazolates were prepared by combining aqueous solutions of thechloride or nitrate metal salts with aqueous solutions of the pyrazoles made basic with sodiumhydroxide or ammonia. The compounds that were prepared are listed in Table 1.2.Table 1.2. Metal pyrazolates prepared by Vos and Groeneveld (37-41)Metal IonPyrazolate Co2 Co3 Ni2 Cu2 Zn2 Ag’ Cd24-Hpz .a • . .3-Mepz4-Mepz • •4-Hdmpz4-Clpz I I4-Brpz I I I I4-Ipz • • I • I4-NO2pz I I IaThe dot indicates that the compound was prepared by Vos and Groeneveld.Characterization of these compounds consisted of infrared, Raman and electronicspectroscopy studies, thermogravimetric measurements, and elemental analyses. Careful23examination of the analytical results presented by Vos and Groeneveld for the 37 binarymetal pyrazolates yields the following observation: 24 of the compounds were subjected onlyto metal analyses and of those the results for 18 compounds lie outside the accepted range(±0.3% absolute deviation); four compounds were subjected to metal, C, H, and N analysesand of those the results for three compounds lie outside the accepted range; nine compoundswere subjected to total elemental analysis and of those the results for eight compounds lieoutside the accepted range. It is doubtful whether many of the samples prepared wereactually pure substances.In 1977, Blake et al. studied the reaction of nickelocene with several pyrazoles inbenzene (42). It was found that both pyrazole and 3-methyl pyrazole react with nickeloceneto form polymeric species with the nominal formulations(C5H)[Ni(4-Hpz)2]5Ni(Cand(C5H)[Ni(3-Mepz)2J9Ni(C respectively. In 1982, Strähle et a!. briefly described thesynthesis of [Fe(4-Hpz)2]from anhydrous FeC13 and lithium pyrazolate in toluene (43). Atthe same time Seelig reported the results of extended HUckel calculations on a polymericchain system of Fe(llI) ions triply-bridged by pyrazolate ligands (44).In the last decade, a number of univalent, coinage metal pyrazolates have beenprepared and structurally characterized. Bonati and co-workers have reported on thesynthesis and structure determination by single crystal, X-ray diffraction of ths[.t-3,5-bis(trifluoromethyl)pyrazolato-N,N’}thgold(I), [Au((CF32pz)1 (45,46). This compound is atrimetallocycle with a central nine-membered ring and each Au(I) centre is linearlycoordinated by two pyrazolate ligands. Other compounds, recently synthesized, whichpossess the same structural motif are [Cu(Ph2pz)j3(Ph2pz = 3,5-diphenylpyrazolate) (47),[Ag(Ph2pz)13 (26), and [Au(Ph2pz)13 (48). In addition, the hexametallic compound[Au(Ph2pz)]6has been prepared and structurally characterized (48).Although, as described above, research into binary, transition metal pyrazolates spansa century, relatively few such compounds have been prepared for which purifies have beenunambiguously ascertained and characterization was thorough. Even fewer of those24compounds have been subjected to single crystal X-ray diffraction studies and, prior to thecurrent work, in the case of divalent metal pyrazolates, not one such study has beenundertaken. Except in the case of [Co(4-Hdmpz)2],studies of the magnetic properties ofsuch compounds have been limited to the determination of room temperature magneticmoments and in no cases have quantitative analyses been undertaken.1.5 SCOPE OF THIS STUDY AND ORGANiZATION OF THE DISSERTATIONNumerous studies of low dimensional paramagnetic compounds have been conductedin the past. These studies can be grouped into two broad categories: the first category mightbe deemed the approach of the physicist and the second that of the chemist. In the formercategory, readily available, robust compounds such as transition metal halides,chalcogenides, sulfates, and nitrates were examined. Because of the robustness and chemicalsimplicity of these compounds, it was often possible to obtain large single crystals of thesubstances; thus complete susceptibility tensors could be determined. This enabled physiciststo make detailed comparisons of experimental results with theoretical models developed topredict magnetic behaviour. The majority of studies to date have been of this type. Alimitation of working with “simple” compounds is that the bridging ligands connectingparamagnetic ions are often monoatomic, for example, halide bridges. In such cases it is notfeasible to modify the bridging ligand to vary systematically a particular structural parameterand then monitor the corresponding variation in magnetic properties.The latter approach, that of the chemist, has been to prepare and examine compoundsof greater chemical sophistication. Often in these studies, the bridging ligands are mono- orpolyfunctional organic molecules. Because such ligands are amenable to systematicderivatization they are ideal for use in magnetic property/structure correlation studies. Thereare drawbacks in the study of organic ligand-based, low dimensional compounds. These25materials are insoluble and infusible, so macroscopic crystals of the materials can be difficult,if not impossible, to prepare. This is a two-fold problem: it prevents the determination ofprecise structural parameters by single crystal X-ray diffraction and prevents determinationof the complete susceptibility tensor in magnetically anisotropic materials. Another drawbackassociated with organic ligand-bridged compounds is that the bridge can be several atomslong. As was indicated in section 1.2.1, exchange effects attenuate very rapidly withincreasing separation of magnetic centres; thus multi-atom bridged magnetic materials maytend to show rather weak exchange interactions.Applying the classification scheme described above, the present study falls into thesecond category. Two series of paramagnetic 1-D compounds were prepared: one seriesbased on the copper(II) ion and the other series based on the cobalt(ll) ion. Both seriesemploy organic, (substituted)pyrazolate anions as bridging ligands. The principal aims of thiswork were to: characterize these compounds structurally, determine their magnetic propertiesas a function of temperature, and attempt to interpret those properties in terms of thestructures and compositions of the compounds. Initially, only binary, first row transitionmetal pyrazolates were to be the focus of the work as previous magnetic characterization ofsuch compounds was very limited (35,36). During the course this study, several O-Doligometallic compounds of (substituted)pyrazolates with cobalt, copper, zinc, andmolybdenum were prepared and characterized as was a poly(copper(H) 4-iodopyrazolate/4-iodopyrazole) compound. Many of these O-D compounds provided valuable informationabout some of the properties of the binary transition metal pyrazolates.This dissertation is divided into seven chapters and an Appendix. The last section ofthe present chapter, which follows, outlines the physical methods used in the characterizationof the compounds prepared during this study. Chapter 2 describes the preparation andcharacterization of a series of binary, linear chain compounds of copper(II) bridged by(substituted)pyrazolate ligands. In Chapter 3, the preparation and characterization ofoligometallic copper(1) and copper(II) pyrazolate derived species is described. Oligometallic26zinc(ll) and cobalt(ll) complexes have been synthesized and characterized. Thesecompounds, along with poly(zinc(II) 3,5-dimethylpyrazolate) and a series of Coal)(substituted)pyrazolates which have been prepared and characterized, are discussed inChapter 4. Chapter 5 describes the preparation and characterization of four miscellaneousnickel, copper, and molybdenum pyrazolate derived compounds. Chapter 6 consists of asummary of the results of this study and suggestions for future research which might beconducted on these compounds. Finally, Chapter 7 describes the experimental detailsregarding the preparation of these compounds, their subsequent physical characterization andunsuccessful attempts at preparing other metal pyrazolate compounds.1.6 PHYSICAL METHODS OF CHARACTERIZATIONA variety of physical techniques was used to gather information about compositions,structures, and magnetic properties of the compounds considered in this work. Most of thetechniques employed are familiar to chemists and it is beyond the scope of this dissertation toprovide detailed accounts of the methods. Instead, each technique is listed, very brieflydescribed, and the information pertinent to this study which the technique yields outlined.Variable temperature magnetic susceptibility studies are not routinely conducted by chemistsand because such studies form the core of this work, magnetic measurement techniques willbe described in modest detail.1.6.1 MAGNETIC PROPERTY MEASUREMENTMethods for magnetic property measurement may be divided into two basiccategories: force methods and induction methods. Of force methods, the two most commonlyemployed techniques are the Gouy method and the Faraday method. The Gouy method uses a27long, rod-shaped, homogeneous sample (usually a powder packed into a tube) suspendedfrom a sensitive balance between the poles of a magnet such that the bottom of the sample ispresent in the homogeneous magnetic field region and the top is in the region of negligiblemagnetic field (Figure 1.9(a)). The force exerted on the sample is a function of the volume(a) (b)FIG. 1.9. Sample configurations for (a) the Gouy method and (b) the Faraday method.occupied by the sample in the region of the field gradient. The force exerted on eachcross-sectional differential volume along z is given in the equationf= KvH!!L [1.11]azwhere ic is the isotropic volume susceptibility, v is the volume of the sample, H is the appliedfield, and aHlaz is the field gradient (there is also a force component along the x-direction,but the apparatus is arranged so that this force is not measured). When the sample ispositioned as in Figure 1.9(a), a paramagnetic substance experiences an increase in weightwhile a diamagnetic substance experiences a decrease in weight. In the Faraday method, asmall sample is placed in a field region where H(aHjaz) is constant (the apparatus is28designed such that H = = aH.,Jaz aHjaz = 0). In order to maximize the region overwhich this condition is met, specially shaped magnet pole faces are employed as shown inFigure 1.9(b). In this case, [1.11] becomesfXgmHx , [1.12]where x8 is the isotropic mass susceptibility and m is the mass of the sample. The Gouytechnique was employed in this study.The second class, induction methods, relies on the fact that when a magnetic materialis inserted in an induction detection coil, then a change in the voltage is induced in the coil.The strength of the induced voltage is given by the equation (13)iN2A diV 413coilwhere v is the voltage, N the number of turns of wire, A the cross sectional area, Scoil thelength of the coil, di/dt the frequency of current oscillation, and ji. the magnetic permeabilityof the magnetic material defined in equation [l.2b]. Thus, the magnitude of the voltage signalmay be related to the magnitude of the susceptibility of the material studied. Two instrumentsbased on induction were used in the present study: a vibrating sample magnetometer (VSM)and a superconducting quantum interference device (SQUID) magnetometer. The VSM wasdeveloped by Foner in the mid-1950’s (49) and operates via the inductive detection of thedipole field of an oscillating magnetic sample. A schematic diagram of a VSM is illustratedin Figure 1.10. The sample is mounted on a rigid rod which is driven in a sinusoidal motion,at a constant frequency, by the transducer. The voltage generated in the sample coils isproportional to the magnetic moment of the sample. Moreover, the reference coils are used tophase detect the signal from the sample coils with respect to the vibration frequency. Thisphase-sensitive detection removes spurious signals and thus provides a relatively sensitive29I transducervacuum housing_____[fi_ reference coilssample coils sample rodmagnet polesUtusampleFIG. 1.10. Schematic diagram of a vibrating sample magnetometer.susceptibility measurement.The SQUID magnetometer relies on the Josephson junction effect in a superconductor(50). A SQUID consists of a superconducting ring with a weak link and this device is capableof amplifying very small changes in magnetic field into large electrical signals. A schematicdiagram of the SQUID magnetometer used in this study is shown in Figure 1.11. The sampleis attached to a rigid rod and drawn through the detector coil, which is wound in asecond-order gradiometer configuration (this configuration reduces the ultimate sensitivity ofthe instrument, but also reduces the effects of extraneous magnetic fields). The signalgenerated by the sample is then amplified by the SQUID and processed by the associatedelectronics. Because the instrument employs superconducting components, immersion in aliquid helium bath is necessary.Of course, it is essential to conduct variable temperature studies with theseinstruments, so their sample zones are equipped with cryostats which permit temperature30stepping motorsample drivesample rodsuperconductingsolenoidFIG. 1.11. Schematic diagram of a SQUID magnetometer.control betweenGerloch’s text (11).<2 K and >300 K. The techniques outlined above are described in detail in1.6.2 X-RAY DIFFRACTION METHODS1.6.2.1 SINGLE CRYSTAL X-RAY DIFFRACTIONFor crystalline materials, single crystal diffraction is the premier technique forobtaining detailed metrical information about those substances. The method is based on thequantitative interpretation of the 3-D interference pattern generated when a periodic atomicarray is subjected to collimated X-radiation. An excellent general treatment of the principlesand application of diffraction methods is provided in West’s text (51).second ordergradiometersample311.6.2.2 POWDER X-RAY DIFFRACTIONThe principles which govern single crystal diffraction also apply to powderdiffraction; however, in this case the sample studied consists of, perhaps, thousands ofrandomly oriented microcrystallites. This has the effect of collapsing the 3-D interferencepattern obtained in single crystal diffraction to only two dimensions in powder diffraction;thus a large amount of information available for unambiguous structure solution in the caseof single crystal diffraction studies is lost in powder diffraction studies. Nonetheless, powderdiffractograms provide useful “fingerprints” for crystalline materials and, in some instances,provide limited information about the structures of compounds (51).1.6.3 SPECTROSCOPIC METHODS1.6.3.1 INFRARED SPECTROSCOPYInfrared (IR) spectra arise from the quantized absorption of infrared radiation due toatomic vibrations within materials. For the compounds considered in this work, the IRspectra mainly reflect vibrational modes of the pyrazolyl moieties present. Thus, theinformation provided by JR spectroscopy is “local”, or molecular; it does not indicate whatkinds of extended structures the compounds might possess. The molecular information isuseful in identifying the nature of the particular pyrazolyl moieties present with regard totheir substitution patterns and the types of metal coordination modes they exhibit in thedifferent complexes. Nakamoto’s text (52) is a source of valuable of information about IRspectroscopy and its application to inorganic coordination compounds.321.6.3.2 ELECTRONIC SPECTROSCOPYElectronic transitions involve the excitation of the ground state electronicconfiguration of a system to higher energy states by absorption of electromagnetic radiationspanning the near-infrared (NIR), visible, and ultraviolet (UV) regions of the spectrum. Inthis study, the electronic transitions of interest are those due to excitation of the transitionmetal 3d electrons. The spectra due to these transitions can provide information about thestrength of the ligand fields about the metal centres and information about thestereochemistries of the metal-ligand chromophores. Furthermore, ligand-field parametersderived from these spectra may be used to account for temperature independentparamagnetism (TIP) in the magnetic properties of these compounds. Lever’s text treats thetheory and application of electronic spectroscopy to transition metal compounds (53). ELECTRON PARAMAGNETIC RESONANCE SPECTROSCOPYWhen a paramagnetic substance is placed in a magnetic field, it has a preferred,ground state alignment of the electronic spin vectors with respect to that field. A spin maytranscend to the excited state (reversed orientation) by absorption of a quantum of energycorresponding to the microwave region; this excitation is termed electron paramagneticresonance (EPR). Experimentally, a sample is irradiated with a fixed microwave energy andthe applied field is varied until the resonance condition is met. Depending on particularconditions, the EPR spectra of paramagnetic transition metal species can provide informationabout the stereochemistries of metal-ligand chromophores, the delocalization of unpairedelectrons in substances and the temperature dependence of such spectra can yield informationabout magnetic exchange properties in such compounds. In the present study, EPR spectrawere obtained for many of the copper compounds to gain stereochemical information, wherepossible, and to compare experimental g values with those derived from magnetic33susceptibility data. The application of EPR spectroscopy to transition metal compounds hasbeen treated in depth (54,55). NUCLEAR MAGNETIC RESONANCE SPECTROSCOPYNuclear magnetic resonance (NMR) spectroscopy is a technique familiar to chemistsand widely employed. NMR is analogous to EPR except that now it is the excitation ofnuclear spin vectors which is of interest and energy required to do this lies in theradiofrequency range. In this study, the use of NMR spectroscopy was restricted solely to thecharacterization of substituted pyrazoles which were prepared in the laboratory to ensure thatthe correct isomers had formed and that they were pure.1.6.4 MISCELLANEOUS METHODS1.6.4.1 ELEMENTAL ANALYSISGenerally, elemental analysis provides information about the empirical formulae ofcompounds; no structural information is obtained. In the present work, many of thecompounds prepared have relatively few atoms in their ImonomericH units and in such caseselemental analysis provides an effective initial screening test to determine if the desiredcompounds have been obtained. THERMAL ANALYSISDifferential scanning calorimetry (DSC) studies were conducted on most of thecompounds discussed in this work. In DSC studies, the sample of interest is placed in a cell34mounted on a heating block along with an identical reference cell. The sample and referenceare heated at a controlled rate and their temperatures monitored. When a thermal event occursin the sample a temperature difference between the reference and sample arises and heat isinput by the calorimeter to the appropriate cell in order to maintain equal temperatures inboth cells. The differential quantity of heat input is plotted as a function of temperature andthe DSC thermogram results. Thus, phase changes and the loss of volatile components frommaterials may be monitored as a function of temperature. Weight measurements of samplesbefore and after thermal events permits a crude form of thermal gravimetric analysis. DSCand related techniques are described in West’s text (56). MASS SPECTROMETRYIn mass spectrometry, a compound is volatilized and ionized (which may causefragmentation into smaller charged species). The ions are then accelerated across a voltagepotential into mass analyzer magnetic fields which cause separation of the differentfragments on the basis of their mass-to-charge ratios (m/z). The relative numbers of theseions are plotted as a function of m/z. This information may be used to determine themolecular formulae of complexes and, depending on fragmentation patterns, what constituentcomponents are present in the molecules (57). SCANNING ELECTRON MICROSCOPYIn scanning electron microscopy (SEM) a focussed electron beam is raster scannedacross a sample and this causes the emission of secondary electrons from the sample. Thesesecondary electrons are collected and focussed using electromagnetic lenses resulting in theformation of a topographical image with resolution down to lO2 A. Such images provide35information about the morphologies and size distributions of particles in powdered materials(58).36CHAPTER 2POLY(COPPER(II) PYRAZOLATES)2.1 INTRODUCTIONThe synthesis and characterization of a series of eleven copper(II) pyrazolates withthe general formula, [Cu(pz*)2],where pz = 4-Hpz, 3-Mepz, 4-Mepz, 4-Clpz (two forms), 4-Brpz, 4-Hdmpz, 4-Medmpz, 4-Cldmpz, 4-Brdmpz, and indz, is described in this chapter. Thecompounds are considered in three groups: the 4-Xpz compounds, the 4-Xdmpz compoundsand finally the pair [Cu(3-Mepz)2]and [Cu(indz)2].This segregation of the complexes isbased on both their properties and the substitution patterns of the pyrazolate anions.2.2 RESULTS AND DISCUSSION2.2.1 COPPER(TI) 4-X-PYRAZOLATES (X = H, Me, Cl, and Br) SYNTHESES, PHYSICAL AND THERMAL PROPERTIESDetailed descriptions of the synthetic procedures which were successfully employedin the preparation of these compounds are given in Chapter 7, Section 7.2.3. The preparationof [Cu(4-Hpz)2jwas the first synthesis attempted in this work and, initially, the methods ofVos and Groeneveld (37) were employed. These methods are straightforward and aresummarized in the equation below.37NaOH(aq) or NH3(aq+ 2pz*H(aq) > [M(pzj,,j, + 2NaX(aq) [2.1]room temp. or2NH4X(aq)where M is a divalent transition metal ion, X is either chloride or nitrate, and pz*H ispyrazole or a substituted pyrazole. Several attempts were made to synthesize [Cu(4-Hpz)2Jvia Vos and Groeneveld’s method and in every case an insoluble black material was obtainedin accord with the findings of the above authors, but analysis of this material indicated that itpossessed the nominal formulationCu2(pz)3OH (Anal. calcd. forC9H10u2N60:C 31.3, H2.9, Cu 36.8, N 24.3,04.6; found: C 31.5, H 2.8, Cu 36.6, N 24.4,04.7).Attempts to prepare [Cu(4-Hpz)2]by Mahier’s method, as described by Trofimenko(26), in which Cu(4-Hpz)4(OAc)2 is pyroiyzed, were also unsuccessful. Fortunately,observations made during this attempted synthesis led to a successful method for preparationof the desired compound. Copper metal shot was reacted with molten pyrazole in thepresence of dioxygen to yield [Cu(4-Hpz)2]as a green solid. The equation for this reaction isshown below.Cu + xs 4-HpzH . > [Cu(4-Hpz)2J .+ H201 + soluble [2.2]18 h, stimng by-productsReaction [2.2] has been generalized to other metals and substituted pyrazoles, as will bediscussed below, and is henceforth referred to as the molten ligand precursor/metal (MLPM)reaction. With regard to the compounds considered in this section, only one other complexwas prepared as a bulk sample using the MLPM reaction; specifically, [Cu(4-Mepz)z] wasprepared as a green powder.Although the MLPM reaction was discovered independently during this work, relatedreactions have previously been examined. The MLPM reaction falls under the more generalgroup of reactions in which a protonic substrate reacts with a metal in the presence of 02 to38yield the corresponding metal complex. The most ubiquitous of such reactions is the rustingof iron. A more closely related reaction to this work occurs in the use of benzotriazole as acorrosion inhibitor for copper and copper based alloys (59-6 1). These studies have shown thatthe benzotriazole forms an impenetrable layer on the surface of copper metal which protectsit from attack by corrosion agents. It has been suggested that the layer formed by thebenzotriazole is a coordination polymer of copper, but the structure of this polymer has notbeen determined, nor has the oxidation state of the copper in this material been firmlyestablished. A more recent study of imidazole and benzimidazole as copper metal corrosioninhibitors has also concluded that copper coordination polymers likely form the protectivefilm on the copper metal (62). In related studies, Gargano et a!. found that copper metalreacts with substrates such as methanol, phenol, nitromethane, and benzoic acid in thepresence of molecular oxygen to form various copper(I) and copper(II) complexes (63).Furthermore, the authors found that the presence of a base (in this case pyridine) wasnecessary in order for these reactions to occur.As was mentioned above, the MLPM reaction was employed to synthesize a numberof compounds studied in this work; however, it was found not to be universally applicable. Amore general synthetic procedure was sought to prepare the desired transition metalpyrazolates. With this in mind, electrochemical synthesis of the binary pyrazolates wasconsidered and the preparation of [Cu(4-Hpz)2]was attempted as a test case. Sodiumpyrazolate was dissolved as an electrolyte in molten pyrazole and a current was passedthrough the stirred solution between copper metal electrodes under a dinitrogen atmosphere.A colourless solid formed at the anode in the molten pyrazole, but no [Cu(4-Hpz)2]wasobserved to form until the electrolysis mixture was exposed to the atmosphere whereupon agreen solid immediately formed from the white material (the white material was probably[Cu(4-Hpz)j). The solid product from the electrolysis in air was isolated and the greenmaterial was identified as [Cu(4-Hpz)2]by its IR spectrum and thermochromic behaviour39(see below). This method of synthesis was not pursued further because examination of thesolid from the reaction revealed the presence of a large number of metallic copper whiskers,probably caused by refining of the electrodes.Application of the MLPM reaction to the syntheses of [Cu(4-CIpz)2]and [Cu(4-Brpz)2]was unsuccessful as this led to intractable mixtures of products; however, variationsof the MLPM reaction were successfully employed in the syntheses of these compounds.[Cu(4-Clpz)2]exhibited polymorphism and was prepared in green and brown forms. Thegreen form of [Cu(4-Clpz)2]was prepared by reacting Cu20 with molten 4-ClpzH in thepresence of dioxygen and a small amount of xylene,02, 130 °CCu20+ excess 4-ClpzH . . > [Cu(4-Clpz)21J. + H201 + soluble [2.3]8.5h stlmng(green) by-productsxyleneThe xylene was present simply to wash volatilized 4-ClpzH back into the reaction vessel. Amixture of the brown and green forms of [Cu(4-Clpz)2jin which the brown form was themajor product (>95%) was prepared by reacting Cu(OH)2 with molten 4-ClpzH in thepresence of a small amount of xylene,02, 135 °CCu(OH)2+ excess 4-ClpzH > [Cu(4-Clpz)2J,j, + H20t + soluble [2.4]6 days, stirringxylene (mostly brown) by-products[Cu(4-Brpz)2]was obtained as a green powder using a reaction similar to [2.4] which isshown below.N2, 145 °CCu(OH)2+ excess 4-BrpzH . . > [Cu(4-Brpz)2],j, + H201 + soluble [2.5]16 h, stirring by-productsacetoneIn addition to providing a means of obtaining bulk samples of [Cu(4-Hpz)2Jand40[Cu(4-Mepz)2J,the MLPM reaction, with suitable modification, also provided a means ofobtaining X-ray diffraction quality single crystals for three of the five compounds discussedabove: [Cu(4-Hpz)2],[Cu(4-Mepz)2j,and [Cu(4-Clpz)2J(green). The modification of themethod involved either initially flushing the reaction apparatus with N2 gas before heatingthe copper metal/(substituted)pyrazole mixture and then allowing air to diffuse slowly intothe system and/or the reaction mixture was left unstirred to promote larger crystal growth.Unfortunately, the MLPM reaction did not afford single crystals of [Cu(4-Brpz)2]rAttemptsto produce such crystals from a quiescent mixture of Cu(OH)2 and molten 4-BrpzH yieldedsingle crystals of [Cu(4-Brpz)2}which, under microscopic examination, were of similarcolour to and possessed the same morphology as the single crystals of the other 4-substitutedpyrazolates; however, the crystals of [Cu(4-Brpz)2Jwere too small for a single crystal X-raydiffraction study.The copper(ll) 4-X-pyrazolates are air-insensitive solids. They are insoluble in waterand all common organic solvents, consistent with polymeric structures. The compoundsdissolve with decomposition in mineral acids and can be hydrolyzed to insoluble degradationproducts with prolonged heating in boiling water.The morphologies of the microcrystaflites for these compounds were examined usingscanning electron microscopy. SEM images of [Cu(4-Hpz)2]and [Cu(4-C1pz)2j(green) areshown in Figures 2.1(a) and (b).The copper(ll) 4-Xpyrazolates are involatile, infusible solids. DSC studies showed thecompounds to decompose, without melting, at the following temperatures: [Cu(4-Hpz)2],270 °C; [Cu(4-Mepz)2j,270 °C; [Cu(4-Clpz)2](green), 290 °C; [Cu(4-Clpz)2j(brown),285 °C; [Cu(4-Brpz)]290 °C. An interesting aspect of the four green copper(ll) 4-X-pyrazolates is that they are thermochromic: at room temperature the compounds are green,but at 77 K they are blue to greenish blue. Of the four compounds, [Cu(4-Hpz)2]showed thegreatest colour variation over this temperature range. The colour variation of these41FIG. 2.1. SEM images of (a) [Cu(4-Hpz)2jand (b) [Cu(4-Clpz),J (green).The white line represents a length of 5 $im.compounds was qualitatively examined by cooling them in various fixed temperature slushbaths. It was determined that the colour changes occurred in the 176 K to 77 K range for the 4-Me, 4-Cl, and 4-Br compounds and that for [Cu(4-Hpz),j the colour change begins at about176 K and is complete by 133 K. [Cu(4-Clpz)2j(brown) was not found to exhibitthermochromism in the room temperature to 77 K range. It is possible that thethermochromism exhibited by the green compounds is connected with structural changes inthe compounds. The results of investigations related to such structural changes are discussedin Section,2.1.2 X-RAY DIFFRACTION STUDIESAs was described in the preceding section, three of the five copper(ll) 4-X-pyrazolates yielded single crystals large enough for single crystal X-ray diffraction studies tobe conducted. In addition, X-ray powder diffractograms were recorded for all five42(a) (b)compounds. These X-ray diffraction studies were conducted at room temperature; however,because the green copper(II) 4-X-pyrazolates exhibit thermochromism, low-temperature (i.e.<150 K) powder diffractogranis were also recorded for these compounds and a lowtemperature single crystal X-ray diffraction study was performed on [Cu(4-Hpz)2]. ROOM TEMPERATURE SINGLE CRYSTAL DIFFRACTION STUDIESSingle crystal diffraction studies have revealed that [Cu(4-Hpz)2], [Cu(4-Mepz)2J,and [Cu(4-Clpz)2](green) are isomorphous and isostructural crystallizing in the space groupIbam. As was anticipated, these pyrazolates are linear chain coordination polymers; thisstructural motif is illustrated for [Cu(4-Hpz)2]in Figure 2.2. The coordination geometryFIG. 2.2. View of the crystal structure of [Cu(4-Hpz)2jdown the crystallographic c-axis.about the copper(H) centres in these compounds is that of a distorted tetrahedron havingexact (crystallographic) D2 symmetry. A stereoscopic view of the coordination geometry ofthe copper(ll) centre in [Cu(4-Clpz)2](green) is shown in Figure 2.3. Crystallographic data,Lb43CL CLFIG. 2.3. Stereoview of the coordination about the Cu(II) centre in [Cu(4-C1pz)2J(green); 50% probability thermal ellipsoids are shown for the non-hydrogen atoms.atomic positional parameters, and bond lengths and angles for [Cu(4-Hpz)2j, [Cu(4-Mepz)2], and [Cu(4-Clpz)2](green) are listed in Appendix I, Tables I-i through 1-10 andFigure 2.4 indicates the atom labelling scheme which is used for these compounds inconjunction with those tables. The Cu—N bond lengths are 1.957(2), 1.962(2), and 1.961(5) Afor the 4-H, 4-Me, and 4-Cl compounds, respectively. Referring back to Figure 2.2, it can beseen that the pyrazolyl rings are not coplanar with the Cu(N—N)2u rings of the chains. Thedihedral angles between these rings are 9. 1O(3)°, 7.70(3)°, and 11 .OO(3)° for [Cu(4-Hpz)2],[Cu(4-Mepz)2],and [Cu(4-Clpz)2](green), respectively. In addition, the obtuse dihedralangles between consecutive Cu(N—N)2u rings are 126.10(3)°, 126.60(3)°, and 128.20(3)°for [Cu(4-Hpz)2], [Cu(4-Mepz)2J, and [Cu(4-C1pz)2](green), respectively. The methylhydrogen atoms in [Cu(4-Mepz)jwere observed to be disordered and this was adequatelyaccounted for by assuming a (1:1) two-fold disordered model for the positions of the atoms.Further description of the structures of these compounds is deferred to Section 2.4. Atom labelling scheme for the crystallographic data on the copper(I1)4-X-pyrazolates: X=H(l) for [Cu(4-Hpz),J, CH3 (C(3), H(2), H(3), and H(4)) for[Cu(4-Mepz)2],,and Cl for [Cu(4-Clpz)2](green). ROOM TEMPERATURE POWDER DIFFRACTION STUDIESThe powder diffractograms for the green copper(H) 4-X-pyrazolates are shown inFigures 2.5(a) through (d). The d-spacings and relative peak intensities are listed in AppendixII, Table 11-1. The structures for [Cu(4-Hpz)2J,[Cu(4-Mepz)2J,and [Cu(4-Clpz)2]are, ofcourse, known from the single crystal studies, so their powder diffractograms serve mainly as“finger prints” for those compounds. In the case of [Cu(4-Brpz)2j,large single crystals couldnot be prepared and its powder diffraction pattern serves a more important role. The X-raypowder diffraction pattern for [Cu(4-Brpz)2jwas indexed to an orthorhombic unit cell.Fifteen unambiguously indexed diffraction peaks were employed in a least squaresrefinement of the unit-cell parameters with the following results: a = 9.44(1), b = 13.32(2), c= 7.70(1) A. Systematically absent reflections in the diffraction pattern are consistent withthe space group Ibam. Miller indices, observed diffraction peak d spacings and their relativeintensities and calculated d spacings based on the least-squares refinement of indexedxH(1)N245(b)15.0 i015 20 25 30 35 40 45 50 55 602eFIG. 2.5. Powder X-ray diffractograms of (a) [Cu(4-Hpz)2],(b) [Cu(4-Mepz)2J,(c) [Cu(4-Clpz)2J(green), and (d) [Cu(4-Brpz)2].reflections are listed in Table 2.1. Diffraction peaks not included in the analysis were thosewhich could not be unambiguously indexed because they consist of several overlappingdiffraction peaks. The fact that the powder diffraction pattern for [Cu(4-Brpz)2Jcould beindexed to the same space group as the other three green copper(II) 4-X-pyrazolates, coupledwith the observations that the crystals of all four green compounds possess the same colour2e0C)2e 2e46TABLE 2.1. X-ray powder diffraction data for [Cu(4-Brpz)2]h k 1 dObS(A) ‘Rel d3(A)1 1 0 7.729 100 7.7020 2 0 6.692 6 6.6642 0 0 4.726 72 4.7201 2 1 4.467 18 4.4461 3 0 4.026 74 4.0 12---3.854 52 -1 1 2 3.469 8 3.4450 2 2 3.336 16 3.335---2.988 12 -2 2 2 2.723 53 -- --2.567 9 -3 1 2 2.403 7 2.3974 0 0 2.358 11 2.360---2.223 7 ----2.142 3 ----2.011 3 -4 2 2 1.926 7 -1 7 0 1.867 4 1.8663 5 2 1.799 3 1.7981 3 4 1.736 4 1.737- --1.628 3 -3 7 0 1.628 2 ----1.571 2 -and morphology and that they exhibit the uncommon property of thermochromism providessubstantial evidence that [Cu(4-Brpz)2}is isomorphous and probably isostructural with[Cu(4-Hpz)2j,[Cu(4-Mepz)2j,and [Cu(4-Clpz),j (green).The powder diffractogram for [Cu(4-Clpz)21(brown) is shown in Figure 2.6. The dspacings and relative intensities of the peaks in this diffraction pattern are listed in AppendixII, Table TI-i. It is apparent from comparison of this figure with Figure 2.5(c) that the brownand green forms of [Cu(4-Clpz)2]are structurally distinct and the peaks at 11.8°, 22.7°, and33.8° 20 in Figure 2.6 indicate that a small amount of the green form (<5%) is present in thepredominantly brown form sample. Precise details of the brown form’s structure are47C,,4-0L)5FIG. 2.6. Powder X-ray diffractogram of [Cu(4-Clpz)2J(brown).unknown; however, comparison of this compound’s thermal and solubility properties with theother structurally characterized 4-substituted pyrazolates suggests a polymeric structure for[Cu(4-Clpz)2](brown) also. Moreover, as will be discussed in Section, the magneticproperties of [Cu(4-Clpz)2](brown) are consistent with an extended chain polymericstructure. LOW TEMPERATURE DIFFRACTION STUDIESIn view of the fact that the green copper(II) 4-X-pyrazolates exhibitedthermochromism upon cooling, low temperature single crystal and powder X-ray diffractionstudies were conducted in order to gain some insight as to the possible source of thistherrnochromism. A low temperature single crystal structure determination of [Cu(4-Hpz)2]was performed at 116(6) K. Selected crystallographic parameters from this low temperatureand the room temperature determination are compared in Table 2.2.2e40 45 50 55 6048TABLE 2.2. Crystallographic parameters for [Cu(4-Hpz)2Jat 294 K and 116 K294K 116KSpace group Ibam Ibama(A) 7.917(1) 7.870(2)b(A) 11.491(2) 11.259(2)c(A) 7.778(1) 7.807(3)Bond lengths (A)Cu—N 1.957(2) 1.959(2)N—N’ 1.359(3) 1.364(3)N—C(1) 1.345(3) 1.349(2)C(1)—H(1) 1.02(3) 1.00(2)C(1)—(2) 1.371(3) 1.386(2)C(2)—H(2) 0.80(6) 0.97(4)Bond angles (°)N—Cu—N2 99.4(1) 99. 19(5)N—Cu—N3 139.5(1) 140.72(5)N—Cu—N4 94.3(1) 93.80(5)C(1)—N—N’ 107.4(1) 107.1(1)C(1)—N—Cu 121.7(1) 121.2(1)N—N—Cu 130.28(5) 130.41(8)H(1)—C(1)—N 128(2) 125(2)H(1)—C(1)—C(2) 122(2) 124(2)N—C(1)—(2) 110.4(2) 110.3(2)H(2)—C(2)—C( 1) 127.4(5) 127.4(3)104.3(3) 104.1(2)aSurscpts refer to the symmetry operations: (1) x, y, -z; (2) 1 -x, -y, z; (3) 1 -x, y, 1/2-z; and (4) x, -y, 1/2-z.One may note that, except for two of the angles, the intrachain bond lengths and angles donot change significantly upon cooling from room temperature to 116 K, but the a- and b-unitcell edge lengths do shorten significantly and there is a slight, but significant, lengthening ofthe c-unit cell edge length. Changes associated with the shortening of the a- and b-axislengths are expansion of the pyrazolyl ring-Cu(N—N)2Curing dihedral angle from 9.1(1)° to9.87(5)° and expansion of the consecutive Cu(N—N)2u rings dihedral angle from 126.1(1)°to 127.62(3)°.Low temperature diffraction studies were extended to the other copper(ll) 4-X-49pyrazolates using the powder method only. The amount of structural information availablefrom the powder method is not nearly as great as the amount from single crystal studies;nonetheless, low temperature lattice parameters were derived from the powder diffractogramsin the manner described for [Cu(4-Brpz)2]in Section Comparisons between theroom temperature and low temperature lattice constants are made in Table 2.3.Ti 2.3. Room temperature and low temperature lattice constants for the green copper(lI) 4-X-pyrazolates with estimated standard deviations in the last digit in parenthesesCompound Temperature a b c(K) (A) (A) (A)[Cu(4-Hpz)2j 294 7.917(1) 11.491(2) 7.778(1)116 7.870(2) 11.259(2) 7.807(3)Change 0.047(2)* •O.232(3)* 0.029(3)*[Cu(4-Mepz)2J 294 9.7436(6) 12.6106(8) 7.7482(6)109 9.765(7) 12.458(6) 7.743(6)Change 0.021(7) -0.1 53(6)* -0.005(6)[Cu(4-C1pz),] 294 9.155(4) 12.968(6) 7.7 17(5)110 8.813(6) 13.036(8) 7.78(1)Change 0.07(1) 0.06(1)[Cu(4-Brpz)2] 295 9.44(1) 13.32(2) 7.70(1)140 9.06(1) 13.49(2) 7.77(1)Change 0.38(1)* 0.17(3) 0.07(1)*Difference greater than 7a.Examination of Table 2.3 reveals that the four green copper(JJ) pyrazolates can be classifiedinto two groups: [Cu(4-Hpz)2]and [Cu(4-Mepz)2Jform one group in which the majorchange in lattice parameters is the contraction of the b-axis; [Cu(4-C1pz)2jand [Cu(4-Brpz)2Jform the other group in which contraction of the a-axis is the major change in latticeparameters. The partitioning of these four compounds into two groups is also manifested in50their magnetic properties as will be discussed in Section SPECTROSCOPIC BEHAVIOUR2. INFRARED SPECTROSCOPYUnassigned JR band frequencies and relative intensities for the copper(ll) 4-substituted pyrazolates appear in Appendix IV, Table TV-i. Assignments of the vibrationalspectra of pyrazole and 4-methylpyrazole have been made previously (41,64). Assignmentsof the vibrational spectra of 4-chloro-and 4-bromopyrazole have been made by Reedijk et at.(65) and Vos and Groeneveld (39). The vibrational spectra of pyrazolate, 4-chloro and 4-bromopyrazolate (39), and 4-methylpyrazolate (41) have also been assigned. Diagnostically,the spectra of the [Cu(4-Xpz)2]compounds provided little information about their structures;however, they were used for detection of neutral pyrazole by the presence of N—H stretchingand bending bands. Principally, the IR spectra of these compounds were used as“fingerprints”. ELECTRONIC SPECTROSCOPYThe electronic spectra for the [Cu(4-Xpz)2jcompounds are presented in Figure 2.7.All four green, 4-substituted pyrazolates exhibit very similar spectra with a single, broad,asymmetric band in the visible region at about 15,500 cm’ (645 nm). This band probablyinvolves all four of the d-d transitions, unresolved, expected for a CuN4 chromophore ofdistorted tetrahedral geometry (66). The higher energy band at about 25,000 cm1 (400 nm) in[Cu(4-Hpz)2J,[Cu(4-Mepz)2],and [Cu(4-Clpz)2](green) and at 27,000 cm’ (370 nm) in[Cu(4-Brpz)21is likely charge transfer in origin. [Cu(4-Clpz)21(brown) shows a very broad51C.0C,C0.0IC0.0FIG. 2.7. Electronic spectra of the Cu(II) 4-Xpyrazolate polymers: (a) X = H,(b) X = Me, (c) X = Cl (green), (d) X = Br, and (e) X Cl (brown).absorption band over the entire visible region with indications of band maxima at 25,600cm (390 nm), 20,000 cm1 (500 nm), and 16,700 cm1 (600 nm). The maximum at 25,600cm is probably charge transfer in origin. The more complex nature of the spectrum in thevisible region and the general shift of the band(s) to higher energy than those observed in thegreen complexes suggests a CuN4 chromophore distorted even further from tetrahedralsymmetry and more towards a square-planar geometry (66). ELECTRON PARAMAGNETIC RESONANCE SPECTROSCOPYEPR spectra for powdered samples of the copper(II) 4-substituted pyrazolates wererecorded at room temperature and in the liquid N2 temperature range. The EPR spectra of theWavelength (nm)52green 4-substituted pyrazolates are quite similar and exhibit the following temperaturedependence: at room temperature the spectra consist of single, moderately broad resonancelines, each with a shoulder apparent on the low field side of the line; at low temperature thelinewidths narrow and the spectra are resolved into axial-type, two g value patterns. The 77 Kspectrum for [Cu(4-Hpz)2]is shown in Figure 2.8 as a typical example of the lowtemperature spectra. Examination of the temperature dependence of the EPR spectrum ofIFIG. 2.8. Powder EPR spectrum of [Cu(4-Hpz),] at 77 K.