Reaction of [NbCl3(TMEDA)]2 with LDA in a stoichiometric ratio of 1:3 Nb/Lithium amide, resulted in a new diamagnetic niobium (II) cluster (TMEDA)3Nb2Cl5Li (2.1) with a very short Nb≡Nb bond. Similarly, reaction of NbCl4(THF) 2 with 2 equiv. of Cy2NLi caused the reduction of the metal center to produce a diamagnetic and tetranuclear cluster, [Nb4Cl 12Li2(THF)8] (2.2). Further reaction of [NbCl3(TMEDA)]2 or (TMEDA)3Nb2Cl 5Li (2.1) with (3,5-Me2Ph)(Ad)NLi in THF yielded the tetravalent, dinuclear and diamagnetic {[(3,5-Me2Ph)(Ad)N]Nb(3,5Me 2Ph)}2(mu-NAd)2.2(ether) (2.3). In the reaction of Nb2Cl6(TMEDA)2 with azobenzene cleavage of the N=N bond results in the formation of the niobium imido complex, NbCl3(C6H5N)(TMEDA) (2.4). The reaction of Nb2Cl6(TMEDA)2 with the potassium salt of tripyrrolylmethane in a 1:1 ratio afforded two products which have been isolated and characterized. The first compound is the dinuclear {[H-C(C4H3N)3]Nb(THF)}2 •2THF (3.1) with two trivalent and diamagnetic metals bridged by two tripyrrolyl trianions. The second product of the reaction is {([H-C(C 4H3N)3]2NbK)2}{Nb4 Cl11[K(THF)2]2}•2THF (3.2 ) which displays some unusual features. The complex is ionic with the cationic {Nb4Cl11[K (THF)2]2} 2+ unit containing the metal in a mixed valence state and the cationic moiety are balanced by two identical anionic {([H-C(C4H3N) 3]2NbK)2}- units each containing Nb in the tetravalent state. A similar reaction carried out with the lithium salt of the tripyrrolide anion led instead to a monomeric tetravalent {[H-C(C 4H3N)3]2Nb}{Li (THF)4} 2 (3.3) complex. Reaction of Nb2Cl5Li(TMEDA)3 with Ph 2NK in THF afforded a complex reaction from which three complex were isolated and characterized. The first compound Cl2Nb 2(mu-Cl)3(TMEDA)2 (4.1) was obtained in analytically pure form as dark purple, paramagnetic and moderately air-sensitive crystals. The second compound is a trivalent and dinuclear {[(Ph2N) 2Nb]2[mu-NPh(mu-eta1:eta2-C 6H4)](mu-H)}{Li(TMEDA)2}.toluene (4.2 ) which was formed via oxidative addition of the two Nb(II) centers to the C--H sigma-bond of one aromatic ring. This species thermally rearranges into the dinuclear and diamagnetic (Ph2N) 2Nb{[mu-NPh(eta6-C6H5)] 2Nb}.ether (4.3). (Abstract shortened by UMI.)

The effect of the support in heterogeneous catalysis is very important but often poorly understood. The reaction of tetraneopentylchromium(IV) with Sylopol 952, a silica gel used industrially as a carrier for alpha-olefin polymerization catalysts, has been investigated and compared with the previously reported reaction on Aerosil 200, a fumed silica often used to model catalyst supports. Grafting, thermolysis, and metathetical exchange reactions of chromium(IV) neopentylidenes were found to be the same on either Aerosil 200 or Sylopol 952. However, reactivity of the grafted organochromium fragments towards ethylene was found to be quite different on the two supports. The research described in this thesis also explores the polymerization activity of bis(neopentyl)chromium(IV) fragments at higher ethylene pressures than were previously investigated. This result has greatly altered thinking about the nature of the active site for olefin polymerization over surface organochromium fragments. The supported bis(neopentyl)chromium(IV) is no longer believed to require thermal activation to create chromium alkylidene active sites. A possible mechanism is postulated. Propylene polymerization by the Aerosil-supported chromium alkylidene was observed as a slow pseudo-first-order reaction. The kinetics were examined at the gas-solid interface by in situ FTIR spectroscopy and the second-order rate constant compared to those previously measured for ethylene and 1-hexene.

In aqueous solution, Hg2+ forms both 1:1 and 1:2 complexes with the sulfite anion. Hg(SO3)22- is redox-stable, but when one SO32- ligand dissociates, the redox-unstable HgSO3 is generated. With Hg 2+(aq) in excess, HgSO3 was observed for the first time. This complex undergoes a highly temperature sensitive intramolecular redox reaction which is kinetically first-order. The activation parameters are consistent with a unimolecular bond cleavage. In the presence of excess sulfite, Hg(SO3)2 2- is formed. Under these conditions, the rate of reduction of Hg2+(aq) to Hg0 depends inversely on the concentration of uncoordinated sulfite, yet remains unaffected by the sulfite liberated through dissociation. Using the kinetics of decomposition of Hg(SO3)22-, we have re-evaluated the sequential binding constants for sulfite to the mercuric ion and find a value for K1 that is four orders of magnitude greater than K 2, which differs from previous estimates in which they were similar in magnitude. The new values greatly alter speciation predictions for atmospheric Hg in global models. Rather than the major mercury species in clouds being stable Hg(SO3)22, we predict that HgSO 3 should predominate in both unpolluted and polluted (by SO2) air. At 25°C and pH 3, HgSO3 decomposes to Hg0 in 5 minutes. If this is in fact the major reduction route for mercury, as has been proposed, very little mercury should be present in rainwater. We find evidence for the existence of a highly soluble Hg0·SO 2 complex. According to our results, the presence of sulfite causes the solubility of Hg(0) to increase by at least three orders of magnitude due to its formation of an SO2 adduct. This has serious implications for atmospheric deposition, since it may be that the major mercury species in precipitation is the elemental form.

Treatment of compounds with general formula RuCl2(PP)(PPh 3) (PP = dppb (3a); binap (3b); dcypb ( 3c)) or [RuCl2(PP)]2 (PP = dppb (4); dcypb (5)) with PhCHN2 generates alkylidene complexes of the type RuCl2(PP)(CHPh) (2a--c), the first such systems to exhibit high ROMP activity without halide or phosphine abstraction. Low polymer polydispersity is found in ring-opening metathesis polymerization (ROMP) of norbornene via these catalysts. Investigations into the role of dissociated PPh3 show that free phosphine acts as a poison, dramatically retarding the rate of polymerization via 3 relative to catalysts derived from 4 or 5. Chelate retention is suggested by the narrow polydispersities and increased cis olefin content