[Cu(4-Hpz)2]at various intermediate temperatures between the extremes revealed only agradual variation from one type of spectrum to the other. The spectrum of [Cu(4-Clpz)2](brown) at room temperature consists of a single, moderately broad absorption line. The lowtemperature spectrum for this compound also consists of a single line; however, the linewidthhas considerably decreased and this spectrum is depicted in Figure 2.9. Although the EPRspectrum for [Cu(4-Clpz)2j(brown) provides little information about the stereochemistry ofthe copper(II) chromophore, such an EPR spectrum is not inconsistent with a square-planarH >53FIG. 2.9. Powder EPR spectrum of [Cu(4-C1pz),] (brown) at 77 K.copper(II) chromophore geometry in which the tetragonal axes are grossly misaligned (67).The g values obtained for these compounds are presented in Table 2.4 where g0 iscalculated from the expressiong0 = 4g11 + [2.6]A note of clarification is proffered at this point. The room temperature g values wereobtained from the crossover points of the EPR spectra, but because some splitting is resolvedon the low field side of the resonance lines in the cases of the green compounds the actual g0values will be somewhat larger than those presented, in closer agreement with the lowtemperature values.H >54TABLE 2.4. Measured g values for [Cu(4-Xpz)2]compoundsaCompound Temperature (K) g02.104[Cu(4-Hpz)2] 293 — —96 2.242 2.078 2.1332.108[Cu(4-Mepz)2J 293 — —87 2.223 2.079 2.1272.108[Cu(4-Brpz)2] 293 — —86 2.238 2.083 2.1352.107[Cu(4-Clpz)2](green) 293 — —86 2.225 2.080 2.1282.084[Cu(4-Clpz)2J(brown) 293—= 2.08492 —aThe uncertainty in these values is less than 1%. MAGNETIC PROPERTIESMagnetic susceptibilities of the five [Cu(4-Xpz)2Jcompounds were measured from 2to 300 K using a SQUID magnetometer.Powder magnetic susceptibility and effectivemagnetic moment versus temperature data for the compounds are tabulated in Appendix III,Table HI-i. The copper(ll) 4-substituted pyrazolates examined here exhibit strongantiferromagnetic coupling. In all cases the magnetic moments of these compounds decreasesignificantly with decreasing temperature. versus temperature plots are shown for thesecompounds in Figures 2.10(a) to (e). It can be seen that all the compounds exhibitsusceptibility maxima: a clear indication of antiferromagnetic coupling. These maxima occurfor the 4-H, 4-Me, 4-Br, 4-Cl (green), and 4-Cl (brown) compounds at 143, 180, 200, 195,and 160 K, respectively. All compounds show an increase in susceptibility as temperaturedecreases at the lowest temperatures studied reflecting the presence of a small amount ofpararnagnetic impurity (likely structural) in the samples.551250SC.)C— 1000Z 750C)C)C!)5000 50 100 150 200 250 300Temperature (K)1250SC)0‘—i 10004.)-oC)C.)Cl)50050 100 150 200 250 300Temperature (K)1250SC)cc0,iooo•01Z3 750C)C)C!)1725000 50 100 150 200 250 300Temperature (K)0 50 100 150 200 250 300Temperature (K)FIG. 2.10 continued overleaf56150001250‘— 1000.0C)Cl)7505000 50 100 150 200 250Temperature (K)300Fic. 2.10. Powder magnetic susceptibility plots for the Cu(II) 4-Xpyrazolate polymers:(a) X = H, (b) X = Me, (c) X = Cl (green), (d) X = Br, and (e) X = Cl (brown). Thelines are the calculated curves for the best fit values of the parameters listed in Table2.5. The solid upper curves represent fits to the full data sets while the lower curvesrepresent fits to only the higher temperature data. The dashed curves represent thebest fit values calculated from the parameters in Table 2.6.A curious feature of the susceptibility data for the 4-Cl (green) and 4-Br derivatives isthe presence of discontinuities in the data at 133 and 108 K, respectively. Thesediscontinuities are shown in the insets of Figures 2.10(c) and (tf. Although small, thediscontinuities are reversible and reproducible.In an effort to quantify the extent of antiferromagnetic coupling in these systems thesusceptibility data have been analyzed using the isotropic Heisenberg model for exchangecoupled linear chains developed by Bonner and Fisher (68) employing the polynomialdeveloped by Hall and others (69,70),57Ng2I.LXchain=kTwhere x = kT4Il. In [2.71 and throughout this dissertation N is Avogadro’s number, g theLands splitting factor, I.tB the Bohr magneton, k is Boltzmann’s constant, T the absolutetemperature, and J the exchange constant. The 1-D Heisenberg model was developed forideal linear chains (recall Figure 1.7 in Chapter 1), so in order to account for the influence ofparamagnetic impurities on the magnetic susceptibilities at the lowest temperatures in the4-substituted pyrazolates, a Curie law term was added to [2.7] of the formNg2.tS(S+1)3kTXcaic =[‘4iXchain1’Xpara [2.9]where lOOP is termed the percent paramagnetic impurity (%P). The best fits of [2.9] wereobtained by minimizing the function0.25 + 0.14995 x4+ 0.30094r21 + 1.9862r’+ 0.68854r2+ 6.0626x3[2.7][2.8]where S is the spin of the magnetic ion (in this case S = 1/2). Thus, the observedsusceptibility for these compounds was modelled with the following expression:F=(Xbs)2[2.10]58where n is the total number of points in the data set. The value of F for the best fit alsoprovides a measure of agreement between theory and experiment. Computer fits of [2.9] tothe data were made over the range 2-300 K and the best fit values off, g, and %P along withthe corresponding F values are presented in Table 2.5.TABLE 2.5. Derived magnetic parameters for the [Cu(4-Xpz)2JcompoundsTemperatureCompound range (K) (cm-1) ga %pab F[Cu(4-Hpz)2] 135-300 81 2.12 — 0.00142-300 82 2.13 0.31 0.0097.[Cu(4-Mepz)2] 135-300 96 2.15 — 0.00262-300 99 2.18 0.54 0.016[Cu(4-Brpz)2] 110-300 105 2.20 — 0.00422-300 106 2.21 0.51 0.022[Cu(4-Clpz),] (green) 140-300 104 2.13 — 0.00202-300 99 2.08 0.42 0.026[Cu(4-C1pz)2](brown) 135-300 88 2.06 — 0.00122-300 90 2.07 0.72 0.020aThe estimated uncertainties inJ, g, and %P are ±2 cm, ±0.02, and ±0.02, respectively.b%p arbitrarily set to zero when analysis involved the high temperature data only.Because of the discontinuities in the susceptibility data for the 4-Cl (green) and 4-Brcompounds, attempts were made to fit only the data above the discontinuity temperatures andsimilar fits of the other 4-substituted compounds were also made for comparison. In thesecases the fits were obtained setting the %P arbitrarily to zero. The upper solid curves inFigures 2. 10(a)-(e) represent the best fit results to the data over the whole temperature rangeand the lower solid curves represent the best fits to the data in the high temperature range. Itis apparent that the derived parameters differ little whether all of the data or only the high59Ng2,.iS(S+1)Xpara 3k(T- 0)-J -eCompound (cm1) g %P (K) F[Cu(4-Hpz)2] 82[Cu(4-Mepz) 100[Cu(4-Brpz)2] 106[Cu(4-C1pz)2J(green) 100[Cu(4-C1pz)21(brown) 91aThe estimated uncertainties in J, g, and %P are ±2 cm’2.13 0.33 0.22.18 0.66 0.62.21 0.70 1.12.08 0.62 1.42.08 0.93 0.9±0.02, and ±0.02, respectively.temperature data are fit to the model. Furthermore, the experimentally determined g values(from EPR spectroscopy) and those calculated from the model are generally in goodagreement. Attempts to fit the low temperature data only to the model were, on the otherhand, not successful leading to significantly higher JI values and, more importantly,unrealistic g values on the order of 3.0.Examination of the best fits of [2.9] to the data for the 4-substituted pyrazolates overthe whole temperature range in Figures 2. 10(a)-(e) reveals that the model predictions tend todeviate most in the low temperature range. In response to this, equation [2.9] was modified sothat the Xpara term was modelled as a Curie-Weiss law paranlagnet,[2.11]instead of a Curie law paramagnet as in [2.8]. Best fit parameters of the modified equation[2.9] to the data for the five compounds are presented in Table 2.6.TABLE 2.6. Derived magnetic parameters for the [Cu(4-Xpz)2]compounds in the 2-300 Krange with %P modelled as a Curie-Weiss law paramagneta0.00870.0110.0140.0210.009260The dashed curves in Figures 2.1O(a)-(e) represent the best fits of [2.9] with Xp&a modelled asin [2.11]. It is apparent that this model provides marginally better fits of the experimentaldata at the lower temperatures, but a fourth parameter is required to achieve thisimprovement. In addition, comparison of the J and g parameters in Table 2.6 with those inTable 2.5 reveals that there are no significant differences between the two sets of results; thusmodelling of the paramagnetic impurities as Curie law paramagnets is adequate forquantification of the 1-D magnetic exchange in these compounds.Having analyzed the magnetic susceptibility of these compounds and quantified themagnitude of their exchange interactions, it is worthwhile to give some consideration to howthe nature of the pyrazolate bridge and structure affect the magnitude of exchange coupling inthese compounds. Recall that the single crystal X-ray diffraction studies showed that the 4-H,4-Me, and 4-Cl (green) compounds are isomorphous and structurally very similar.Furthermore, powder X-ray diffraction studies showed that the 4-Br compound isisomorphous and very probably isostructural with the other three green pyrazolates.Considering the structural similarities of these compounds, it is not surprising that these fourcompounds exhibit very similar magnetic properties. In the following discussion, the valuesof J considered ase those listed in Table 2.5 corresponding to the high temperature fits. Theabsolute values of J obtained for the 4-Cl and 4-Br derivatives are the same withinexperimental error and slightly larger than for the 4-Me derivative. The value of IJI for [Cu(4-Hpz)2] is significantly smaller than those obtained for the other three compounds. Thisvariation does not correlate with simple ideas based on relative ligand basicity, since onewould expect the halo-substituted pyrazolate ligands to be least basic (as a measure of this:PKa’S for the 4-XpzH2 cations are 3.09, 2.53, 0.64, and 0.60 for X = Me, H, Br, and Clrespectively (7 1-73)), form the weakest bonds with copper, and therefore to provide thepoorest rather than the most efficient pathways for exchange.Studies by Richardson and Hatfield (74) and Inoue (75) established that in the chain61compound [Cu(pyz)(NO3)2](pyz = pyrazine) (76-78) the it-orbitals of the bridging pyrazineligands pmvide important pathways for propagation of magnetic exchange between thecopper(I1) centres. Consideration was therefore given to the possibility that the ic-system ofthe pyrazolate ligand may also provide a significant contribution to the overall pathway forexchange in the compounds considered in this work. A HAM molecular orbital calculation(79) on the pyrazolate ion reveals the highest occupied molecular ic-orbital (HOMO) and thisis illustrated in Figure 2.11(a). The orientation of this orbital with respect to the magneticorbital on copper in [Cu(4-Hpz)2Jis shown in Figures 2.11(b) and 2.11(c). Defining theangle between the plane of the pyrazolyl ring and the fused Cu(N—N)2u ring as a and theobtuse dihedral angle between consecutive Cu(N—N)2urings as (3 (Figure 2.11(b)) it is clearthat increased overlap between the HOMO and the magnetic orbital and therefore increasedexchange (assuming the HOMO provides an important superexchange pathway) would beexpected as 13 increases or a decreases. The room temperature cx and (3 angles for the threecompounds for which the structures are known and the low temperature a and (3 angles for[Cu(4-Hpz)2Jare presented in Table 2.7.TABLE 2.7. [Cu(4-Xpz)]structural parameters (see Figure 2.11(b))Compound Temperature (K) cx (°) (3 (°)[Cu(4-Hpz)2j 294 9.10(3) 126.10(3)116 9.87(5) 127.62(3)[Cu(4-Mepz)2J 294 7.70(3) 126.60(3)[Cu(4-C1pz)2J(green) 294 11.00(3) 128.20(3)No obvious correlation is evident from the data presented in Table 2.7. Comparison of theroom temperature angles shows that on going from [Cu(4-Hpz)2]to [Cu(4-Mepz)], (3increases slightly and a decreases significantly. This is consistent with the larger 1.1162(a)(b)(c)FIG. 2.11. (a) HOMO for the pyrazolate ion. (b) and (c): two different viewsof the overlap of the magnetic orbital on copper with the pyrazolate HOMO.possessed by the 4-Me derivative over that of the 4-H compound (see Table 2.5).Unfortunately, there are no other unambiguous comparisons. For example, on going from the4-H to the 4-Cl compound both and a increase, thus having opposing effects on orbitaloverlap and the magnitude of exchange. There is an even greater obstacle to deriving amagneto-structural correlation from Tables 2.5 and 2.7. Application of the it-orbitalsuperexchange paradigm outlined above using only a single J value for each compound ispredicated on the temperature invariance of the structures of these compounds. Clearly, theresults from Tables 2.3 for all four compounds and from Table 2.7 for [Cu(4-Hpz)2]bringinto question the assumption of structural temperature invariance because the intracompound63variations of a and (3 for [Cu(4-Hpz)2},over the temperature range studied, are comparableto the intercompound variations of a and f3 observed at room temperature.The possibility of using the it-orbital overlap model to propose a self-consistentrationale for those facts which are known about the magnetic properties of the individualcopper(II) 4-substituted pyrazolates is considered presently. The structural studies of thegreen 4-substituted pyrazolates demonstrated that the unit cell parameters of thesecompounds vary on cooling from room temperature to low temperatures. Furthermore, in thecase of [Cu(4-Hpz)2Jthe unit cell changes are known to be concomitant with changes in thea and f3 dihedral angles and that, within uncertainty, intrachain bond lengths do not change.What is not known is the structural behaviour of these compounds at intermediatetemperatures: are the changes observed phase transitions, i.e. do they occur at specific criticaltemperatures, or are the changes gradual occurring continuously over the temperature intervalspanned? Experiments to measure the intermediate temperature structural dependence ofthese compounds would have yielded valuable information; however, such experiments werenot attempted as they are extremely time consuming and prohibitively expensive. Thethermochromic behaviour of the 4-substituted pyrazolates at intermediate temperatures maybe used as a piece of evidence to suggest that the structural changes in these compounds aregradual if it is assumed that the thermochromism arises from the structural changes.Presumably, the structural changes induced in these compounds upon cooling arisefrom “quenching” of vibrational motion which permits alteration of the complex intra- andinterchain van der Waals interactions present. This is the source of a second uncertaintybecause, a priori, it is impossible to predict how this multitude of changing interactions willalter the a and (3 dihedral angles in the complexes. In the cases of [Cu(4-Hpz)2]and [Cu(4-Mepz)2J, the principal structural change in cooling from room temperature is a compressionof the b-axis (see Figure 2.2). Table 2.7 shows that, in [Cu(4-Hpz)2], cooling of thecompound results in a small increase in a and a larger increase in (3. One might assume thatsimilar changes occur in these angles in [Cu(4-Mepz)2J.Increasing a and (3 angles, within64the overlap paradigm suggested here, have opposing effects on the magnitude of exchangeand it is possible that the two effects cancel leading to a constant J and the absence ofdiscontinuities in the versus temperature piots for the 4-H and 4-Me compounds. [Cu(4-Clpz)2](green) and [Cu(4-Brpz)2]show discontinuities in their versus temperature plotsand the principal structural change in cooling from room temperature is a compression of thea-axis. The magnitudes of the susceptibilities for these compounds below their discontinuitytemperatures suggests that their exchange interactions have lessened. Within the overlapmodel proposed here, this weakening of exchange may be due to increases of thea anglesand/or decreases of the f3 angles in these compounds as the temperature is lowered. Thedecrease in the a-axis lengths in these compounds could be accomodated by decreases intheir f3 values. However, the results for the 4-H derivative would suggest that changes in 13should coincide with significant changes in the b-axes and this is not seen here. Thedecreases in the a-axes might also be accomodated by increases in the a angles with little orno change in the f3 angles. This could account for the relatively constant b-axis lengths inthese compounds. As precise values of a and f3 at low temperatures have been determined forneither the 4-Cl (green) nor the 4-Br species, the proposed angular changes in thesecompounds must be regarded as tentative. Regarding the discontinuities themselves, it shouldbe recognized that the quantitative dependence of overlap and exchange upon the a and 13angles individually is not known and, for that matter, not specified within the modelconsidered here. It is entirely possible that the discontinuities observed are due to the rapidsupersedence, over a narrow temperature range, of the effects of one type of angular changecompared with others.Although the pyrazolate HOMO may be an important medium for magneticexchange between the copper(ll) centres, it is unlikely that the HOMO is the sole propagatorof exchange. Surely other ligand orbitals contribute to the magnetic exchange process;however their contributions are not as conveniently assessed. As a final comment in this65discussion, although the precise discernment of the orbital influences on the compoundsconsidered here may not be possible, there is no doubt that orbital interactions do control theextent of exchange observed. As a case in point, consider the green and brown forms of [Cu(4-Clpz)2]. Their magnetic susceptibility data model well as linear chain antiferromagnets, butshow that the two forms have significantly different J values. The compounds arecompositionally equivalent, but the powder diffraction patterns for the two forms indicatethat they possess different structures and the electronic spectra show them to have differentcopper chromophores; thus it is most likely that the specific copper ion/ligand orbitaloverlaps differ between the two forms.It was stated above that the copper(II) 4-substituted pyrazolate polymers exhibitstrong l-D antiferromagnetic exchange coupling. It is appropriate to put the magnitude ofexchange coupling in these pyrazolates into context by comparison with other copper(II)linear chain compounds. Mono-, di-, and ti-i-halide bridged copper(II) chain compounds havebeen extensively studied (3,80-82) and I]] in these compounds ranges from —0 to —130 cm1,but this range of coupling magnitudes requires a more explicit treatment. In the vast majorityof halide bridged copper(II) chain compounds IJj37 cm1. Only KCuF3 shows very strong 1-D antiferromagnetic coupling with a 1.11 of —130 cm1 (83-85). There are examples ofcopper(II) linear chain compounds with many other kinds of bridging ligands, such as H20(1), alkoxides (86), and oxo-anions (1). lu for these compounds is generally 10 cm1. Othercopper(II) chain compounds prepared and studied in this laboratory by other workers areailcyiphosphinate and pyrazine complexes. The phosphinates involve ti-i-atomic O—P—Obridges between copper(II) centres and the strongest antiferromagnetic exchange observedpreviously in these systems is in the a-forms of copper(ll) n-decyl and n-dodecylphosphinatewhere [It values of 29 cm have been reported (87). The copper(II) chain compounds ofpyrazine and substituted pyrazines, which involve the tetra-atomic N—C—C—N bridge, showvalues of [Jj ranging up to a maximum of 6 cm1 (74,78,88). Thus, along with KCuF3 and afew oxide bridged species (89), the 4-substituted pyrazolates, with J values ranging from -8166to -105 cm-1, are amongst the strongest 1-D antiferromagnetically coupled copper(ll) chainsystems known.2.2.2 COPPER(II) 4-X-3,5-DIMETHYLPYRAZOLATES (X = H, Me, Cl, Br) SYNTHESES, PHYSICAL AND THERMAL PROPERTIESDetailed descriptions of the synthetic procedures which were successfully employedin the preparation of these compounds are given in Chapter 7, Section 7.2.3. Success with theMLPM reaction in the syntheses of [Cu(4-Hpz)2]and [Cu(4-Mepz)2]led to the applicationof the reaction to the syntheses of the copper(ll) 4-X-3,5-dimethylpyrazolates, where X = H,Me, Cl, and Br. The MLPM reaction did yield samples of the desired compounds,unfortunately with accompanying by-products which could not be separated from the binarycopper(II) dimethylpyrazolates (in the case of [Cu(4-Hdmpz)2]the impurities were detectedin the EPR spectra for the sample). This finding led to an interesting aside connected with theMLPM reaction. The reaction uses molecular oxygen as an oxidant and a brief investigationwas undertaken to see how other oxidants might cause the reaction to behave. Specifically,3,5-dimethylpyrazole and copper metal were combined in aqueous 30% hydrogen peroxide.A vigorous reaction ensued which subsequently yielded beautiful, large, blue-green dichroiccrystals. X-Ray diffraction and elemental analysis revealed the crystals to be the dimeric[Cu(OAc)2(H0)];one of the seminal compounds in magnetic studies of moleculartransition metal complexes (90)! The 4-X-3,5-dimethylpyrazolyl moieties do not seem topossess as high a resistance to oxidation as that of the 4-X-pyrazolyl moieties. It was thoughtthat the prolonged reaction times at elevated temperatures and the presence of the potentiallycatalytic copper metal surface required by the MLPM reaction might have promoted theformation of intractable oxidation by-products during attempts to synthesize [Cu(4-Xdmpz)2}.’ To circumvent the formation of intractable by-products a reaction was sought67which eliminated the presence of metallic copper and reduced the reaction time required.This was achieved by modification of the MLPM reaction. The altered reaction, whichafforded pure samples of the four compounds discussed in this section, involved the oxidationof [Cu(4-Xdmpz)]3 (the syntheses and characterization of these copper(I) species aredescribed in Chapter 3) in the presence of the corresponding molten4-X-3,5-dimethylpyrazole and dioxygen to yield the desired copper(ll) compound,02[Cu(4-Xdmpz)]3+xs 4-Xdmpzll115445 0C[Cu(4-Xdmpz)2],j,+H20t [2.12]2-8 h, stirring + soluble by-productsIt is interesting to note that another oxidation product of [Cu(dmpz)]3 has recently beencharacterized. Arclizzoia et al. have prepared a cyclic, octameric, copper(II) complex bridgedby 4-Hdmpz and hydroxide from the aerial oxidation of [Cu(4-HdmpzXJ3in wet CH2I,pyridine, or nitrobenzene (9 1,92).The 4-substituted dimethylpyrazolates were prepared as fine powders ranging incolour from reddish-brown in the case of [Cu(4-Hdmpz)2]to brown for [Cu(4-Medmpz)2jand [Cu(4-Cldmpz)2jand to dark brown for [Cu(4-Brdmpz)2].Like the 4-substitutedpyrazolates, the copper(ll) 4-substituted dimethylpyrazolates are insoluble in commonorganic solvents with a range of polarities and can be dissolved in mineral acids withdecomposition. [Cu(4-Medmpz)2],[Cu(4-Cldmpz)2],and [Cu(4-Brdmpz)2]dissolve withdecomposition in water and wet solvents yielding green soluble species and small amounts ofgreyish-green insoluble species. The soluble hydrolysis products do not crystallize, but formfilms upon drying. [Cu(4-Hdmpz)2]is apparently resistant to hydrolysis in water and watercontaining solvents at room temperature.1The characterization of such oxidation by-products forms part of Chapter 3.68The 4-Xdmpz copper(11) complexes are thermally robust and DSC studies show themto decompose without melting at the following temperatures: [Cu(4-Hdmpz)2]245 °C;[Cu(4-Medrnpz)2J, 233 °C, [Cu(4-Cldmpz)2], 245 °C; and [Cu(4-Brdmpz) 245 °C.Unlike the copper(TI) 4-Xpz compounds, the 4-Xdrnpz compounds do not exhibitthermochromism over the 300-77 K temperature range.2.2.12 X-RAY DIFFRACTION STUDIESSuccessful preparation of X-ray diffraction quality single crystals for some of the 4-substituted copper(II) pyrazolates lead to attempts to prepare such crystals of the 4-substituted dimethylpyrazolates. All attempts failed to provide macroscopic single crystalsand as a result, only powder X-ray diffraction studies were carried out on these compounds.The powder X-ray diffractograms for these compounds are shown in Figures 2.12(a) through(d). The d-spacings and relative intensities of the peaks in these diffraction patterns are listedin Appendix II, Table 11-2. The diffractograms for these compounds provide no informationabout the precise structural details, but they do indicate the complexes are microcrystallinematerials. SPECTROSCOPIC BEHAVIOUR2. INFRARED SPECTROSCOPYUnassigned JR band frequencies and relative intensities for the copper(II) 4-substituted 3,5-dimethylpyrazolates appear in Appendix IV, Table IV-2. The IR spectra ofthese complexes were recorded primarily as a preliminary test of product purity. The IRspectra of the copper(I) 4-Xdmpz compounds exhibit a sharp band between 700 and 8006910U CUFIG. 2.12. Powder X-ray diffractograms of the Cu(ll) 4-X-3,5-dimethyl-pyrazolate complexes: (a) X = H, (b) X = Me, (c) X = Cl, and (d) X = Br.cm1. These bands shift 10-15 cm lower in energy in the corresponding copper(II)compounds, thus permitting these bands to be used as a check of whether conversion of thecopper(1) to copper(II) compounds was complete during syntheses of the latter complexes.Although the IR spectra for these compounds cannot provide detailed informationCC(b)(a)•, I.. ..__.....A.%.....___(c)5. 10. 15. 20. 25. 30. 35. 0. 45. 50. 55. 50.2e5. 00. 55. 20. 35. 30. 35. 40. 45. 50. 55. 55.2e70about their structures, they do provide some information about the types of pyrazolylmoieties in the compounds. It has been reported that the in-plane Vrmg mode of 4-HdmpzH,which occurs at 1,596 cm in solid 4-HdmpzH and CC14 solutions of the ligand (93), shifts tolower energy upon coordination to a metal, for example: to 1,556 cm1 in Co(4-HdmpzH)21(36), 1,573 cm1 in Au(4-HdrnpzH)(PPh3) BF,and 1,575 cm1 in Au(4-HdmpzH)Cl andAu(4-HdmpzH)2C1 (94). This band occurs at even lower frequencies in the coordinateddeprotonated ligand: 1,530 cm1 in [Cu(4-Hdmpz)]3(42) (1,525 cm4 in Appendix IV, TableIV-4), 1,512 cm1 in [Co(4-Hdmpz)2J(42) (1,523 cm1 in Appendix IV, Table IV-8).Examination of the JR spectra for all of the 4-X-3,5-dimethylpyrazolyl complexes studied inthis work indicates that the coordination shift observed in the 4-H derivative occurs in all ofthe 4-X-3,5-dimethyl species (these bands occur in solid samples of the pure ligands at: 1,596cm4 in 4-MedmpzH, 1,598 cm4 in 4-CldmpzH, 1,584 cm-’ in 4-BrdmpzH, and 1,577 cm4 in4-IdmpzH). For the [Cu(4-Xdmpz)2]species, the Vñflg bands occur at 1,523 cm4, 1,508 cm-1,1,513 cm-1, and 1,506 cm for the 4-H, 4-Me, 4-Cl, and 4-Br derivatives, respectively. Theshift to lower frequencies of the Vrjng mode in these compounds is likely due to theuncoupling of the Vg mode from the 13NH mode of the protonated ligand rather thancoordination per se because in the case of the ionic compound Na[4-Hdmpz], the band inquestion occurs at 1,512 cm-’.Another JR spectral correlation with coordination mode in the methylated pyrazolylspecies has been identified by making use of the vibrational assignments of the 4-Xpz anions(X = H, Me, Cl, Br, and I) made by Vos and Groeneveld (39,41). To facilitate discussion ofthis correlation, the labelling scheme depicted in Figure 2.13 for the methyl groups of thepyrazolyl ring will be used here and throughout the dissertation. Of course, in the case ofbridging pyrazolates, by symmetry, only ortho- and meta-methyl groups may be present.Complexes containing only bridging, methylated pyrazolate ligands (Chapters 2, 3, and 4)exhibit shifts from the corresponding free ligand values of 35-50 cm to higher energy for71metapara H orthoHFIG. 2.13. Methyl group labelling scheme for discussion of bands.the C-CH3 modes of their ortho-methyl groups. 4-Medmpz containing complexes exhibit asecond, higher energy 3C.CH3 band, presumably due to the mew-methyl group which does notshift greatly upon coordination. This shift is attributed to increased steric interaction of theortho-methyl group with other adjacent groups upon coordination. Complexes which containterminal methylated pyrazole ligands (Chapters 3 and 4) exhibit two C-CH3 bands, oneshifted approximately 40-50 cm higher in energy and the other shifted only 10-20 cm1higher in energy. The former band is attributed to the ortho-methyl group and the latter bandto the para-methyl group as its interaction with the coordination centre is likely to be muchweaker. Table 2.8 lists the positions of the bands in the [Cu(4-Xdmpz)2]complexesalong with the corresponding bands in the free neutral ligands in the solid state. That theseshifts require coordination of the ligand and not simply deprotonation in order to occur isgiven further support by the fact that in Na[4-Hdmpz] the 13CCH3 band appears at 417 cm4whereas in solid 4-HdmpzH it appears at 416 cm1.H HHH72TABLE 2.8. Positions of the bands in the [Cu(4-Xdmpz)2Jcomplexes and the 4-XdmpzH moleculesBand positionCompound (cm-1)4-Hdmpzll 416[Cu(4-Hdmpz),J 4634-MedmpzH 482581[Cu(4-Medmpz)2] 522 5794-CldmpzH 479[Cu(4-Cldmpz)2j 5144-BrdmpzH 473[Cu(4-Brdmpz)2] 5242. ELECTRONIC SPECTROSCOPYThe electronic spectra for the [Cu(4-Xdmpz)J compounds are presented in Figure2.14. In the visible region, two band maxima are observed for each of the compounds asfollows: X = H, 19,000 cm1 (526 nm) and 23,100 cm1(sh) (433 nm); X = Me, 18,300 cm1(546 nm) and 21,700 cmt(sh) (461 nm); X = Cl, 17,800 cm1(sh) (562 nm) and 20,700 cm1(483 nm); X = Br, 19,000 c&(sh) (526 nm) and 21,700 cm (461 nm). In addition, all fourcompounds show the presence of strong bands in the UV region, presumably charge-transferin origin (Figure 2.14), as well as a weak band in the NW region at 10,500 cm (952 nm).The bands in the visible region which likely arise from d-d transitions are at significantlyhigher energies than the corresponding bands described in Section for the analogous[Cu(4-Xpz)] compounds. According to Hathaway (66), complexes with CuN4chromophores which exhibit d-d bands in the 18,000 to 20,000 cm1 region, as observed forthe 4-Xdmpz complexes here, are likely to have square-coplanar stereochemistries. This73U,CDL0a,C.)C0-L0U,.0spectroscopic criterion of stereochemistry has been successfully applied to pyrazolatecomplexes previously (95). The monometallic complexes [Me2Ga(4-Hdmpz)]Cu and[Me2Ga(4-Hpz)]Cuhave CuN4 chromophores with compressed tetrahedral coordination inthe former and square-coplanar coordination in the latter (96). The former complex exhibits dd bands at 9,400 cm1 and 12,000 cm-1 (sh) while the corresponding bands in the lattercomplex are in the 17,000 to 20,000 cm1 region (95). ELECTRON PARAMAGNETIC RESONANCE SPECTROSCOPYIn an attempt to gain more insight into the nature of the chromophores of thecopper(II) 4-Xdmpz compounds, their powder EPR spectra were recorded at 90 K and in thecase of [Cu(4-Brdmpz)2]also at room temperature. The complexes give broad, asymmetric,single resonance lines which, unfortunately, lack the features necessary for the determination74500 1000 1500Wavelength (nm)FIG. 2.14. Electronic spectra of the Cu(II) 4-X-3,5-dimethylpyrazolatecomplexes: (a) X = H, (b) X = Me, (c) X = Cl, and (d) X = Br.of g values directly from the spectra. Furthermore, in the case of [Cu(4-Brdmpz)2]the roomtemperature resonance line is even broader than the 90 K line. The EPR spectrum for [Cu(4-Hdmpz)2jis typical of the spectra for all four compounds and is shown in Figure 2.15.H>Fic. 2.15. EPR spectrum of [Cu(4-Hdmpz)2]at -90 K.Comparing the spectra of the 4-Xpz complexes with those of the 4-Xdmpz complexes, thereappears to be less exchange narrowing in the latter series of compounds than in the formerseries, due to weaker antiferromagnetic coupling (discussed below). MAGNETIC PROPERTIESBefore discussing the magnetic properties of the copper(II) 4-Xdmpz complexes it isnecessary to consolidate the information on these compounds presented above with results tobe given presently and use this information to make inferences about the structures of thecomplexes. The empirical formulae for the 4-Xdmpz compounds do not support monomericspecies, unless one assumes either the unlikely coordination number of two for copper(II) or75the equally unlikely formation of three-membered chelate rings from endobidentatepyrazolate rings. A possible thmetallic structure for the copper(II) 4-Xdmpz compoundsbased on the type suggested by Blake et at. for nickel(II) pyrazolates (42) is shown in Figure2.16; however, a physical model of this structure indicates unfavourable interactions betweenadjacent methyl groups. A more likely alternative to the structure depicted in Figure 2.16 isFIG. 2.16. Possible trimetallic structure for [Cu(4-Xdmpz)2}compounds.that the compounds form chain polymers as do the copper(II) 4-Xpz species. Like the 4-Xpzcompounds, the 4-Xdmpz compounds are crystalline, insoluble, infusible solids and althoughthese observations are not positive proof that the 4-Xdmpz compounds are polymeric, theyare consistent with such a structure type. SEM studies provide further evidence for theformation of [Cu(4-Xdmpz)2]chain compounds. Figure 2.17 is a SEM image of [Cu(4-Hdmpz)2], typical of the copper(ll) 4-Xdmpz compounds, showing that the compoundcrystallizes in long, fibrous strands; a morphology consistent with chain polymer materials.76‘NCuFIG. 2.17. SEM image of [Cu(4-Hdmpz)2J,Although it is thought that the compounds considered in this section are polymeric,they are not likely to be isostructural with the copper(ll) 4-Xpz compounds because in such astructure the methyl substituents on adjacent pyrazolate rings would experience considerablesteric crowding. This steric crowding could be relieved in one of two ways: either the chaincompounds could adopt a more regular tetrahedral CuN4 chromophore, or the chain couldadopt a stepped structure as depicted in Figure 2.18. The EPR and electronic spectroscopyfindings indicate that the chromophores for the copper(II) 4-Xdmpz complexes are, as agroup, distinct from those of the copper(ll) 4-Xpz complexes. In particular, the electronicspectra of the 4-Xdmpz complexes indicate that they possess square-coplanar chromophores,thus the stepped chain structure depicted in Figure 2.18 which is based on the structures of[(ON)Ni(4-Hdmpz)2Ji and [(C3H5)Ni 4-Hdmpz)2]i (97,98) seems to be the mostplausible structure for these compounds.Magnetic susceptibilities of the four [Cu(4-Xdmpz)2jcompounds were measured77N — N — N N — NCu Cu Cu CuI”,.FIG. 2.18. Stepped linear chain model for the [Cu(4-Xdmpz)2}compounds.from 2 to 300 K using a SQUID magnetometer. Powder magnetic susceptibility and effectivemagnetic moment versus temperature data for the compounds are tabulated in Appendix III,Table 111-2. The complexes examined here exhibit strong antiferromagnetic coupling, thoughnot as strong as the exchange observed in the [Cu(4-Xpz)2]compounds. In all cases themagnetic moments decrease significantly with decreasing temperature. versus temperaturepiots (Figures 2.19(a) to (d)) reveal that all the compounds exhibit susceptibility maximawhich occur in the range 105 to 120 K. All compounds show an increase in susceptibility astemperature decreases at the lowest temperatures studied reflecting the presence of smallamounts of structural paramagnetic impurity in the samples as was observed for the [Cu(4-Xpz)2jcompounds. In accord with the structure type proposed above, the magnetic data forthe [Cu(4-Xdmpz) compounds were analyzed using the isotropic Heisenberg model forexchange coupled linear chains modified to account for the presence of paramagneticimpurities as described in Section Reasonably good fits between experimental dataand theory were obtained for the X = H, Me, and Br compounds and best fit values of the J,g, and %P parameters along with the corresponding F values are given in Table 2.9.782500 2000020001500C)C.)U)10000C.,C.)1000-*,C)C)CCCl)500FIG. 2.19. Powder susceptibility versus temperature plots for the Cu(II) 4-X-3,5-dimethyl pyrazolate complexes: (a) X = H, (b) X = Me, (c) X = Cl, (d) X = Br.The lines are the calculated best fit curves using the parameters in Table 2.9:solid lines, fits to high temperature data only; dashed lines, fits to whole temperaturerange; dotted lines, fits to whole temperature range with a Curie-Weiss law paramagnetic impurity term.0 50 100 150 200 250 300Temperature (K)15000 500B 1750BVcc15001250C)C.)U)100025000B20001500-oa)C.)U)1000100 150 200 250 300Temperature (K)0 50 100 150 200 250 300Temperature (K)0 50 100 150 200 250 300Temperature (K)79TABLE 2.9. Derived magnetic parameters for the [Cu(4-Xdmpz)2]compounds with estimatedstandard deviation in the last digit in parenthesesTemperature -J -eCompound range (K) (cm-1) g %P (K) FX = H 2-300 56(1) 2.22(2) 0.28(1) — 0.0232-300 58(1) 2.24(2) 0.75(6) 4.4(4) 0.007840-300 58(1) 2.24(1) 0.55(4) — 0.0018X=CH3 2-300 61(1) 2.26(2) 0.31(1) — 0.0122-300 61(1) 2.26(2) 0.33(2) 0.1(2) 0.01240-300 60(1) 2.25(2) oa — 0.0040X = Cl 2-300 8 1(10) 2.3(2) oa — 0.1840-300 67(1) 2.24(1) oa — 0.0048X = Br 2-300 65(1) 2.2 1(2) 0.50(2) — 0.0142-300 66(1) 2.22(2) 0.65(2) 0.9(1) 0.007240-300 67(1) 2.22(1) 0.67(4) — 0.0023a%P arbitrarily fixed at zero.Better agreement between the experimental data and theory, particularly in the region of thesusceptibility maximum, is obtained when only data at and above 40 K are used in the fittingprocedure. Values of J and g obtained from these high temperature data fits are also given inTable 2.9. The dashed lines in Figure 2.19 represent the best fit of theory to the data over thewhole temperature range and the solid lines represent the best fits of theory to the hightemperature data. Because the paramagnetic impurity conthbutes more to the lowtemperature susceptibility data the improvement in fit on using the high temperature dataprobably reflects inadequacies in modelling the paramagnetic component. Thus, as with the[Cu(4-Xpz)2]complexes, the paramagnetic components in the [Cu(4-Xdmpz)2]complexeshave also been modelled as Curie-Weiss law paramagnets. Addition of the fourth parameter,e (see [2.11]), produced an excellent fit over the entire temperature range (dotted lines inFigures 2.19(a), (b), and (d)) studied for the X = H and Br complexes. The value of 0 in the80X = CH3 case is not significantly different from zero. The J and g values obtained with thismodel agree well with those obtained using fits to the high temperature data only (Table 2.9).[Cu(4-Cldmpz)2]exhibits rather unique magnetic properties. Above 40 K it behavesmuch like the other 4-Xdmpz complexes, exhibiting a broad maximum in its susceptibilityversus temperature plot at 120 K. The magnetic data above 40 K are reasonably wellmodelled according to the isotropic Heisenberg model for exchange coupled linear chainsand the best fit values of J and g obtained using this model, ignoring paramagnetic impurity,are comparable to the values obtained for the other 4-Xdmpz compounds (Table 2.9). Below40 K, however, the susceptibility drops abruptly until about 8 K, below which the effects ofparamagnetic impurity are seen. The magnetic properties for this compound cannot bemodelled well below 40 K using the linear chain model with or without paramagneticimpurity. The abrupt decrease in susceptibility below 40 K is similar behaviour to thatobserved for CuSb2O6(99) which possesses the trirutile structure and behaves magneticallyas a linear chain antiferromagnet above 9 K. Nakua et a!. (99) attributed the rapid decrease insusceptibility below 9 K in CuSb2O6as a transition to long range antiferromagnetic order.One of the pieces of evidence they used to support this conclusion was an examination of thefield dependence of CuSb2O6.Nakua et a!. found that the magnetization of the compound at2 K (below T) over the applied field range 0-30 kOe exhibited two linear regimes with atransition region between. This behaviour is characteristic of a spin flop transition inpolycrystalline antiferromagnets. At 11 K (above T) they found the magnetization to becompletely linear over the same field range which is characteristic of the paramagnetic state.Noting the similarity between the temperature dependence of the susceptibility ofCuSb2O6 and [Cu(4-Cldmpz)2j, a study of the magnetic field dependence of [Cu(4-Cldmpz)2jwas undertaken. The compound’s applied field dependence was examined in therange 0-55 kOe at 5 K, 10 K, 15 K and 20 K. The results of this study are depicted in Figure2.20 (the magnetization data are listed in Table 111-3 in Appendix III). It is apparent that the81400E30E0Co 200N0Cc10000FIG. 2.20. Isothermal magnetization data for [Cu(4-Cldmpz)2]magnetization data for [Cu(4-Cldmpz)2]do not exhibit the spin flop transition evident in thedata for CuSb2O6.The magnetization plots at 10 K and especially 5 K for [Cu(4-C1dmpz)2]show some deviation from linearity at the highest fields. This is most likely due to magneticsaturation effects as the Brillouin function r value at 5 K and 55 kOe is 1.65 for thiscompound. Even though [Cu(4-Cldmpz)2]does not exhibit field dependence indicative of aspin flop transition in the range studied, this does not rule out the possibility of such atransition occurring in the compound. The molecular field theory (1) description of spin flopin antiferromagnets predicts that the transition field, HSF, at T = 0 K is dependent on both theanisotropy field and the antiferromagnetic exchange energy. If the anisotropy fields forCuSb2O6and [Cu(4-Cldmpz)2]are similar, then the HSF would be expected to be greater forthe compound with the larger antiferromagnetic exchange. [Cu(4-Cldmpz)2]exhibits itsanomalous decrease in susceptibility for an antiferromagnetically coupled linear chain in therange 30-40 K, whereas CuSb2O6exhibits this feature between 8-9 K. This would imply that8220000 40000 60000Applied Field (Oe)if long-range antiferromagnetic ordering is occurring in [Cu(4-Cldmpz)2]then the magnitudeof the interaction is much larger in this compound than in CuSb2O6and, as discussed above,one would expect HSF to be higher in the former compound than in the latter, perhaps beyondthe field range accessible in this study.The field dependence studies conducted here do not preclude the possibility of long-range order accounting for the unusual susceptibility data for [Cu(4-Cldmpz)2],but neitherdo they support such a conclusion. Other possibilities which might account for the behaviourof this compound are: a transition to 2-D ordering, a structural phase transition whichfacilitates much stronger l-D antiferromagnetic exchange, or a chain dimerization processresulting in the formation of an antiferromagnetically coupled alternating chain.Turning to a consideration of the magnitude of the exchange coupling in these [Cu(4-Xdrnpz)2Jcomplexes, some interesting comparisons can be made with the [Cu(4-Xpz)2]analogues. In both sets of compounds the IJI values for the 4-Cl and 4-Br derivatives are thesame within experimental uncertainty and are slightly, but significantly, greater than thevalues for the 4-H and 4-Me derivatives. An argument was presented in Section thatthe variation in observed coupling constants does not correlate with simple ideas based onrelative ligand basicities also applies to the 4-Xdmpz series of compounds. Whether the factthat the 1.11 values order in the same way in both the 4-Xpz and 4-Xdmpz compounds issimply coincidental or due to the intrinsic donor/acceptor properties of the X functionalgroups is not known. The experimental evidence suggests that the two classes of chaincompounds have very different intrachain structures and this probably accounts for the factthat the [Cu(4-Xpz)21complexes exhibit 1./1 values (range 81 to 105 cm1) which areconsistently higher than those of the [Cu(4-Xdmpz)2]complexes (range 58 to 67 cm1).832.2.3 MISCELLANEOUS POLY(COPPER(II) PYRAZOLATES) SYNTHESES, PHYSICAL AND THERMAL PROPERTIESDetails of the syntheses of [Cu(3-Mepz)2]and [Cu(indz)2]are described in Chapter7, Section 7.2.3. The MLPM reaction was successfully employed in the synthesis of [Cu(3-Mepz)2]as a green powder. [Cu(indz)2],on the other hand, was prepared as a brownish-black powder via reaction [2.12] using [Cu(indz)]3 (discussed in Chapter 3) as a startingmaterial.Like the other copper(II) pyrazolates discussed in this chapter, [Cu(3-Mepz)2]and[Cu(indz)21are insoluble in water and all common organic solvents and both compounds canbe dissolved with decomposition in mineral acids. However, as with three of the [Cu(4-Xdmpz)2]compounds, [Cu(indz)2Jdecomposes slowly if left in wet solvents. DSC studiesrevealed that [Cu(3-Mepz)2Jand [Cu(indz),] decompose, without melting, at 240 and 220°C, respectively. Neither of these compounds exhibited thermochromism over the 300-77 Ktemperature range.A SEM image of [Cu(indz)2Jis shown in Figure 2.21. It is interesting to note that inthis case the compound fails to show the individual crystallites present in the images of theother copper(ll) pyrazolates. Instead, rather amorphous looking masses appear. Themicroscopic morphology of [Cu(indz)2]is consistent with the fact that attempts to preparemacroscopic single crystals of the compound (as well as [Cu(3-Mepz)2])failed in everyinstance. X-RAY DIFFRACTION STUDIESSingle crystals suitable for X-ray diffraction studies were not prepared for [Cu(3-Mepz)2] and [Cu(indz)2Jso only powder X-ray diffractograms were recorded for these84compounds. These powder patterns are shown in Figures 2.22(a) and (b). The d-spacings andrelative intensities of the peaks in these diffraction patterns are listed in Appendix II, Table II-3. Their diffractograms demonstrate that the compounds are microcrystalline, but notisomorphous nor are they isomorphous with any of the other copper(II) pyrazolatesconsidered in this chapter. SPECTROSCOPIC BEHAVIOUR2. INFRARED SPECTROSCOPYUnassigned JR band frequencies and relative intensities for [Cu(3-Mepz)2jand[Cu(indz)21appear in Appendix IV, Table IV-3. The JR spectra of these compounds do notprovide information on the details of their structures, except insofar as they demonstrate theFIG. 2.21. SEM image of [Cu(indz)2J.The white line represents a length of 5 jim.85(a)05.025 35 40 45 50 55 6026(b)5.0 10 15 20 25 30 35 40 45 50 55 6026FIG. 2.22. Powder diffractograms of (a) [Cu(3-Mepz)jand (b) [Cu(indz)2].presence of the particular pyrazolate ion and the absence of N—H functional groups in thecompounds. Comparison of the isomeric copper(ll) 3-methyl- and 4-methylpyrazolatesreveals that the two compounds have similar JR spectra though there appears to be a greatersplitting of bands in the 3-Me derivative. This increased splitting of bands is consistent withthe lower nominal symmetry of the 3-Mepz ion, C, compared to that of the 4-Mepz ion, C2,.862. SPECTROSCOPYThe electronic spectra for [Cu(3-Mepz)2]and [Cu(indz)2]are shown in Figure 2.23.FIG. 2.23. Electronic spectra of (a) [Cu(indz)2Jand (b) [Cu(3-Mepz)2].As the colours of the two compounds would suggest, their electronic spectra are quitedifferent. [Cu(3-Mepz)2jexhibits a spectrum similar to those of the [Cu(4-Xpz)2Jcompounds with a broad, asymmetric band maximizing at 14,750 cm1 (678 nm) and a higherenergy band at 25,000 cm-’ (400 nm). The lower energy band is likely due to d-d transitionswhile the higher energy band is probably charge transfer in origin. This spectrum suggeststhat [Cu(3-Mepz)2jpossesses a distorted tetrahedral CuN4 chromophore like that of the[Cu(4-Xpz)2}compounds. [Cu(indz)2Jis a brownish-black material and in accord with thisit absorbs strongly throughout the visible region (a very dilute mull was used to obtain thespectrum shown in Figure 2.23). A band maximum is present at 25,000 cm4 (400 nm) and the500 1000 1500Wavelength (nm)87shoulder of a band occurs at approximately 18,000 cm1 (555 nm). In addition, a weak bandappears in the MR region at 10,000 cm1 (1,000 nm). A similar band was observed in theelectronic spectra of the [Cu(4-Xdmpz)2jcompounds. The strong, rather featureless,absorption across the visible spectrum makes interpretation of the spectrum difficult,although the chromophore seems likely to be different from that of the 3-Mepz derivative. ELECTRON PARAMAGNETIC RESONANCE SPECTROSCOPYEPR spectra for [Cu(3-Mepz)21and [Cu(indz)2jat 77 K are presented in Figures2.24(a) and (b). In both cases the spectrum consists of a single, broad, asymmetric line withthe linewidth of [Cu(3-Mepz)2jbeing considerably greater than that of [Cu(indz)2].Theroom temperature spectra of these compounds are similar; however the breadth of theresonance lines increase somewhat over those at 77 K. The spectra of [Cu(3-Mepz)2]and[Cu(indz)2J are generally similar to those of the [Cu(4-Xdmpz)2]compounds andconsequently they distinguish the two compounds under discussion here from the green [Cu(4-Xpz)2] compounds. Unfortunately, the spectra of these compounds do not provideinformation for definitive assignment of chromophore stereochemistry. MAGNETIC PROPERTIESAs with the [Cu(4-Xdmpz)2]compounds, it is useful to consider what tentativestructures may be assigned to [Cu(3-Mepz)2}and [Cu(indz)2]based on the informationpresented above before discussing their magnetic properties. Following the arguments usedfor the [Cu(4-Xdmpz)2Jcompounds in Section, the empirical formulae of [Cu(3-Mepz)2] and [Cu(indz)2]do not support the formation of monomeric compounds. Theinvolatile, infusible, insoluble nature of the compounds is consistent with the formation ofextended polymeric materials. The compounds are most likely chain polymers as are the88IIFIG. 2.24. EPR spectra of (a) [Cu(indz)2Jand (b) [Cu(3-Mepz)2]at 77 K.structurally characterized [Cu(4-Xpz)2] compounds, but steric considerations andspectroscopic evidence suggest the chromophores of [Cu(3-Mepz)2]and [Cu(indz)2Jdiffersomewhat from those of the [Cu(4-Xpz)2jcompounds.Electronic spectroscopy suggests that [Cu(3-Mepz)2]has the flattened tetrahedral(b)2OO G-l__LJ flhr_l89geometry of the green copper(ll) 4-substituted pyrazolates; however, the appearance of the dd transition band at slightly lower energy than those of the [Cu(4-Xpz)2]compoundssuggests a more regular tetrahedral geometry in the former compound. This supposition isconsistent with the steric requirements of the 3-methylpyrazolate ligand. Because 3-Mepzpossesses a substituent only at the 3-position of the ring, a copper(II) chain compound of thematerial need not necessarily adopt the stepped structure proposed for the [Cu(4-Xdmpz)2]complexes. [Cu(3-Mepz)2]could adopt a structure similar to the type possessed by the [Cu(4-Xpz)2]species (Figure 2.2), but in order reduce interaction between the methyl group of thepyrazolate ring and the face of an adjacent ring in the chain, the acute dihedral angle betweenadjacent Cu(N—N)2u rings would have to increase to a value greater than those determinedfor the [Cu(4-Xpz)2]complexes. In other words, in [Cu(3-Mepz)2]the CuN4 chromophorewould have to adopt a more regular tetrahedral stereochemistry. Such an alteration of thechromophore from those of the green 4-substituted copper(ll) pyrazolates might well cause achange in the electronic ground state of the copper(IT) ion. This may account for the differentappearance of the EPR spectrum of the copper(ll) 3-methylpyrazolate with respect to thespectra of the [Cu(4-Xpz)2]complexes.The electronic spectrum of [Cu(indz)2]suggests that the compound could possess asquare-coplanar stereochemistry. That being the case, the stepped chain structure proposedfor the [Cu(4-Xdmpz)2]complexes could be equally well proposed for the copper(II)indazolate. Additional support for this comes from the appearance of the EPR spectrum for[Cu(indz)2]which is similar to those observed for the [Cu(4-Xdmpz)2jcomplexes.Magnetic susceptibilities of [Cu(3-Mepz)21and [Cu(indz)2]were measured from 2to 300 K using a SQUID magnetometer. Powder magnetic susceptibility and effectivemagnetic moment versus temperature data for the compounds are tabulated in Appendix III,Table 111-4. The two compounds considered in this section exhibit strong antiferromagneticcoupling, in the range observed for the copper(ll) pyrazolates considered above. In both casesthe magnetic moments of these compounds decrease significantly with decreasing90temperature. x versus temperature plots, shown in Figures 2.25(a) and (b), revealI I I I 2000 I I I(a)- (b)cDC) ‘ 1500-: 10000.) 1000Cie ICl) 0 50 100 150 200 250 300500 I I I 500 I I I0 50 100 150 200 250 300 0 50 100 150 200 250 300Temperature (K) Temperature (K)FIG. 2.25. Powder susceptibility versus temperature plots for (a) [Cu(indz)2jand (b) [Cu(3-Mepz).The lines represent the best fits to theory and are calculated from the parameters in Table 2.10. The dotted lines are best fits to the hightemperature data only. The solid lines are fits to the full data ranges and thedashed line in (a) is the fit with a Curie-Weiss law paramagnetic impurity term.susceptibility maxima at 110 K and 160 K for the 3-Mepz and indz derivatives, respectively.Low temperature susceptibility behaviour in both compounds indicates the presence ofstructural paramagnetic impurity as was observed in the [Cu(4-Xpz)2]and [Cu(4-Xdmpz)2]91compounds. The amount of paramagnetic impurity in [Cu(indz)2]is substantially greaterthan in the other copper(I1) pyrazolates considered in this work as can be seen in the inset ofFigure 2.25(a). The large and unsymmetrical nature of the bridging indazolate ligand maycontribute to this relatively large paramagnetic impurity content through the increasedformation of structural defects. The SEM study of [Cu(indz)2]concurs with this propositionas it indicated a very small microcrystallite size for the complex. In accord with the structuretypes proposed above, the magnetic data for [Cu(3-Mepz)2Jand [Cu(indz)2]were analyzedusing the isotropic Heisenberg model for exchange coupled linear chains modified to accountfor the presence of paramagnetic impurities as described in Section Reasonable fitsbetween experimental data and theory were obtained for the compounds and best fit values ofthe J, g, and %P parameters along with the corresponding F values are given in Table 2.10.The calculated curves for these fits are shown in Figures 2.25(a) and (b).TABLE 2.10. Derived magnetic parameters for [Cu(3-Mepz)2jand [Cu(indz)2jwithestimated standard deviation in the last digit in parenthesesTemperature -J -eCompound range (K) (cm-’) g %P (K) F[Cu(3-Mepz)2} 2-300 62(1) 2.12(1) 0.58(2) — 0.0252-300 62(1) 211(1) 0.55(4) 0.2(2) 0.02540-300 62(1) 2.12(1) oa — 0.014[Cu(indz)2] 2-300 92(1) 2.12(2) 1.65(2) — 0.0152-300 93(1) 2.12(1) 1.78(2) 0.31(3) 0.007740-300 82(2) 2.04(3) oa — 0.036a%p arbitrarily fixed at zero.Inspection of the best fit parameters in Table 2.10 indicates that modelling of theparamagnetic impurity component of the susceptibility in [Cu(3-Mepz)2]as a Curie-Weisslaw paramagnet yielded no significant change in the best fit parameters nor an improvementof the fit quality factor, F. Modelling of the high temperature data in the case of [Cu(3-92Mepz)2}yielded essentially the same J and g values as modelling of the data over the fulltemperature range. In the case of [Cu(indz)2], inclusion of the Weiss parameter in theparamagnetic impurity term did significantly improve the quality of the theoretical fit to thedata; however, it caused no significant change in the values of J and g. Moreover, thepresence of a greater amount of paramagnetic impurity in [Cu(indz)2Jthan in the othercopper(ll) pyrazolates studied here is borne out by the value of %P. The best fit to the hightemperature data for [Cu(indz)2]is significantly poorer than the other two fits over the fulltemperature range and the cause of this anomaly is the presence of a large paramagneticimpurity component in the material. In modelling the high temperature data %P was fixed atzero; however, in the case of [Cu(indz)21the paramagnetic impurity still makes a substantialcontribution to the total susceptibility at temperatures above 40 K thus accounting for thepoor fit.If chromophore geometry is an important determinant of the magnitude of exchangein these chain copper(II) pyrazolates then, based on what is known and supposed about the[Cu(4-Xpz)2]and [Cu(4-Xdmpz)2Jcomplexes, one might have expected [Cu(3-Mepz)2jtoshow stronger exchange coupling than [Cu(indz)2j;the opposite is observed experimentally.That [Cu(3-Mepz)2]exhibits a 1.11 value much lower than any of those determined for the[Cu(4-Xpz)2]compounds could be accounted for using the model discussed in Section2.2.1.4 and depicted in Figure 2.11(b). It was proposed above in this section that [Cu(3-Mepz)21possesses a structure similar those of the [Cu(4-Xpz)2]complexes, but that the 13angle (Figure 2.11(b)) is smaller in [Cu(3-Mepz)2]than in the latter complexes. Accordingto the exchange mechanism proposed in Section a smaller value for 13 shouldcorrespond to a weaker exchange interaction as is, in fact, observed for [Cu(3-Mepz)2J.[Cu(indz)2]is proposed to possess a structure similar to the type proposed for the [Cu(4-Xdmpz)2]compounds yet it exhibits a strength of exchange much greater than the rangefound for the latter compounds. In these cases the precise structural details are not known for93the compounds and an orbital pathway has not been defined so no structural argument is putforward to account for the much stronger exchange. It is possible that the extended it-systemof the indazolate ligand compared with the other pyrazolates examined in this work is thesource of this increased antiferromagnetic exchange interaction.2.3 SUMMARY AND CONCLUSIONSA series of eleven copper(ll) (substituted)pyrazolates has been prepared andcharacterized by a variety of physical techniques. The structures of three of these compounds,namely [Cu(4-Hpz)2],[Cu(4-Mepz)2J,and [Cu(4-Clpz)2](green), have been determined bysingle crystal X-ray diffraction and they were found to be isomorphous and isostructural.These three compounds consist of linear chains of distorted tetrahedrally coordinatedcopper(II) ions doubly-bridged by pyrazolate anions. Powder X-ray diffraction studies andthe physical properties of [Cu(4-Brpz)2]indicate that it is isomorphous and probablyisostructural with the other green [Cu(4-Xpz)2]compounds. The green [Cu(4-Xpz)2]compounds exhibit thermochromism at low temperatures and investigation of the temperaturedependence of the structures of these compounds revealed that they segregate into two groupsin terms of the structural transformations. [Cu(4-Clpz)2]was prepared as green and brownpolymorphs which exhibited different crystal structures and physical properties. Studies ofthe other copper(ll) substituted pyrazolates by indirect methods led to the proposal that theyare also chain compounds of copper(II) ions doubly-bridged by pyrazolate anions. [Cu(3-Mepz)2Jis thought to possess a distorted tetrahedral CuN4 chromophore like the green [Cu(4-Xpz)2] complexes while the [Cu(pz*)2]x (pz* = 4-Xdmpz, indz, and 4-Clpz (brown))complexes are thought to possess square-coplanar CuN4 chromophores.Studies of the magnetic properties of the copper(II) (substituted)pyrazolates revealedthat all of the compounds exhibit strong, one-dimensional antiferromagnetic coupling with Jvalues ranging from -58 to -105 cm1. [Cu(4-Brpz)2jand [Cu(4-Clpz)2J(green) exhibit94discontinuities in their versus temperature plots which are probably due to structuraltransformations in the complexes. [Cu(4-C1dmpz)2]shows an anomalously rapid decrease inmagnetic susceptibility at low temperatures for an antiferromagnetically coupled linear chaincompound suggestive of a structural transition or a magnetic transition to a higherdimensionality.95CHAPTER 3OLIGOMETALLIC COPPER PYRAZOLATES3.1 INTRODUCTIONDuring efforts to synthesize macroscopic single crystals of some of the [Cu(4-Xdmpz)2]complexes, several unexpected oligometallic copper(I), copper(ll) and mixedvalence copper(I/II) complexes were obtained. The discovery of these compounds led todevelopment of a rational synthesis for a number of trimeric copper(I) pyrazolates. Thischapter describes the preparation of these compounds and their characterization.3.2 RESULTS AND DISCUSSION3.2.1 COPPER(1) PYRAZOLATES3.2.1.1 SYNTHESES, PHYSICAL AND THERMAL PROPERTIESPrior to this work a number of univalent copper and silver (substituted)pyrazolateshad been synthesized. This earlier work was reviewed in Chapter 1. Details of the synthesesof the compounds prepared in this current study are provided in Chapter 7, Section 7.2.2.Univalent coinage metal ions are diamagnetic, so the pursuit of coordination compounds ofsuch ions was not motivated by their potential magnetic properties. In fact, the current studyof copper(I) pyrazolates was engaged as a result of the unintentional preparation of copper(I)3,5-dimethylpyrazolate. During an attempt to synthesize single crystals of [Cu(4-Hdmpz)2Jfrom Cu(OH)2 in molten 3,5-dimethylpyrazole, large single crystals of colourless [Cu(4-96Hdmpz)13 formed as the major product (-‘-70%) and [Cu(4-Hdmpz)2]formed as the minorcopper containing constituent (—30%):N2, 130 °CCu(OH)2+xs 4-HdmpzH 10 days >[Cu(4-Hdmpz)] + [Cu(4-Hdmpz)2+? [3.1]—30%It is apparent that the copper(II) starting material is partially reduced in this reaction andconsequently the 3,5-dimethylpyrazole reactant must be oxidized. The oxidation products of3,5-dimethylpyrazole have not been determined, but oxidation of pyrazoles, albeit with morecommon oxidants, has been documented (100). The great value of reaction [3.1] is that ityielded large, X-ray diffraction quality single crystals of [Cu(4-Hdmpz)]3which had notpreviously been obtained; however, a more amenable synthetic route was necessary in orderto conveniently prepare bulk samples of this and other copper(I) pyrazolates. Such a routewas developed as a variation of the synthesis of powdered [Cu(4-Hdmpz)]3as described byArdizzoia et a!. (101). CuT is dissolved in acetonitrile and combined with a stoichiometricamount of some pyrazole, then a stoichiometric amount of triethylamine is added, rapidlyyielding the desired copper(I) pyrazolate as a white powder:N2, R. T.3 Cul + 3 pz*H + 3EtN— [Cu(pz*)]3.J, + 3 [EtNHJ[l] [3.2]CH3Nwhere pz = 4-Hdmpz, 4-Medmpz, 4-Cldmpz, 4-Brdmpz, 4-Idmpz, and indz. The motivationfor preparation of these compounds was the discovery that the copper(I) pyrazolates wereexcellent starting materials for the preparation of a number of copper(IT) pyrazolates whichcould not be prepared in pure form using the MLPM reaction (see Chapter 2). [Cu(4-Hdrnpz)]3 was not the only copper(I) pyrazolate for which X-ray diffraction quality singlecrystals were prepared; crystals of [Cu(4-Medmpz)13were synthesized via the reaction97presented below:limited 02Cu + xs 4-MedmpzH > [Cu(4-Medmpz)]3 —50% [3.3]140°C,40h +[Cu(3-COdnipz)(4-Medmpzfl)]u --5O9The second product from reaction [3.3] is discussed in Section 3.2.2.The [Cu(4-Xdmpz)]3compounds and [Cu(indz)] are insoluble, or at best, sparinglysoluble in common organic solvents. They are colourless solids and are insensitive to 02when dry; however, in water containing solvents they are oxidized to yellow-green solublespecies. Such an aerial oxidation product of [Cu(4-Hdmpz)]3was characterized by Ardizzoiaet al. (9 1,92) and mentioned in Chapter 2. The [Cu(4-Xdmpz)]3complexes are thermallyrobust. The X = Me complex melts at 308 °C and thermally decomposes at 365 °C. The X =H, Cl, Br, and I compounds decompose, without melting, at 300 °C, 350 °C, 300 °C, and 300°C, respectively. X-RAY DIFFRACTION STUDIESPreviously, workers had concluded that [Cu(4-Hdmpz)13 is a polymeric material(36,101). The preparation of this compound and [Cu(4-Medmpz)] during this work in singlecrystal form has permitted an unequivocal determination of the structures of both compoundsby X-ray diffraction and revealed them to possess the trimeric ring structures illustrated inFigures 3.1 and 3.2. Crystallographic data, atomic coordinates, and bond lengths and anglesappear in Appendix I, Tables I-li through 1-19. As can be seen in Figures 3.1 and 3.2, thecompounds are planar trimetallocycles with central Cu3N6 rings (the rings are planar towithin 0.07 A and 0.06 A in [Cu(4-Hdmpz)]3 and [Cu(4-Medmpz)],respectively). The98FIG. 3.1. ORTEP diagrams of the two crystallographically independent molecules of[Cu(4-Hdmpz)13;50% probability thermal ellipsoids are shown for the non-hydrogen atoms.copper(I) centres in these molecules are approximately linearly coordinated by two pyrazolylnitrogen atoms. The N—Cu—N bond angles range from 173.4-175.3° (average 174.3°) in[Cu(4-Hdmpz)]3and 173.9-174.7° (average 174.4°) in [Cu(4-Medmpz)1.A curious featureof both compounds is that they are found associated in pairs in their crystal lattices. Thisassociation, illustrated in Figure 3.3, may be mediated by weak Cu... Cu interactions and will99FIG. 3.2. ORTEP diagram of [Cu(4-Medmpz)J3;50% probability thermal ellipsoidsshown for the non-hydrogen atoms.FIG. 3.3. Stereoscopic view of [Cu(4-Hdmpz)13showing the thmer-trimer Cu... Cuinteractionsbe discussed in more detail below.100A number of related Cu(1), Ag(1), and Au(I) compounds have been reported recentlyin the literature (46-48). Selected structural parameters from these earlier reports and those of[Cu(4-Hdmpz)]3 and [Cu(4-Medmpz)13are given in Table 3.1 and relate to the generaltrimetallocycle formulation:RNRN- MNQ11 ROf the six compounds listed in Table 3.1, only [Cu(4-Hdmpz)13,[Cu(4-Medmpz)J3,and[Au(3,5-(CF3),pz)] exhibit essentially planar central nine-membered M3N6 rings whereasthe other three compounds show a marked puckering in their central rings. It is interesting tonote that the molecular arrangement depicted above for the [Cu(4-Xdmpz)13complexes isalso found in the structure of 3,5-dirnethylpyrazole itself (103). The structure of 4-HdmpzHconsists of planar trimeric units with strongly H-bonded hydrogen atoms between the threepyrazole molecules replacing the Cu atoms of the [Cu(4-Xdmpz)13 metallocycles.Comparison of the Cu—N bond lengths of [Cu(4-Hdmpz)J3and [Cu(4- Medmpz)13with thoseof [Cu(3,5-Ph,pz)] shows the bond lengths of the former two compounds to be considerablyshorter than those of latter compound. Further comparisons with the Cu—N bond lengths inother copper(J) complexes linearly coordinated by aromatic nitrogen heterocycles (104)indicate that the Cu—N bond lengths in the [Cu(4-Hdmpz)]3 and [Cu(4-Medrnpz)13complexes are not atypical.RR101TABLE 3.1. Selected structural parameters for [M(4-X-3,5-R2pz)}3species (distances in A)Compound M-N M-N M-N M. . .M M.. .M M.. .MM R X obsa avg est intra intrav inter Reference1.845 3.195Cu Me H to 1.852 2.04 to 3.2 18 2.946 This work1.858 3.2571.848 3.114Cu Me to 1.853 2.04 to 3.205 3.069 This work1.855 3.2672.041 3.280Cu Ph H to 2.08 1 2.04 to 3.339 472.105 3.4062.08 3.305Ag Ph H to 2.085 2.19 to 3.388 482.09 3.486Au Ph H 1.978 2.06 3.368 7.567 481.89 3.344Au CF3 H to 1.93 2.06 to 3.348 3.998 461.96 3.355average; est, estimated (102); intra, intramolecular; intrav, averageaAbbreviations: obs, observed; avg,intramolecular; inter, intermolecular.Returning to a consideration of the inter-trimer pairing mentioned above, it may benoted that of the compounds listed in Table 3.1 only the two compounds studied in this workexhibit inter-trimer metal atom separations which are smaller than the corresponding intratrimer metal atom separations. In the two [Cu(4-Xdmpz)j3 complexes studied here, thetrimers are arranged in pairs about crystallographic inversion centres such that two copperatoms from one Cu3N6metallocycle are positioned close to two copper atoms from a secondtrimer with the third copper atom in each case being positioned above the centre of thepyrazolyl ring at the opposite end of the inversion-related trimer unit. The distances of thisthird copper atom from the mean plane of the adjacent pyrazolyl ring in [Cu(4-Hdmpz)]3and102[Cu(4-Medmpz)13are 3.272 A and 3.28 1 A, respectively. The inter-trimer separations in[Cu(4-Hdmpz)] and [Cu(4-Medmpz)1 (Table 3.1) are slightly longer than the sum of thevan der Waals radii (2.86 A) (102) and thus the interaction between trimers must necessarilybe very weak. The only related trimeric structure to show similar inter-trimer interactions isthe [AuC(OEt)=N(p-tolyl)]3complex (105). In this structure the intra-trimer Au... Audistances are reported at 3.224, 3.288, and 3.299 A with one inter-thmer Au... Au distance of3.224 A responsible for joining the trimers in pairs. In this example the nature of theinteraction is not specified, but again is likely to be weak since the distance observed is of thesame order as the sum of the van der Waals radii (3.22 A) (102). Whether the close approachof the triniers in the structures of [Cu(4-Hdmpz)]3and [Cu(4-Medmpz)]3is the result of atrue metal-metal interaction or simply an artifact of the crystal packing forces in thesecompounds is not known. SPECTROSCOPIC STUDIES3. INFRARED SPECTROSCOPYThe JR spectra for the copper(J) pyrazolate complexes were recorded principally asinitial checks of product purity. As was explained in Chapter 2, Section, diagnosticbands in the 700-800 cm1 region of the spectra were used to determine when reactionsemploying the copper(I) pyrazolates as starting materials were complete. Unassigned JR bandfrequencies and relative intensities for the compounds appear in Appendix IV, Table IV-4. Inthe dimethyl species, the highest energy, in-plane Vflflg modes of the pyrazolyl moieties occurat 1,525 cm1, 1,500 cm1, 1,520 cm1, 1,512 cm1, and 1,499 cm1 for the 4-H, 4-Me, 4-Cl, 4-Br, and 4-I derivatives, respectively. These values are consistent with the presence of 4-X-3,5-dimethylpyrazolate anions in the compounds. Furthermore, the 13CCH3 vibrations in the 4-H, 4-Me, 4-Cl, 4-Br, and 4-I derivatives appear at 464 cm1, 530 cm1 (581 cm1 for the meta103methyl group), 532 cm-1, 521 cm, and 516 cm-1, respectively. All of these values are shiftedto significantly higher energies in comparison to the corresponding neutral pyrazoles (seeChapter 2, Section This is consistent with the presence of only ortho methylgroups on the pyrazolate rings and, consequently, the presence of only bridging, coordinatedpyrazolate ligands. MASS SPECTROMETRYElectron impact mass spectra were recorded for all of the copper(I) pyrazolates andthe relative intensity versus rnlz plots appear in Appendix V, Figures V-i through V-6. All ofthe copper(I) compounds studied here exhibit electron impact mass spectra in which the mostintense peak is that due to the molecular ion expected for the thmer. The isotopic patterns arein excellent agreement with those predicted. This coupled with the facts that the 4-Hdmpzand 4-Medmpz derivatives were determined to have trimeric structures (Section andthat, with the exception of the compound hexakis(p.-3 ,5-diphenylpyrazolato-N,N’)hexagold(I)(48), all other X-ray structurally characterized univalent coinage metal pyrazolates aretrimers (Table 3.1) leads to the conclusion that the copper(I) 4-halo-3,5-dimethylpyrazolatesand copper(I) indazolate are all cyclic trimeric compounds.3.2.2 COPPER(I) AND COPPER(II) CARBOXYLPYRAZOLATE COMPLEXES3.2.2.1 SYNTHESES, PHYSICAL AND THERMAL PROPERTIESEfforts to synthesize single crystals of some copper(ll) 4-substituted 3,5-dimethylpyrazolates led, instead, to the isolation of unexpected oxidation products. Detaileddescriptions of the syntheses of these complexes are provided in Chapter 7, Section 7.2.4.One such compound appears in reaction equation [3.3]. The copper(I/II) mixed oxidation104state trimetallic species, [Cu(3-CO2dmpz)(4-MedmpzH)]u, was prepared as pale purplecrystals. The compound is insoluble in common organic solvents and thus could not beseparated from the other product of the reaction, [Cu(4-Medmpz)13.A few of the purplecrystals were separated manually from the reaction mixture for X-ray diffraction studies. Itshould be noted that examination of the reaction mixture from [3.31 prior to separation of theproducts from the unreacted copper metal shot revealed that the crystals of [Cu(4-Medmpz)]3formed a layer immediately surrounding the copper beads and [Cu(3-CO2dmpz)(4-MedmpzH)]2Cu formed a second layer surrounding the layer of [Cu(4-Medmpz)13.Thisobservation as well as the presence of copper in both the +1 and +2 oxidation states in theproducts suggests a stepwise oxidation of copper metal and also suggests that [Cu(4-Medmpz)13 may be a precursor of [Cu(3-CO,drnpz)(4-MedmpzH)J2u. Conversion ofcopper(I) to copper(II) in such complexes was amply demonstrated in the syntheses ofseveral copper(II) pyrazolate complexes from the corresponding copper(I) compounds asdescribed in Chapter 2. If [Cu(4-Medmpz)]3 is the precursor of [Cu(3-CO2dmpz)(4-Medmpz)j2Cu, and this is by no means certain, then the oxidation of the methyl groups seenunder the conditions employed here presumably takes place on coordinated 4-Medmpz. Theoxidation of free 4-MedmpzH itself cannot be ruled out; however, the presence of metalliccopper does seem to be required, acting perhaps as a catalyst, since no carboxylate containingproducts were observed in the reaction of [Cu(4-Medrnpz)]3with dioxygen in the absence ofmetallic copper (see Chapter 2).Two conformers of another carboxylate containing complex were preparedunintentionally during the attempted preparation of macroscopic single crystals of [Cu(4-Brdmpz)2]Reaction of copper shot with molten 4-BrdmpzH in air, with vigorous stirring,yields over a 24 h period, a brown powder identified by elemental analysis to be largely[Cu(4-Brdmpz)2].An easily recognized impurity in the brown powder is a small amount of agreen substance present as prisms. By carrying out the reaction with a quiescent mixture ofreactants the green crystals became the major product. The green crystals (two forms) were105separated manually from minor amounts of unidentified by-products. The crystals wereidentified as the copper(II) complex [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)].The reaction ispresented below:140 °CCu + 4-BrdmpzH 3 days> [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)]+ H20 +? [3.4]two formsAs stated above, the complex occurs in two slightly different conformations as a result ofcrystallization in a monoclinic and a thclinic modifications. The two modifications could notbe separated in bulk samples and subsequent studies were conducted with a mixture of thetwo forms; however, it is estimated that the mixture is >99% triclinic form. IR bandfrequencies and relative intensities for this material are in Appendix IV, Table IV-5.Although the conditions employed here were similar to those described in [3.3] for the 4-MedmpzH reaction, no copper(J) species were seen to form in any significant amounts in thiscase.DSC measurements on [Cu(4-Br-3-CO,Mepz)(4-BrdmpzH)2],in air (Figure 3.4),show the onset of an endothermic event followed immediately by exotherrnic decompositionat -‘-250 °C. To investigate the possibility that the endothermic event corresponds to loss ofneutral 4-BrdmpzH, and to identify possible products of such a dissociation reaction, asample of the complex was heated at 165 °C under vacuum for 9 h. During the heatingprocess, the compound changed colour from green to olive-green and a white solid was seento sublime from the sample. The white product was identified as 4-BrdmpzH by meltingpoint (117 to 118 °C) and by ‘H-NMR (& 2.21 (s, 6, CH3 11.2 (br s, 1, NH)). The sample of[Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)],underwent a mass loss corresponding to the loss of2.5 moles of 4-BrdmpzH per mole of compound. These observations are consistent with theformation of a compound of nominal compositionCu4(4-Br-3-CO2Mepz)- rdmp H)3106I I I I —LI I I I I50 100 150 200 250 300 350Temperature (°C)FIG. 3.4. DSC thermogram of [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)2].Anal. calcd. forC35H3Br7u4N14O8:C 26.4, H 2.1, N 12.3; found: C 26.1, H 2.3, N 12.4.Finally, with regard to the thermal properties of [Cu(4-Br-3-CO,Mepz)(4-BrdmpzH)] it isnoted that, like the copper(II) 4-substituted pyrazolates discussed in Chapter 2, the complexconsidered here also exhibits thermochromism; it turns from green at room temperature tobluish-green at 77 K. X-RAY DIFFRACTION STUDIESAttempts to find a method for synthesis of bulk samples of [Cu(3-CO2dmpz)(4-MedmpzH)]2Cu in pure form were not made and because a bulk sample of the compoundwas not isolated, its characterization was limited to a single crystal X-ray diffraction studydescribed here and an EPR study discussed in Section A stereoview of [Cu(3-107CO2dmpz)(4-MedmpzH)]uis shown in Figure 3.5. The complex is centrosymmetric andFIG. 3.5. Stereoview of [Cu(3-CO2dmpz)(4-MedmpzH)j,Cu; 33% probabilitythermal ellipsoids are shown for the non-hydrogen atoms.the entire molecule, apart from the hydrogen atoms, is planar to within 0.2 A.Crystallographic data, atomic coordinates, and bond lengths and angles are listed inAppendix I, Tables 1-20 through 1-24. The molecules stack in the solid state and this is shownin Figure 3.6. The shortest intermolecular distance is Cu(2)—O(1)(x, y, z-1) = 3.145(6) A. The3-CO2dmpz ligands chelate the central copper(ll) ion (Cu(1)) via the carboxylate oxygenatom (0(1)) and the adjacent ring nitrogen atom and also form bridges between Cu(1) andboth of the copper(I) centres (Cu(2) and Cu(2)*): one via a pyrazolate -N—N- bridge and theother via a single-atom bridge involving 0(1). This latter bridge involves a very weak copperoxygen interaction in which the distance involved (Cu(2)—0(1) = 2.638(5) A) is long for anormal bond between copper(I) and oxygen but is significantly shorter than the sum of vander Waals radii (2.95 A) (103). The second oxygen atom of the carboxylate group is not108FIG. 3.6. Stereoview of the packing arrangement for [Cu(3-C02dmpz)(4-MedmpzH)]u.involved in coordination to the metal ions but contributes to the stability of the complex byhydrogen-bonding to the N—H hydrogen atom of the terminal monodentate 4-MedmpzHligand coordinated to Cu(2) (N(4)—H(1). . .0(2), H.. .0 = 1.88 A, N.. .0 = 2.859 A).The central copper(H) ion, Cu(1), has a trans square-planar coordination geometrywith Cu—0 = 1.978(5) A, Cu—N = 1.889(5) A. The 0—Cu—N angle within the five-memberedchelate is 83.0(2)°. The copper(I) ion, Cu(2), has a T-shaped coordination geometry featuringa nearly linear N—Cu—N unit with strong Cu—N bonds (Cu—N = 1.860(6) A and 1.869(6) A,N—Cu—N = 173.8(3)°) perpendicular to the very weak Cu—0 bond involving theasymmetrically bridging carboxylate oxygen atom, 0(1) (N—Cu—0 = 90.8(2) and 93.5(2)°).[Cu(4-Br-3-CO2Mepz)(4- rdmpzH),1,shown in Figures 3.7(a) and (b), is a dimericcopper(II) complex in which the 4-Br-3-CO2Mepz ligand is both bridging (Cu—N—N—Cu)and chelating (one carboxylate oxygen atom and the adjacent ring nitrogen atom).109(b)FIG. 3.7. Stereoviews of [Cu(4-Br-3-CO2Mepz)(4- rdmpzH)1;(a) monoclinic form,(b) triclinic form.Crystallographic data, atomic coordinates, and bond lengths and angles for the twopolymorphs of [Cu(4-Br-3-CO,Mepz)(4-BrdmpzH)21,are listed in Appendix I, Tables 1-25through 1-33. In the triclinic form of [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)jthe moleculepossesses a centre of symmetry and in the monoclinic form it possesses a two-fold axis. Thecentral Cu(N—N),Cu six-membered ring is exactly pianar in the triclinic form while in themonoclinic form it is non-planar, having a boat conformation. The maximum deviation fromthe mean Cu(N—N),Cu plane in the monoclinic form is 0.330(5) A for N(2).The coordination geometry about the copper centres in both forms of [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)j1is axially compressed trigonal bipyramidal with the carboxylate110oxygen atom and a nitrogen atom of the bridging 4-Br-3-CO2Mepz ligand occupying theaxial positions. The chelating nitrogen atoms of the 4-Br-3-COMepz ligand and nitrogenatoms of the two terminal 4-Brdmpzll ligands comprise the equatorial ligand set. There issome distortion of the thgonal bipyramid towards a square-pyramidal geometry in whichN(5) is apical. This may be responsible, along with steric factors, for the significantlengthening of the Cu—N(5) bond in both polymorphs. The axial bond lengths in the twoforms do not differ significantly (Cu—O = 1.992(5) A and 1.982(4) A, Cu—N = 1.949(5) A,respectively for the monoclinic and triclinic forms). The three equatorial Cu—N bonds in bothforms are significantly different from one another. For the terminal 4-BrdmpzH ligands, thecorresponding bond lengths in the two crystal forms also differ significantly. The averageequatorial Cu—N distances in the two forms are, however, nearly identical: 2.100 A for themonoclinic form and 2.107 A for the triclinic form.Even though the molecular symmetry of [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)1differs in the two polymorphs, the hydrogen bonding motif remains the same. In bothcrystalline modifications of this complex the molecules are linked about crystallographicinversion centres by N—H.. .0 hydrogen bonds involving the N—H hydrogen atoms of theterminal 4-Brdmpzll ligands to form infmite chains extending along the crystallographicc-axes; this is shown in Figures 3.8(a) and (b). Each linkage involves two weak and twostrong N—H.. .0 hydrogen bonds. The strong hydrogen bonds involve the N(4) hydrogen atomand the uncoordinated carboxylate oxygen atom, 0(2). The weak hydrogen bonds are formedby the N(6) hydrogen atom and the metal-coordinated oxygen atom, 0(1), these atoms beinginvolved in both intra- and intermolecular hydrogen bonds. ELECTRON PARAMAGNETIC RESONANCE SPECTROSCOPYAs was mentioned in Section, [Cu(3-C0dmpz)(4-MedmpzH)]ucould notbe separated in bulk from [Cu(4-Medmpz)13.Nonetheless, since [Cu(4-Medmpz)] is EPR111(a)(b)FIG. 3.8. Stereoviews of the packing arrangement in [Cu(4-Br-3-CO2Mepz)(4- dmpzH)];(a) monoclinic form, (b) triclinic form.silent it was possible to obtain EPR spectra of [Cu(3-CO,dmpz)(4-MedmpzH)]2uemploying the mixture of the two complexes ground to a fine powder. The room temperaturespectrum is shown in Figure 3.9. The spectrum shows two regions of absorption centred atand gj. Nuclear hyperfine splitting of g11 is observed with two of the four componentsresolved and the other two obscured by the g1 absorption. There is no observable splitting ofg1. The EPR parameters, calculated to first order are g11 = 2.202, g1 = 2.048 and 1A1 = 132 x1O cm* There is no significant change in the spectrum at 77 K. The two g value spectrumis as expected for square-coplanar copper(ll) with negligible exchange coupling (67).Exchange effects normally broaden the EPR spectra of ‘concentrated” copper(II) complexes112IFIG. 3.9. Room temperature EPR spectrum of [Cu(3-CO2dmpz)(4-MedmpzH)]u.in the solid state to such an extent that hyperfine data are masked. The observation ofhyperfine splitting of g11 in powdered samples is relatively rare; however, it has been observedpreviously, for example, in a number of tetrakis(pyridine) complexes of copper(II) (106). MAGNETIC PROPERTIESThe magnetic susceptibility of [Cu(4-Br-3-.CO,Mepz)(4-BrdmpzH)2]was measuredfrom 2 to 300 K using a SQUID magnetometer. Powder magnetic susceptibility and effectivemagnetic moment versus temperature data for the compound are tabulated in Appendix III,Table 111-4. A plot of versus temperature (20-300 K) is shown in Figure 3.10. Thesusceptibility passes through a broad maximum at 130 K, typical for twoantiferromagnetically coupled copper(I1) centres. As was observed in the magnetic behaviourof the compounds discussed in Chapter 2, the susceptibility rises rapidly at the lowestH >113200001500o 1000C)U)zCl)50050 100 150 200 250 300Temperature (K)FIG. 3.10. Plot of versus temperature for [Cu(4-Br-3-COMepz)(4-BrdmpzH)].Lines represent the best fit to the data of the Bleaney-Bowers equation with a Curie lawparamagnetic impurity term (dotted line) and a Curie-Weiss law paramagnetic impurityterm (solid line).temperatures studied due to the presence of paramagnetic impurity. The magneticsusceptibilities (2-300 K) were fitted to the theoretical expression given by Bleaney andBowers for isotropically exchange coupled dimeric copper(JI) systems (107),Ng2.tXbb kT [3.5]In addition, a term was included to account for the paramagnetic component as was done inChapter 2. Thus the actual expression used to fit the experimental data was [2.9] where [3.5]114replaced Xchajn in [2.9] and [2.8] was used to model the paramagnetic component. The best fitwas achieved by minimizing the F value defined in equation [2.10]. This modellingprocedure gave best fit values of J, g, and %P of -74.0(6) cm1, 2.11(1), and 5.15(7)%,respectively with the fitting function F = 0.029. A better fit was obtained by modelling theparamagnetic component as a Curie-Weiss paramagnet, equation [2.1 1]. This yielded best fitvalues of J, g, %P, and 0 of -75.4(1) cm’, 2.123(2), 5.59(2)%, and -0.46(2) K, respectively(F = 0.0061). These parameters are considered valid for the triclinic form of [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)]This assumption is considered reasonable because the similarstructures of the monoclinic and triclinic forms of the complex, particularly in terms of bondlengths and angles associated with the bridging atoms, are not likely to yield widely differingmagnetic properties. Hence, the presence of <1% of the monoclinic form in the sample wouldnot be expected to cause the observed susceptibilities to be measurably different from thoseof the pure triclinic form.Studies on eleven copper(ll) (substituted)pyrazolate chain compounds (Chapter 2)revealed J values ranging from -58 cm1 to -105 cm’. The value of J = -75.4 cm1 obtainedhere for [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)]lies in the middle of this range. It isinteresting to note that in spite of the fact that [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)]isdimeric and has five-coordinate copper(I1) centres, the exchange between the metal centres isnot greatly different from that observed in the polymers. It should be pointed out that J fromthe Bleaney-Bowers equation and J from the Hall expression for the Bonner-Fisher model areonly quasi-comparable quantities. The former represents half of the energy separationbetween the ground state (in this case a singlet state) and the excited state (triplet), while thelatter is a measure of the spread of paramagnetic states in a 1-D energy band where the spinpaired states lie lowest. A more useful comparison might be made between [Cu(4-Br-3-CO2Mepz)(4-BrdrnpzH)],and other copper dimetallic complexes containing the bridgingpyrazolate moiety.There appears to be no previous report of a purely pyrazolate bridged dimetallic115copper(II) complex which has been characterized magnetically and by X-ray diffraction.There are a number of copper(II) complexes which contain bridging pyrazolates plus otherbridging ligands which have been characterized by single crystal X-ray diffraction studies.Mazurek (108), Nishida (109), Doman (110) and co-workers have reported on a series ofpyrazolate bridged, Schiff-base, square planar copper(II) dimetallic species where the Schiffbase, 2, is derived from a diamino straight chain alcohol backbone and salicylaldehyde (in 2,m= 1 or2andn= 1 or2).H\(CH) (CH)2rn 2n2In these complexes the copper(II) ions are also bridged by an alkoxide group. The J values inthese complexes range from -120 cm1 to -297.5 cm1. It must be recognized that the alkoxidebridge is very effective in propagating magnetic exchange and is likely to provide animportant exchange pathway in these complexes making direct comparisons of these J valueswith that obtained for [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)],inappropriate. The dimetalliccomplex ICu(L-S)(4-Hpz)(CH3OHYI (L-S = 3) has been reported to be only weaklyantiferromagnetic with J = -1.8 cm (111). In addition to a pyrazolate bridge this compoundhas a bridging thiophenolate ligand and it seems likely that the latter is having a significanteffect on the exchange, again making comparisons with the compound studied in the presentwork inappropriate.H1160-NPerhaps the most useful comparisons to be made with [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)21are with the dimetallic complexes reported by Ajó, Bencini, and Mani (112)and by Prins et a!. (113). The former authors reported magnetic studies on the species[(HB(4-Hpz),)Cu(4-Hpz)(X)C ((4- pz))](X = Cl, Br) where square-pyramidallycoordinated copper(ll) ions are triply bridged by two pyrazolate ligands and one halide ion. Jvalues of -120.5 cm1 and -122 cm were found for the chloride and bromide complexes,respectively. The authors argue that because the exchange constants are insignificantlydifferent between the chloride and bromide complexes the primary pathway for exchangemust be the bridging pyrazolate ligands. Prins et al. reported studies on the related bistriazolate bridged [Cu(bpt)(CF3SO) H20)](bpt = bis(pyridin-2-yl)-1,2,4-triazolate). In thiscomplex, the copper(ll) ions are present in distorted octahedral coordination environmentsand J = -118 cm4. Several years ago, Barraclough and co-workers reported on thecharacterization by indirect methods of a dimetallic copper(II) complex doubly bridged bypyrazolate ligands and end-capped by acetylacetonate ligands (114). It was suggested that thecopper ions in this complex are coordinated in a square-planar fashion. The magneticN3117susceptibility of the complex was measured between 92 K and room temperature and J wasdetermined to be -35 cm’; however, over the temperature range studied, no maximum in thesusceptibility plot was observed. The J value determined in this case is much smaller thanthose measured for the compounds studied by Ajó and co-workers (112) and Prins and coworkers (113) in spite of the fact that Barraclough et at. proposed the same kind of magneticorbital-azolate pathway as suggested by the former authors. In view of this and the facts thatthe structure of Cu(4-Hpz)(acac) was not determined nor were magnetic data collected tosufficiently low temperature it is suggested that the structure proposed by Barraclough et at.may be incorrect.A number of years ago, a purely pyrazolate-bridged copper(ll) dimer was prepared inthis laboratory and structurally characterized (115). [Cu(4-Hpz)(Me2Ga(4-Hpz)(OCH2CHN(Me)) 1 (Me2Ga(4-Hpz)(OCH2CHN(Me) = dimethyl(N,dimethylethanolarnino)(1-pyrazolyl)gallate) is a dimeric complex in which five-coordinate,square-pyramidal copper(II) chromophores are linked by a double pyrazolate bridge. Eachbridging pyrazolate coordinates at the apex of one copper(II) chromophore and the basalplane of the other copper(II) chromophore. The complex is depicted in Figure 3.11. As thismolecule is an obvious candidate for comparison with the [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)2]complex, a magnetic susceptibility study was undertaken. The results of thestudy are presented graphically in Figure 3.12 and the data are tabulated in Appendix III,Table 111-4. The maximum in versus temperature for [Cu(4-Hpz)(Me2Ga(4-Hpz)(OCH,CH2N(Me)) jwhich occurs at --4.5 K indicates antiferromagnetic coupling andthe magnitude of the susceptibility is much higher than that of [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)2](Figure 3.10), thus the former complex is much more weakly coupled than thelatter. The magnetic data for [Cu(4-Hpz)(Me2Ga(4-Hpz)(OCHN Me)))]weremodelled using the Bleaney-Bowers expression and a paramagnetic impurity term asdescribed above for {Cu(4-Br-3-CO2Mepz)(4- dmpzH)1.This yielded best fit values for J,118___/2z7----—- N.,1Ga-..IFIG. 3.11. Bis-p.-pyrazolato-(N( 1 ),N(2))-bis[dimethyl(N,N-dimethylethanol-amino)-( 1 -pyrazolyl)gallato(N(2),O,N)copper(TI)J.g, and %P of -2.63(1) cm4, 2.013(1), and 3.8(1)% with F = 0.0068.The preceding discussion described pyrazolate bridged, dimetallic complexes whichexhibit J values varying over two orders of magnitude. The magnitudes of exchange in thethree complexes studied by Ajó (112) and Prins (113) and co-workers are virtually identicaland significantly greater than in [Cu(4-Br-3-COMepz)(4 rdmpzH),J,. Some or all of thisdifference may be accounted for by differences of the metal-ligand chromophores. In thepreviously studied complexes, the coordination geometry is either square-pyramidal ortetragonally elongated octahedral with the d22 orbital as the magnetic orbital and thebridging azolate ligands bonded in the xy plane of each copper ion. In [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH),], the geometry is trigonal bipyramidal. The magnetic orbital is thed2 orbital and the pyrazolates bridge such that one nitrogen occupies an axial site on onecopper while the other nitrogen occupies an equatorial site on the second copper. Sinceoverlap of the magnetic d2 orbital with the ligand orbitals in the xy plane is expected to be1190.04-4I—0EC)0.02I0.010.00250 300FIG. 3.12. Magnetic susceptibility versus temperature plot for[Cu(4-Hpz) (Me,Ga(4-Hpz)(OCH,CH,N(MejThe solid line representsthe best fit to the data of the Bleaney-Bowers equation including a Curielaw paramagnetic impurity term.weaker than overlap with ligand orbitals on the z axis, the exchange pathway provided in thiscomplex would not be expected to be very efficient and this could account for the weakerexchange coupling observed in [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)].A similar overlapmismatch occurs in [Cu(4-Hpz)(MeGa(4-Hpz)(OCHCH,N(Me)))1The chromophoregeometry in this complex is square-pyramidal and the magnetic orbital is expected to be thed22 orbital in such a case. Here the pyrazolate ligand bridges such that one nitrogen atomoccupies the apical site of one copper chromophore while the other nitrogen atom occupies abasal plane site on the second copper chromophore. Overlap of the magnetic d22 orbital0 50 100 150 200Temperature (K)120with the ligand orbitals along the apical z axis is expected to be weaker than overlap withligand orbitals in the basal xy plane and the situation is analogous to the preceding one; theexchange pathway is relatively poor. This scheme may account for the differences inexchange between the former three complexes and the latter two. However, the questionremains as to why there is such a large difference in exchange between [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)] and [Cu(4-Hpz)(Me2Ga(4Hpz)(OCHN Me)))] Thedifference may lie in the nature of the magnetic orbitals of the two complexes. Theprincipally d2 character of the magnetic orbital in [Cu(4-Br-3-CO2Mepz)(4-BrdrnpzH)]means that the magnetic orbital has appreciable density along all three Cartesian coordinateaxes of the copper(ll) chromophore while in [Cu(4-Hpz)(Me2Ga(4-Hpz)(OCH2CHN(Me)) ] the magnetic orbital is of character which is localizedmainly in the xy plane and has little electron density along the z axis of the chromophore. Asecond possible explanation of the difference is that the expected weak magnetic exchangedue to orbital overlap mismatch in [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)]relative to [Cu(4-Hpz)(Me2Ga(4-Hpz)(OCHCN(Me)))1is ameliorated by the presence of the carboxylateligand in the former compound. The carboxylate group in 4-Br-3-CO2Mepz is almostcoplanar with the pyrazolyl ring, thus one may envisage orbitals delocalized over both thering and the carboxylate group. Such orbitals may act as moderately efficient exchangepathways because the carboxylate oxygen coordinates the copper axially permitting effectiveoverlap with the d2 magnetic orbital.3.3 SUMMARY AND CONCLUSIONSA number of oligometallic complexes containing copper in the +1 and +2 oxidationstates have been synthesized and fall into two categories. The first category consists of sixcyclic copper(I) pyrazolate trimers which contain a central Cu3N6ring. The structures of twoof these complexes, [Cu(4-Hdmpz)]3and [Cu(4-Medmpz)]3,have been determined by single121crystal X-ray diffraction and on the basis of mass spectral data and literature reports on otherunivalent coinage metal pyrazolates it was concluded that the remaining four complexespossess similar structures. As discussed in Chapter 2, the copper(I) complexes were found tobe excellent starting materials for preparation, in pure form, of a number of copper(II)pyrazolate chain compounds. The second group consists of three compounds which share thepresence of 4-X-3,5-dimethylpyrazolyl moieties in which the 3-methyl substituent has beenoxidized to a carboxylate group. The trimetallic mixed valence copper(I/H) compound [Cu(3-CO2dmpz)(4-MedmpzH)j,Cu was characterized by a single crystal X-ray diffraction studyand EPR spectroscopy. The dimetallic copper(ll) complex [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)2] was characterized by single crystal X-ray diffraction studies and wasdiscovered to crystallize in triclinic and monoclinic forms. The complex may, betherrnolyzed to yield a species of nominal formulation Cu4(4-Br-3-CO2Mepz)4BrdmpzH)3.Susceptibility measurements revealed the copper(IJ) ions in the dimer to beantiferromagnetically coupled with J = -75.4 cm1.122CHAPTER 4ZINC(II) AND COBALT(II) PYRAZOLATES4.1 INTRODUCTIONIn Chapter 1 previous studies on zinc and cobalt pyrazolate complexes were reviewed.The present work initially focused on the poly(cobak(II) pyrazolates) and a series of sixcomplexes were successfully prepared. Attempts to grow macroscopic single crystals of thesecompounds were unsuccessful, though during these endeavours interesting dimetallic,trimetallic and tetrametallic species were discovered. The work on zinc(II) pyrazolatesstemmed from the knowledge that zinc(ll) can often isomorphically replace cobalt(ll) in four-coordinate complexes (116) and it was hoped that single crystals of isomorphous zinc(II)pyrazolates might be prepared. This objective was not met, however, single crystals of adimeric species and samples of microcrystalline oligometallic and polymeric zinc(II)pyrazolates were prepared. Although chronologically most of the synthetic work on thecobalt(ll) complexes preceded that on the zinc(II) complexes, the latter are considered first inorder to ease comparison between the two groups.4.2 RESULTS AND DISCUSSION4.2.1 ZINC(II) 3,5-DIMETHYLPYRAZOLATES4.2.1.1 SYNTHESES, PHYSICAL AND THERMAL PROPERTIESDetailed descriptions of the syntheses discussed in this section are provided inChapter 7, Section 7.2.7. Vos and Groeneveld (41) reported the synthesis of [Zn(4-Hdmpz)2]a number of years ago using reaction [2.1], however, their characterization of the complex123consisted only of a zinc content determination which was in poor agreement with theexpected result. As the desire here was to prepare single crystals of [Zn(4-Hdmpz)2Jand inlight of the fact that Vos and Groeneveld were unable to produce single crystals of any metalpyrazolates (37-41), their synthetic methods were not employed. The MLPM reaction wasemployed in an attempt to grow single crystals of [Zn(4-Hdmpz)2], however, anotherproduct was obtained by this reaction as is shown below:Air, 115Zn + xs 4-HdmpzH 5 days> [Zn(4-Hdmpz)2(4-HdmpzH)] + H201 [4.1]The dimeric zinc species [Zn(4-Hdmpz)2(4-HdmpzH)] was obtained in low yield from thisreaction as colourless, X-ray diffraction quality single crystals. Slight modification ofreaction [4.1] yielded samples of the complex in microcrystalline form,02, 90 °CZn + xs 4-HdmpzH > [Zn(4-Hdmpz)2(4-HdmpzH)] + H201 [4.2]acetone45 h, stirringThe melting point of 3,5-dimethylpyrazole (108 °C) is above the temperature used in thereaction necessitating the use of a solvent; hence, a small amount of acetone was added inreaction [4.2] solely to dissolve the 3,5-dimethylpyrazole. Reaction [4.11 was performed athigher temperature (130 °C), with stirring, to see if [Zn(4-Hdmpz)2]could be obtained. Alow yield of a white, microcrystalline material was produced which elemental analysisindicated to have the nominal formulation Zn(4-Hdmpz).0.323(4-H pzH). Evidencepresented below indicates that this material is best described as a dimeric/oligometallicmixture and thus it will be referred to as such.Interestingly, both the dimeric and dimeric/oligometallic zinc complexes describedabove can be converted into the polymeric material [Zn(4-Hdmpz)2], although the124conditions differ for doing so. The DSC thermogram for [Zn(4-Hdmpz)2(4-HdmpzH)],shown in Figure 4.1, indicates an endothermic event at 230 °C (zH = 150 kJ mo11). Weight50 100 150 200 250 300 350 400Temperature (°C)FIG. 4.1. DSC thermogram of [Zn(4-Hdmpz)(4-HdmpzH)1difference measurements show a weight loss over the event of 28% which is what would beexpected for the loss of two neutral 4-HdmpzH molecules per mole of dimer. Elementalanalysis of the thermolysis product confirmed it to have the [Zn(4-Hdmpz)2]formulation.On the other hand, thermolysis of Zn(4-Hdmpz)2.0.323(4-Hd pzH shows an endothermicevent at 178 °C (with no event at 230 °C) and elemental analysis of the product alsoconfirmed the loss of neutral (4-HdmpzH). The difference in thermal behaviour between thetwo materials receives further comment in Section Further heating of the [Zn(4-Hdmpz)2Jmaterial results in the onset of decomposition without melting at —330 °C.125X-RAY DIFFRACTiON STUDIESSINGLE CRYSTAL X-RAY DIFFRACTIONFIG. 4.2. Stereoscopic view of the [Zn(4-Hdmpz)2(4-HdmpzH)] molecule; 50%probability thermal ellipsoids are shown for the non-hydrogen atoms.Crystallographic data, atomic coordinates, and bond lengths and angles are tabulated inAppendix I, Tables 1-34 through 1-38. The molecular structure consists of two zinc ionsbridged by two 4-Hdmpz ions with each zinc ion end-capped by a 4-Hdmpz ion and a neutral4-HdmpzH molecule. Strong hydrogen bonding occurs between the capping ligands(N(2)—H(1). . .N(2) (x, l/4-y, 1/4-z), H.. .N = 1.77 A, N.. .N = 2.612(7) A, N—H.. .N = 141°)with the unique hydrogen atom being disordered between two possible sites. The moleculehas exact (crystallographic) D2 and approximate D2h symmetry (with crystallographic twofold axes passing through the zinc ions and C(7) atoms). The zinc ions are coordinated in apseudo-tetrahedral fashion (N—Zn—N = 99.6(2)-i 13.8(2)°) and are separated by a non-bonded4. 1.2. 1The molecular structure of [Zn(4-Hdmpz)2(4-HdmpzH)] is shown in Figure 4.2.126distance of 3.748(1) A. The central Zn(N—N)2n ring is approximately planar (to within0.064(3) A). The dihedral angle between the normals of the central and terminal Zn(N—N)2nplanes is 87.9°. The bridging Zn—N distance of 1.991(3) A is significantly shorter than theterminal Zn—N distance of 2.025(3) A. The Zn—N distances are similar to the values of1.981(5)-2.053(5) A reported for other Zn(II) pyrazolate and pyrazole structures (117-119). POWDER X-RAY DIFFRACTION STUDIESThe thermal properties of the dimeric complex and the dimeric/oligometallic materialseemed to rule out the possibility that the latter is simply a mixture of the dimer and thepolymer, since if this were the case then the thermogram would be expected to show anendothermic event at 230 °C. As a result, further evidence for the presence of another phase(or phases) was sought in the powder X-ray diffractograms for the complexes. Diffractionpatterns for the dimeric zinc 3,5-dimethylpyrazolate, the polymeric zinc 3,5-dimethylpyrazolate, and the dirneric/oligometallic material are presented in Figures 4.3(a) to(c). The d-spacings and relative intensities for the peaks in these patterns are listed in Table II-4 in Appendix II. The pattern for the dimeric/oligometallic mixture clearly shows thepresence of the dirneric phase, however, the peaks corresponding to those of the polymericphase are all shifted from those of the true polymer. It is thought that the shifted peakscorrespond to the presence of oligometallic species. INFRARED SPECTROSCOPYInfrared spectra of the zinc(II) 3,5-dimethylpyrazolates reveal some interestingfeatures about these complexes, particularly in the case of the dimeric species. Infrared datafor the complexes are tabulated in Appendix IV, Table IV-6 and the spectra for [Zn(4-12700c-)FIG. 4.3. X-Ray powder diffractograms of (a) [Zn(4-Hdmpz)2(4-HdmpzH)],(b) [Zn(4-Hdmpz)2j,and (c) the dimeric/oligometallic mixture.Hdmpz)(4-Hd pzH)]., and [Zn(4-Hdmpz)2]are shown in Figures 4.4(a) and (b). The4-2e128C)ct5I4000Wavenumbers (cm1000 800 600 400 200FIG. 4.4. Infrared spectra of (a) [Zn(4-Hdmpz)2(4-HdmpzH)],(b) [Zn(4-Hdmpz)2],and (c) the dimeric/oligometallic mixture. Spectra obtained from solid state Nujol mulls.spectrum for [Zn(4-Hdmpz)2(4-HdmpzH)] reflects the presence of strong hydrogen bonding3000 2000 15001)129in the molecule with the appearance of two broad v vibration bands centred at —2,380 and—1,860 cm1.These bands are significantly shifted from the broad band centred at 3,000 cm1in solid 4-HdmpzH and the band at 3,484 cm1 in dilute CC14 solutions of the ligand (120).Similar band shifts, involving VOH vibrations have long been recognized as occurring in metaldimethyiglyoxime complexes containing strong intramolecular hydrogen bonding (121).Bellamy and co-workers(122,123) and Novak (124) have examined in some detailcorrelations between infrared frequency shifts and hydrogen bond distances in crystals. Theempirical correlation of Bellamy and Owen (123) predicts a frequency shift of about 2,000cm4 for an intermolecular hydrogen bond with a N—H.. .N distance on 2.612 A. The shiftsobserved in [Zn(4-Hdmpz)2(4-HdmpzH)] are less than this, which is no doubt at least partlydue to the non-linear intramolecular N—H.. .N hydrogen bond. In addition, it should be notedthat there is considerable uncertainty in the predicted shift due to its strong dependence ondistance at short N.. .N distances. In [Zn(4-Hdmpz)2]these v bands are missing, asexpected, as is the broad medium intensity band occurring at 870 cm1 in the dimer. Thislatter band is likely an out-of-plane bending vibration, y, although it has been assignedpreviously to a C—H bending out-of-plane mode (120). Consistent with the y assignment,this band is also missing in the spectrum of Na[4-Hdmpzj.The in-plane v5 vibration, discussed in Chapter 2, Section, appears at1,529 cm1 in [Zn(4-Hdmpz),], consistent with the presence of the 4-Hdmpz anion. In spiteof the fact that there are two crystallographically unique 3,5-dimethylpyrazolyl rings in [Zn(4-Hdmpz)2(4-Hd pzH)J, and, moreover, that formally there are both neutral and deprotonatedligands present in the complex, only one band assignable to this v5 mode is observed at1,525 cm1. On the basis of the low frequency observed for this vibration, it may beconcluded that all of the ligands are effectively anionic. This is somewhat surprising in thecase of the capping ligands and is likely a consequence of the very strong hydrogen bondingassociated with these ligands. The remainder of the JR spectrum of the dimer is quite similarto that of the polymer, [Zn(4-Hdmpz)2].However, there are differences in detail which130warrant mentioning as they will become important in the discussion of the cobalt(fl)complexes. The 1,700-1,500 cm region of the dimer spectrum consists of two broadshoulders at approximately 1,630 and 1,570 cm1 and the strong ‘Vrfflg band at 1,525 cm1 witha prominent shoulder to the low energy side whereas the polymer exhibits a weak band at1,690 cm and a resolved peak at 1,546 cm1 with the Vng band at 1,529 cm1 with only aweak shoulder to the low energy side. The dimer shows two bands in the 1,350-1,300 cmregion at 1,344 and 1,328 cm1 with the lower energy band of considerably weaker intensity.The polymer, on the other hand shows only one strong band, split into a doublet with. peaks at1,337 and 1,325 cm-1. The relative intensities of the bands at approximately 1,088 and 979cm1 are reversed in the two compounds. Moreover, the splittings of bands in the 800-750cm region are quite different in the two samples. The final difference is associated with theweak, bands below 500 cm* The polymer exhibits one band at 458 cm and in thedimer this appears as two bands at 439 and 461 cm-1, the former frequency presumablyassociated with the para-methyl groups of the terminal ligands and the latter with the orthomethyl groups of the bridging and terminal ligands. The band occurs at 412 cm1 in solid 4-HdmpzH.Finally, the JR spectrum of the dimenc/oligometallic mixture is considered Thespectrum is shown in Figure 4.4(c) and close examination reveals it to be a composite ofthose of the dimer and the polymer where the 1,700-1,500 cm-’ region reflects the spectrumof the polymer, the 1,350-1,300 cm regions is similar to that of the dimer, and the otherregions appear to be a mixture of both the dimer and the polymer. It was previouslyconcluded that this material is not, in fact, a dimer/polymer mixture, but rather adimeric/oligometallic mixture. The JR spectroscopy results can only be explained byconcluding that the spectrum of the oligometallic species, in the presence of the dimer, isindistinguishable from the spectrum of the polymer, in the presence of the dimer.1314.2.2 OLIGOMETALLIC COBALT(II) 3 ,5-DIMETHYLPYRAZOLATES4.2.2.1 SYNTHESES, PHYSICAL AND THERMAL PROPERTIESAttempts to prepare X-ray diffraction quality single crystals of a number ofpoly(cobalt(II) pyrazolates) were unsuccessful. However, these attempts yielded singlecrystals of two oligometallic complexes which were useful as models for the polymericmaterials. Detailed descriptions of the syntheses of these compounds are given in Chapter 7,Section 7.2.5. Initial efforts for producing diffraction quality single crystals concentrated onthe MLPM reaction as this method was successfully employed in the preparation ofpoly(copper(II) pyrazolate) crystals. However, once it was determined that the method wasnot amenable to the synthesis of macroscopic single crystals of the Co(ll) complexes,variants of the MLPM reaction were tried. The first of these was a reaction in which[Co(acac)214was heated in molten 4-HdmpzH for 2 h at 145 °C. A small amount of insolublepurple powder was obtained which analyzed as impure [Co(4-Hdmpz)2].This insolublematerial was extricated from the cooled reaction mixture (which was mainly solid 4-HdmpzH) using CH21, and then suction filtering the resulting slurry. The filtrate wasconcentrated and subjected to further heating at 155 °C for 16 h as a quiescent solution.Large, diamond shaped, bluish purple single crystals of the trimetallic species, [Co(4-Hdmpz)2C1(4-HdmpzH ]o,were prepared in this manner. Once it was recognized that sucha complex could be prepared, the reaction conditions were modified to intentionally preparethis compound in greater quantities. This reaction scheme is shown below:1321.N2, 160 °C,2h[Co(acac)2]4+ xs 4-HdmpzH > pink solid2. dissolve in CH21 + [4.3]4-HdmpzHfilter, collect filtrateand concentrate.1. N/CHCl vapour130 °C, 18 h2. Acetone extractionand methanol wash[Co(4-Hdmpz),Cl(4-HdmpzH)]2o+soluble green and blue by-productsThis procedure yielded [Co(4-Hdmpz)2C1(4-HdmpzH)]oin 23% yield. The compound isair-insensitive and sparingly soluble in CH21 and acetone. The DSC thermogram of thecomplex shows an endothermic event at 293 °C (probably melting) followed immediately byexothermic decomposition.The second variant of the MLPM reaction involved the heating of CoCO3.xH2Oin anextremely concentrated solution of 4-HdmpzH in acetone at 100 °C (just sufficient acetonewas used to melt the 4-HdmpzH at this temperature). This yielded large, well formed, darkpurple crystals of a compound which elemental analysis indicated to have the nominalformulation Co(4-Hdmpz)2.0.34 (4-Hd pzH) and this is summarized below:N2, 100 °CCoCO3.xH2O+ xs 4-HdmpzHh> Co(4-Hdmpz),.0.344(4-Hdmpz) [4.41+ C02+H20This material is unique in its properties and behaviour and is of particular relevance to the133study of the polymeric Coal) pyrazolates. The product from reaction [4.4] exhibitsinteresting thermal behaviour. As the differential scanning calorimeter was unavailable foruse during the study of this compound a small sample of crystals of Co(4-Hdmpz)2.0.344(4-HdmpzH) was heated in a melting point apparatus and observed. In the 180-200 °C range,colourless crystals were observed to have formed in the cool end of the capillary tube, thematerial having sublimed from the sample (the colourless crystals were most probably 4-HdmpzH). No decomposition of the sample crystals was observed at that point. Heating wascontinued to the apparatus limit of 360 °C with no further visible change in the sample.Microscopic examination of the sample before the heating process revealed it to consist oflarge, almost opaque, dark purple crystals whose morphology is best described as elongatedoctahedral. The crystal facets were shiny and the crystal edges sharp. After heating,microscopic examination showed that the crystal facets had now become dull and chalky,however, the crystal edges were still sharp and the crystals had maintained their mechanicalintegrity; they could be manipulated without damage. Evidently, the thermolysis of thismaterial, resulting in the volatilization of 4-HdmpzH, left the original crystals largelyunaffected. The results of a single crystal diffraction study of the compound, which will bediscussed in greater detail in Section, are of even greater interest. The crystal selectedfor study was found to be [Co(4-Hdmpz)2(4-HdmpzH)],isomorphous and isostructural with[Zn(4-Hdmpz)2(4-HdmpzH)1.This is surprising considering that the bulk formulation for thecompound is Co(4-Hdmpz).0.34 (4-Hd pzH) and not Co(4-Hdmpz)2.(4-HdmpzH aswould be expected for pure dimer. This inconsistency is discussed in the following section.Another Co(ll) 3,5-dimethylpyrazolate species was prepared unexpectedly duringattempts to synthesize the mixed cobalt/zinc polymer [Co113Zn2(4-Hdmpz)][Zn(4-Hclmpz)2(4-Hd pzH)] was reacted with [Co(CH3N)6][BF42(125) and excesstriethylamine in the hope of obtaining [CoiZn2(4-Hdmpz)j.Instead, the reaction yieldedtwo products: a light purple powder which, by JR spectroscopy and elemental analysis, isthought to be impure [Co1..Zn(4-Hdmpz)2](0 < n < 1) and a small amount of well formed,134macroscopic, bluish purple crystals of[Co4(4-Hdmpz)60].[Co(CH3N)6][BF42+ [Zn(4-Hdmpz)2(4-HdmpzH)1 + 10 Et3N [451N2CH3N[Co4(4-Hdmpz)60]+ impure [Coi..Zn(4-Hdmpz)2]< R T4 daysCrystals of the tetrametallic compound, [Co4(4-Hdmpz)60j, were difficult to separate inquantity from the insoluble purple powder, so characterization of the complex was limited toa single crystal X-ray diffraction structure determination and the recording of an IR spectrumof the compound. The diffraction study revealed the presence of oxygen in [Co4(4-Hdmpz)601 in spite of the fact that the reaction was performed under anhydrous, anaerobicconditions as outlined in equation [4.5]. The most likely source of oxygen in the reaction wasthe starting material, [Co(CH3N)6][BF4,.This complex was prepared according to themethod of Hathaway, Holah, and Underhill (125) and it is a very thermally sensitive andhygroscopic compound. Although C, H, and N analyses of the [Co(CH3N)6J[BF4]2indicated it to be pure, its JR spectrum exhibited a weak band in the VOH region suggestingthe presence of a small amount of H20 in the complex. Crystals of [Co4(4-Hdmpz)60]arestable in air and the complex is sparingly soluble in CH3N and acetone. X-RAY DIFFRACTION STUDIESAs mentioned in the preceding section, the three oligometallic Co(II) speciesdiscussed here were prepared in macroscopic, single crystal form thus permitting singlecrystal diffraction studies to be carried out on them. These diffraction studies have been135invaluable in the characterization of these complexes. However, only in the case of [Co(4-Hdmpz)2Cl(4-HdmpzH ]ohave the diffraction results yielded a thoroughly unambiguousstructure determination.[Co(4-Hdmpz)Cl(4-HdmpzH)]ocrystallizes in the space group Pt as a trimetallicmolecule with the Coal) ions in an approximately linear arrangement. The terminal Coal)ions are linked to the central cobalt by double 3,5-dimethylpyrazolate bridges. The complexis illustrated in Figure 4.5. Crystallographic data, atomic coordinates, and bond lengths andFIG. 4.5. Stereoscopic ORTEP diagram of the [Co(4-Hdmpz)2Cl(4-HdmpzH)]omolecule; 33% probability thermal ellipsoids are shown for the non-hydrogen atoms.angles are tabulated in Tables 1-39 through 1-43 in Appendix I. The central cobalt iscoordinated by four pyrazolate nitrogen atoms in a pseudo-tetrahedral fashion with Co—Nbond lengths ranging from 1.988(4)-2.001(4) A (average = 1.993 A), N—Co—N bond anglesranging from 107.1(2)-i 13.9(2)°, and a dihedral angle between the fused N—Co—N planesrunning along the approximate molecular Co—Co—Co axis of 90.2° (the angle formed bythese three cobalt centres is 169.2°). The terminal Co(II) ions are each coordinated by twoCIGC17C26C29CI2C17CIIdoCIIdoC21Casce136bridging pyrazolate nitrogen atoms, one terminal 3,5-dimethylpyrazole nitrogen atom andone terminal chloride ion in a distorted tetrahedral manner. The average terminal Co—N bondlength is somewhat longer than the average bridging Co—N bond length (2.030(5) A versus1.981(5) A, respectively) and the average Co—Cl bond length is 2.25(1) A. BothCo(N—N)2o rings in the complex exhibit a half-boat conformation in which the terminalcobalt is displaced from the mean plane defined by the central cobalt and the four nitrogenatoms. The average distance between cobalt centres within the Co(N—N)2o rings is non-bonding at 3.627 A. There are two molecules per unit cell related by an inversion centre. The“pyrrolic’ hydrogen atom of the terminal 4-HdmpzH ligand is hydrogen bonded to thechloride ligand of an adjacent molecule and vice versa (average N... Cl = 3.376 A) creatingchains of the trimetallic complex running along the a-direction of the crystal lattice asdepicted in Figure 4.6.FIG. 4.6. Stereoscopic view of the unit cell packing and intermolecular hydrogenbonding in the crystal structure of [Co(4-Hdmpz),Cl(4-HdmpzH)]2o.137As was mentioned in Section, [Co(4-Hdmpz)(4-HdmpzH)] is isomorphousand isostructural with the zinc analogue. Crystallographic data, atomic coordinates, and bondlengths and angles are tabulated in Table 1-34 and Tables 1-44 through 1-47 in Appendix I.Because of its similarity to the zinc dimer an illustration of an individual cobalt dimer is notpresented here; however, Figure 4.7 illustrates the unit cell packing of [Co(4-Hdrnpz)2( -HdmpzH)12 in its crystal structure. The figure shows how the individual molecules lie inFIG. 4.7. Stereoscopic view of the unit cell packing in the crystal structure of[Co(4-Hdmpz)2(4-HdmpzH)]sheets parallel to the ab-plane of the crystal lattice such that all of their terminal ligands arecoplanar. Within each sheet, Co—Co vectors of individual molecules align in a coparallelmanner and, moreover, chains of Co—Co vectors running along the a-axis are not onlycoparallel, but collinear. Adjacent sheets are oriented such that the dimer molecules in theneighbouring sheets are staggered by translation of one dimer length along the a-axis. Thisresults in the presence of columns, parallel to the c-axis, of overlapped terminal pyrazolylrings with the ring planes normal to the c-axis. Thus it may be envisaged that the dimer138forms, in the solid state, a most suitable precursor structure for the formation of polymeric[Co(4-Hdmpz)2Jby condensation of the dimeric molecules into longer chains via extrusionof neutral 4-HdmpzH. In fact, the experimental observations of the sample from which thisstructure was determined, to be discussed presently, suggest that some form of incipientpolymer formation has, indeed, occurred.In Section it was stated that elemental analyses suggest a nominal formulationof Co(4-Hdmpz)2.O.34 (4-Hd pzH) for the material and not Co(4-Hdmpz)2.(4-HdmpzH aswould be expected for pure dimer. If a structure consisting of a doubly pyrazolate-bridgedchain of Co(II) ions with end-capping neutral ligands, as in the cases of the zinc and cobaltdimers, is assumed for this compound then the nominal formulation corresponds to a meanchain length of 5.8 Co(II) ions (for comparison, such a formula for the Zn(II)dimeric/oligometallic mixture discussed above would possess a mean chain length of 6.2Zn(II) ions). Of course, if such a process is occurring in this system then the nominalformulation only provides information about the mean chain length; it indicates nothingabout the distribution of chain lengths. One possible composition is a biphasic distribution:the bulk compound consists of a mixture of the dimeric phase (well characterized in the caseof the Zn(II) complex) and the infinite polymer phase such that their ratio yields the observedempirical formula. Electronic spectroscopy indicates that such a dispersion is very unlikely.Rationally synthesized pure [Co(4-Hdmpz)2]exhibits an electronic spectrum (see Section4. markedly different from bulk Co(4-Hdmpz)2.O.34 (4-Hd pzH). The spectrum ofthis latter compound indicates the presence of a fairly regular tetrahedral CoN4 chromophore(see Section, consistent with the type structurally characterized in the dimericcomponent of the compound. The biphasic distribution model requires the sample to be -.-66% (by mole Co(ll) ions) polymer phase, however, there is no indication of this phase in thespectrum of the compound. Moreover, microscopic examination of the sample revealed it toconsist entirely of macroscopic crystals, all of the same colour and morphology. Most of the139crystals were opaque, but a few of the smaller crystals were quite transparent. One mightexpect the presence of two crystal morphologies in a biphasic distribution.It is also important to note that the crystal selected for study by diffraction formed onehalf of a larger, elongated octahedral crystal. The half-crystal was chosen for examination bythe crystallographer because it was much more transparent than the remainder of the wholecrystal or the other crystals in the sample. A well formed, but opaque crystal which had beenselected previously for study was found to diffract X-rays too poorly to be useful. In otherwords, the crystal structure of the dimeric species appears to be compatible with the bulksample, but it is not necessarily representative of the molecular contents of most of thecrystals in the bulk sample. Though it was not possible to conduct a DSC study on thesample, qualitative observation of its thermal behaviour indicated the loss of 4-HdmpzH fromthe sample in the same temperature range as the zinc dimeric/oligometallic mixture,consistent with polymerization, and after evolution of the 4-HdmpzH the sample crystalsremained largely intact. This suggests that either the crystal lattice of this material is tolerantof considerable structural reorganization (not a likely possibility) or that relatively littlestructural reorganization occurs during this proposed polymerization. The opacity of themajority of crystals and poor diffraction characteristics of the one such opaque crystal whichwas chosen for study are most likely related and probably arise from disorder within thecrystal lattice. An oligometallic Co(II) pyrazolate system consisting of a distribution of chainlengths would likely exhibit such disorder.The tetrametallic complex [Co4(4-Hdmpz)60] consists of a central oxygen atomaround which the four Co(ll) ions are pseudo-tetrahedrally arrayed and the six edges of thetetrahedron are bridged by 4-Hdmpz ligands. The molecule is shown in Figure 4.8.Crystallographic data, atomic coordinates, and bond lengths and angles are tabulated in Table1-39 and Tables 1-48 through 1-51 in Appendix I. The molecule crystallizes in the Pt spacegroup with two molecules per unit cell related by an inversion centre. The bond length fromthe Co(ll) ions to the central oxygen range from l.936(2)-l.953(2) A (average Co—O = 1.944140cx? Cx?cso C30FIG. 4.8. Stereoscopic ORTEP diagram of the [Co4(4-Hdmpz)60]molecule;33% probability thermal ellipsoids are shown for the non-hydrogen atoms. Co(4)lies directly behind the oxygen atom.A) and the Co—0—Co angles range from 108.1(l)-il 1.8(l)°. The Co(ll) centres are separatedby an average non-bonding distance of 3.21 A (range = 3. 148-3.307 A). Co—N bond lengthsrange from l.987(3)-2.017(3) A (average Co—N = 1.998 A). The average 0—Co—N bondangle is 98.3° and the average N—Co—N bond angle is 118.00; thus the coordination geometryabout the cobalt centres is best described as trigonal pyramidal with the oxygen atom formingthe apex of the coordination polyhedron. With regard to the 3,5-dimethylpyrazolate ligandsthemselves, it was determined during the refinement procedure that the hydrogen atoms offour of the methyl groups were two-fold disordered, so in each such case the site occupancyfactors were fixed at 50% for each orientation. The [Co4(4-Hdmpz)60] complex wassynthesized using [Zn(4-Hdmpz)2(4-HdmpzH)] as a starting material and, as a result, zincatoms may be present in some of the molecules via isomorphous replacement of the cobalt(II)141ions. If Zn(ll) ions are randomly distributed over the cobalt sites in the lattice, it is estimatedthat as many as 10-20% of the cobalt sites could be occupied by zinc before the diffractiondata would indicate the presence of zinc.The structure of [Co4(4-Hdmpz)60]is strongly reminiscent of the well known basicberyllium carboxylates, [Be(O2CR)0](where R is an alkyl group) (126,127), and Zn(IJ) isalso known to form complexes of the type [Zn4(O2CR)60] (128). The basic berylliumacetates are quite resistant to hydrolysis while the zinc analogues are readily hydrolysed. It isbelieved that this reactivity difference is due to the fact Be(II) in a tetrahedral environment iscoordinatively saturated while Zn(II) can exhibit coordination numbers higher than four. Thelimited observations that have been made on [Co4(4-Hdmpz)60]seem to indicate that it ismoderately resistant to hydrolysis in spite of the fact that it is coordinatively unsaturated. Thestructure of this compound provides a clue to this behaviour (Figure 4.8) in that the opencoordination site on each cobalt centre is blocked by the three methyl groups of the 4-Hdmpzligands. SPECTROSCOPIC BEHAVIOUR4. INFRARED SPECTROSCOPYJR spectra were recorded for all three of the oligometallic Co(ll) species describedabove and the positions and relative intensities of unassigned bands have been tabulated inTable IV-7 in Appendix IV. [Co(4-Hdmpz)2Cl(4-HdmpzH)] exhibits a strong, relativelysharp band at 3,265 cm1, attesting to the moderate hydrogen bonding present betweenmolecules of this compound. Unlike the JR spectrum of the zinc dimer, the spectrum of [Co(4-Hdmpz)2C1(4-HdmpzH ]o reveals the presence of two types of 3,5-dimethylpyrazolylrings in the molecule. There are two aromatic VCH bands present at 3,121 cm1 and 3,143 cm1and by comparison with the spectra of other complexes the former band is due to the bridging142rings and the latter is due to the terminal rings. There are two Vg modes exhibited in thespectrum at 1,525 cmt and 1,568 cm4 due to the bridging and terminal rings, respectively.The 600-1,400 cm1 region generally conforms to the pattern expected for the3,5-dimethylpyrazolyl ring. However, the number of bands is much greater than what wouldbe expected for such a ring in a single chemical environment. This is consistent withdistinction of the terminal and bridging rings. Finally, there are two bands observed inthe 400-500 cm region at 432 c& and 459 cmt corresponding to the para-methyl groupsof the terminal ligands and or:ho-methyl groups of the bridging and tenninal ligands,respectively.The JR spectrum of Co(4-Hdmpz)2.0.34 (4-Hd pzH) is practically identical to thatof the zinc dimer. The strongly hydrogen bonded v bands are present in the spectrum andthe presence of two bands at 444 c& and 462 cmt are consistent with terminal andbridging ligands, respectively. The only notable difference between the two spectra is that inthe present compound the relative intensities of the bands at 1,323 cm and 1,337 cm1 arereversed with respect to the corresponding bands in the zinc dimer. Consequently, JRspectroscopy does not allow one to distinguish between the Co(U) dimer and theoligometallic species proposed to be present in this material.The JR spectrum of [Co4(4-Hdmpz)60]is as expected for a complex possessing onlybridging 3,5-dimethylpyrazolyl moieties. There is only one aromatic VCH band present at3,125 cm4. The single Vg band appears at 1,526 c&, consistent with an anionicmethylated ligand. Moreover, there is only one I3c..cband at 458 cm, consistent with thepresence of only para-methyl groups and, consequently, the presence of only bridgingligands. There is one notable difference between the spectrum of this complex and the spectraof the two previous cobalt species: [Co4(4-Hdmpz)60] exhibits a medium intensity,moderately broad band at 555 cm1. As the band appears only in this compound, it issuggested that the band is due to a vibrational mode of the molecule’s central Co40core.1434. ELECTRONIC SPECTROSCOPYThe electronic spectra of [Co(4-Hdmpz)2C1(4-HdmpzH)Jo and Co(4-Hdmpz)2.0.344(4-HdmpzH) were obtained (solid state, Nujol mulls) and are shown in Figure4.9. The band positions in wavelength and wavenumbers are presented in Table 4.1 alongwith tentative assigments for the transitions. In a tetrahedral ligand field, the electronicground state term for Co(II) is 4A2 and three transitions to the excited states 4T2,4T1(F), and4T1(P) (v1, v2, and v3, respectively) may be observed. In practice, usually only two groups ofbands are observed, one in the near infrared, the other in the visible region. v2 is usuallyobserved in the range 5,000-11,000 cm1 and v3 is observed in the range 13,500-19,500 cm-1.The v1 mode is often difficult to detect (129) as it occurs around 5,000 cm1, or lower, andcan suffer interference from vibrational overtone and combination bands (the v1 band was notobserved in any of the electronic spectra recorded for the Co(II) species studied in this work).Figure 4.9 and Table 4.1 show that [Co(4-Hdmpz),Cl(4-HdmpzH)]2opossessesconsiderable structure in its formally v2 and v3 transitions with bandwidths of the order of6,000 cm1 and 2,500 cm1 in the former and latter transitions, respectively. The structure andbandwidths arise from three sources: spin-orbit coupling, vibrational structure, and thepresence of a lower than cubic symmetry ligand field around the Co(1I) ion. According toLever (130), spin-orbit coupling alone is insufficient to account for bandwidths as large asthose observed for [Co(4-Hdmpz)2C1(4-HdmpzH)] (theoretically, to first order, v2 shouldbe 4 wide, where is the spin orbit coupling constant, if spin-orbit coupling is the onlycontributor to bandwidth). Vibrational structure is unlikely to account for the remainder ofthe bandwidth and structure in the v2 band, so that leaves low-symmetry ligand fields asimportant contributors. The v3 mode in the visible region of the spectrum, in addition tohaving low-symmetry contributions to its width and structure may also show band splittingdue to coupling with nearby spin forbidden quartet-doublet transitions (130). The largebandwidths and band structure in the electronic spectrum of [Co(4-Hdmpz)21(4-144C)C)ct50rJ)•0FIG. 4.9. Electronic spectra of (a) Co(4-Hdmpz)2.O.34 (4-Hd pzH) and (b)[Do(4-Hdmpz)2Cl(4-HdmpzH)].HdmpzH)]2Coare principally due to the presence of two different chromophores revealed bythe X-ray diffraction study on the complex. [Co(4-Hdmpz)2C1(4-HdmpzH)] possesses a500 1000 1500 2000Wavelength (nm)145central CoN4 chromophore of fairly regular tetrahedral symmetry and two terminal CoN31chromophores of rather low symmetry; the crystallographic symmetry of the CoN31chromophores is C1 and they have very approximate C3, point symmetry (see Figure 4.5).TABLE 4.1. Electronic spectral data for [Co(4-Hdmpz)2C1(4-HdmpzH)]oand Co(4-Hdmpz)2.0.344(4-Hdmp H °Band position”Compound (nm) (cm-’) Assignmentc[Co(4-Hdmpz)2C1(4-HdmpzH)]o 1,695 5,900 w,sh1,418 7,050 m,sh1,235 8,100 m 4T1(F)—A2)995 10,050m (v2)814 12,280 w,sh625 16,000 s,sh597 16,750 s541 18,480 s,sh (v3)455 21,980w SF288 34,720 vs252 39,680 vs IL/CF217 46,080 vsCo(4-Hdmpz)2.0.344(4-HdmpzH) 1,260 7,940 m,sh1,138 8,790 m1,015 9,850m (v2)591 16,920s577 17,330 s532 18,800 s,sh (v3)405 24,690 w,sh SF285 35,090 vs IL/CT216 46,300 vsaRerd in the solid state using Nujol mulls.bAbbreviations: w, wealc m, medium; s, strong; vs. very strong; sh, shoulder.CAssignmen are based on the tetrahedral formalism. Abbreviations: SF, spin-forbidden; IL/Cl’, internal ligand orcharge transfer.146Co(4-Hdmpz)2.0.34 (4-Hd pzH) also shows the v2 and v3 bands consistent with atetrahedral chromophore. In this case the bandwidths are of the order of 2,000 cm1 and 1,900cm for the v2 and v3 bands, respectively. These narrower bandwidths are consistent with thepresence of only one type of chromophore in the complex and they suggest a less distortedCo(ll) chromophore than those observed in the case of [Co(4-Hdmpz)2Cl(4-HdnipzH)]o.In concordance with this observation, the X-ray diffraction study of [Co(4-Hdmpz)2( -HdmpzH)]2,which is a component of Co(4-Hdmpz)2.0.34 (4-H pzH , showed the dimericspecies to possess crystallographically exact D2 point symmetry and fairly regular tetrahedralstereochemistry about the Co(H) centres: this may well represent the stereochemistry of thebulk compound.Within the context of ligand field theory and the tetrahedral formalism, the values ofDq and B for a Co(ll) complex may be calculated from a knowledge of v2 and v3 (v1 = lODq)using the following equations (131):Dq =[9(v2+v3)-(85(v-4 1/21/340 [4.6]B =(v+-3ODq)/15 [4.7]Equations [4.6] and [4.7] were applied to the spectral data for the two compounds consideredin this section and the results are listed in Table 4.2. Of course, the values calculated in thesecases are only crude approximations as neither of the complexes have truly tetrahedralchromophores and, furthermore, [Co(4-Hdmpz)2Cl(4-HdmpzH)] possesses two differenttypes of chromophore.The end of this section brings to a close discussion of experimental studies on Co(4-Hdmpz)2.344(4-Hdmp H), so it is appropriate to make final comments about the structuralnature of this material. A single crystal X-ray diffraction study showed that [Co(4-Hdmpz)2( -HdmpzH)]2is a component of the material, but the bulk formula of the substance, its147TLE 4.2. Ligand field parameters for [Co(4-Hdmpz)2C1(4-HdmpzH ]oand Co(4-Hdmpz)2.0.34 (4-Hd pzH(estimated uncertainties in parentheses)aDq BCompound (cm-’) (cnr’)[Co(4-Hdmpz)2C1(4-HdmpzH)]Do 520(20) 665(60)Co(4-Hdmpz).0.34 (4-Hd pzH 515(10) 720(20)The effective v2 and v3 values, v2 and v3, used were 8,800 and16,800 cm4 for [Co(4-Hdmpz>C1(4-HdmpzH)]oand 8,770 cm’ and 17.540cm4 for Co(4-Hdmpz.0.344(4-HdmpzH).appearence and electronic spectrum indicate that the dimer comprises much less than onethird of the sample. On the other hand, the polymeric phase, [Co(4-Hdmpz)2},is not presentin an amount large enough to be detected by electronic spectroscopy. This suggests that mostof the sample consists of oligometaflic species with the general formula,[Co(4-Hdmpz)J(4-HdmpzH)2.As discussed in Section, the crystal structure of thedimeric phase appears particularly amenable to polymerization of the dimer molecules. Thismay occur through a process of molecular dislocation in which a dimer molecule from oneab-sheet in the lattice moves into the gap between dimer molecules in an adjacent sheet (seeFigure 4.7) with the concomitant release of four molecules of 4-HdmpzH. TI such a processoccurs in this material (and in the analogous zinc species), then one might expect chainlengths to increase by steps of four. In other words, oligometaffic chains would be present inthe bulk material containing 2, 6, 10, etc. metal ions. That both the zinc and cobaltoligometallic species possess mean chain lengths close to six (6.2 and 5.8, respectively) isconsistent with (but by no means proof for) such a process.1484.2.2.4 MAGNETIC PROPERTIESThe magnetic studies described in Chapters 2 and 3 involved compounds of Cu(II)ions. The Cu(II) ion is a d9 system and is the example of a transition metal Jahn-Teller ionpar excellence. As a result, the electronic ground states of Cu(II) complexes are generallyorbital singlet states, thus orbital momentum in the ion is essentially “quenched’1 (aside froma small amount of orbital angular momentum due to coupling with excited states).Furthermore, the Cu(ll) ion is an S = 1/2 system, so it does not suffer zero-field splittingeffects. This means that the single-ion magnetic properties of the Cu(ll) ion arestraightforward and the observation of a temperature dependent magnetic moment in apolymetallic Cu(ll) species is indicative of inter-ion interactions, hence, the popularity, ofCu(ll) in the study of polymetallic exchange effects.Co(ll), on the other hand, does not possess such straightforward single-ioncharacteristics. Co(II) is a d7 system and octahedrally coordinated Co(II) has an orbitallydegenerate 4Tig ground state and thus the ion possesses orbital angular momentum in itsground state. Spin-orbit coupling and geometric distortion of the octahedral chromophoresplit the ground state yielding a Kramers’ doublet of lowest energy and the splitting is of theorder of kT (in the range studied experimentally). Consequently, the magnetic moment ofeven single-ion octahedral Co(ll) complexes typically shows considerable temperaturedependence. If, in addition to this, inter-ion effects are introduced to the system, then thequantitative modelling of the magnetic behaviour of these systems becomes a very difficultproblem and it becomes necessary to resort to approximations.Tetrahedral Co(II) complexes are somewhat more tractable systems. As mentioned inthe preceding section, the ground state of tetrahedral Co(II) is 4A2 and thus it is an orbitallynon-degenerate state. Little or no temperature dependence of the magnetic moment isexpected in these systems. The p value for isolated S = 3/2 ions is 3.87 11B but, because of149the admixing of a small amount of orbital momentum due to spin-orbit coupling of theground state with excited states, experimental room temperature teff values for suchcomplexes are observed in the range 3.98-4.82 I1B (129). Lower than octahedral symmetryligand fields, through second-order spin-orbit coupling, cause zero-field splitting (ZFS) of theground quartet spin state into m = ±1/2 and m = ±3/2 states where 2D represents thesplitting of these two states if the ligand field is of axial symmetry. ZFS causes greater orlesser dependence of the magnetic moment on temperature depending on the magnitude ofthe splitting.Although three oligometallic cobalt(11) compounds were prepared, magneticsusceptibility measurements were made only on [Co(4-Hdmpz)2C1(4-HdmpzH)]o.Magnetic susceptibility was measured as a function of temperature in the range 2-300 K witha powdered sample of the complex. Susceptibility and effective magnetic moment data aretabulated in Table 111-5 in Appendix Ill. A magnetic moment versus temperature plot isshown in Figure 4.10. The effective moment of [Co(4-Hdmpz)-,Cl(4-HdmpzH)]2oat 300 Kis 7.44 B’ consistent with the presence of three high spin (S = 3/2) Co(II) ions in themolecule. As the temperature is lowered, the magnetic moment drops to a value of 2.96 11B at2 K. This decrease in moment as the sample is cooled may reflect one or both of ZFS, andinter-ion antiferromagnetic exchange effects.The zero-field spin Hamiltonian appropriate to a trimetallic Co(II) complex is:Hspin = F1ex + Hsi + Hzfs , [4.8]where the various terms are defined in Chapter 1. As mentioned above, in octahedral Co(II)complexes the second and third terms have substantial influence over the magneticsusceptibility and in tetrahedral complexes the third term has influence over the magneticsusceptibility. The solution to [4.8] in which the exchange, spin-orbit coupling, and zero-fieldsplitting are treated simultaneously has not yet been found. Moreover, such a solution would1508cu0C)-jcvci20 50 100 150 200 250 300Temperature (K)FIG. 4.10. Plot of effective magnetic moment versus temperature for [Co(4-Hdmpz)2-Cl(4-HdmpzH)],Co. The solid line represents the best fit of the trimetallic Heisenbergmodel with J and J13 variable. The dashed line represents the same model with J13 fixedat zero. The dotted line represents the preceding model coupled with a molecular field term.inquire far more information than is available from powder susceptibility studies alone inorder to yield meaningful results. To circumvent this difficulty, the explicit treatment of thesecond and third terms in equation [4.8] is often neglected. This is not as bad as it sounds atfirst. If [11 is very much less than the splitting due to single ion anisotropy and if one restrictsone’s self to a low temperature range, then essentially only the lowest Kramers’ doublet ispopulated and the Co(II) may be described as an anisotropic effective spin 1/2 (5’ = 1/2) ionsubject only to exchange coupling. At high temperatures where kT is much greater than thesingle ion splitting and if [11 is also large then S = 3/2 behaviour is approximated. Of course,if the exchange interaction and single-ion splittings are comparable in magnitude then they151both must be treated explicitly in the Hamiltonian in order to quantitatively modelsusceptibility data.If single-ion anisotropy effects are neglected, then the Hamiltonian appropriate fordescribing exchange effects in trimetallic complexes is:.2J[aS1ZS+ 3(S1x2+S1YS2) [4.9]+ czS2z3+ (S2x3+S2YS3)]2Ji3[czSizS ÷(S1x3+S1YS3)]where the subscripts 1 and 3 denote the terminal Co(II) ions and the subscript 2 denotes thecentral Co(ll) ion. When a = = 1, the isotropic Heisenberg model is obtained and when a =1 and = 0, the anisoiropic Ising model results. An assumption which is often made in thedevelopment of magnetic exchange models for extended systems, be they 1-, 2-, or3-dimensional, is that only the exchange interaction between a given spin site and its nearestneighbours needs to be considered; interaction between next-nearest neighbours or beyond isvery small and can be neglected. The assumption of nearest neighbour only interactionsgreatly simplifies the development of mathematical expressions for such models whichwould otherwise be very difficult, or altogether intractable tasks. Linear, thmetalliccomplexes permit the validity of this assumption to be tested because both J (J = = J23 isthe nearest neighbour interaction) and J13 (next-nearest neighbour interaction) in [4.9] maybe accounted for explicitly.Before discussing efforts to model the magnetic data for [Co(4-Hdmpz)21(4-HdmpzH)]2Co, previous work on related trimetallic azole bridged complexes will bedescribed. In 1978, Mackey and Martin reported a study of the magnetic properties of[(Ni(1,2,4-triazole)30HNi][N06.2 (132). In this complex the Ni(ll) ions are inoctahedral coordination environments and thus they formally possess the 3A2g ground state;such ions are known to be quite isotropic. The compound exhibited relatively little ZFS and152the authors analyzed their magnetic susceptibility data employing an expression due toGinsberg er a!. based on the Heisenberg form of Haniiltonian [4.9] (133). Mackey and Martinfound that two different sets of paranters gave comparable good quality fits to the data. Inone set of parametersJ13-’— J1, with .113 antiferromagnetic, and in the other set J—2J13,withJ antiferromagnetic. Further magnetization and EPR studies led the authors to conclude thatthe latter set of coupling constants were more appropriate. With regard to the discussion ofthe preceding paragraph, one sees that this trimetallic azole bridged Ni(ll) complexexperiences a next-nearest neighbour exchange coupling with a magnitude -‘-50% of that ofthe nearest neighbour exchange.Starting about ten years ago, Reedijk and co-workers commenced a study oftrimetallic and dimetallic, first row transition metal complexes in which the bridging ligandswere alkylated 1,2,4-triazolyl moieties (134-141) and several of these reports describe workperformed on Co(II) species. The Co(Il) compounds studied by Reedijk and co-workers arerepresented schematically in Figure 4.11 along with their derived magnetic parameters inTable 4.3. These workers neglected explicit treatment of single ion effects in modelling themagnetic properties of their complexes. Figure 4.11 shows that all of the complexes, save 3,possess Co(ll) ions in pseudo-octahedral environments and the opening discussion of thissection described the highly anisotropic nature of octahedral Co(II) with regard to itssingle-ion magnetic properties. Reedijk and co-workers dealt with this anisotropy implicitlyby analyzing only the low temperature susceptibilities of these complexes using Hamiltonian[4.9] in its Heisenberg and Ising forms in the effective spin 1/2 (S’ = 1/2) formalism.Furthermore, in all cases J13 was assumed to be zero. The authors did not provide fit qualityfactors for their magnetic modelling studies, but from examination of the plots that wereprovided it is concluded that the fits were generally of moderate quality. Examination of thederived magnetic parameters in Table 4.3 indicates that complexes in which the Co(ll) ionsare triply bridged by triazoles possess values which are larger (1.11 9-15 cm1) than153SCS S7FIG. 4.11. Polymetallic Co(ll) complexes prepared by Reedijk and co-workers . Theblack dots represent Co(II) ions and the circles represent various alkylated 1,2,4-triazoles.complexes which are only doubly bridged (ignoring thiocynates) by triazole ligands (J 5-9cm). The larger g values required in the Ising model fits are consistent with the implicitassumption in using the Ising model that represents essentially only x11 (142). In somecases the Ising model provided a better fit to the data and in other cases the Heisenbergmodel provided the better fit, in spite of the fact that the bridging arrangements of most of theC SSCS 23 45 6154compounds are, superficially, quite similar. The reason for this is most likely that neithermodel adequately describes the true magnetic behaviour of these complexes.TABLE 4.3. Derived magnetic parameters for Reedijk and co-worker’s trimetallictriazolyl Co(II) complexes-JStructure typea Modelb (cm-1) g’1 HTS’l/2 7.0(3) 5.69(2)1 HTS’1/2 4.7(3) 5.16(3)2 ITS’l/2 13.2 7.8HTS’1/2 9.0 4.53 JTS’1/2 8.5 9.0HTS’1/2 5.6 5.04 IDS’l/2 10.0 7.16(zJ’+0.02)5 ITS’l/2 14.7 6.2HTS’1/2 11.2 5.l(fixed)5 HTS’l/2 11(1) 4.5(3)6 HTS’l/2 12.0(1) 4.5(3)7 HTS’l/2 14.4 4.7aSee Figure 4.11. In the cases of 1 and 5, two compounds with those structure types were prepared.bAbbreviations: ITS’1/2, Ising trimetallic S’ = 1/2; HTS’1/2, Heisenberg trimetallic S’ = 1/2; IDS’1f2,Ising dimetallic S’ = 1/2.Returning to consideration of the nearest neighbour/next-nearest neighbour exchangequestion, Reedijk and co-workers presented two reasons justifying neglect of J13 in theirmodelling procedures. Firstly, the susceptibility data for related Ni(II) and Mn(II) complexescould be fit well by ignoring the parameter (138,140). Secondly, in a related Fe(H) complexthe central iron ion undergoes a spin-crossover transition and in the low-spin state the155terminal ions exhibit no coupling (135,137). In a very recent communication, Antolini et al.have apparently laid to rest the question of J13 exchange in trimetallic cobalt complexes(143). The authors described the structural and magnetic characterization of a mixed valenceCo(ll)ICo(llI) complex in which a central Co(III) ion is connected to two terminal Co(ll) ionsby triple triazolyl bridges, analogous to the structures described above. The terminal Co(II)ions of the mixed valence complex were shown to be magnetically uncoupled. This apparentresolution of the problem will shortly be clouded through consideration of the results of themagnetic modelling studies on [Co(4-Hdmpz)2Cl(4-HdmpzH)lo.Table 4.3 shows that Reedijk and co-workers uniformly applied the S’ = 1/2formalism to the complexes they studied. This treatment was justified for they consideredonly the low temperature data in octahedral coordination complexes. In the case of [Co(4-Hdmpz)2Cl(4-HdmpzH ]o, the Co(ll) centres are tetrahedrally coordinated and theintention was to model the full 2-300 K data range, so, for the reasons discussed above,application of an S = 3/2 model was considered appropriate. The Heisenberg form ofHamiltonian [4.9] was used in combination with the spin vector ccp1ing scheme firstproposed by Kambe for trimetallic systems (144), and later generalized by Sinn (145), toderive the following expression for the magnetic susceptibility for an exchange coupled, S =312 linear trimetallic system:156Ng22 495exp[(9J+4.5J13)/kfl+252 [exP[4.5Jikfl+exp[(- 1.5Ji3)/k]3kT r I [4.10]+ 105 [exP[(4.5J13-J)/kT]+exp[(- 1.5J3-J)IkT]+exp[(3J-5.5J13)/kT]J+30 xp[(4.5J13-1 2.J)IkT] +exp[(-6J- 1 .5J3)/kTj+30 exp[(-2J-5.5J13)/kTl+exp[-7.5J13/kTJ+3 exp[(-9J- 1. 5J13)IkT]+exp[(-5J-5.5J13)/kTJ2oexp[(9J+4.5J13)IkTj+16 exp[4.5J3IkTj÷exp[(6 - 1.5J3)/kT]+12 exp[(4.5J3-J)IkT]+exp[(-1.5J3-J)/kT]+exp[(3J-5.5J) kT]+8 exp[(4.5J13-12J)/kTj+exp[(-6J- 1.5J3)/kTj+8 exp[(-2J-5.5J13)IkT]+exp[-7.5Tj+4 exp[(-9J- 1.5J3)/kTj+exp[(-5J-5.5J13)/kTjThis equation assumes no difference between central and terminal Co(ll) g values and allowsone to vary J and J13 independently. In addition to applying equation [4.101, fits to varioustemperature ranges of data were made using a thmetallic S’ = 1/2 Ising model (138), an S’ =1/2 Heisenberg model (146), and equation [4.10] combined with the molecular field equation(13):157XcaicXcaic‘ 2 2 [4.11]1 — (2zJ B ) xiwhere z is the number of interacting neighbouring molecules and J’ is the magnitude ofexchange coupling with those molecules. Equation [4.111 compensates for secondary,intermolecular interactions. The data were fit by minimizing the function [2.10] as describedin Chapter 2 and the results of these modelling efforts are listed in Table 4.4.TABLE 4.4. Derived magnetic parameters for [Co(4-Hdmpz)2Cl(4-HdmpzH)Jowithuncertainties in the last digit in parenthesesTemperature -J -J1 -zJ’Modela range (K) (cm1) (cm) (cm1) g FHTS’1/2 2-300 9(1) 0” — 4.63(8) 0.192-40 4.9(9) 0b — 3.9(1) 0.141TS’1/2 2-50 7(2) — — 11.9(5) 0.17HTS3/2 2-40 1.2(4) 0” — 1.72(8) 0.152-300 2.3(4) 0” — 2.08(4) 0.2012-300 3.6(1) — 2.287(9) 0.0272-300 2.84(3) 1.76(2) — 2.292(3) 0.011HTS3/2+MF 2-300 2.48(5) 0” 0.75(1) 2.304(5) 0.01512-300 2.3(3) 0” 0.8(2) 2.305(5) 0.013aAbbreviations: HTS ‘1/2, Heisenberg trimetallic S’ = 112; iTS’ 1/2, Ising trimetallic S’ = 1/2; HTS3/2, Heisenbergtrimetallic S = 3/2; HTS3/2+MF, Heisenberg irimetallic S = 3/2 plus molecular field exchange.at zero.Table 4.4 shows clearly that the models based on the S’ = 1/2 formalism yield very poor fits,even over a narrow temperature range. Though not shown here, allowing J13 to vary in thesemodels causes only a marginal improvement in the quality of the fits, but at the expense of anindeterminate J13. It is only the S = 3/2 based models which have yielded good fits to thesusceptibility data. A good quality fit was achieved for the data from 12-300 K by neglecting158J13 (Figure 4.10). As the molecule’s spin ground state for such a coupling scheme is S’ = 3/2,the limiting moment, for the modelled g value, as absolute zero is approached is 4.43 LB. Atthe lowest temperatures, however, the observed magnetic moment can be seen (Figure 4.10)to drop well below the moment predicted by the model. Recalling that the molecular structure(Section demonstrates considerable distortion of the terminal Co(ll) ions away fromideal tetrahedral symmetry and that the electronic spectrum for the complex supports thisobservation (Section, it is possible that the low observed magnetic moments at lowtemperatures are due to zero-field splitting. To investigate this possibility, the lowesttemperature data (<10 K) were modelled with the axial distortion, ZFS expression for S = 3/2ions (13):ii2 2ivgt 1+9eXII=kT 4(l+e)N22gB 4+(3/x)(1—e) [4.12]X±kT 4(1+e-’)2Xcaic —! Xii + Xiwhere x = D/kT (the expression for in Ref. 13 is incorrect). Fitting of the low temperaturedata to [4.12] yielded a poor fit (F = 0.10) and an indeterminate D, thus ZFS does not seem toaccount for the observed low temperature susceptibility behaviour.The best fit was achieved by allowing both J and J13 to vary and this yielded a verygood fit to the full data range as shown in Figure 4.10. The magnitude of J from this fit iscomparable to those derived before for the triazole bridged complexes because exchangeconstants determined for Co(II) as an 5’ = 1/2 ion are four times greater than the true values(147). Figure 4.12 shows the spin state energy level diagram for [Co(4-Hdmpz)2l(4-159HdmpzH)]2Cobased on the best fit and it is apparent that the ground state of the complex is100 I I80 -3D2DQD40 2D1D•11D20 2Q .23D.0•12D•2.1.30- .2 3D -0.5 1.5 2.5 35 45Total SpinFIG. 4.12. Spin state energy level diagram for [Co(4-Hdmpz)2C1(4-HdmpzH)]ocorresponding to the best fit parameters of the trimetallic Heisenberg model with ()J andJ13 variable and (ci) J13 fixed at zero. The number beside each level is the sumof the spins of the terminal Co(II) ions.S’ =1/2. If the best fit results are to be believed then [Co(4-Hdmpz)2C1(4-HdmpzH)loexhibits a next-nearest neighbour exchange interaction which is —60% of the nearestneighbour interaction. This result contradicts the conclusions of Reedijk et al.(135,137,138,140) arid Antolini eta!. (143).160Consideration of the crystal structure of [Co(4-Hdmpz)2C1(4-HdmpzH)Josuggested an alternative interpretation of the exchange interactions in the complex. In thecrystal lattice each [Co(4-Hdrnpz)2Cl(4-HdmpzH)] molecule is linked to two adjacentneighbours through hydrogen bonds (Figure 4.6). Although these hydrogen bonds are ofmoderate strength, it is possible that they might provide an intermolecular pathway forexchange. After all, in this scheme there are only four intervening atoms between terminalCo(ll) ions (—N—N—H... Cl—) whereas there are five intervening atoms in the intramolecularcoupling scheme (—N--N—Co—N—N—). Furthermore, there is considerable precedence for themediation of magnetic exchange via hydrogen bonded bridges (148). To explore thispossibility, the magnetic susceptibility of the complex was modelled using equation [4.101with J13 fixed at zero and coupled with the molecular field expression [4.111 to account forintermolecular interaction. The results of this procedure are presented in the last two lines ofTable 4.4. This model yields a good fit to the data over the whole temperature range and avery good fit if only the data above 12 K are considered (see Figure 4.10). However,application of the molecular field correction is justifiable only if J/zJ’>5-10 (13) and theresults in Table 4.4 correspond to much smaller ratios. The small J/zJ’ ratio in [Co(4-Hdmpz),C1(4-HdmpzH)j2Co suggests that if an intermolecular exchange interaction isoperative then its magnitude is such that it must be treated explicitly in the exchangeHamiltonian. One might envisage a magnetic alternating chain of trimetallic clusters.Although models have been developed for analogous alternating chain dimetallic clusters ofCu(ll) (3), no such model has been developed for the system described here.There appear to be two models which could account for the magnetic properties of[Co(4-Hdmpz)2C1(4-HdmpzH)]o, but each with its own attendant difficulties. On onehand, intramolecular coupling provides the best fit over the full temperature range, but itrequires a large next-nearest neighbour interaction to achieve that fit. On the other hand,intermolecular coupling provides an excellent fit over most of the temperature range, but the161magnitude of the intramolecular versus intermolecular coupling makes the model appliedinappropriate. Further experimentation might allow one to distinguish between the twomodels. The inset of Figure 4.10 shows Peff predicted by the two models extrapolated to verylow temperature. If susceptibility measurements could be made to these low temperatures,then the two modes of coupling could possibly be distinguished. Furthermore, magnetizationstudies could also shed light on the problem. A caveat here, though, is that if paramagneticimpurity is present in the sample it may be masking the true low-temperature behaviour ofthe complex as paramagnetic impurity contributions to magnetic properties are greatest at thelowest temperatures. Nonetheless, efforts to model the temperature dependence of themagnetic susceptibility of [Co(4-Hdmpz)2C1(4-HdmpzH)]2Co have revealed the presence ofantiferromagnetic exchange coupling between the terminal and central Co(ll) ions in themolecule and, furthermore, that this exchange coupling alone is insufficient to account for themagnetic behaviour of this molecule at low (ca. liquid helium) temperatures.4.2.3 POLY(COBALT(II) PYRAZOLATES) SYNTHESES, PHYSICAL AND THERMAL PROPERTIESDetailed descriptions of the synthetic procedures employed in the preparation ofcompounds discussed in this section are provided in Chapter 7, Section 7.2.5. Efforts toprepare and study binary copper(JI) (substituted)pyrazolates met with considerable successand as a result of this, efforts were undertaken to prepare a similar series of cobalt(II)complexes. Moreover, the success of the MLPM reaction with copper inspired efforts toprepare the cobalt species this way and preparation of [Co(4-Hpz)2]was the first synthesisattempted. The reaction was, however, unsuccessful as oxidation of the cobalt metal did notstop at the divalent state, but proceeded on to an orange Co(III) material which elementalanalyses suggested was a mixture of compounds. This result was consistent with the162observations of Vos and Groeneveld as they indicated that [Co(4-Hpz)2]was sensitive tooxidation in the presence of moisture (37).In light of this, the synthesis of [Co(4-Hpz)2]was effected using CoC12, pyrazole,and NaOH in an aqueous medium:NaOH, H20CoC12+ xs 4-HpzH > [Co(4-Hpz)2]J, +2 NaC1 + 2 H20 [4.13]N2, R.T.[Co(4-Hpz)2]was obtained as a purple powder. As discussed in Chapter 2, the metalsalt/aqueous base method of synthesis was ineffective in preparing the binary copper(II)pyrazolates and, although it did yield pure [Co(4-Hpz)2],the method was not effective inpreparing other cobalt(II) pyrazolates despite literature reports to the contrary (33,35). Twoliterature reports describe the synthesis of [Co(4-Hdmpz)2]from the condensation of[Co(acac)2J4and hydrazine (35,36). Attempts to prepare the complex via this method wereunsuccessful as judged by elemental analyses and powder X-ray diffraction. It turns out thatthe MLPM reaction is well suited to the preparation of several cobalt(II) pyrazolates fromsubstituted pyrazoles and this reaction is summarized below:02Co + xs pz*H > [Co(pz)2]+ 2 H,O1 [4.14]4’ 100-130°CI 18 h - 6 days + soluble by-productsmassiveorpowderedIn reaction [4.14], pz*H represents 3-MepzH, 4-HdmpzH, 4-MedmpzH, 4-CldmpzH, and 4-BrdmpzH. In the cases of [Co(3-Mepz)2], [Co(4-Hdmpz)2],and [Co(4-Medmpz)2]theproducts were prepared from powdered cobalt metal (a large metal surface area was requiredto obtain useful yields of these complexes) and the halogenated products were prepared from163massive cobalt metal. Presumably, further oxidation of the cobalt centres is inhibited by thepresence of the 3(5)-methyl groups on the pyrazolate ligands.All of the poly(Co(II) pyrazolates) were obtained as finely divided, purple powders.Except for [Co(4-Hpz)2j,the compounds are air-insensitive. [Co(4-Hpz)2]is moderatelysensitive to dioxygen in the presence of moisture. All of the compounds are insoluble incommon organic solvents and can only be dissolved with decomposition in mineral acids.DSC and melting point studies showed that the compounds are thermally robust (much moreso than the poly(Cu(II) pyrazolates)). In the case of [Co(3-Mepz)2],the complex undergoesan endothermic event at 405 °C with concomitant decomposition. The DSC instrument wasnot available for study of the other Co(II) complexes. However, visual observation revealedthat the compounds can be heated to temperatures >360 °C without decomposing.The particle morphologies of these compounds were examined by SEM and images ofthe [Co(4-Hpz)2J[Co(3-Mepz)2J,[Co(4-Hdmpz)2J,and [Co(4-Medmpz)2]complexes arepresented in Figures 4.13(a)-(d). It is apparent from these images that the 4-Hpz, 4-Hdrnpz,and 4-Medmpz derivatives possess elongated, crystallite morphologies, whereas the 3-Mepzderivative consists of particles which are quasi-spherical agglomerations of microcrystallites.As well, it is noted that in the case of [Co(4-Hdmpz)2]the particles are extremely small,with dimensions in the sub-micron regime. These morphologies become important in thediscussion of the magnetic properties of these complexes. X-RAY DIFFRACTION STUDIESAs stated in the introduction to this chapter, no X-ray diffraction quality singlecrystals of the cobalt(I1) pyrazolates were prepared, however, some powder diffractionstudies were undertaken with the compounds. Studies were initially undertaken using adiffractometer equipped with a copper anode in the X-ray generator and this caused technicaldifficulties in the recording of powder X-ray diffractograms for these complexes. The X-ray164FIG. 4.13. SEM images of (a) [Co(4-Hpz)2],(b) [Co(3-Mepz)2],(c) [Co(4-Hdmpz)2j,(d) [Co(4-Medmpz)2j.The white bar in each image represents a length of 5 pm.absorption edge of cobalt is coincident with the Cu Kcx X-radiation generated by the(b)(d)165instrument. Because of this, the cobalt in these compounds undergoes X-ray fluorescence(149) and high-intensity, noisy backgrounds result in the recorded diffractograms. This effectis illustrated for the case of [Co(4-Hpz)2]in Figure 10 15 20 25 30 35 40 45 50 55 602&FIG. 4.14. Powder diffractogram of [Co(4-Hpz)2jrecorded using a Cu KaX-radiation diffractometer.Because of the poor quality of such diffractograms and the possibility that the highbackground contribution might be masking an amorphous component of the materials,diffractograms were recorded using a diffractometer equipped with an iron anode X-raygenerator. Diffractograms for the alkylated Co(II) pyrazolates were recorded using thisinstrument (there were insufficient amounts of the halogenated Co(II) pyrazolates to conductsuch studies) and showed these materials to be highly crystalline. The d-spacings and relativeintensities of the diffraction lines for these compounds are listed in Appendix II, Table 11-5.Furthermore, comparison of the diffractogram of [Co(4-Hdmpz)2]with that of [Zn(4-Hdmpz)2]reveals the two complexes to be isomorphous.Referring back to the SEM images of the preceding section, one will note that the166compounds with elongated crystallite morphologies tend to have their crystallites orientedsuch that the longitudinal axes of the particles are arrayed perpendicularly to the direction ofthe original packing force (in this case, gravity). Of course, such behaviour would beexpected intuitively: when a handful of tooth-picks is dropped how do the tooth-picks lie?Advantage was taken of the geometly of the iron anode, X-ray diffractometer to see if thispreferred orientation of microcrystallites occurs in the bulk compounds. Two diffractionsamples of [Co(4-Medmpz)2]were prepared (for details see Chapter 7, Section suchthat the packing force used in preparation was in one case perpendicular to the samplesubstrate (a glass slide) and in the other case parallel to the substrate. The correspondingdiffractograms are shown Figures 4.15(a) and (b). The first indication of preferred orientationcomes from the relative intensity of the most intense line at 11.2° 20. In the perpendicularpacked sample the most intense line is relatively more intense compared to the lesser linesthan it is in the parallel packed sample. It should be noted, however, that the diffractometerused was constructed in the Debye-Scherrer geometry, in which case the relative intensitiesof low 20 diffraction lines may be affected by geometric differences between samples. Thusthis first observation should be regarded with caution. This is not a problem because there areother observations which unambiguously demonstrate the presence of preferred orientation inthis material. Comparison between the two diffractograms of the pairs of diffraction lines at20 = 14.9° and 15.75°, 20 = 24.25° and 25.05°, and 20 = 33.3° and 33.75° shows that theirrelative intensities are reversed, hence preferred orientation of the crystallites is present inthis compound. Studies of magnetic properties conducted on the Coal)(substituted)pyrazolates, which are discussed below, revealed orientation effects. The resultsof these X-ray diffraction experiments are important because, at least in the case of [Co(4-Medmpz)2], they provide an independent experimental verification of the presence ofpreferred particle orientation.167I I I I IC)CC/2JLCJ2(b)‘II I I60 50 40 30 20 102eFIG. 4.15. Powder diffractograms of [Co(4-Medmpz)2]with the sample packingforce (a) parallel and (b) perpendicular to the sample substrate.1684.2.3.3 SPECTROSCOPIC BEHAVIOUR4. INFRARED SPECTROSCOPYUnassigned band frequencies and their relative intensities for the cobalt(II)(substituted)pyrazolates are tabulated in Appendix IV, Table P1-8. N—H stretching andbending bands are absent from the spectra of all of the binary cobalt(II) complexes indicatingthat only the (substituted)pyrazolate anions are present.Above 600 cm1, the IR spectra for [Co(4-Hpz)2J,[Cu(4-Hpz)2},and [Ni(4-Hpz)2](Chapter 5) are very similar and this is mainly due to the fact that, in the 600-4,000 cmtregion of the spectrum, the bands observed are due principally to the internal modes of thepyrazolyl moiety and not to the extended structures of the complexes. This rationale is furthersupported by the observation that the spectrum of ionic Na[4-HpzJ is very similar to those ofthe aforementioned binary, divalent metal pyrazolates. The [M(4-Hpz),] CM = Co, Ni, Cu)complexes each show two strong bands in the 250-600 cm1. Na[4-Hpz] shows noabsorptions in the 250-600 cm1 region, so it is suggested that the bands observed for the[M(4-Hpz)2jcompounds in this region are due principally to M—N stretching and bendingmodes. In the cases of [Co(4-Hpz)2]and [Cu(4-Hpz),] these bands occur at similarpositions (334 and 322 cm1 in the Co(ll) species and 362 and 342 cm1 in the Cu(II) species)with the same relative intensities (the higher energy band in each pair being of lowerintensity). On the other hand, the M—N bands in [Ni(4-Hpz)2joccur at substantially higherenergies (459 and 414 cm1) and their relative intensities are reversed with respect to theformer two compounds. This difference in the low frequency JR bands of the nickel complexand the similarity of the spectra of the cobalt and copper complexes suggests that thecoordination environment in the nickel complex is quite different from the environments ofthe cobalt and copper complexes and that the coordination environments of the former two169complexes are similar. As will be discussed in Chapter 5, the nickel complex is thought topossess a square-planar coordination geometry. Thus it may be tentatively suggested that[Co(4-Hpz)2]possesses a chromophore similar to that of [Cu(4-Hpz)2j:a pseudo- or moreregular tetrahedral stereochemistry.The IR spectrum of [Co(3-Mepz)2Jis quite similar to that of [Cu(3-Mepz)2].Thec-ca mode of the 3-methylpyrazolate has shifted from the free ligand value of 358 cm1 to404 cm1 which is consistent with coordinated 3-methylpyrazolate.The [Co(4-Xdrnpz)2J(X = H, Me, Cl, Br) compounds all exhibit spectracharacteristic of the corresponding anionic, bridging 4-Xdmpz ligands (see Chapter 2,Section their high energy Vnng bands appear at 1,523, 1,511, 1,514, and 1,510 cm4and their para-methyl bands appear at 455, 519 (meta-methyl at 581 cm-1), 518, and519 cm-’ for the X = H, Me, Cl, and Br derivatives, respectively. Finally, it should be notedthat the IR spectrum of [Co(4-Hdmpz)2]is virtually identical with that of [Zn(4-Hdmpz)2]. ELECTROMC SPECTROSCOPYThe electronic spectra of the cobalt(II) (substituted)pyrazolates are shown in Figures4.16(a) and (b). It is apparent from these figures that the compounds can be divided into twogroups: the 3,5-dimethyl derivatives forming one group and [Co(4-Hpz)2]and [Co(3-Mepz)2J forming. the other group. With regard to possible chromophore geometries, thestructured band at -.550 nm (v3) in these compounds is too low in energy for an octahedralcomplex with aromatic, nitrogen donor ligands (129); the complexes are more likely topossess four-coordinate Co(II) ions. The spectra do not have the general appearance of thespectra of square-planar Co(II) species (150) and the intense colour of the cobalt(II)(substituted)pyrazolates suggests a non-centrosymrnetric chromophore. In addition, thespectra for both groups are similar in general appearance to the spectra of the oligometallicCo(ll) complexes described in Section Therefore, it is concluded that the170clC)0(I)11)ct1)C)$-i0C),.0FIG. 4.16. Electronic spectra of the Co(I1) (substituted)pyrazolates; (a) the 4-Xdmpzderivatives and (b) Co(4-Hpz)2]and [Co(3-Mepz)21.500 1000 1500 2000Wavelength (nm)500 1000 1500 2000Wavelength (nm)171[Co(pz*)2](pz* = 4-Hpz, 3-Mepz, 4-Xdmpz) complexes possess (pseudo-)tetrahedral CoN4chromophores. Band positions and relative intensities for the binary cobalt(II) pyrazolateshave been tabulated in Tables 4.5 and 4.6.TABLE 4.5. Electronic spectral data for [Co(4-Hpz)2}and [Co(3-Mepz)2]Band position’Compound (nm) (cm1) Assignmentb[Co(4-Hpz)2] 1,235 8,100 m,sh1,095 9,130 m 4T1(F).-A2)960 10,400 m,sh (v2)591 16,920s566 17,670 s,sh 4T1(P).-A2F)530 18,870 s,sh (v3)300 33,330 s IL/CT235 42,550 s[Co(3-Mepz)2] 1,235 8,100 m,sh1,080 9,250 m990 10,100 m,sh (v2)591 16,890s573 17,450 s,sh 4T1(P)—A2F)528 18,940 s,sh (v3)290 34,480 vs IL/CT222 45,050 vsaAbbreviations. w, weak; m, medium; s, strong; vs, very strong; sh, shoulder.bAssigen are based on the tetrahedral formalism. Abbreviations: SF, spin-forbidden; IL/CT, internal ligand orcharge transfer.172TABLE 4.6. Electronic spectral data for the [Co(4-Xdmpz)2](X = H, Me, Cl, Br) complexesBand positionaCompound (nm) (cm-1) Assignments’[Co(4-Hdmpz)2] 1,370 7,300 m,sh1,290 7,750 m1,135 8,810 m,sh (v2)904 11,060m617 16,210s558 17,920s522 19,160 s,sh (v3)487 20,500 w,sh SF285 35,090 vs IL/CT214 46,730 vs[Co(4-Medmpz)2j 1,405 7,120 m,sh1,295 7,720 m1,145 8,730 m,sh (v2)903 11,070 m,sh618 16,180s562 17,790 s 4T1(P)—A2F)524 19,080 s,sh (v3)311 32,150 vs IL/CT239 41,840 vs[Co(4-C1dmpz)2j 1,375 7,270 m,sh1,262 7,920 m 4T1(F)—A2)1,130 8,850 m,sh (v2)975 10,260 w,sh946 10,570m618 16,180s559 17,890 s523 19,120 s,sh (v3)486 20,580 w,sh SF312 32,050 vs258 38,760 vs,sh IL/CT228 43,860 vstable continued overleaf173TABLE 4.6. continuedBand positionCompound (nm) (cm’) Assignment[Co(4-Brdmpz)2] 1,380 7,2500 m,sh1,255 7,970 m1,135 8,810 m,sh (v2)932 10,7300m,620 16,130s562 17,790 s 4T1(P)—A2F)523 19,120 s,sh (v3)310 32,260 vs IL/CT233 42,920 vsaAbbreviations: w, weak; m, medium; s, strong; vs, very strong; sh, shoulder.bAssigents are based on the tetrahedral formalism. Abbreviations: SF, spin-forbidden; IL/CT, internal ligand orcharge transfer.Using the tetrahedral formalism, equations [4.6] and [4.7] were used to calculate Dq and Bvalues for the complexes considered in this section: these values are listed in Table 4.7.TABLE 4.7. Ligand field parameters for the [Co(pz*)2]complexes (pz* = 4-Hpz, 3-Mepz, 4-Hdmpz,4-Medmpz, 4-Cldmpz, 4-Brdmpz)(numbers in parentheses represent estimateduncertainties)Dq B V2*a v3*Compound (cm) (cm’) (cm-1) (cm-’)[Co(4-Hpz)2] 540(10) 705(20) 9,130 17,650[Co(3-Mepz) 550(10) 680(20) 9,260 17,450[Co(4-Hdmpz)2] 485(20) 775(40) 8,330 17,890[Co(4-Medmpz) 485(20) 780(40) 8,330 17,760[Co(4-Cldmpz)2] 485(20) 770(40) 8,330 17,820[Co(4-Brdmpz) 485(20) 765(40) 8,330 17,730a2* and are the visually estimated band centres for and v3 used in the calculations of Dq and B.174The Dq and B values for [Co(4-Hpz)2]and [Co(3-Mepz)2]are comparable with those listedin Table 4.2 for the oligometallic species. Moreover, the Dq values for these complexes areessentially the same as those found for a number of [Co(py*)4]+cations (py* = pyridine, 4-methylpyridine, and 3-methylpyridine) while the B values are somewhat larger in thepyrazolate complexes than in the pyridine complexes (150). The ligand field parametersderived for the [Co(4-Xdmpz)2Jcomplexes are the same within uncertainty, but as a groupthese complexes possess smaller values for Dq and larger values for B than those of [Co(4-Hpz)2Jand [Co(3-Mepz)2J.This is consistent with a weaker interaction between the Co(II)ions and the nitrogen donor atoms of the ligands in the cases of the 4-X-3,5-dimethylpyrazolates than in the other two pyrazolate complexes. Such a finding contradictswhat one might expect intuitively in this series of complexes. Alkyl groups are electrondonors to aromatic rings, thus the dimethylated pyrazolates might be expected to interactwith the Co(1T) ion more strongly than pz or 3-Mepz.Following the discussion of Section with regard to bandwidths and bandstructure, it is proposed that the cobalt(ll) 4-X-3,5-dimethylpyrazolates possess CoN4chromophores with pseudo-tetrahedral geometries which are substantially distorted from theregular tetrahedral geometry and ICo(4-Hpz)2] and [Co(3-Mepz)2J have CoN4chromophores which are much closer to a regular tetrahedral geometry. It is this basicdifference in chromophore geometry which likely contributes most to the differences in thecalculated “tetrahedral” ligand field parameters between these two groups of compounds aswill be discussed later. PROPOSED STRUCTURES AND MAGNETIC PROPERTIESAs with the [Cu(4-Xdmpz)2]complexes, preparation of the cobalt analogues did not175afford diffraction quality single crystals of these compounds, so discussion of their magneticproperties is preceded by a consideration of their probable structures. The cobalt(II)(substituted)pyrazolates are insoluble, infusible, involatile materials and such properties areconsistent with polymeric structures. The empirical formulae of the compounds could besatisfied if their structures consisted of chains of Co(ll) ions doubly bridged by(substituted)pyrazolate ligands as confirmed in the [Cu(4-Xpz)2]compounds and suggestedin the [Cu(4-Xdmpz)2]species, [Cu(3-Mepz)2],and [Cu(indz)2](Chapter 2). The electronicspectra of the Co(II) complexes indicate that they possess (pseudo-)tetrahedral CoN4chromophores and, furthermore, that the 4-X-3,5-dimethylpyrazolate compounds havegeometries quite distorted from the ideal tetrahedral stereochemistry. Such chromophorespreclude the formation of oligometallic species analogous to the one depicted in Figure 2.16.There are other indirect pieces of evidence which suggest that the Co(ll) compoundsconsidered here consist of chains of (pseudo-)tetrahedrally coordinated metal ions. One mayeasily envisage the structures of the dimetallic and trimetallic Co(II) 3,5-dimethylpyrazolates,determined by X-ray diffraction, as being small segments of an extended linear chain.Finally, [Zn(4-Hdmpz),] most certainly consists of chains of tetrahedrally coordinated Zn(II)ions as this ion contains a spherically symmetric, filled 3d electronic shell and the polymer isderived from the dimeric species in which the Zn(II) ions are tetrahedrally coordinated. Thecobalt analogues of these zinc compounds are isomorphous with the latter and this lendsfurther credence to the proposed structures for the binary cobalt pyrazolates. Thus, a pictureemerges of the general structure of these compounds and it is illustrated in Figure 4.17.Recall that electronic spectroscopy indicated the 4-X-3,5-dimethylpyrazolates ofcobalt(ll) possess more distorted tetrahedral geometries than [Co(4-Hpz)2]or [Co(3-Mepz)2] The structure proposed in Figure 4.17 may account for this experimentalobservation. When the R groups on each pyrazolate ring in the figure are both hydrogenatoms (4-Hpz) or if only one R group per ring is a methyl group (3-Mepz), molecularmodels indicate that there is negligible steric interaction between one pyrazolate ring and its176x x xRRFIG. 4.17. Proposed structure for the Co(II) (substituted)pyrazolate polymers(where R is H or Me and X is H, Me, Cl, or Br).adjacent, translational equivalent. In these two cases there is no obvious intra-chain, interligand effect to cause distortion of the chromophores away from the ideal Td symmetry inwhich the angle, a (Figure 4.17), would be 109.47° and the dihedral angle between fusedCo(N—N)2orings would be 900. However, if the CoN4 chromophores are strictly tetrahedraland both R groups on each pyrazolate ring are methyl substituents, then there would be asubstantial, repulsive steric interaction between adjacent, translationally equivalentpyrazolate ligands. The structure could minimize this repulsion by lengthening the Co—Nbonds and/or undergoing an S4 elongation of the chromophore along the chain axis. The latterprocess would entail a compression of a to a value less than the tetrahedral angle and as bondangle deformations are generally lower energy processes than bond length changes, the aangle compression is likely to account for most of the deformation described above. If someCo—N bond lengthening has also occurred in the dimethyl derivatives relative to the othertwo cobalt polymer derivatives, then this may account for the lower values of Dq and highervalues of B observed in the former compounds compared to the latter two species. Thex x x177proposed large chromophore distortion in the 4-X-3,5-dimethylpyrazolates relative to thepyrazolate and 3-methylpyrazolate species is also supported by the magnetic properties ofthese compounds which are considered in the following section.Magnetic susceptibilities of the [Co(4-Xdmpz)2]compounds and [Co(3-Mepz)2]were measured from 2 to 300 K using a SQUID magnetometer. In the case of [Co(4-Hpz)],magnetic susceptibility versus temperature was measured over the range 4.2-82 K using aVSM magnetometer. Magnetic measurements were also made over a wider temperaturerange for this compound using a SQUID magnetometer, however, in this case the sample wasexposed to the atmosphere for several hours (recall that [Co(4-Hpz)2]is moderatelydioxygen sensitive in the presence of moisture). Although no visible decomposition wasobserved, it is possible that some decomposition of the sample occurred during this timewhich may have affected the SQUID magnetometer derived susceptibility values and,accordingly, these values should be regarded with a certain degree of caution. [Co(4-Hpz)2J,[Co(3-Mepz)2], [Co(4-Hdmpz)2J, and [Co(4-Medmpz)2]exhibit maxima in theirsusceptibility versus temperature plots at 38, 25, 11, 15 K, respectively. The halogenateddimethyl cobalt(II) derivatives show incipient susceptibility maxima at 10 K. These maximaindicate the presence of antiferromagnetic coupling in all of the complexes. The versustemperature plots for the cobalt complexes are shown in Figure 4.18(a). The plotting of thesedata on a single graph emphasizes the point that at higher temperatures [Co(4-Hdmpz)2j,[Co(4-C1dmpz)2],and [Co(4-Brdmpz)2]exhibit similar magnetic susceptibilities whereas[Co(4-Medmpz) and [Co(3-Mepz) show substantially higher susceptibilities. In light ofprevious discussions noting the similarities between the four 4-X-3,5-dimethylpyrazolates thehigher susceptibility in the 4-methyl derivative seemed unusual. Furthermore, as mentionedin Section, the room temperature magnetic moments found for “magnetically dilute”,tetrahedral Co(II) compounds range from 3.98-4.82 while magnetic moments for [Co(4-Medmpz)2]and [Co(3-Mepz)2]are 4.93 J.tB and 5.14 B’ respectively. These moments areunusually high, especially considering that the temperature dependence of their178-40EC,,SC)>0.02FIG. 4.18. Powder susceptibility versus temperature plots for the Co(ll) (substituted)pyrazolate polymers: (a) uncorrected and (b) corrected susceptibility values in the cases of[Co(3-Mepz)2]and [Co(4-Medmpz)2]./DC9I I I0C,,SC)-4C)riD0. [Co(4—Hpz)2J• [Co(3—Mepz)• [Co(4—Hdmpz)][Co(4—Medmpz)2[Co(4—C1dmpz)J[Co(4—Brdmpz)2••.I8I88II.4I0 50 100 150 200Temperature (K)I I I250 3000.04 01(b)o [Co(4—Hpz)2]• [Co(3—Mepz)• [Co(4—Hdmpz)2]8 [Co(4—Medmpz)° [Co(4—C1dmpz)2][Co(4—Brdmpz)Ii,II,,t,iII I I100 150 200 250 300Temperature (K)0.000 50179susceptibilities indicates antiferromagnetic coupling in the compounds! Because of theseobservations, there was a concern that the samples of [Co(4-Medmpz)2]and [Co(3-Mepz)2]were contaminated by some other agent.As described in Section, all of the binary cobalt(ll) (substituted)pyrazolates,save [Co(4-Hpz)2J, were synthesized via the MLPM reaction. In the cases of [Co(4-Hdmpz)2], [Co(4-Medmpz)2],and [Co(3-Mepz)2]the compounds were prepared usingcobalt metal powder and only a small fraction of the cobalt metal was consumed during thereaction, thus a mixture of the finely divided cobalt(ll) pyrazolate derivative and excesscobalt metal is obtained at the end of the reaction period. Cobalt metal is ferromagnetic andthus it was separated from the product mixture by repeatedly dredging a solvent slurry of theproduct mixture with a magnet until no more cobalt metal was observed on the magnet. It ispossible that in the cases of the two compounds which show unusually high magneticmoments cobalt metal was not completely removed from the samples. To investigate thispossibility, field dependence studies were conducted on the two compounds in question. Low-dimensional antiferromagnetically coupled systems possess only short-range spin order andthus may be described as existing in the paramagnetic state. As such, these materialsgenerally do not exhibit an applied field dependence of their susceptibilities in moderatefields, i.e., their M versus H plots are linear. On the other hand, cobalt metal is a long-rangeordered ferromagnet and below its saturation field (—9,000 Oe at room temperature) cobaltshows non-linear field dependence levelling-off to a constant magnetization above thesaturation field. Thus, if a small amount of cobalt metal is present in [Co(4-Medmpz)2]and[Co(3-Mepz)2},then magnetization versus applied field plots for these compounds shouldexhibit non-linear behaviour below 9,000 Oe and linear behaviour above 9,000 Oe whichextrapolates to non-zero magnetizations at zero applied field. These intercept values can beused to back-calculate the amount of cobalt metal present in the samples. Magnetization plotsare shown for [Co(4-Medmpz)2]and [Co(3-Mepz)2jin Figures 4.19(a) and (b), respectively(magnetization data for these compounds are tabulated in Appendix III, Table 111-6). The1803.5 I I(a)3.0- :: lOOK‘25- 0.40.26200 400200 K01.5-0)ctS0.5 -0.01 I IC 10000 20000 30000 40000 50000Applied field (Oe)5— 4C.)S0-z20)0 10000 20000 30000 40000 50000Applied field (Oe)FIG. 4.19. Isothermal magnetization plots for (a) [Co(4-Medmpz)2jand (b) [Co(3-Mepz)z].1810 2000 4000 0000 8000 10000(b)100200 K300 Kbehaviour predicted if small amounts of cobalt were present in the samples is precisely whatis observed in these plots and it is concluded that the suspect samples contained minuteamounts of cobalt. Based on these plots, it was calculated that the samples of [Co(4-Medmpz)2]and [Co(3-Mepz)2]were, respectively, 99.94% and 99.89%, by weight, pure.Obviously, cobalt impurities at these levels would not have been detected by the standardcharacterization techniques applied to these compounds before magnetic studies wereundertaken. Further support for these results comes from a similar field dependence studyconducted on [Co(4-Hpz)2J.This complex was not synthesized by the MLPM reaction and,so, was not expected to exhibit any non-linear field dependence of its magnetization. Thisexpectation was met; [Co(4-Hpz)2Jshowed linear magnetization behaviour over the entirefield range studied. Because the variable temperature susceptibility studies on thesecompounds were conducted in an applied field of 10,000 Oe, the cobalt contaminant wasmagnetically saturated over the temperature range studied and made an effectivelytemperature independent contribution to the observed susceptibilities. The original data werethus corrected by subtraction of the constant contribution from the cobalt metal and thecorrected data are shown in Figure 4.18(b) along with data from the other complexes. It isapparent that the correction has brought the higher temperature susceptibilities of [Co(3-Mepz)21and ICo(4-Medmpz)21in line with the other three dimethyl derivatives.During initial magnetic studies of these compounds discrepancies were observedbetween the data collected using the VSM and SQUID magnetometers. Both instrumentssubject the samples to homogeneous applied fields, but the orientations of the fields withrespect to the way in which the samples are packed into their holders are orthogonal: in theVSM magnetometer the applied field is perpendicular to the sample packing force directionand in the SQUID magnetometer the field is parallel to the sample packing force direction.As discussed in Section, Co(II) complexes often exhibit considerably anisotropicmagnetic properties. The elongated crystallite morphologies observed in the SEM images of[Co(4-Hpz)2], [Co(4-Hdmpz)2], and [Co(4-Medmpz),] and the preferred orientation182observed in the powder X-ray diffraction study of [Co(4-Medmpz)2]led to the hypothesisthat the crystallites in the samples used for magnetic studies were subject to a certain degreeof preferred orientation which was partially resolving the anisotropic magnetic properties ofthese complexes. To test this hypothesis pellets of the samples were pressed under highpressure in the hope that this would maximize the preferred orientation of the sampleparticles. Based on the morphologies of the crystallites, the model depicted in Figure 4.20 isproposed for the preferred orientation exhibited by these compounds. If one assumes that theDirection of packing forceSample pellet.FIG. 4.20. Diagram showing the preferred orientation of crystallites in samplesused for the magnetic studies of the Co(II) (substituted)pyrazolates.polymer chain axes of these complexes run parallel to the long axes of the crystallites, thenone recognizes that application of an external field normal to the pellet face depicted inFigure 4.20 (referred to as the perpendicular direction) will permit an enhanced sampling ofthe magnetic properties of these compounds in the plane normal to the chain axis.Application of the external field in the direction parallel to the pellet face (referred to as therandom orientation) should actually result in random sampling of crystallite orientationsbecause there is little difference in the crystallite edge lengths perpendicular to the long axisHrandom183and, of course, a random distribution of long axis orientations within the plane perpendicularto the packing force direction.Susceptibility versus temperature studies were conducted (with the SQUIDmagnetometer) using pellets of each of the cobalt(1I) pyrazolate derivatives, except for [Co(4-Hpz)2], in the two orientations depicted in Figure 4.20 and the results are shown in Figures4.21(a) through (J). Magnetic susceptibility and effective magnetic moment versustemperature data for the compounds are tabulated in Appendix Ill, Table 111-7. The 4-X-3,5-dimethylpyrazolates show essentially no difference between the random and perpendicularorientation susceptibilities from room temperature down to approximately 30 K, howeverbelow this temperature the susceptibilities of the two orientations begin to diverge, with theperpendicular orientation values larger than the random orientation values in all four cases. Inthe case of [Co(3-Mepz)2](Figure 4.21(f)), there is little difference between the measuredsusceptibilities in the two orientations at the lower temperatures and the divergence that isobserved is the opposite of what was observed in Figures 4.21(a) through (d); theperpendicular orientation susceptibilities lie somewhat lower than the random orientationvalues. The poorer resolution of susceptibilities in the two orientations with this complexmay be due either to a smaller anisotropy in its magnetic properties and/or the that fact thatthe particles of this material appear to be roughly spherical aggregates of smallermicrocrystallites (see Figure 4.13(b)). As a result, they are not likely to exhibit muchpreferred orientation. The orientation studies on [Co(4-Hpz),] (Figure 4.21(f)) should beregarded with a modicum of caution. The susceptibilities for the random orientation arereliable, but only span the temperature range 4.2-82 K. The perpendicular orientation valueswere recorded using a sample which had been exposed to the atmosphere, as discussed at thebeginning of the section. The decomposition product of the compound is most certainly adiamagnetic cobak(llI) species, so partial decomposition of [Co(4-Hpz)21would tend toresult in lower overall susceptibility values. Having stated this, it is very unlikely that theapparent anisotropy exhibited by this compound is totally due to decomposition because such184a degree of decomposition would be visually detected. These oriented pellet studies haveFIG. 4.21 continued overleafI I0.04 - .‘% (a) -‘ •.5 V..C,,o0.03o 0•0880.02 - a.aCl)0.01 I I0 20 40 60 80 100Temperature (K)I I0.06 (c).5C,,E0040.- 0 •-:5VC) 8.C.)0.02- 1.. —I I0 20 40 60 80 100Temperature (K)004 I I.— • (b).5C’) 0.03o •.I0.02 -C)C.)••(I)0.01 I0 20 40 60 80 100Temperature (K)0.05— 0(d)I.50.04eC’)I: 0.03—Saz I.C) II..C.)0.02- —Cl)0.01 I I0 20 40 60 80 100Temperature (K)1850.030 I 0.04 I I(e)—. (f)0.025 - 0 0 000 000 0.030 00 •:, 0.020 i 0 • 0 0 .‘ 0• 0 ..0 ‘ • .00.02-0.0150.010 I I I 0,01 I I0 20 40 60 80 100 0 20 40 60 80 100Temperature (K) Temperature (K)FIG. 4.21. Oriented pellet susceptibility versus temperature plots for the Co(II)(substituted)pyrazolate polymers: (a) [Co(4-Hdmpz).,j, (b) [Co(4-Medmpz)2],(c)[Co(4-Cldmpz)2],(d) [Co(4-Brdmpz)21,(e) [Co(4-Hpz),] and (f) [Co(3-Mepz)2]The open circles represent the random orientation and the solid circles represent theperpendicular orientation.clearly demonstrated anisotropy in the magnetic properties of four of the cobalt(II)(substituted)pyrazolates considered in this section; however, the studies do not yieldquantitative information about the magnitudes of anisotropies because whether the pressedpellets resolve preferred orientation to the extent of 1%, 10%, or 50% is not known(macroscopic single crystals are required for a full determination of magnetic anisotropy incompounds).The reader will recall from the discussion in Section that quantitativemodelling of the magnetic properties of cobalt(ll) species can be a daunting task. Thepolymeric cobalt(II) (substituted)pyrazolates are no exception to that general statement. The186results of efforts to model the susceptibility data for [Co(4-Hdmpz)2C1(4-HdmpzH)]o,though not completely conclusive, seemed to indicate that Heisenberg, antiferromagneticexchange between nearest neighbours was operative in the trimetallic complex. Using this asa starting point, attempts to model the susceptibility data for the polymeric Co(II) complexeswere made employing the S = 3/2, Heisenberg model for antiferromagnetically coupled linearchains developed by Hiller et a!. (152) based on the numerical calculations of Weng (153):N2 2rI 1.2500 + 17.041xXchain= I [4.15]kT [1 +6.7360x+238.47xwhere x = !j/kT. As with the polymeric copper(II) pyrazolates, equation [4.15] was combinedwith the paramagnetic impurity term [2.8] into equation [2.9] and the random orientationsusceptibility data were fit to equation [2.9] by minimization of function [2.10]. The derivedbest fit parameters over various temperature ranges are listed in Table 4.8. The parametervalues listed in Table 4.8 correspond to best fits which vary in quality from very poor toexcellent depending on the compound and temperature range considered. The curvescalculated from these fits are shown in Figures 4.22(a) through (f. In the cases of [Co(4-Hpz)2j and [Co(3-Mepz)2], fits to the data over the full temperature ranges studiedreproduce reasonably well the temperatures at which the maxima in the susceptibilities areobserved, however, the fits do not adequately model the sharpness of the maxima. For the[Co(4-Xdmpz)2]complexes, the Heisenberg chain theory models the low temperature datapoorly when applied to the full data range as the temperatures at which the calculatedsusceptibility maxima appear are too high. Modelling of the high temperature data rangesyield very good fits with only two parameters, J and g, but extrapolation of these fits to lowtemperatures, as shown in the figures, reveals shortcomings. In the cases of the [Co(4-Xdmpz)2}species, the predicted susceptibility maxima are too low, too broad, and appear187TABLE 4.8. Derived magnetic parameters for the [Co(pz*)2]complexes (pz* = 4-Hpz, 3-Mepz, 4-Hdmpz, 4-Medmpz, 4-Cldmpz, 4-Brdmpz) using the Heisenberg, S = 3/2,antiferromagnetically coupled chain model with estimated standard deviation in the last digitin parenthesesTemperature -JCompound range (K) (cm-1) g F[Co(4-Hpz)2] 4.2-41 5(3) 2.1(7) oa 0.1635-82 5.00(8) 2.35(1) oa 0.00774.2-82 7(1) 2.5(2) 0 0.15[Co(3-Mepz)2] 2-35 5(2) 2.4(4) 0.4(2) 0.05750-300 3.68(2) 2.242(1) 0(1 0.0272-300 4.43(8) 2.29(1) 0.46(7) 0.042[Co(4-Hdmpz)21 2-20 1.89(6) 1.64(2) 0.30(8) 0.01650-300 4.6(1) 2.24(1) oa 0.0242-300 3.31(9) 2.15(2) 0.5(1) 0.062[Co(4-Medmpz),1 2-20 2.63(9) 1.76(3) 1.73(8) 0.009150-300 5.74(8) 2.341(7) oa 0.0152-300 4.4(1) 2.24(2) 1.2(1) 0.051[Co(4-Cldmpz)21 2-20 2.2(1) 1.62(4) 6.3(3) 0.01450-300 5.08(3) 2.238(2) oa 0.00482-300 4.3(1) 2.17(1) 3.9(1) 0.043[Co(4-Brdmpz)2] 2-20 2.16(5) 1.63(2) 4.06(9) 0.007550-300 4.42(9) 2.168(7) oa 0.0162-300 3.75(8) 2.11(1) 2.7(1) 0.041Fixed at zero.at much higher temperatures than the observed maxima. The fits to the [Co(4-Hpz)2]and[Co(3-Mepz)2]data are better, except that now the observed susceptibilities decrease morerapidly below the maxima than the model predicts. Fits to the low temperature data for [Co(4-Hpz)2]and [Co(3-Mepz)2],, are rather poor. Again, the model does not accommodate therapid decrease of the observed susceptibilities at temperatures below those of thesusceptibility maxima. The quality of the low temperature fits to the data for the [Co(4-Xdmpz)2]compounds is excellent. However, the calculated g values are unrealistically low,188so the fits may simply be fortuitous.7 0.030SS00.02.0Q),,S..2. 0.030.02a)00.01FIG. 4.22 continued overleaf7 0.030SC.,SV0.02.00.a)C.)0.010.060S0. 50 100 150 200 250 300 0 50 100 150 200 250 300Temperature (K) Temperature (K)0 50 100 150 200 250 300Temperature (K)0.000 50 100 150 200 250 300Temperature (K)1890.0300.030,025 -.‘ 0.020 . 0.02-0.015CI) CI) 0.010.010 I I I I0 20 40 60 80 100 0 50 100 150 200 250 300Temperature (K) Temperature (K)FIG. 4.22. Susceptibility versus temperature plots for the Co(II) (substituted)pyrazolate polymers: (a) [Co(4-Hdmpz)2J,(b) [Co(4-Medmpz)2],(c) [Co(4-Cldmpz)2],(d) [Co(4-Brdnipz)2J,(e) [Co(4-Hpz)J,and (I) [Co(3-Mepz)1.The lines are the calculated curves for the best fit parameters listed in Tables 4.7 through 4.9. The solidlines correspond to the Heisenberg model and the dotted lines correspond to the Isingmodel.The S = 3/2, Heisenberg model for antiferromagnetically coupled chains accounts forthe magnetic behaviour of the binary cobalt(ll) (substituted)pyrazolates over limitedtemperature ranges (particularly at higher temperatures), but it does not describe thebehaviour of the complexes over the full temperature ranges examined. In considering thisinadequacy, the observed temperature dependence of the susceptibilities of these complexesneeds to be compared with what would be expected qualitatively. As mentioned previously,all six compounds exhibit maxima or incipient maxima in their susceptibility plots.Furthermore, over the 2-300 K range, Figure 4.18(b) shows that the compounds exhibitsusceptibilities of roughly similar magnitudes. Generally, given a group of similar systems,I I I(e)190studies of magnetic materials indicate that stronger antiferromagnetic exchange interactionslead to maxima at higher temperatures and lower overall susceptibilities (for example,consider the copper(II) (substituted)pyrazolates). In spite of the fact that the present groupexhibits susceptibilities of similar overall magnitude, the Co(ll) 4-X-3,5-dimethylpyrazolatesshow maxima at significantly lower temperatures than [Co(4-Hpz)2]or [Co(3-Mepz)2].This inconsistency may be due to zero-field splitting of the Co(II) single-ion susceptibilitiesinduced by distortion of the pyrazolate ligand field away from tetrahedral symmetry. ZFSmanifests itself in powder susceptibility measurements as a reduction of the observedsusceptibility (which becomes more pronounced at lower temperatures) from what would beexpected from the corresponding Curie law behaviour. Electronic spectroscopy and structuralconsiderations support this supposition (see above). Referring to Figure 4.17, it is suggestedthat a is smaller in the 4-X-3,5-dimethylpyrazolates than in [Co(4-Hpz)2]or [Co(3-Mepz)2]If compression of the a-angle is the only significant distortion in these complexes,then an axially symmetric CoN4 chromophore results and single-ion susceptibilities areinfluenced by the ZFS parameter, D, in expression [4.12]. If D is positive, then the m = ± 1/2state is stabilized relative to the m = ±3/2 state and vice versa if D is negative. A priori, it isdifficult to predict the sign of D, but angular overlap model calculations performed onidealized Co(ll) chromophores (129) predict that for values of a lower then the tetrahedralangle, D should be negative. Thus, one might expect m = ±3/2 to be the ground state in theCo(H) 4-X-3,5-dimethylpyrazolate polymers. Such a ZFS effect would have importantconsequences for exchange coupling in these systems because at temperatures where kT<2Dthe ,n = ±1/2 state becomes depopulated and the spin vector is essentially restricted toalignment along the principal axis of the CoN4 chromophore (in the case of the 4-X-3,5-dimethylpyrazolate compounds it has been proposed that the principal chromophore axis isparallel to the polymer chain axis, but it is not a requirement that this generally be the case).191If the spins on the individual Co(II) ions can orient only along one chromophore axis and ifexchange coupling between ions occurs, clearly, the coupling can occur only through thatsingle component of the spin vector. In such a case, the single ion anisotropy will havecaused the coupling to appear Ising-like. Conversely, if D is positive, then the m = ±1/2 statebecomes the ground state and in an exchange coupled system, the coupling will appear to beXY-like at temperatures where kT<2D. Thus, attempts to quantify exchange interactions weremade employing available anisotropic models.Smith and Friedberg (154) have developed expressions for modelling susceptibilitydata which account for axial ZFS and Heisenberg exchange coupling simultaneously. Theseexpressions are valid only when [11>IDI. It is possible to gain some estimate of whether thatcondition is met in the present compounds by considering the Heisenberg chain modellingresults and the oriented pellet studies together. The Heisenberg modelling studies suggest thatthe I]] values for the [Co(4-Xdmpz)2Jcomplexes lie in the range 1.5-6 cm-1. Assuming thatthe anisotropy exhibited in the oriented pellet studies is due to single-ion, ZFS effects, theobserved behaviour (which is a lower limit to the magnitude of the anisotropy) indicates thatIDI is of the order of 10 cm* Clearly, the condition required to apply the expressions ofSmith and Friedberg (154) is not met in these compounds. If the magnitude of D is muchgreater than the lower limit suggested by the oriented pellet studies, then at low temperaturesonly the lowest Kramers’ doublet of the electronic states manifold is populated and the Co(II)ions in these compounds will behave as S’ = 1/2 ions. Anisotropically coupled, S = 1/2, linearchain models have been studied theoretically. Bonner and Fisher have used numericalmethods to calculate the susceptibility behaviour of exchange coupled finite rings of S = 1/2spin centres and then extrapolated those results to infmite chains (68). The coupling of thesesystems are governed by the Harniltonian:= 2J2(SjzSjz) + a(SxSx + SYSY) [4.16]ij192where a varies between zero and one. When a = 1 the Heisenberg model is obtained andwhen a = 0 the Ising model results. Closed-form expressions have been derived for themagnetic susceptibility of Ising coupled, S = 1/2, linear chains (68,155) and they arepresented below:i..i 2 2ivg11 PB‘“kT= [4.17a]4kTXi= Ng2t[tanh + z)ech2fj [4. 17b1X 4X + hi [4.17c]In Hamiltonian [4.16], if the z component term is set to zero and a is set equal to one then theX-Y exchange coupling Hamiltonian obtains. Katsura (156) derived the following expressionfor the perpendicular susceptibility resulting from X-Y exchange coupling in S = 112 linearchains:(Iti 2 2 Iivg 11B I dwXi iickT J cosh[(J/kl)cosw]0where o = 2nk/N.It was suggested above that the proposed distortion of the chromophore in the 4-X-3,5-dimethylpyrazolates might induce Ising-like exchange coupling between Co(ll) centres in thepolymers at low temperatures. To investigate this possibility, fitting of the expressions in193[4.17] to the low temperature susceptibility data was attempted. Before presenting the resultsof these modelling efforts, it is important to consider how the expressions in [4.17] might bereasonably applied. Implicit in the assumption that S’ = 1/2 models apply to the behaviour ofthese compounds is the condition that g1 and g1 are anisotropic effective components of thetrue g-tensor for the S = 3/2, Co(II) ion. In an axially distorted system, the relationshipsbetween the effective, g’, components and the true g values are as follows (129): g11’ = 3g11 andg1’ = 0 (rhombic distortions of the chromophore may alter the values of the effective g’tensor components). Thus, if Ising-like exchange is present in these compounds, use of[4. 17a] alone in [4. 17c] should provide an adequate fit to the observed data. Recall thatFigure 1.7 in Chapter 1 illustrates the susceptibility behaviour of the S = 1/2 Ising model inreduced coordinates. It is important to note that, according to the model, X11-0 as T-.O.Contrary to this prediction, the experimentally measured susceptibilities show an increase inmagnitude as the lowest temperatures are reached. In Chapter 2 this behaviour was ascribedto a small amount of paramagnetic impurity in the Cu(II) (substituted)pyrazolates and it wasaccounted for by inclusion of a Curie law or Curie-Weiss law term in the susceptibilitymodel. In the corresponding Co(H) species considered here, experimental observationsindicate the presence of anisotropy in these compounds so that inclusion of a Curie law termmay be inadequate to account for the lowest temperature susceptibility data. Consequently,the low temperature data for the compounds were fit using [4. 17c] with fixed at zero andthis expression was combined with a Curie-Weiss law paramagnetic impurity term as inequation [2.11] where Xp&a was modelled using expression [4.12] (g = 4g11’). The best fitresults are presented in Table 4.9. The F values in Table 4.9, along with the correspondingcalculated curves in Figures 4.22(a) through (1), show this model to provide an excellent fit tothe low temperature data for these complexes. Comparison with Table 4.8 shows that theIsing parallel susceptibility model coupled with a Curie-Weiss law pararriagnetic impurityterm yields higher quality fits to the observed data, particularly in the cases of [Co(4-Hpz)2]194TABLE 4.9. Derived magnetic parameters for low temperature data fits of the S = 1/2, Ising x11model coupled with a Curie-Weiss law paramagnetic impurity component for the [Co(pz*)](pz* = 4-Hpz, 3-Mepz, 4-Hdmpz, 4-Medmpz, 4-Cldmpz, 4-Brdmpz) complexes. Estimatedstandard deviation in the last digit in parenthesesTemperature -J -eCompound range (K) (cm-1) g11’ (K) F[Co(4-Hpz)2] 4.2-41 32.6(2) 9.34(5) 18(2) 51(5) 0.0044[Co(3-Mepz) 2-35 28.6(4) 9.58(8) 15.3(4) 13(1) 0.0061[Co(4-Hdmpz)2] 2-20 12.5(1) 6.06(2) 24(2) 13(1) 0.0041[Co(4-Medmpz) 2-20 16.6(2) 6.42(3) 13.8(3) 6.8(2) 0.0031[Co(4-Cldmpz)2} 2-20 16.2(1) 6.27(2) 14.2(1) 3.1(1) 0.0020[Co(4-Brdmpz)J 2-20 15.0(2) 6.22(3) 14.1(3) 4.1(2) 0.0039and [Co(3-Mepz)2J,than does the Heisenberg model with a Curie law paramagnetic impurityterm. Of course, the former is a four-parameter model while the latter is only a three-parameter model. Fits to the low temperature data using the Heisenberg model coupled withthe Curie-Weiss paramagnetic impurity term only yield marginal decreases in the F values atthe expense of indeterminate %P. It should be noted that the %P values and Weiss constantsobtained with the Ising parallel susceptibility model are rather large in magnitude. The largevalues of the paramagnetic impurity term parameters demonstrate the limitation in applying aclassical model to account for the low temperature magnetic behaviour of these compounds.In an attempt to factor-out the contribution made by the paramagnetic impurity termsto the quality of the 1-D Ising and Heisenberg models, the data in the region of thesusceptibility maxima were modelled without the benefit of paramagnetic impurity terms.This was done only for the [Co(4-Hpz)2]., [Co(3-Mepz)2j,[Co(4-Hdmpz)2],and [Co(4-Medmpz)2]species because the halogenated derivatives do not exhibit discrete susceptibilitymaxima. The results are presented in Table 4.10.195TABLE 4.10. Derived magnetic parameters for low temperature data fits of the S = 1/2, Isingand S = 3/2, Heisenberg models to the data for the [Co(pz*)2](pz = 4-Hpz, 3-Mepz, 4-Hdmpz, 4-Medmpz) complexes. Estimated standard deviation in last digit in parenthesesTemperature -JCompound range (K) Model (cm-1) g or g11’ F[Co(4-Hpz)2] 19-54 Ising 24.0(2) 8.86(5) 0.0097Heisenberg 5.9(5) 2.5(1) 0.033[Co(3-Mepz)2] 20-45 Ising 19.8(1) 8.74(3) 0.0033Heisenberg 4.3(1) 2.36(3) 0.0097[Co(4-Hdmpz)2] 10-20 Ising 8.6(2) 6.09(6) 0.011Heisenberg 1.81(3) 1.62(1) 0.0068[Co(4-Medmpz)2J 12-20 Ising 10.4(3) 6.12(9) 0.0089Heisenberg 2.15(4) 1.62(2) 0.0072The calculated curves corresponding to the best fit parameters in Table 4.10 are shown inFigures 4.22(a) through (f) (these curves span only the susceptibility maxima). Only in thecase of [Co(4-Hpz)2Jdoes the Ising model provide a substantially better fit than theHeisenberg model. In fact, the Heisenberg model provides a marginally better fit to the datafor the 4-Hdmpz and 4-Medmpz derivatives; however, the resulting g values areunrealistically low. Some support for application of the Ising parallel susceptibility model tothese compounds comes from the results of EPR studies, conducted at 4.2 K, onmonometallic analogues. Compounds with the general formula Co[R2B(pz*)] (R = H orphenyl and pz = 4-Hpz, 4-Hdmpz, or 4-Medmpz) were found to exhibitg11>’gj (157). In thecases of [Co(4-Hpz)2jand [Co(3-Mepz)2],wide temperature ranges were modelled byallowing a contribution due to in equation [4. 17c}. In the case of [Co(4-Hpz)2j,the fullrandom orientation powder susceptibility data set was modelled allowing J, g11, and g1’ tovary with the resulting best fit values of -39.5(2) cm, 7.56(2), and 3.59(1), respectively (F =0.0080). The perpendicular orientation data were modelled over the range 2-300 K with196inclusion of a Curie law paramagnetic impurity term. The derived values of J, g11’, g1’, and%P were -43(1) cm’, 5.9(1), 3.61(9), and 0.86(6)%, respectively (F = 0.024). With [Co(3-Mepz)2],data over the range 2-300 K were modelled in a similar fashion. The derived valuesoff, g11’, g1’, and %P were -38.3(5) cmt, 6.19(7), 4.24(5), and 0.82(3)%, respectively (F =0.010). When such modelling efforts were made with the data for the [Co(4-Xdmpz)2]species, g11 invariably tended to zero and for the halogenated species, J became indeterminate.Characterization of the Co(ll) (substituted)pyrazolates was also attempted using theanisotropic X-Y model expressed in open-form in equation [4.18] and for which an exactexpression is not currently available. Although Duxbury et a!. (158) developed numericalapproximations to [4.18], they did not provide explicit expressions for those approximations.In order to apply [4.18] to the experimental data, a crude numerical expression was derived inthe form of a forty-one term Simpson’s rule approximation. The expression is large and is notshown here. Comparison of the behaviour of this approximate function with the results ofKatsura (156) indicates that the approximation is reasonably reliable down to values of kT/J0.35. Application of the model to limited temperature ranges of data encompassing theobserved susceptibility maxima generally resulted in best fits of poor to fair quality; the datafor [Co(4-Hpz)2]and [Co(3-Mepz)2]yielding the highest quality fits. An importantshortcoming in the application of the X-Y model to these data is that no expression for inlinear chain systems is available, but within the framework of an axially distorted, S = 3/2chromophore, g1’ = g and gj’ = 2g (129). Thus, neglect of x1 as was done here, incursconsiderable error in the fitting procedure.The preceding discussion makes clear the fact that no single model currently availablefor quantification of exchange interactions in the Co(II) (substituted)pyrazolates adequatelydescribes their observed magnetic properties over wide temperature ranges. In spite of thisdifficulty, the modelling exercises facilitate the qualitative interpretation of the magneticproperties of these compounds. The high temperature susceptibility data for all of the197compounds can be adequately fit with the antiferromagnetically coupled S = 3/2 linear chainmodel. This procedure yields true g values in the range expected for (pseudo)tetrahedrallycoordinated Co(ll) ions. At temperatures below those of the maxima in susceptibility, theobserved susceptibilities in these compounds decrease more rapidly than can be accountedfor by the Heisenberg exchange model. This fact, along with the results of the oriented pelletstudies reveals the presence of anisotropy effects in the compounds. Application of the S’ =1/2, Ising parallel susceptibility model to the low temperature data for these compoundsyields excellent fits and reasonable g11’ values. Moreover, application of the X-Y model,albeit in incomplete form, showed that the model does not account for the magneticproperties of these compounds at temperatures below those corresponding to the maxima insusceptibility. The derived J values from the Ising and Heisenberg models can be used toqualitatively rank the magnitude of exchange coupling in these complexes as follows:4-Hpz> 3-Mepz>> 4-Medmpz 4-Cldmpz 4-Brdmpz> 4-HdmpzThis loose division of the compounds into two exchange strength groups also correlates withtheir ligand field properties and proposed chromophore distortions. Whether thesecorrelations are genuine or simply coincidental is not known, however, a weaker ligand fieldis consistent with the type of distortion proposed in the 4-X-3,5-dimethyl species relative tothe other two complexes. The pyrazolate ligands are most certainly the medium of magneticexchange in these complexes and one might reasonably expect that a weaker interactionbetween the ligands and Co(ll) centres would manifest itself in smaller magnitudes ofexchange. As a final comment on the nature of the exchange interaction in the Co(ll) species,although the modelling results indicate that low temperature, l-D, Ising-like exchange maybe manifest in these compounds, the true exchange behaviour is probably of the Heisenbergtype and the anomalously sharp susceptibility maxima result from single-ion, ZFS effects.Further interpretation of the order of the magnitudes of exchange coupling or its nature is not198warranted without detailed knowledge of the structures of these compounds.4.3 SUMMARY AND CONCLUSIONSA number of oligometallic and polymeric, divalent, zinc and cobalt(substituted)pyrazolates have been prepared. The dimeric species [Zn(4-Hdmpz)2( -HdmpzH)]2can be isolated in pure form from a variant of the MLPM reaction. Variation insynthetic conditions and thermolysis studies with this compound have shown that it can bepolymerized to [Zn(4-Hdmpz)2jand furthermore, that the polymerization process can bearrested at the incipient polymer stage.Polymeric Co(II) species, analogous to [Zn(4-Hdmpz)21,have also been preparedwith the general formula [Co(pz*)2](pz* = 4-Hpz, 3-Mepz, 4-Hdmpz, 4-Medmpz, 4-Cldmpz, 4-Brdmpz). No macroscopic single crystals of these compounds were prepared andcharacterization of the Co(II) species by indirect methods lead to the conclusion that theyconsist of linear chains of (pseudo)tetrahedrally coordinated Co(H) ions doubly bridged bypyrazolate ligands. Moreover, the six compounds differentiate into two groups. The firstgroup consists of [Co(4-Hpz)2]and [Co(3-Mepz),] with fairly regular tetrahedral CoN4chromophores. The second group consists of the [Co(4-Xdmpz)2Jspecies (X = H, Me, Cl,Br) in which steric effects distort the CoN4 chromophores, probably byS4-elongation, awayfrom regular tetrahedral symmetry. The magnetic properties of the polymers differentiate inthis same manner. The former group exhibits fairly strong antiferromagnetic exchangecoupling, while magnetic exchange in the latter group appears to be weaker. In all six.compounds, oriented pellet studies, the general appearance of the temperature dependence ofsusceptibility, and tentative theoretical modelling efforts suggest that the compoundsexperience anisotropy effects which cause their low temperature magnetic behaviours toappear Ising-like.199A group of three oligometallic Co(II) compounds were prepared unexpectedly duringefforts to synthesize single crystals of the polymeric Coal) compounds. [Co(4-Hdmpz)21(4-HdmpzH)]2Co, [Co(4-Hdmpz)2(4-HdmpzH)] and [Co4(4-Hdmpz)60] have beensynthesized in macroscopic single crystal form and characterized by single crystal X-raydiffraction, amongst other techniques. Magnetic studies conducted on [Co(4-Hdmpz)21(4-HdmpzH)]2Co show it to exhibit nearest-neighbour, antiferromagnetic Heisenberg exchange,however there are additional magnetic phenomena present in this material which have notbeen clearly delineated. Though [Co(4-Hdmpz)2(4-HdmpzH)] has been characterizedcrystallographically, the bulk material from which it was obtained has the empirical formulaCo(4-Hdmpz)2.O.34 (4-Hcl pzH . This material, as with the analogous Zn(II) species, seemsto show incipient polymer formation and has provided a model for the structure of thecorresponding polymeric complex.200CHAPTER 5MISCELLANEOUS COMPOUNDS5.1 INTRODUCTIONA number of compounds were prepared which did not fall into the categoriesconsidered in the previous chapters. The success of the MLPM reaction in yieldingpyrazolate compounds of cobalt, copper, and zinc inspired attempts to prepare other metalpyrazolate compounds. These investigations lead to the preparation of poly(nickel(II)pyrazolate) and two octametallic molybdenum clusters. Also considered here is a Cu(II) 4-iodopyrazolate compound which contains neutral 4-iodopyrazole.5.2 RESULTS AND DISCUSSION5.2.1 NTCKEL(II) PYRAZOLATE5.2.1.1 SYNTHESIS AND PHYSICAL PROPERTIESAs with the previously discussed Cu(ll) and Co(II) (substituted)pyrazolates, the initialpreparation of [Ni(4-Hpz)2]was attempted using the method of Vos and Groeneveld (37).Elemental analyses indicated that the product resulting from the attempt was, at best, impure[Ni(4-Hpz)2].The compound was successfully prepared via the MLPM reaction outlinedbelow:Ni + xs 4-HpzH > [Ni(4-Hpz)2J+H201 [5.1]48 h201The detailed description of this synthesis is given in Chapter 7, Section 7.2.6. The rate ofreaction [5.11 with massive nickel metal was impractically slow, consequently finelypowdered metal was used as the source of nickel. Nickel metal is ferromagnetic andunreacted nickel metal was separated from the desired product in the same manner used forpurification of the MLPM derived Co(II) compounds. The efficacy of the magnetic dredgingtechnique for purifying [Ni(4-Hpz)2]receives comment below.[Ni(4-Hpz),J was obtained as a yellowish-orange, finely divided powder. Thecompound is insoluble in water and all common organic solvents and can only be dissolvedwith decomposition in mineral acids. A DSC study of the compound showed that itdecomposes without melting at 380 °C. An SEM image of the compound is shown in Figure5.1 revealing an elongated particle morphology.FIG. 5.1. SEM image of [Ni(4HpZ)2JrThe white line represents alength of 2 jim.2025.2.1.2 CHARACTERIZATION AND PROPOSED STRUCTUREA powder X-ray diffractogram was obtained for [Ni(4-Hpz)2]and is shown in Figure5.2. The d-spacings and relative intensities of the peaks in this pattern are listed in Appendix0ib 2b 3b 4b 5b sb 7b eb 90FiG. 5.2. Powder X-ray diffractogram of [Ni(4-Hpz)2]11, Table 11-6. The diffractogram demonstrates that [Ni(4-Hpz)2]is microcrystalline, butcomparison with the diffraction patterns of [Cu(4-Hpz)2j(Figure 2.5(a)) and [Co(4-Hpz)2](Figure 4.14) indicates that [Ni(4-Hpz)2Jis isomorphous with neither the Cu(II) nor theCo(11) species.The IR spectrum for [Ni(4-Hpz)2]above 600 cm4 is consistent with the presence ofonly anionic pyrazolyl moieties in the compound. Below 600 cm two strong absorptions arepresent at 459 and 414 cm. Since no bands are observed in the 250-600 cm1 region ofspectrum of Na[4-Hpz], it is proposed that the bands in this region in [Ni(4-Hpz)2jare due tometal-ligand stretching and bending modes. The positions and relative intensities of203unassigned bands are listed in Table IV-9 in Appendix IV.The solid state, Nujol mull electronic spectrum of [Ni(4-Hpz)2]is shown in Figure5.3. No bands are visible in the near-infrared region of the spectrum. In the visible region twoI I I I I I I I I IrizC.)0I I I500 1000 1500 2000Wavelength (nm)FIG. 5.3. Electronic spectrum of [Ni(4-Hpz)2J.bands, due to d-d transitions, are discernible: a very weak shoulder at 16,950 cm (590 nm)and a strong absorption at 23,200 cm1 (431 nm). In the ultraviolet region of the spectrum twovery strong bands are observed at 33,300 cm1 (300 nm) and 45,660 cm4 (219 nm). Theultraviolet bands are ascribed to charge transfer and/or internal ligand transitions. Theelectronic spectrum of this compound in the visible region is characteristic of a square-planarstereochemistry for the Ni(ll) centres (159).In further support of this assignment for the NiN4 chromophore, magneticmeasurements made over the temperature range 78-300 K, using a Gouy balance,demonstrated the complex to possess only residual paramagnetism (on the order of 100 x 10.6cm3mol4),but it was temperature independent. This residual paramagnetism may be due to204second order Zeeman effects (TIP). Square-planar Ni(II) complexes are generallydiamagnetic. That an effectively diamagnetic complex was isolated from a reaction mixturecontaining nickel metal powder using the technique referred to in Section, istestament to the effectiveness of the product slurry/magnetic dredging technique.Electronic spectroscopy and magnetic susceptibility measurements indicate that [Ni(4-Hpz)2Jpossesses a square-planar NiN4 chromophore. The SEM image shows that the samplemicrocrystallites possess an elongated morphology consistent with a polymeric structure.Therefore, it is proposed that [Ni(4-Hpz)2]consists of chains of square-planar coordinatedNi(II) ions doubly bridged by pyrazolate ligands. These chains likely possess the steppedmotif found in [(ON)Ni(4-Hdmpz)2]iand[(C3H5)Ni(4-Hdmpz)2]i(97,98) and proposedfor the [Cu(4-Xdmpz)2]species (Chapter 2).5.2.2 A COPPER(II) 4-IODOPYRAZOLYL COMPOUND5.2.2.1 SYNTHESIS AND PHYSICAL PROPERTIESDuring an attempt to prepare [Cu(4-Ipz)2]from [Cu(4-Ipz)] (29) and molten 4-iodopyrazole an unexpected product was obtained. This reaction is presented below:02, 110 °C[Cu(4-Ipz)j + xs 4-IpzH11 h> Cu(4-Ipz)2.(4-IpzH + +H20t [5.2]Cu(4-Ipz)2.(4-IpzH was isolated from reaction [5.2] as a finely divided, olive greenpowder. A detailed description of this synthesis is provided in Chapter 7, Section Thematerial is insoluble in water and a range of organic solvents. Elemental analysis suggestedthe nominal formulation for the compound and this was supported by the 98% yield ofmaterial obtained from the reaction. The JR spectrum for the compound exhibits the bands205expected for the 4-iodopyrazolyl moiety and, in addition, a very sharp, strong band at 3,405cm1 which is consistent with the presence of 4-iodopyrazole in the material. The positionand sharpness of this band indicates that the N—H group of the 4-iodopyrazole is involved inlittle, or no hydrogen bonding. Unassigned IR bands and their relative intensities are listed inTable IV-9 in Appendix IV. Evidence that the neutral 4-iodopyrazole in this compound isstrongly bound, possibly through coordination to the copper centres, comes from theobservation that heating of a sample of this compound in vaciw at 125 °C for 20 h resulted ina mere 2% reduction in the mass of the sample (if all of the neutral pyrazole had beenvolatilized during this time, then an 18% mass loss would have been observed). MAGNETIC PROPERTIESThe powder magnetic susceptibility of Cu(4-Ipz)2.(4-IpzH), as a function oftemperature, was measured in the range 5-300 K using a SQUID magnetometer.Temperature, magnetic susceptibility, and magnetic moment values are listed in AppendixIII, Table 111-7. A plot of these data is shown in Figure 5.4. The data exhibit a maximum inversus temperature at 150 K indicative of the presence of antiferromagnetic exchangecoupling. As the lowest temperature is approached the susceptibility begins to rise rapidly.This increase is likely due to the presence of a small amount of paramagnetic impurity in thesample. The susceptibility data were analyzed using the antiferromagnetically coupled,Heisenberg linear chain model in same manner as were the data for the [Cu(4-Xpz)2]complexes in Chapter 2, Section Analysis of the full temperature range employing aCurie law paramagnetic impurity term yielded values for J, g, and %P of - 82.3(8) cm1,2.33(1), and 1.16(2)%, respectively (F = 0.0 14). The calculated curve for this fit is shown asa dotted line in Figure 5.4. If only the data in the 100-300 K range are modelled then J, g, and%P are -84.4(4) cm1, 2.345(3), and 1.3(1)%, respectively (F = 0.0019). This fit is shown asthe solid curve in Figure 5.4. Finally, if data are fit over the full temperature range and the2062000C.,C)1750C1500C)C)12501 0000 50 100 150 200Temperature (K)FIG. 5.4. Magnetic susceptibility plot versus temperature for Cu(4-Ipz)2.(4-IpzH).The solid represents the best fit of the Heisenberg, 1-D, antiferromagnetic couplingmodel to the data in the 100-300 K range. The dotted and dashed curves are the bestfits of the same model to the full data range including Curie and Curie-Weiss lawparamagnetic impurity terms, respectively.paramagnetic impurity is accounted for using a Curie-Weiss law term then J, g, %P, and eare -84.2(6) cm4, 2.341(7), 1.41(3)%, and -1.5(2) K, respectively (F = 0.0081). The dashedcurve in Figure 5.4 represents this last fit. The data in the region of 70 K deviate from thesusceptibilities predicted by the model. This behaviour is reminiscent of the susceptibilitiesof [Cu(4-Clpz)2]and [Cu(4-Brpz)2jand may, in the case of the iodinated species, also bedue to some structural transformation. On the whole, however, the magnetic data for Cu(4-Ipz)2.+(4-IpzH) can be well accounted for using the Heisenberg linear chain model. Thissuggests that Cu(4-Ipz)2.4(4-IpzH Consists of linear chains of Cu(I1) ions and consideringthe large magnitude of J, the Cu(II) centres are presumably bridged by 4-Ipz ligands. Why250 300207this compound should incorporate an extra half mole of 4-iodopyrazole per Cu(II) ion whilethe [Cu(4-Xpz)2Jcompounds do not is not understood. Perhaps the large iodine atom of thepyrazolyl ring creates large enough gaps between chains to permit inclusion of the neutralmolecules in the compound’s lattice.5.2.3 OCTAMETALLTC MOLYBDENUM OXO-PYRAZOLATE CLUSTERS5.2.3.1 SYNTHESES, PHYSICAL AND THERMAL PROPERTIESDetails of the procedures employed in the preparation of compounds discussed in thissection are given in Chapter 7, Section 7.2.8. The motivation for examination of molybdenumpyrazolates was not the potential elucidation of their magnetic properties; it was simply adesire to see if molybdenum and pyrazole would react. Consequently, finely dividedmolybdenum powder was combined with molten pyrazole and heated in air as outlinedbelow:Air, 90°CMo + xs 4-HpzH >[Mo8(4-Hpz)618-HpzHj.2(4-H zH) [5.3]18 h-i-H20At cessation of the reaction, the pyrazole had evaporated from the heated zone of the reactionvessel exposing the molybdenum powder, most of which had not reacted. Mixed with themolybdenum powder were numerous very small, well formed, red single crystals of [Mo8(4-Hpz)6018(4-Hp H)].2(4-HpzH (this formula was determined by single crystal X-raydiffraction). There were too few crystals to conduct any sort of characterization studies otherthan the X-ray diffraction study which is discussed in detail below.The initial X-ray diffraction structure determination of [Mo8(4-Hpz)60184-HpzH)6].2(4-HpzH) did not permit an unambiguous assignment of the oxidation states of all208the molybdenum ions in the cluster because it was not certain whether the disordered solvatepyrazolyl moieties in the lattice were neutral molecules, pyrazolate anions, or pyrazoliumcations. During this period of uncertainty, efforts to rationally prepare bulk samples of thiscluster were undertaken. Working under the assumption that all of the molybdenum centreswere hexavalent, an attempt was made to synthesize the cluster using MoO3 as a source ofmolybdenum. The reaction of MoO3 with pyrazole yielded another octarnetallic species andthe reaction is outlined below:darknessair, 100°CMoO3+ xs 4-HpzH >[Mo8(4-Hpz)621-HpzH].3(4-HpzH).H2O4, [5.4]8 days+H201The product was obtained as a pale yellow powder (if the reaction mixture is left quiescentthen large single crystals are formed). The compound is photosensitive, turning brown onprolonged exposure to light and this process cannot be reversed by subsequent storage of thecompound in the dark. This compound (or some variant) exhibits slight solubility inconcentrated pyrazole/tetrahydrofuran (THF) solution because the THF filtrate from theisolation of the product of reaction [5.4] yielded more [Mo8(4-Hpz)6021-HpzH].3(4-HpzH).-H2Oupon concentration and further heating. The colour change which occurs in thecompound upon photolysis is not accompanied by any marked change in the material’scrystal structure because powder X-ray diffractograms recorded before and after severalhours of irradiation of a sample (using a low pressure Hg arc lamp) were identical.Furthermore, the crystal used in a single crystal diffraction study of this compound (discussedbelow) was exposed to ambient light during the several days of data collection. The crystalgradually darkened over this time, but the diffraction data indicated that no degradation of thecrystal took place.A DSC study was conducted on a sample of [Mo8(4-Hpz)6021-HpzH1.3(4-209HpzH).4H20from room temperature up to 350 °C (Figure 5.5). The thermogram shows50 100 150 200 250 300 350Temperature (°C)FIG. 5.5. DSC thermogram of [Mo8(4-Hpz)6021(4-HpzH)6].3(4-HpzH).+H20.several endothermic events in the room temperature to 300 °C range followed by anexothermic event at 330 °C. The sample exhibited a mass loss of 45% over this temperaturerange (the lattice water and all pyrazolyl moieties in the compound comprise 48% of itsmass). The complicated DSC thermogram no doubt results from the presence of both solvatemolecules and coordinated neutral pyrazole ligands, some or all of which may be volatilizedupon thermolysis.[Mo8(4-Hpz)6O21-HpzH].3(4- zH).H is air-insensitive in the solid state. Thecompound’s resistance to aerial oxidation is not surprising because the molybdenum ions arein the hexavalent state. In light of this, a preliminary investigation was conducted on thebehaviour of this compound in a reducing environment. [Mo8(4-Hpz)6021-HpzH].3(4-HpzH).4H20was combined with dry, degassed methanol under a dinitrogen atmosphere and210the mixture was stirred at room temperature for 24 h. No change was apparent in the mixtureafter this time; the compound neither dissolved nor changed colour. As well, subsequentexposure to dry dioxygen caused no change in the mixture. However, when the mixture wasexposed to water vapour (with or without 02 present), the methanol became bluish-greenwithin a couple of hours. In a related experiment, [Mo8(4-Hpz)6021-HpzH].3(4-HpzH).4H20was combined with ethanol and the mixture refluxed for 16 h, in air. Thecomplex completely dissolved during this time yielding a dark blue solution. The ethanol wasflash evaporated from the solution yielding a blue powder.[Mo8(4-Hpz)6021-HpzH].3(4-HpzH).+H20reacts with alcohols and water seems to be necessary for this reaction to takeplace.A brief effort was made to see if W03 would yield an analogous cluster compound;however, no reaction took place between the W03 and pyrazole.Reaction [5.4] is not suitable for preparation of [Mo8(4-Hpz)6018-HpzH.2(4-HpzH), which is a mixed-valence Mo(V)fMo(VI) species. This suggested that a startingmaterial with molybdenum present in an oxidation state lower than six might afford [Mo8(4-Hpz)6018(4-Hp H)].2(4-Hpz ). This hypothesis was tested by reacting “molybdenum blue”,Mo2O5H, (160) with molten pyrazole, in air, at 108 °C for 20 h. The solid product mixturefrom this reaction mixture was extracted with dry THF yielding an insoluble light orange-brown solid and a dark orange solution (which, of course, contained excess pyrazole). Thesolution phase was allowed to stand for several hours and during this time a light yellow solidprecipitated from solution leaving a pale orange supernatant. The supernatant wasconcentrated to a viscous liquid and heated at 100 °C for 72 h. This yielded an approximatelyequal mixture of large, well formed red crystals of[Mo8(4-Hpz)618-HpzH.2(4-H zH)and pale yellow crystals (most likely [ o(4-Hpz)21-HpzHJ.3(4- zH).H0)Although this synthetic procedure did yield the mixed-valence cluster, the compound was notobtained in pure form. A second, higher quality single crystal diffraction study (described211below) showed that the mixed-valence cluster contains six Mo(V) ions and two Mo(VI) ions,thus possessing a mean molybdenum oxidation state of 5.25. The starting material,Mo2O5H, possesses a mean oxidation state of 5.5. Considering that no other reducing agentwas present in the reaction system, it is not surprising that pure [Mo8(4-Hpz)60184-HpzH)6].2(4-HpzH) was not obtained. It may be possible to prepare the mixed-valencecluster in bulk form using a molybdenum starting material with a lower mean oxidation state,for example, MoO2. Unfortunately, time constraints precluded further study of these clustersystems. X-RAY DIFFRACTION STUDIESTwo single crystal diffraction studies were conducted on [Mo8(4-Hpz)60184-HpzH)6j.2(4-HpzH . The first study was performed with a rather small crystal, so a secondstudy was undertaken with a much larger crystal (prepared from the “molybdenum blue”reaction described above) yielding a superior diffraction data set. The results of this secondstudy are described in this section. Crystallographic data, atomic positional parameters, andbond lengths and angles are tabulated in Tables 1-52 through 1-56 in Appendix I. Themolecular structure of [Mo8(4-Hpz)6018-HpzH],as its formula would suggest, is rathercomplex. Stereoviews of the molecule in two orientations are shown in Figure 5.6. Thecluster crystallizes in the monoclinic space group C21c. The structure of this molecule ismore easily understood (and compared with the molybdenum cluster to be discussed below)if it is considered in terms of skeletal fragments. The metal centres in this molecule form theoctametallic framework which may be loosely described as two basally joined trigonalpyramids and thus the framework possesses an approximate three-fold rotational symmetryaxis (the true, crystallographic, symmetry allows the cluster only a two-fold rotational axis).This then divides the molybdenum centres into two types: two apical molybdenums and sixbasal molybdenums. Each apical molybdenum centre is a formally hexavalent, d° ion and is212FIG. 5.6. Two stereoscopic ORTEP views of[Mo8(4-Hpz)6018-HpzH;32% probability thermal ellipsoids are shown for the non-hydrogen atoms.connected to three equatorial molybdenum centres by a bridging pyrazolate ligand and abridging oxo ligand. Hence, the apical centres have MoN3O chromophores. The basalmolybdenums are formally pentavalent, d’ ions. Each basal Mo(V) ion is coordinated by aterminal oxo anion, a terminal neutral pyrazole ligand, and two bridging oxo anions (which213connect a given basal Mo(V) to an adjacent Mo(V)). The average terminal Mo—O bondlength is 1.685 A, consistent with multiple bond character. Furthermore, the average distanceseparating adjacent basal Mo(V) centres is only 2.574 A. This distance is intermediatebetween the Mo—Mo distance in molybdenum metal (2.78 A) and twice the ionic radius ofMo(V) (1.18 A) ions (161). Furthermore, Mo—O—Mo angles between bridged Mo(V) centresare much less than 900, ranging from 82.78(6)-83.46(7)° (average Mo—O—Mo angle = 83.0°).These observations are suggestive of Mo—Mo bond formation between pairs of basal MoW)ions. Thus the basal molybdenum ions are seven-coordinate with MoN2O4(Mo)chromophores.The terminal neutral pyrazole ligands are intramolecularly hydrogen bonded tobridging oxo anions associated with adjacent basal Mo(V) pairs. There are two solvatepyrazole molecules present in the crystal lattice per molybdenum cluster and these solvatemolecules occur as hydrogen bonded pairs with the N—H proton disordered between the twonitrogen atoms on a given ring. The solvate pyrazole molecules and the crystal packing of themolybdenum clusters are shown in Figure 5.7.Crystallographic data, atomic positional parameters, and bond lengths and angles for[Mo8(4-Hpz)6O21-HpzH].3(4- zH).H are tabulated in Tables 1-52 and 1-57 through1-60 in Appendix I. A stereoview of[Mo8(4-Hpz)6021-HpzH1is shown in Figure 5.8 andit is apparent that the cluster is structurally similar to [Mo8(4-Hpz)6018-HpzH].Thecompound crystallizes in the trigonal space group R3(h) and, as with the mixed-valencecompound, this cluster possesses an octametallic framework which may be described as twobasally joined trigonal pyramids. In this case, however, the cluster possesses true three-foldrotational symmetry. Furthermore, in [Mo8(4-Hpz)6021-HpzH],there are three additionaloxygen atoms present. Consequently, all of the molybdenums in the cluster are formallyhexavalent. The apical sites consist of MoN3O chromophores bridged to the basalmolybdenum ions by a single pyrazolate ligand and a single oxo ligand, as in the mixedvalence cluster. The principle difference between the two complexes arises with the basal214FIG. 5.7. Stereoscopic unit cell packing diagram for[Mog(4-Hpz)6018(4-HpzH.2(4-HpzH showing the positions of thedisordered solvate pyrazole molecules.molybdenum centres. In the single valence complex, the basal molybdenum ions arehexavalent species linked to an adjacent basal molybdenum ion by a single, bridging oxoligand; there are no Mo—Mo bonds present. Moreover, each basal molybdenum ion iscoordinated to two terninal oxo ligands. The coordination spheres around the basalmolybdenum ions are completed by neutral pyrazole ligands yielding MoN2O4chromophores.The terminal neutral pyrazole ligands in the cluster are intramolecularly hydrogenbonded to the terminal oxo ligands of the basal Mo(VI) ions. The pyrazoles exhibit two typesof hydrogen bonding: three of the ligands form only a single hydrogen bond to an oxo speciesand the other three pyrazoles form bifurcated hydrogen bonds to two different oxo species.There are also three solvate pyrazole molecules and one-half of a water molecule permolybdenum cluster in the crystal lattice of this material. The solvate pyrazole molecules are215czFIG. 5.8. Stereoscopic ORTEP view of[Mo8(4-Hpz)6021-HpzH];33% probability thermal ellipsoids are shown for the non-hydrogen atoms.FIG. 5.9. Stereoscopic ORTEP view of the disordered pyrazole and watermolecules in[M8(4-Hpz)621-HpzHj.3(4- zH).HO.216two-fold disordered and surround the water molecule as shown in Figure 5.9.The structural similarities of the single-valence and mixed-valence, octametallicmolybdenum clusters are emphasized in Figure 5.10.Mo/\0 O—\/MoMo(V/VI)/Mo—OMo Mo\E— 0/MoMo(VI) 5.10. Framework for the mixed-valence and single-valenceoctametallic molybdenum clusters.INFRARED SPECTROSCOPYJR spectra were obtained for [Mog(4-Hpz)6021(4-HpzH1.3(4-Hp ).H and theblue product obtained from the reaction of this compound with ethanol. Unassigned bandfrequencies and their relative intensities are listed in Table IV-9 in Appendix IV. The bandsin these spectra are consistent with the presence of pyrazolyl moieties. In both spectra there isa broad, structured band in the 700-900 cm1 region of the spectrum due to both pyrazolyl andMo—0 vibrations (162).2175.2.4 FURTHER COMMENT ON THE OC1’AMETALLIC MOLYBDENUMCLUSTERSThe two molybdenum clusters discussed here are ancillary to the principle thrust ofthis study; however, they are unique compounds and deserve to be placed in an appropriatecontext. The octametallic molybdenum clusters might best be related to the well knownpolyoxometalates. Polyoxometalates have a venerable history, Berzelius having reported thefirst such compound of molybdenum in 1826 (163). The transition metal elements whichreadily form polyoxometalates are tungsten, molybdenum and, to a lesser extent, vanadium,niobium, and tantalum. Polyoxometalates are composed of these elements in high oxidationstates at the centres of MO units which are corner- or edge-linked to form symmetrical,polyhedral clusters. Related species, the heteropolyoxometalates, contain, in addition to aparticular metal ion, another metal or main group element. The archetypal example of thesecompounds is[H3PW12040.6H2,the first polyoxometalate to have its structure determinedcrystallographically (164). At present, there are literally hundreds of such compounds knownand in many cases their structures have been determined crystallographically. A recentreview by Pope and Muller outlines the development and applications of polyoxometalatechemistry (165). Polyoxometalates have important applications as industrial oxidation andacid catalysts (166). Furthermore, photocatalytic oxidation of organic substrates by thesecompounds is currently under active investigation (167,168). The great facility ofpolyoxometalate complexes in catalytic processes stems from their abilities to act as “oxygenrelays” and/or electron relays (167) manifested by the ready reduction of these compounds tomixed-valence species with characteristic intense blue or brown colours (165). Preliminaryinvestigation of the properties of the hexavalent molybdenum cluster prepared in this studyshows that it may behave similarly to the classical polyoxomolybdates. [Mo8(4-Hpz)60,14-HpzH)6].3(4-HpzH).4H,O is photosensitive like the polyoxometalates. Moreover, thecompound reacts with oxidizable substrates (methanol or ethanol) to form intensely blue218coloured species, suggesting that the molybdenum cluster may form mixed-valence speciessimilar to those of the polyoxometalates (that such reduction is possible is indisputable;[Mo(4-Hpz)6O18(4-HpzH].2(4-HpzH is formally a six electron reduction product of thehexavalent cluster).The feature which makes the molybdenum clusters prepared in this study unique isthe pyrawlyl moieties incorporated in their structures. A sub-branch of polyoxometalatechemistry which has developed in the last decade involves the synthesis of polyoxometalatesincorporating organic functional groups. One strategy to prepare such compounds hasinvolved the binding of organometallic fragments such as (ri5-RCH4)T (where R is afunctionalized organic chain) with lacunary polyoxometalate fragments yielding such speciesas(i5-RCH4)TiPW11O39 and (ii5-RCH4)TiP2W17O617(169,170). Another strategyinvolves substitution of terminal oxo ligands on complete polyoxometalates with an organicligand. A recent example of this is the preparation of the imido species, [Mo5018(MoNtol)]2(Ntol = p-tolylimide) from [Mo6O19]2(171). The present octametallic molybdenum clusterspossess readily derivatized pyrazolate (and for that matter, pyrazole) rings which would beexpected to be hydrolytically stable (a feature lacking in the iinido complex described above).One might expect such organic functionalization to afford some control over the solubiitycharacteristics and reactivities of these clusters permitting a wide ranging investigation oftheir potential chemistry.5.3 SUMMARY AN]) CONCLUSIONS[Ni(4-Hpz)2]has been prepared via the MLPM reaction. Spectroscopic evidencesuggests that the compound possesses square-planar NiN4 chromophores. The compound isdiamagnetic which also supports a square-planar stereochemistry for the Ni(ll) ions.Accordingly, it is proposed that compound’s extended structure consists of stepped linear219chains.A copper(I1) compound with the empirical formula Cu(4-Ipz)2.(4-IpzH wasobtained from the reaction of molten 4-iodopyrazole with [Cu(4-Ipz)]. JR spectroscopyindicates that the N—H hydrogen atom is engaged in little or no hydrogen bonding. Themagnetic susceptibility data for the compound are modelled reasonably well by theHeisenberg, linear chain antiferromagnetic coupling theory which yields a J of -84 cm1.Based on the magnetic behaviour, it is tentatively suggested that the compound has a linearchain structure.Two unusual octametallic molybdenum oxo-pyrazolate clusters were prepared andcharacterized by single crystal X-ray diffraction. One of the clusters, [Mo8(4-Hpz)60184-HpzH)6], is a mixed-valence species containing molybdenum in the penta-and hexavalentstates and the other cluster, [Mo8(4-Hpz)6021-HpzH],contains molybdenum in thehexavalent state oniy. Preliminary studies of the properties of this latter compound show it topossess some similarity to the well known polyoxornetalates.220CHAPTER 6GE’ERAL SUMMARY AND SUGGESTIONS FOR FUTURE WORK6.1 GENERAL SUMMARYIn this study, the synthesis and physical characterization, with particular emphasis onthe magnetic properties, of low-dimensional, transition metal complexes incorporatingpyrazolate and pyrazolate derived ligands as bridging species has been explored.A series of eleven binary copper(ll) (substituted)pyrazolates of the general formula,[Cu(pz*)2](pz* = 4-Hpz, 3-Mepz, 4-Mepz, 4-Clpz (two forms), 4-Brpz, 4-Hdmpz, 4-Medmpz, 4-Cldmpz, 4-Brdmpz, indz), was prepared. Three of these complexes were studiedusing single crystal X-ray diffraction (pz* = 4-Hpz, 4-Mepz, 4-Clpz (green)) which revealedthem to contain rhombically distorted tetrahedral CuN4 chromophores, linked by doublebridges of pyrazolate ligands, thus forming linear chains. Considerable indirect evidenceindicated that the remaining eight complexes also crystallize with a linear chain motif thoughit is thought that six of these complexes have square-planar CuN4 chromophores (pz* = 4..Clpz (brown), indz, 4-Hdmpz, 4-Medmpz, 4-Cldmpz, 4-Brdmpz). Temperature dependencestudies of the powder magnetic susceptibilities of these compounds demonstrated that all ofthe compounds behave as Heisenberg linear chain systems with strong antiferromagneticcoupling (J values ranging from -58 cm1 to -105 cm1) and it is thought that the it-orbitals ofthe pyrazolate ligands provide an important medium for propagation of this strong exchange.Three of the compounds (pz* = 4-Clpz (green), 4-Brpz, 4-Cldmpz) exhibit anomalies in theirversus temperature plots. Low temperature structural studies on the 4-Xpz speciesindicates that these compounds undergo significant structural changes which may beresponsible for the observed magnetic anomalies. The 4-Cldmpz derivative shows an221unusually rapid decrease in magnetic susceptibility at low temperatures for anantiferromagnetically coupled linear chain compound which is suggestive of a structuraltransition or a magnetic transition to a higher dimensionality. In addition to the binarycopper(fl) pyrazolates a copper(II) 4-iodopyrazolate containing neutral 4-iodopyrazole wasalso prepared. The magnetic properties of this species indicate that it also is a stronglyanti ferromagnetically coupled Heisenberg linear chain system.A series of six binary cobalt(II) (substituted)pyrazolate compounds has been prepared(j)z* = 4-Hpz, 3-Mepz, 4-Hdmpz, 4-Medmpz, 4-Cldmpz, 4-Brdmpz). Indirect evidence andcomparison with oligometallic Co(H) and Zn(II) complexes of known structure and thesupposedly polymeric species, [Zn(4-Hdmpz)2J,lead to the proposal that these polymericCo(ll) complexes consist of pseudo-tetrahedral CoN4 chromophores linked together bydouble pyrazolate bridges to form linear chains in analogy with the Cu(II) compounds.Magnetic studies have shown that the Co(II) species are also antiferromagnetically coupled;however, these studies have also shown that the compounds are magnetically anisotropic, aphenomenon often observed in Co(JI) complexes. This anisotropy complicates thequantitative interpretation of the magnetic properties of these compounds, but it appears thatthe [Co(4-Hpz)21and [Co(3-Mepz)2]experience much stronger antiferromagnetic couplingthan the other four derivatives. The more regular tetrahedral chromophore proposed for thetwo former species may be responsible for the stronger exchange observed in thesecompounds.Attempts were made to prepare [Ni(pz*)21species; however, their isolation in pureform was fraught with difficulty. Moreover, the only nickel(ll) complex that was prepared inpure form, [Ni(4-Hpz)2],is diamagnetic.Although the initial motivation for this work was the preparation and characterizationof binary linear chain complexes of transition metal pyrazolates, fourteen oligometallic O-D,or molecular, pyrazolate complexes incorporating, cobalt, copper, zinc, or molybdenum wereprepared. A series of six trimetallic copper(I) complexes with the general formula, [Cu(pz*)]3222(,pz* = 4-Hdmpz, 4-Mednipz, 4-Cldmpz, 4-Brdmpz, 4-Idnipz, indz), were prepared. Two ofthese complexes, the 4-Hdnipz and 4-Medrnpz derivatives, were characterized by singlecrystal X-ray diffraction. All of these complexes, save the 4-Idmpz species, served asexcellent starting materials for the preparation of the corresponding copper(II)(substituted)pyrazolates. The trimetallic, mixed-valence Cu(I)ICu(ll) complex, [Cu(3-CO2dmpz)(4-Medmpz)]u, was characterized by single crystal X-ray diffraction and EPRspectroscopy. Observations suggest that this compound forms via oxidation of [Cu(4-Medmpz)]3 in the presence of 4-MedmpzH. The dimeric species, [Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)2],crystallizes in tridinic and monoclinic modifications; both have beencharacterized by X-ray diffraction. Magnetic studies conducted on this compound revealedthe Cu(II) centres to experience moderate antiferromagnetic coupling (J = -75.4 cm1).Two isomorphous, dimeric species with the formula [M(4-Hdmpz)2(4-HdmpzH)] (M= Co, Zn) have been characterized by single crystal X-ray diffraction. The zinc dimer wasprepared in pure form, however, the cobalt dimer occurred as part of a bulk sample with thenominal composition, Co(4-Hdmpz)2.O.34 (4-Hd pzH). Observations indicate that incipientpolymer formation occurs in both the zinc and cobalt species. The trimetallic complex, [Co(4-Hdmpz)2Cl(4-HdmpzH ]o, and the tetrametallic complex, [Co4(4-Hdmpz)601, have beenprepared and characterized crystallographically. Magnetic studies on [Co(4-Hdmpz),Cl(4-HdmpzH)]2Co demonstrate that the compound experiences intramolecular, nearest neighbourantiferromagnetic coupling and additional effects, perhaps significant next-nearest neighbourantiferromagnetic exchange.Finally, two novel octametallic molybdenum oxo-pyrazolate complexes wereprepared and characterized crystallographically. One of the complexes is a mixed-valence,Mo(V)/Mo(VI) cluster and has the formulation[Mo8(4-Hpz)6018-HpzH].2(4-H z . Themolecular structure of this complex indicates the presence of metal-metal interactionsbetween pairs of adjacent Mo(V) centres. The other complex contains only hexavalent223molybdenum ions and preliminary investigation of the chemical behaviour of this cluster hasrevealed similarities with the well known polyoxometalates.6.2 SUGGESTIONS FOR FUTURE WORKThe compounds which have been prepared and examined in this study present manyavenues for future research. These possibilities will be considered in turn for each of thegroups discussed in Section 6.1.Regarding the copper(II) (substituted)pyrazolates considered in Chapter 2, it ispossible that single crystals large enough for X-ray diffraction studies could be obtained for[Cu(4-Brpz),J with continued preparative efforts. The precise structural parameters derivedfrom such a study would be of great value in furthering magneto-structural correlation forthis class of compounds. A related endeavour would be the preparation of [Cu(4-Fpz)2]inboth powder and single crystal form. This compound would likely exhibit the same structureas the other four 4-Xpz species and a synthetic route for 4-fluoropyrazole has been previouslypublished (172,173). Some initial work has been conducted regarding another fluorinatedpyrazolyl system and preliminary results from that work are recounted next.In further exploration of the effects of different substituents on the properties of the[Cu(pz*)2Jspecies, some initial effort was made to prepare the compound [Cu((CF3)2pz)](where (CF3)2pz = 3,5-bis(trifluommethyl)pyrazolate), analogous to the [Cu(4-Hdmpz)species. 3,5-Bis(trifluoromethyl)pyrazole was prepared as described previously (174) andreacted with copper via the MLPM reaction. Large, dark green crystals were obtained and themass balance of the reaction suggested that the product has a Cu:(CF3)2pz moiety ratio of3:5. The compound is very soluble in a wide range of solvents and when in solution it reactsvery rapidly with water. Its IR spectrum shows bands consistent with the presence of the3,5-bis(trifluoromethyl)pyrazolyl moiety, but no N—H bands were observed. This indicatesthat the ligand is present only in its anionic form. Single crystal X-ray diffraction studies224conducted on the compound revealed that, although the compound exists in macroscopiccrystalline form, the crystal structure has severe disorder. Consequently, only a limitedamount of information could be extracted from the diffraction data. The data indicate that thecompound consists of a trimetallic ring of copper centres bridged by a total of five 3,5-bis(trifluoromethyl)pyrazolyl moieties; this is consistent with the mass balance result. If all ofthe pyrazolyl rings are anionic, as the IR spectrum suggests, then there are formally twoCu(ll) ions and one Cu(I) ion in this complex. It may be that this complex represents thefluorinated equivalent of an intermediate in the formation of the polymeric copper(II)(substituted)pyrazolates. Further study of this compound is certainly warranted.Low temperature, single crystal, crystallographic studies of [Cu(4-Mepz)2]and [Cu(4-Clpz)2] would be fruitful as such studies would yield the precise values of the lowtemperature dihedral angles, ci. and 13, in these complexes. This might permit a more thoroughunderstanding of how the temperature dependence of the structure of these compoundsinfluences their magnetic properties.The magnetic properties of these compounds, especially the [Cu(4-Xpz)21complexes, bear further investigation. If larger single crystals of the [Cu(4-Xpz)2Jspeciescould be prepared, then the full crystal susceptibility tensors for the compounds could bedetermined. It would also be informative to probe for the presence of long-range ordering inthese compounds. The powder susceptibility data provide no indication of long-rangeordering down to -‘-15 K (kT/J 0.02). Below this temperature, the effects of paramagneticimpurity begin to dominate the susceptibility behaviour of these complexes and mightobscure any indication of transitions to long-range order. Other techniques, e.g., heat capacitymeasurements or neutron diffraction studies, if conducted at sufficiently low temperatures,could reveal transitions to 3-D ordering.With respect to the binary cobak(II) (substituted)pyrazolates (Chapter 4), the mostimportant work to be pursued would be the preparation of macroscopic single crystals of any225or all of the complexes studied here. This would, of course, permit single crystal X-raydiffraction determinations of the structures of these compounds establishing with certaintywhether these compounds possess the linear chain structural motif proposed. Moreover, theprecise nature and magnitudes of distortions thought to be present in the chromophores ofthese compounds would be elucidated. Equally important, single crystals of these compoundswould allow the complete determination of their crystal susceptibility tensors. Suchinformation would be invaluable for an understanding of the nature of the magneticanisotropy already demonstrated in these compounds.The dimeric and oligometallic 3,5-dimethylpyrazolyl species of cobalt and zinc areinteresting systems. Investigation of how reaction conditions affect the sample compositionsfor these systems and mean chain lengths might shed light on how these systems polymerize.It would also be interesting to see if such dimeric species could be formed from the other 4-X-dimethylpyrazoles. Low temperature magnetization studies on the trimetallic complex, [Co(4-Hdmpz)2C1(4-HdmpzH ]o, might assist in determining whether substantial next-nearestneighbour magnetic exchange is present in this complex. Such findings would be relevant tothe current controversy about the nature of exchange in related compounds (143).Development of a rational synthesis of[Co4(4-Hdmpz)60}which precludes uncertainty aboutzinc contamination of the complex would permit characterization of its magnetic properties.The octametallic molybdenum clusters prepared during this study (Chapter 5) areinteresting species and deserve further examination. Immediate attention should probably begiven to the development of a synthetic route to the mixed valence Mo(V)/Mo(VI) in purebulk form. Also, preparation of C-substituted derivatives of these complexes with theintention of improving their solubility properties (particularly in the case of the Mo(VI)compound) is a potential area of investigation. If such species could be prepared, then itwould be worthwhile to examine them for potential catalytic activity.Efforts made toward preparation of Ni(II) 3,5-dimethylpyrazolate suggested that itmay be possible to prepare the compound in pure form; further work could be conducted in226this area (e.g., attempt purification of the mixture formed from the reaction of [rh(C5H)Ni(CO)12with 3,5-dimethylpyrazole, described in Chapter 7, using the Mond process).Finally, efforts could be made to prepare and characterize other transition metal pyrazolatecomplexes, such as manganese and iron derivatives.227CHAPTER 7EXPERIMENTAL7.1 INTRODUCTIONDescribed here are the experimental details of this work. The materials and methodsused in the syntheses of compounds and in unsuccessful preparations are described. Theapparatus and instruments used for physical studies are also described.7.2 SYNTHESESUnless otherwise stated, the materials used in this work were of reagent grade andwere used without further purification. The majority of compounds discussed in this workwere air-stable, hence they were prepared without special precautions. Compounds which didexhibit sensitivity towards air or moisture were prepared using Schienk techniques andhandled in a Vacuum Atmospheres Corporation Model HE 43-2 Dri-Lab glovebox under adinitrogen atmosphere. When it was necessary that solvents used during synthesis be free ofwater, they were dried using the following methods: THF was refluxed with sodium andbenzophenone and distilled under dinitrogen; methanol was refluxed with magnesiummethoxide and distilled under dinitrogen; acetonitrile was refluxed with phosphoruspentoxide, distilled onto calcium hydride then refluxed and distilled under dinitrogen; anddiethyl ether was refluxed with calcium hydride and distilled under dinitrogen. When it wasnecessary that solvents be deoxygenated this was achieved by freeze-pump-thaw cycles on avacuum line or by sparging with dinitrogen gas.2287.2.1 PYRAZOLE DERIVATWESPyrazole (4-HpzH), 3-methylpyrazole (3-MepzH), 4-methylpyrazole (4-MepzH),3,5-dimethylpyrazole (4-HdmpzH) and indazole (indzH) were obtained commercially(Aldrich). The remainder of the substituted pyrazoles employed in this work were preparedusing literature methods or variations of those methods. 4-CHLOROPYRAZOLE4-ClpzH4-Chioropyrazole was prepared using a variation of the method of Hüttel, Schafer,and WeIzel (175). Pyrazole (13.6 g, 0.200 mol) was dissolved in 100 mL of glacial aceticacid. To this solution was added NaOC1 (aq) (256 mL of 0.780 M, 0.200 mol) with stirring.The resulting solution was left for 18 h at room temperature and then neutralized withNa2CO3.The neutralized solution was extracted with 3x250 mL portions of CH21. TheCH21 extracts were combined, condensed under vacuum and washed with dilute NaOH(aq). The NaOH solution was extracted with 2 x 100 mL portions of CH21.The CH21extracts were combined and dried under vacuum yielding 17.4 g (85%) of 4- chloropyrazoleas a white powder, mp 72-74 °C (lit.’ 76-77 °C). Anal. calcd. forC3H1N2:C 35.1, H 3.0, N27.3; found: C 35.2, H 2.9, N 27.5. NMR spectroscopy (‘H in CDC13): 6 7.57 (s, 2, CH),11.13(brs, 1,NH). 4-BROMOPYRAZOLE4-BrpzHPyrazole (13.6 g, 0.200 mol) was dissolved in 50 mL of warm water. Bromine (32.0 g,0.200 mol) was added dropwise, with stirring, to this solution until the yellow colour of Br2229persisted. The resulting solution was refluxed for a period of 30 mm and cooled.Concentrated NaOH (aq) was added to the reaction solution until it was slightly basic,resulting in precipitation of an off-white solid. The mixture was suction filtered, the solidwashed with cold water and air dried. The solid was then dissolved in a minimum amount ofdiethyl ether and suction filtered to remove any entrained NaBr. The diethyl ether wasremoved from the filtrate under vacuum yielding 4-bromopyrazole as an off-white powder,mp 90.5-92.5 °C (recrystallized from petroleum ether)(lit. (176) 96-97 °C). Anal. calcd. forC3HBrN2:C 24.5, H 2.1, N 19.1; found: C 23.9, H 2.0, N 18.8. NMR spectroscopy (1H inCDC13): 67.63 (s, 2, CH), 12.1 (br s, 1, NH). 4-IODOPYRAZOLE4-IpzH4-lodopyrazole was prepared according to the method of Huttel, Schafer, and Jochum(177). Pyrazole (28.5 g, 0.419 mol) and sodium acetate (126 g, 1.53 mol) were dissolved in250 mL of water and then heated to reflux. To this was added dropwise, with stirring, asolution of KI (210 g, 1.26 mol) and 12 (112 g, 0.440 mol) dissolved in 100 mL of water.Addition of the solution was ceased when the colour of the 13 persisted in the solution. Thereaction solution was then stored overnight at 5°C. Needle-shaped, colourless crystalsprecipitated from the solution. These were suction filtered, washed with water, and thendissolved in a minimum amount of CH21.The CH21 solution was filtered to remove anyentrained solids and then flash evaporated yielding 5.7 g (20%) 4-iodopyrazole as fine whiteneedles, mp 108-108.5 °C (lit. (177) 108.5 °C). Anal. calcd. forC3H1N2:C 18.6, H 1.6, N14.4; found: C 18.4, H 1.5, N 14.3. NMR spectroscopy (‘H in CDC13): 6 7.62 (s, 2, CH),12.00 (br s, 1, NH).2307.2.1.4 4-NITROPYRAZOLE4-NO2pzH4-Nitropyrazole was synthesized by means similar to the method of Morgan andAckerman (178) for the synthesis of 4-nitro-3,5-dimethylpyrazole. Pyrazole (14.2 g, 0.209mol) was dissolved in 40 mL of concentrated H2S04 at 0°C. To this was added 24 mL ofconcentrated HNO3.The solution took on a faint yellow colour and was heated on a hot plateat low heat for 24 h. After heating, the solution was allowed to cool to room temperature andthen it was poured onto 300 mL of ice. NaOH pellets (-.- 180 g, 4.5 mol) were slowly added,with constant stirring, to neutralize the solution. The solution fumed as it was neutralized anda white precipitate formed. The precipitate was suction filtered and washed with 400 mL ofwater. The resulting white solid was dissolved in 150 mL of acetone and filtered to remove asmall amount of entrained solid. The filtrate was flash evaporated and the resulting productdried in air at 110°C for two hours yielding 15.4 g (65%) of 4-nitropyrazole as white powder,mp 162-163 °C (lit. (179) 162 °C). Anal. calcd. forC3H2N0:C 31.9, H 2.7, N 37.2; found:C 31.9, H 2.8, N 37.3. NMR spectroscopy (‘H ind6-acetone): 8 8.40 (s, 2, CH), 5.30 (br s, 1,NH). 3,4,5-TRIMETHYLPYRAZOLE4-MedmpzH3,4,5-Trimethylpyrazole was synthesized using the procedure of Chambers et al.(180). Sodium pieces (29.9 g, 1.30 mol) were added slowly to 400 mL of anhydrous methanolunder a dinitrogen atmosphere. Once reaction of the sodium metal was complete,pentane-2,4-dione (129 mL, 1.25 mol) was added to the methanolic solution and the resultingsolution was allowed to cool. lodomethane (86 mL, 1.39 mol) was added dropwise, withstirring, over the course of one hour. Methanol was removed from the solution by fractional231distillation (bp 60 °C at 540 mmHg). A small forerun of the product was discarded and thenthe product was collected by distillation (bp 161 °C at 540 mmHg) The product was dried atroom temperature for 12 h over H2S04 and this yielded 70.0 g (49%) of3-methylpentane-2,4-dione as a pale yellow liquid.The above product (54.2 g, 0.475 mol) was added dropwise to a stirred solution ofhydrazine hydrate (80-90%, 30 mL, 0.48- 0.54 mol), 1.5 mL of glacial acetic acid, and 100mL of water held at 10-15 °C. A white precipitate formed from the reaction. The resultingmixture was cooled to 5 °C for 3 h, suction filtered, the solid washed with ice cold water, anddried under vacuum for 12 h yielding 36.1 g (69%) of 3,4,5-trimethylpyrazole, mp 131-132.5°C (lit. (181) 137-138 °C). Anal. calcd. forC6H10N2:C 65.4, H 9.2, N 25.5; found: C 65.2, H9.0, N 25.5. NMR spectroscopy (‘H in CDC13): 8 1.85 (s, 3, CH3), 2.15 (s, 6, CH3), 10.45 (brs, 1,NH). 4-CHLORO-3,5-DIMETHYLPYRAZOLE4-Cldmpzll4-Chloro-3,5-dimethylpyrazole was prepared using a method similar to that employedin the synthesis of 4-chloropyrazole above (175). 3,5-Dimethylpyrazole (9.23 g, 0.0960 mol)was dissolved in 100 mL of water and glacial acetic acid (15 mL, 0.26 mol) was added to thesolution. NaOC1 (aq) (200 mL of 0.482 M, 0.0964 mol) was added to the solution and it wasstirred at room temperature for 2 h (an off-white solid formed during that time). The mixturewas then neutralized with Na2CO3and made slightly basic with a few drops of concentratedammonia solution. The resulting mixture was cooled to 5 °C and suction filtered. The solidcollected was washed with 200 mL of cold water, and dried under vacuum at roomtemperature for 4 h yielding 10.9 g (87%) of 4-chloro-3,5-dimethylpyrazole, mp 112.5-113.5°C (lit. (182) 117.5-118.5 °C). Anal. calcd. forC5H71N2:C 46.0, H 5.4, N 21.4; found: C23245.8, H 5.3, N 21.5. NMR spectroscopy (‘H in CDC13): 6 2.26 (s, 6, CH3), 9.93 (br s, 1, NH). 4-BROMO-3,5-DIMETHYLPYRAZOLE4-BrdmpzH3,5-Dimethylpyrazole (9.6 g, 0.10 mol) was dissolved in 300 mL of water. Bromine(5.2 mL, 0.10 mol) was added to the solution dropwise, with stirring, the brominedecolourizing almost immediately and a small amount of pale yellow oil separated. Theresulting mixture was neutralized with concentrated NaOH (aq) and then allowed to settle.The aqueous supernatant was cooled to 5 °C overnight and a white solid precipitated whichwas suction filtered, washed with cold water and air-dried. The oily component was driedunder vacuum resulting in the formation of a tan coloured solid. Both solid fractions werecombined, recrystallized from toluene (-•.- 20 mL), suction filtered, washed with ice coldpetroleum ether, and dried under vacuum at room temperature yielding 11.2 g (64%) of4-bromo-3,5-dimethylpyrazole as a white solid, mp 115-117 °C (lit. (178) 118 °C). Anal.calcd. forC5H7BrN2:C 34.3, H 4.0, N 16.0; found: C 34.6, H 4.1, N 16.3. NMR spectroscopy(1H in CDC13): 62.21 (s, 6, CH3), 11.2 (br s, 1, NH). 4-IODO-3,5-DIMETHYLPYRAZOLE4-IdmpzH3,5-Dimethyl-4-iodopyrazole was synthesized using a variation of the method ofMorgan and Ackerman (178). 3,5-Dimethylpyrazole (9.0 g, 0.094 mol) and Na2CO3(0.53 g,0.050 mol) were dissolved in 500 mL of water and the solution was heated to reflux withstirring. To this solution was added dropwise a solution of KI (31.0 g, 0.186 mol) and ‘2 (23.0g, 0.0905 mol) dissolved in 80 mL of water. The reaction was essentially instantaneous withdisappearance of the 13 colour. When addition of the K13 solution was complete, the reaction233mixture was cooled to 5 °C and left overnight. A fibrous white solid precipitated from thesolution. The precipitate was suction filtered and washed with 300 mL of ice cold water. Thefiltercake was dissolved in 200 mL of CH2I and extracted with 200 mL of a dilute aqueoussolution of Na2CO3 to remove any remaining HI. The CH21 phase was collected, flashevaporated, and the residue dried at room temperature under vacuum for 5 h yielding 19.0 g(94%) of 3,5-dimethyl-4-iodopyrazole as a white solid, mp 133.5-134.5 °C (lit. (178) 137°C). Anal. calcd. forC5H71N2:C 27.0, H 3.2, N 12.6; found: C 27.1, H 3.2, N 12.6. NMRspectroscopy (1H in CDC13): ö 2.23 (s, 6, CH3), 11.65 (br s, 1, NH). 4-NITRO-3,5-DIMETHYLPYRAZOLE4-NO2dmpzH3,5-Dimethyl-4-nitropyrazole was synthesized using the method of Morgan andAckerman (178). 3,5-Dimethylpyrazole (15.0 g, 0.156 mol) was dissolved in 30 mL ofconcentrated H2S04 at 0°C. Then 18 mL of concentrated HNO3 was added slowly, withstirring, and a further 60 mL of concentrated H2S04 was added. The solution was leftstanding at room temperature for 8 h and then it was heated on a hot plate to 100 °C for afurther 4 h. During the heating some red fumes were evolved. The solution was then allowedto cool to room temperature and poured onto 300 mL of ice and placed into an ice bath. Themixture was neutralized by dropwise addition of a concentrated NaOH (aq) solution whichresulted in formation of a white precipitate. The mixture was suction filtered and theprecipitate washed with 100 mL of water. The precipitate was then dissolved in 110 mL ofCH21 and the solution filtered and flash evaporated. The resulting white solid was driedunder vacuum for 8 h at room temperature yielding 22.0 g (95%) of4-nitro-3,5-dimethylpyrazole, mp 121-122 °C (lit. (178) 126 °C). Anal. calcd. forC5H7N302:C 42.6, H 5.0, N 29.8; found: C 42.6, H 4.9, N 29.9. NMR spectroscopy (1H in CDC13): ö2342.62 (s, 6, CH3), 10.46 (br s, 1, NH).7.2.2 COPPER (I) PYRAZOLATESFor all of the copper(I) pyrazolate syntheses listed below CuT was used as the coppersource for preparation of bulk samples. The CuT was prepared from aqueous solutions ofCu(N03)2.2H0and K]. TRTS(t-3,5-DIMETHYLPYRAZOLATO-N,N’)TRICOPPER(I)[Cu(4-Hdmpz)]3This copper(I) trimer was prepared employing a more expedient variation of themethod of Attilio et al. (101). Cuprous iodide (1.90 g, 10.0 mmol) and 3,5-dimethylpyrazole(0.96 g, 10 mmol) were dissolved in 200 mL of degassed acetonitrile under a dinitrogenatmosphere. To this solution was added triethylamine (2.0 mL, 14 mmol) which resulted inthe immediate precipitation of a fine white powder. The mixture was stirred at roomtemperature for 20 mm, then suction filtered, the filtercake washed with 50 mL of acetone,and air dried at 110 °C for 30 mm. This yielded 1.25 g (78%) of [Cu(4-Hdmpz)]3as a finewhite powder. Anal. calcd. forC15H21u3N6:C 37.8, H 4.4, N 17.6; found: C 38.0, H 4.4, N17.8.Single crystals of [Cu(4-Hdmpz)]3of X-ray diffraction quality were grown in thefollowing manner. Copper(TI) hydroxide (-0.3 g, 3 mmol) was combined with3,5-dimethylpyrazole (4.0 g, 42 mmol) in a 10 mL round bottomed flask with attachedcondenser and heated at 130°C for 10 d under a dinitrogen atmosphere without stirring.Almost immediately, the mixture turned a dark reddish-brown. After 24 h, colourless crystalshad formed from the mixture. At the end of the reaction period, the product wasapproximately 70% colourless crystals and 30% fine reddish-brown powder. The excess2353,5-dimethylpyrazole was removed by extracting the mixture with acetone and most of thereddish-brown powder was separated from the crystals by suspension of the powder inacetone and subsequent decantation of the suspension. The crystals of [Cu(4-Hdmpz)13werelarge enough that an analytically pure sample of the product was obtained from the mixtureby manual separation. TRIS (p.-3,4,5-TRIMETHYLPYRAZOLATO-N,N’)TRICOPPER(I)[Cu(4-Medmpz)]3This was the first of the Cu(I) pyrazolates prepared in this study and, at the time, itwas not recognized that these compounds are air-sensitive when in contact with solvents.[Cu(4-Medmpz)]3,unlike the other compounds in this series, was prepared in air. Cuprousiodide (1.90 g, 10.0 mmol) was dissolved in 70 mL of acetonitrile. To this was added asolution of 3,4,5-trimethylpyrazole (2.20 g, 20.0 mmol) dissolved in 60 mL of acetone.Within a few minutes the solution had turned brown. Triethylamine (2.8 mL, 20 mmol) wasadded to the solution and a white solid immediately precipitated. The mixture was stirred for1 h and then suction filtered resulting in a white filtercake and a green filtrate. The filtercakewas washed with acetone and petroleum ether and dried in air at 110 °C for 15 mm yielding1.13 g (66%) of [Cu(4-Medmpz)j3as a white powder. Anal. calcd. forC18H27u3N6:C 41.7,H 5.2, N 16.2; found: C 41.3, H 5.2, N 16.0.X-ray diffraction quality single crystals of [Cu(4-Medmpz)]3 were preparedunexpectedly in a reaction described in section TRIS (ii-4-CHLORO-3,5-DIMETHYLPYRAZOLATO-N,N’)TRICOPPER(I)[Cu(4-Cldmpz)]3Cuprous iodide (1.90 g, 10.0 mmol) was dissolved in 80 mL of deoxygenatedacetonitrile under a dinitrogen atmosphere. 4-Chloro-3,5-dimethylpyrazole (1.30 g, 10.0mmol) was added to the solution and a flocculent white precipitate immediately formed.Finally, triethylamine (1.6 mL, 11 mmol) was added to the mixture and more whiteprecipitate formed. The mixture was stirred for 2 h, suction filtered, and the filtercake washedwith 50 mL of acetonitrile and 20 mL of acetone. The filtercake was dried for 30 mm at 110°C in air yielding 1.81 g (94%) of [Cu(4-Cldmpz)]3 as a white solid. Anal. calcd. for15H8C13uN6:C 31.1, H 3.1, N 14.5; found: C 31.1, H 3.2, N TRIS (ji-4-BROMO-3,5-DIMETHYLPYRAZOLATO-N,N’)-TRICOPPER(I)[Cu(4-Brdmpz)]3Cuprous iodide (0.95 g, 5.0 mmol) was dissolved in 80 mL of deoxygenatedacetonitrile under a dinitrogen atmosphere. 4-Bromo-3,5-dimethylpyrazole (0.88 g, 5.0mmol) was added to the solution and a flocculent white precipitate formed. Triethylamine(1.0 mL, 7.1 mmol) was added to the mixture and the mixture was stirred for 2 h. Themixture was then suction filtered, the fikercake washed with 25 mL of acetone, and dried at110 °C for 15 mm in air. This yielded 1.14 g (96%) of [Cu(4-Brdmpz)j3as a white solid.Anal. calcd. forC15H8Br3uN6:C 25.3, H 2.5, N 11.8; found: C 25.1, H 2.5, N 11.8.2377.2.2.5 TRIS (i-4-IODO-3,5-DIMETHYL-PYRAZOLATO-N,N’)TRICOPPER(I)[Cu(4-Idmpz)]3Cuprous iodide (1.90 g, 10.0 mmol) and 3,5-dimethyl-4-iodopyrazole (2.22 g, 10.0mmol) were added to 200 mL of deoxygenated acetonitrile under a dinitrogen atmosphere. Aflocculent white precipitate formed as the reactants dissolved. Triethylamine (2.0 mL, 14mmol) was then added to the mixture and more white precipitate formed. The mixture wasstirred for 5 h, then suction filtered, and the filtercake washed with 50 mL of acetone. Thefiltercake was dried at 110 °C in air for 15 mm. This yielded 2.70 g (95%) of [Cu(4-Idmpz)]3as a fibrous white solid. Anal. calcd. forC15H8u3IN6:C 21.1, H 2.1, N 9.8; found: C 21.3,H 2.2, N TRIS (j.t-4-IODOPYRAZOLATO-N,N’)TRICOPPER(I)[Cu(4-Ipz)}3Cuprous iodide (1.90 g, 10.0 mmol) and 4-iodopyrazole (1.94 g, 10.0 mmol) werecombined in 125 mL of deoxygenated acetonitrile under a dinitrogen atmosphere and stirredfor 20 mm during which time a flocculent white precipitate formed. To this mixture wasadded triethylamine (1.8 mL, 13 mmol) and immediately the solid present turned a lightyellowish-grey. The mixture was stirred for a further 45 mm, filtered under dinitrogen, thefiltercake washed with 100 mL of deoxygenated acetonitrile, and dried under a stream ofnitrogen. This yielded 2.29 g (89%) of [Cu(4-Ipz)] as a yellowish grey powder and it wascharacterized by JR spectroscopy (29).2387.2.2.7 TRIS(p.-INDAZOLATO-N,N’)TRICOPPER(I)[Cu(indz)]3Cuprous iodide (1.90 g, 10.0 mmol) and indazole (1.18 g, 10.0 mmol) were dissolvedin 100 mL of deoxygenated acetonitrile under a diniirogen atmosphere. To this solution wasadded triethylamine (1.5 mL, 11 mmol). The resulting solution was initially a pale yellowand after 0.5 h a white precipitate was seen forming from the solution. The mixture wasstirred at room temperature for 3 h during which time a large amount of flocculent whitesolid deposited from the solution. After this time the mixture was suction filtered underdinitrogen, the filtercake washed with 50 mL of deoxygenated acetone, and dried undervacuum at room temperature for 12 h. This afforded 1.12 g (62%) of [Cu(indz)]3 as a whitesolid. Anal. calcd. forC7H5uN2:C 46.5, H 2.8, N 15.5; found: C 46.3, H 3.0, N COPPER(II) PYRAZOLATESIn the syntheses described in this section, wherever copper is employed as a reactantthis refers to copper metal beads (3-5 mm diameter) which had been cleaned by washing with12 M HC1, water, and finally acetone prior to use. POLY-BIS(p.-PYRAZOLATO-N,N’)COPPER(II)[Cu(4-Hpz)2jCopper (8.13 g, 128 mmol) and pyrazole (10.0 g, 147 mmol) were placed in a 100 mLround bottomed flask fitted with a condenser. The reaction mixture was heated, with stirring,to 110 °C and air was bubbled into the mixture via a Pyrex tube for 18 h. A green solid beganto form within minutes. Sublimed pyrazole was periodically scraped back into the reactionflask during the reaction. The mixture, which had solidified on cooling, was extracted with239CH21,suction filtered, the filtercake further washed with CH21 and petroleum ether, andair dried at room temperature. This yielded 3.80 g (79% based on copper reacted) of[Cu(4-Hpz)2]as a fine green powder. The remaining reacted copper apparently formed abrown CH21 soluble material (in fact, soluble by-products formed during the syntheses ofall the copper(ll) pyrazolates discussed here). Anal. calcd. forC6HuN4:Cu 32.2, C 36.4, H3.1, N 28.3; found: Cu 32.3, C 36.7, H 3.2, N 28.5.Single crystals of [Cu(4-Hpz)2] suitable for X-ray diffraction studies weresynthesized in a manner similar to that described above except that only —0.5 g of copper wasused and the reaction mixture was not stirred. In addition, air was not bubbled through themixture, but instead allowed to diffuse in slowly from the top of the reflux condenser. Thereaction was allowed to proceed for 5 days after which the cooled mixture was washed withwater and then acetone yielding dark green crystals. POLY-BIS (.t-4-METHYLPYRAZOLATO-N,N’)COPPER(ll)[Cu(4-Mepz)2}Approximately 3 g of copper were combined with 4-methylpyrazole (3 g, 36 mmol) ina 5 mL round bottomed flask fitted with a condenser and gas-tee. The mixture was heated at115 °C with stirring, for 18 h, under a dioxygen atmosphere. The mixture was then allowed tocool, washed with CH21 to both extract the mixture from the flask and to dissolve solublecomponents. The resulting slurry was then suction filtered and washed with acetone yielding[Cu(4-Mepz)2]as a fine green powder. The product was dried in air at 110 °C for 30 mm.Anal. calcd. forC8H10uN4:Cu 28.2, C 42.6, H 4.5, N 24.8; found: Cu 28.2, C 42.57, H 4.4,N 24.8.Single crystals suitable for X-ray diffraction studies were prepared by a methodsimilar to that used for preparation of the bulk compound. The conditions differed in that the240mixture was not stirred, the apparatus was flushed with dinitrogen gas and air was allowed todiffuse into the reaction vessel via a pinhole leak, and the reaction was allowed to proceed for3 days. The reaction mixture was washed with acetone yielding dark green crystals of[Cu(4-Mepz)2] POLY-BIS (i.t-4-CHLOROPYRAZOLATO-N,N’)COPPER(II)[Cu(4-Clpz)2]Cu20 (0.358 g, 5.00 mmol), 4-chioropyrazole (4.39 g, 42.8 mmol), and 1.5 mL ofxylene were combined in a 10 mL round bottomed flask fitted with a condenser and gas-tee.The mixture was heated at 130 °C for 8.5 h with vigorous stirring under a dioxygenatmosphere. The mixture was then allowed to cool, extracted with CH21, suction filtered,the filtercake washed with CH21 and finally acetone. The resulting fine green powder wasdried in air at 110 °C for 30 mm yielding 1.000 g (75%) of [Cu(4-Clpz)2].Anal. calcd. forC6H412uN:Cu 23.8, C 27.0, H 1.5, N 21.0; found: Cu 23.9, C 27.2, H 1.6, N 21.2.X-Ray diffraction quality single crystals of [Cu(4-C1pz)2jwere grown by combiningcopper and 4-chloropyrazole in a test tube, then heating the mixture to 120 °C, withoutstirring, for 3 days. The reaction mixture was washed with acetone yielding dark greencrystals of the desired product in addition to some brown and bluish-grey crystals.A brown form of [Cu(4-Clpz)2]was prepared as follows. Freshly prepared Cu(OH)2(0.73 g, 7.5 mmol), 4-chloropyrazole (4.6 g, 45 mmol), and 2 niL of xylene were combined ina 10 niL round bottomed flask fitted with a condenser and heated at 135 °C in air for 6 days,with stirring. The mixture was extracted from the flask with acetone, suction filtered, thefiltercake washed with acetone, and air dried yielding a fine powder which consisted mostlyof a brown material with a trace of green material. Anal. calcd. forC6H412uN:C 27.0, H1.5, N 21.0; found: C 27.1, H 1.5, N 21.2.2417.2.3.4 POLY-BIS (.t-4-BROMOPYRAZOLATO-N,N’)COPPER(II)[Cu(4-Brpz)2JCu(OH)2 (0.487g. 5.00 mmol), freshly prepared from Cu(N03).2H0,was combinedwith 4-bromopyrazole (4.41 g, 30.0 mmol) in a 10 mL round bottomed flask fitted with acondenser and gas-tee. The mixture was heated to 145 °C, with stirring, for 16 h under adinitrogen atmosphere. After this time the reaction mixture was cooled and 5 mL of acetoneadded. The reaction mixture was then refluxed for 3 h causing sublimed 4-bromopyrazole tobe washed back into the flask. At the cessation of the reaction the mixture was allowed tocool and solidify. The solid was then extracted with ethanol which also dissolved solublecomponents. The resulting slurry was suction filtered, the filtercake washed with water,ethanol, and CH21.The resulting green powder was dried under vacuum for 2 h at roomtemperature yielding 0.89 g (50%) of [Cu(4-Brpz)2J.Anal. calcd. forC6H4Br2uN:Cu17.8, C 20.3, H 1.1, N 15.8; found: Cu 18.1, C 20.3, H 1.1, N 15.9.Attempts to grow single crystals of this compound were made by repeating thereaction without stirring. Single crystals did form, however, they were too small for X-raydiffraction studies. BIS (.t-4-IODOPYRAZOLATO-N,N’)HEMI(4-IODOPYRAZOLE)COPPER(II)Cu(4-Ipz)2.4(4-IpzH)[Cu(4-Ipz)J (0.256 g, 1.00 mmol) and 4-iodopyrazole (1.90 g, 9.80 mmol) werecombined in a 10 mL round bottomed flask and heated, with stirring, at 110 °C under adioxygen atmosphere for 11 h. The reaction mixture initially turned dark brown and later242became olive green. The mixture was extracted with TI-IF, suction filtered, the filtercakewashed with THF, and dried in air at room temperature. This yielded 0.533 g (98%) of[Cu2(4-Ipz)-IpzH)] as an olive green powder. Anal. calcd. asC15H1u2I5N0:C 16.5, H1.0, N 12.8; found: C 16.6, H 1.1, N POLY-BIS (j.t-3,5-DIMETHYLPYRAZOLATO-N,N’)COPPER(II)[Cu(4-Hdmpz)2][Cu(4-Hdmpz)2Jwas synthesized via two methods. In the first method copper (3.15g, 49.6 rol) and 3,5-dimethylpyrazole (11.0 g, 114 mmol) were combined in a 50 mLround bottomed flask fitted with a condenser and heated in air at 115 °C, with stirring, for 4days. The mixture was then cooled to room temperature and extracted from the flask withCH21.It was then suction filtered, the filtercake washed with petroleum ether, and dried inair at 110 °C for 4 h. This yielded approximately 0.5 g of [Cu(4-Hdmpz)2]as areddish-brown powder. Anal. calcd. forC10H4uN4:Cu 25.0, C 47.3, H 5.6, N 22.1; foundCu 25.2, C 47.1, H 5.4, N 22.3.In the second method [Cu(4-Hdmpz)13 (0.3 19 g, 2.01 mmol Cu(I)) and3,5-dimethylpyrazole (1.92 g, 20.0 mmol) were combined in a small vessel and heated undera dioxygen atmosphere at 124 °C for 3 h (reaction was probably complete after 1 h). Themixture was then cooled to room temperature and extracted with dry THF and suctionfiltered. The filtercake was washed with more TI-IF and dried at room temperature. Thisyielded 0.509 g (99%) of [Cu(4-Hdmpz)2].Anal. calcd. forC10H4uN4:C 47.3, H 5.6, N22.1; found: C 47.3, H 5.6, N 22.2.2437.2.3.7 POLY-BIS(t-3,4,5-TRIMETHYLPYRAZOLATO-N,N’)COPPER(H)[Cu(4-Medmpz)2][Cu(4-Medmpz)]3(0.345 g, 2.00 mmol) and 3,4,5-trimethylpyrazole (1.76 g, 16.0mmol) were combined in a 10 mL round bottomed flask fitted with a condenser and gas-tee.The mixture was heated to 140 °C, with stirring, under a dioxygen atmosphere for 2 h. Thesample was then cooled to room temperature and dried under vacuum for 10 h. The mixturewas then extracted from the flask under dinitrogen with dry, deoxygenated THF. Theresulting mixture was suction filtered under dinitrogen, the filtercake washed with THF, anddried at room temperature under vacuum for 5 h. This yielded 0.540 g (96%) of[Cu(4-Medmpz)2Jas a fine brown powder. Anal. calcd. forC12H8uN4:Cu 22.5, C 51.1, H6.4, N 19.9; found: Cu 22.5, C 51.0, H 6.4, N POLY-BIS(-4-CHLORO-3,5-DIMETHYLPYRAZOLATO-N,N’)-COPPER(II)[Cu(4-Cldmpz)2]1Cu(4-C1dmpz)13(0.386 g, 2.00 mmol) and 4-chloro-3,5-dimethylpyrazole (2.61 g,20.0 mmol) were combined in a 10 mL round bottomed flask fitted with a condenser andgas-tee. The reaction mixture was heated at 140 °C, with stirring, under a dioxygenatmosphere for 4 h. At cessation of the reaction the mixture was cooled to room temperatureand dried under vacuum for 24 h. The mixture was then extracted from the flask under adinitrogen atmosphere with 50 mL of dry, deoxygenated THF. This mixture was then suctionfiltered under dinitrogen, the filtercake washed with 50 mL of THF, and dried under vacuumat room temperature for 2 h. Thus was obtained 0.561 g (87%) of [Cu(4-Cldmpz)2]as a finebrown powder. Anal. calcd. forC10H212uN4:Cu 19.7, C 37.2, H 3.8, N 17.4; found: Cu24419.7, C 37.2, H 3.7, N POLY-BIS(.t-4-BROMO-3,5-DIMETHYLPYRAZOLATO-N,N’)-COPPER(II)[Cu(4-Brdmpz)2][Cu(4-Brdmpz)]3(0.7 13 g, 3.00 mmol), 4-bromo-3,5-dimethylpyrazole (4.02 g, 22.9mmol), and 2 mL of xylene were combined in a 10 mL round bottomed flask fitted with acondenser and gas-tee. The mixture was heated at 145 °C, with stirring, for 7 h under adioxygen atmosphere. The mixture was then cooled to room temperature and dried undervacuum for 16 h. Once dry, the mixture was extracted from the flask with 20 mL of dry,deoxygenated THF under a dinitrogen atmosphere. The resulting slurry was suction filteredunder dinitrogen, the filtercake washed with 30 mL of THF, and dried under vacuum at roomtemperature for 15 h. This yielded 1.08 g (88%) of [Cu(4-Brdmpz)2]as a dark brown solid.Anal. calcd. forC10H2Br2uN4:Cu 15.4, C 29.2, H 2.9, N 13.6; found: Cu 15.5, C 29.4, H3.0, N POLY-BIS(.t-3-METHYLPYRAZOLATO-N,N’)COPPER(ll)[Cu(3-Mepz)2]Copper (--5 g) and 3-methylpyrazole (-40 mL) were combined in a 50 mL roundbottomed flask fitted with a condenser and heated in air at 110 °C for 20 h, with stirring. Thereaction mixture was extracted from the flask with CH21, suction filtered, the filtercakewashed with CH21 and ethanol, and dried in air at 110 °C for 18 h. This yieldedapproximately 5 g of [Cu(3-Mepz)2Jas a light green powder. Anal. calcd. forC8H10uN4:Cu 28.2, C 42.6, H 4.5, N 24.8; found: Cu 28.2, C 42.8, H 4.5, N 25.0.2457.2.3.11 POLY-BIS(jt-INDAZOLATO-N,N’)COPPER(II)[Cu(indz)2J[Cu(indz)J3(0.298 g, 1.65 mmol) and indazole (2.00 g, 17.0 mmol) were combined ina 10 mL round bottomed flask fitted with a condenser and gas-tee. The mixture was heated at157 °C for 2.25 h, with stirring, under a dioxygen atmosphere. The mixture was allowed tocool to room temperature and was dried under vacuum for 13 h and then extracted from theflask with 35 mL of dry, deoxygenated THF under a dinitrogen atmosphere. The resultingslurry was suction filtered under dinitrogen, the filtercake washed with 50 mL of THF, anddried under vacuum for 2 h at room temperature. This yielded 0.440 g (90%) of [Cu(indz)2]as a fine brownish black powder. Anal. calcd. forC14H0uN4:C 56.5, H 3.4, N 18.8; found:C 56.5, H 3.4, N COPPER CARBOXYLPYRAZOLATESThe compounds discussed in this section were unexpected by-products producedduring attempts to grow X-ray diffraction quality single crystals of substituted3 ,5-dimethylpyrazolates of copper(II). BIS(ji-4-BROMO-3-CARBOXY-5-METHYLPYRAZOLATO-N,N’,O)-TETRAKIS (4-BROMO-3,5-DIMETHYLPYRAZOLE)DICOPPER(II)[Cu(4-Br-3-COMepz)(4-BrdmpzH)]Copper (-‘4 g) and 4-BrdmpzH (-2.5 g) were combined in a 10 mL round bottomedflask fitted with a reflux condenser and the mixture was heated at 140 °C for three days,without stirring. After this time, the reaction mixture was allowed to cool and solidify. The246solid was extracted with CH21.The product consisted of a green, CH21 soluble fraction,a minor fraction of yellow, insoluble material, a minor fraction of bluish, grey insolublematerial, and the majority of the product consisted of well-formed green crystals of[Cu(4-Br-3-CO2Mepz)(4-BrdmpzH)].The yellow and bluish-grey insoluble componentswere removed from the compound by manual separation. The green crystals occur in twosubtly different morphologies. This is a consequence of the fact that the molecule crystallizesin a monoclinic and in a triclinic modification. The vast majority of the crystals are of thetriclinic form. Anal. calcd. forC30HBr6u2N12O4:Cu 10.3, C 29.2, H 2.8, N 13.6; found:Cu 10.3, C 29.4, H 2.8, N BIS[Q.t-3-CARBOXY-4,5-DIMETHYLPYRAZOLATO-N,N’,O)(3,4,5-TRIMETHYLPYRAZOLE)CUPRATE(I)]COPPER(II)[Cu(3-COdmpz)(4-MedmpzH)]uANDTRIS (ii.-3,4,5-TRIMETHYLPYRAZOLATO-N,N’)TRICOPPER(I)[Cu(4-Medmpz)]3Copper (8.18 g, 129 mmol) was combined with 3,4,5-trimethylpyrazole (4.73 g, 42.9mmol) in a 10 mL round bottomed flask, fitted with a condenser, and the mixture was heated,without stirring, in air for 40 h at 142 °C. After the first 18 h of heating, the neck of thereaction vessel had become plugged with sublimed 3,4,5-trimethylpyrazole which limitedfurther contact of the reaction mixture with atmospheric dioxygen. At cessation of thereaction, the mixture was allowed to cool and solidify. The mixture was then extracted with150 mL of dry THF and suction filtered. This yielded a brown filtrate and 0.08 g of filtercakewhich consisted of an approximately equal mixture of colourless crystals of[Cu(4-Medmpz)]3 and pale purple crystals of [Cu(3-CO2dmpz)(4-MedmpzH)]u. Closeexamination of the beads of copper shot under a microscope revealed that the crystals of247[Cu(4-Medmpz)]3 formed a layer on the surface of the bead and the[Cu(3-CO2dmpz)(4-MedmpzH)]u formed a second layer surrounding the[Cu(4-Medmpz)]3layer. Unfortunately, both compounds are insoluble in a variety of solventsand because of the small crystal size, it was not feasible to separate bulk samples of thecompounds manually.7.2.5 COBALT PYRAZOLATESFor the syntheses described in this section, wherever cobalt metal is listed as areagent, unless stated otherwise, this means cobalt metal powder. Yields of the productsobtained from cobalt metal powder could not be determined as percentages because,invariably, substantial amounts of the products remained adhered to the cobalt powder whenit was separated from the reaction mixture. POLY-BIS(ji-PYRAZOLATO-N,N’)COBALT(II)[Co(4-Hpz)2JCoC12.xHO(3.31 g, <14 mmol) and pyrazole (6.08 g, 89.3 mmol) were dissolved in80 mL of water. The solution was degassed and maintained under a dinitrogen atmosphere.NaOH (1.60 g, 40.0 mmol) was dissolved in 40 mL of water and the resulting solution wasdegassed and kept under a dinitrogen atmosphere. The NaOH solution was added dropwise tothe vigorously stirred CoCl2/pyrazole solution at room temperature. The reaction mixture wassuction filtered under dinitrogen and washed sequentially with degassed water, degassedethanol and, finally, degassed petroleum ether. This afforded a purple filtercake which wasdried under vacuum for 1.5 h at room temperature. This yielded —2 g of [Co(4-Hpz)2]as afine purple powder. Anal. calcd. forC3H6oN4:C 37.3, H 3.1, N 29.0; found: C 37.5, H 3.2,248N POLY-BIS (.t-3-METhYLPYRAZOLATO-N,N’)COBALT(II)[Co(3-Mepz)2]Cobalt metal (0.8 g, 14 mmol) and 3-methylpyrazole (6 g, 75 mmol) were combinedin a 100 mL round bottomed flask fitted with a condenser and heated, with stirring, at 100 °Cfor 18 h in air. After this time, CH21 was added to the mixture and unreacted cobalt metalwas removed from the mixture by repeated dredging of the slurry with a large magnet untilno more cobalt metal was visible on the magnet. The mixture was then suction filtered andwashed with CH21 yielding a magenta filtrate and a purple filtercake. The filtercake wasdried at 110 °C in air for 1 h. Thus --3 g of [Co(3-Mepz)2]was obtained as a fine purplepowder. Anal. calcd. forC8H10oN4:C 43.5, H 4.6, N 25.3; found: C 44.1, H 4.7, N POLY-BIS(j.t,-3,5-DIMEThYLPYRAZOLATO-N,N’)COBALT(II)[Co(4-Hdmpz)2]Cobalt metal (2.32 g, 39.4 mmol), 3,5-dimethylpyrazole (6.00 g, 62.4 nimol), and 5mL of xylene were combined in a 50 niL round bottomed flask fitted with a condenser and agas-tee. The reaction mixture was heated, with stirring, at 115 °C for 6 days under a dioxygenatmosphere. The resulting mixture was allowed to cool and solidify. The mixture was thenextracted from the reaction flask with THF and unreacted cobalt metal was removed from themixture by repeated dredging of the slurry with a large magnet until no more cobalt metalwas visible on the magnet. The slurry was then suction filtered and the filtercake washed withseveral portions of THF and then air dried at room temperature. This yielded 0.85 1 of[Co(4-Hdmpz)2]as a fme purple powder. Anal. calcd. forC10H4oN4:C 48.2, H 5.7, N22.5; found: C 48.0, H 5.6, N 22.3.2497.2.5.4 POLY-BIS (i.t-3,4,5-TRJMETHYLPYRAZOLATO-N,N’)COBALT(ll)[Co(4-Medmpz)2]Cobalt metal (2.00 g, 33.9 mmol), 3,4,5-trimethylpyrazole (5.51 g, 50.0 mmol), and 5mL of xylene were combined in a 50 mL round bottomed flask fitted with a condenser and agas-tee. The reaction mixture was heated, with stirring, at 128 °C for 24 h under a dioxygenatmosphere. After this time the reaction flask was filled with a dark purple paste which wasallowed to cool and solidify. The mixture was then extracted from the reaction flask with 150mL of THF, in small portions, and unreacted cobalt metal was removed from the mixture byrepeated dredging of the slurry with a large magnet until no more cobalt metal was visible onthe magnet. The slurry was then suction filtered yielding a magenta filtrate and purplefiltercake. The filtercake was washed with several portions of THF and then air dried at roomtemperature. Thus was obtained 4.00 g of [Co(4-Medmpz)2]as a fine purple powder. Anal.calcd. forC12H80N4:C 52.0, H 6.5, N 20.2; found: C 52.2, H 6.5, N POLY-BIS(t-4-CHLORO-3,5-DIMETHYLPYRAZOLATO-N,N’)-COBALT(U)[Co(4-Cldmpz)2]A cobalt metal disc (4.623 g, 78.44 mmol) and 4-chloro-3,5- dimethyl pyrazole (2.7 g,21 mmol) were combined in a small flask fitted with a condenser. The reaction mixture washeated, with stirring, at 130 °C for 48 h in a dioxygen atmosphere. This resulted in formationof light purple solid and a blue, molten ligand soluble component. At cessation of thereaction, the mixture was allowed to cool and solidify. The mixture was then dried undervacuum for 20 h at room temperature after which time it was extracted from the flask with 20250mL of dry THF under a dinitrogen atmosphere. The mixture was suction filtered and thefiltercake washed with 50 mL of dry THF under a dinitrogen atmosphere. This yielded a bluefiltrate and a light purple filtercake. The filtercake was dried under vacuum at roomtemperature for 1 h. Thus was obtained 0.180 g (57%) of [Co(4-Cldmpz)2]as a fine purplepowder. Anal. calcd. forC10H212oN4:C 37.8, H 3.8, N 17.6; found C 37.5, H 3.8, N POLY-BIS(i-4-BROMO-3,5-DIMETHYLPYRAZOLATO-N,N’)COBALT(ll)[Co(4-Brdmpz)2]A cobalt metal disc (4.566 g, 77.48 mmol) and 4-bromo-3,5-dimethyl pyrazole (6.79g, 38.8 mmol) were combined in a small flask fitted with a condenser. The reaction mixturewas heated, with vigorous stirring, at 130 °C for 40 h in a dioxygen atmosphere. This resultedin formation of a light purple solid and a blue, molten ligand soluble component. At cessationof the reaction, the mixture was allowed to cool and solidify. The mixture was then driedunder vacuum for 24 h at room temperature after which time it was extracted from the flaskwith 20 mL of dry THF under a dinitrogen atmosphere. The mixture was suction filtered andthe filtercake washed with 20 mL of dry THF under a dinitrogen atmosphere. This yielded aturquoise filtrate and a light purple filtercake. The filtercake was dried under vacuum at roomtemperature for 24 h. This yielded 0.3 12 g (8 1%) of [Co(4-Cldmpz)2jas a fine purplepowder. Anal. calcd. forC10H2Br2oN4:C 29.5, H 3.0, N 13.8; found C 29.8, H 3.0, N BIS[DI-ji-3,5-DIMETHYLPYRAZOLATO-N,N’-CHLORO-(3,5-DIMETHYLPYRAZOLE)COBALTATE(II)ICOBALT(II)[Co(4-Hdmpz)( l)(4-HdmpzH)]o[Co(acac)2]4(0.55 1 g, 2.14 mmol) and 3,5-dimethylpyrazole were combined in asmall flask fitted with a condenser. The mixture was heated, with stirring, at 160 °C for 2 h251under a dinitrogen atmosphere. During this time, a small amount of light purple solid formed.After this period of heating, the mixture was cooled to room temperature and solidified. Thesolid was extracted from the flask with 40 mL of CH21. This resulted in a purplesuspension which was suction filtered yielding a light purple filtercake and a pink filtrate.The solvent was removed from the filtrate under vacuum yielding a light pink solid. Thepink solid was loaded into a reaction vessel in which it was heated at 130 °C (above themixture’s melting point), without stirring, for 18 h. During this heating period the reactionsolution was exposed to a slow flow of dinitrogen gas carrying CH21 vapour. By the end ofthe reaction time -.-95% of the excess 3,5-dimethylpyrazole present had sublimed and liquidacetylacetone was visible above the sublimed material. The sublimate was manuallyseparated from the remainder of the reaction mixture. The reaction mixture consisted of adark purple plug at the bottom of the reaction vessel. From the plug, product was extracted bywashing with 80 mL of acetone in small portions. The initial washings were dark green.Subsequent washings were dark blue. The final 30 mL of washings were suction filteredyielding a dark blue filtrate, and a filtercake which consisted primarily of large, well-formedpurple crystals and a few light blue rod shaped crystals. The rod shaped crystals wereremoved from the filtercake by washing with methanol. The fikercake was then dried in airand this yielded 0.142 g (23%) of [Co(4-Hdmpz)( l)(4-HdmpzH)jo as dark purplecrystals. Anal. calcd. forC30UCloN12:C 43.9, H 5.4, N 20.5; found: C 44.2, H 5.5, N20. DIMERIC/OLIGOMETALLIC COBALT(U) 3,5-DIMETHYLPYRAZOLATE-3,5-DIMETHYLPYRAZOLECo(4-Hdmpz)2.0.34 (4-Hd pzH)CoCO3.xH2O(—0.1 g) and 3,5-dimethylpyrazole (—4 g) were combined in a 10 mL252round bottomed flask fitted with a reflux condenser. The mixture was heated at 100 °C andsufficient acetone was added to the mixture in order to dissolve the 3,5-dimethylpyrazole(—1-2 mL). The resulting mixture was allowed to react, without stirring, for 40 h. Atcessation of the reaction, the mixture was allowed to cool to room temperature and solidify.The solid was then extracted from the flask by washing with a total of 80 mL of acetone insmall portions. The extract consisted of a mixture of the acetone solution of3,5-dimethylpyrazole, large dark purple crystals, and a small amount of light purple powder.The crystals were separated from the powder by suspension of the powder in acetone anddecantation. This yielded 0.18 g of crystals with the nominal formulationCo(4-Hdmpz)2.0.34 (4-H pzH). Anal. calcd. forC117H68oN47:C 49.9, H 6.0, N 23.3;found: C 49.9, H 6.2, N,5-DIMETHYLPYRAZOLATO-N,N’)-TETRACOBALT(II)[Co4(4-Hdmpz)60]All manipulations were performed in a dinitrogen atmosphere. [Co(CH3N)6][BF412(0.148 g, 0.309 mmol), prepared by the method of Hathaway, Holah, and Underhill (125),and triethylamine (0.304 g, 3.00 mmol) were combined in 5 mL of dry acetonitrile. To thissolution was added a slurry of [Zn(4-Hdmpz)2(4-HdmpzH)1 (0.211 g, 0.300 mmol) in 20 mLof dry acetonitrile. The mixture was then stirred for 2 h at room temperature which resultedin formation of a dark bluish purple solution and a small amount of light purple solid. Themixture left standing for four days at room temperature in a closed vessel. During this timelarge, well formed bluish purple crystals grew in the solution. A few of these crystals werecollected for X-ray diffraction analysis and they were thus identified as [Co4(4-Hdmpz)601.2537.2.6 NICKEL PYRAZOLATE7.2.6.1 POLY-BIS(i.t-PYRAZOLATO-N,N’)NICKEL(II)[Ni(4-Hpz)2]Nickel metal powder (--1 g) and pyrazole (—5 g) were combined in a 100 mL roundbottomed flask fitted with a condenser. The mixture was heated in air, with stirring, at 110 °Cfor 48 h. The reaction mixture was then allowed to cool and solidify. The excess pyrazolepresent was dissolved by adding CH21 to the mixture. The resulting slurry was thendredged with a large magnet to remove unreacted nickel metal powder and then the mixturewas suction filtered and washed with CH21 and ethanol. The filtercake was dried at 110 °C,in air, for 18 h. This yielded —1 g of [Ni(4-Hpz)2jas a fine yellowish-orange powder. Anal.calcd. forC6HN4i: C 37.4, H 3.1, N 29.0; found: C 37.7, H 3.1, N ZINC PYRAZOLATES7.2.7.1 BIS(i-3,5-DIMETHYLPYRAZOLATO-N,N’-3,5-DIMETHYLPYRAZOLATO-3,5-DIMETHYLPYRAZOLE)DIZINC(II)[Zn(4-Hdmpz)2(4-HdmpzH)]Zinc metal shot (3.50 g, 53.6 mmol) and 3,5-dimethylpyrazole (3.34 g, 34.8 mmol)were combined and heated, with stirring, at 90 °C under dioxygen in a 10 mL round bottomedflask, fitted with a condenser. Sufficient acetone was added dropwise to the reaction mixturefor the 3,5-dimethylpyrazole to dissolve completely, but without allowing the resultantsolution to reflux (—3 mL acetone). After approximately 2 h a white precipitate was visibleand this accumulated during the reaction. The reaction was allowed to proceed for 45 h.254Acetone was added periodically throughout the reaction to maintain the 3,5-dimethylpyrazolein solution. The reaction mixture was then allowed to cool to room temperature and solidify.The mixture was extracted from the flask with 40 mL of acetone in small portions, theresulting slurry suction filtered, the flitercake washed with a further 200 mL of acetone andthen dried under vacuum at room temperature for —30 mm yielding 0315 g (99% yield basedon zinc reacted) of [Zn(4-Hdmpz)2(4-HdmpzH)] as a white powder. Anal. calcd. forC30HN12Zn2:C 51.2, H 6.3, N 23.9; found: C 50.9, H 6.4, N 23.9.Crystals of [Zn(4-Hdmpz)(4-HdmpzH)1 suitable for single crystal X-ray diffractionstudies were obtained by reacting zinc metal with neat molten 3,5-dimethylpyrazole, withoutstirring, in the presence of limited dioxygen. A 2 g piece of zinc metal shot was covered withexcess 3,5-dimethylpyrazole in a test tube. The tube was flushed with N2 gas and then sealedwith Paraflim into which a pinhole was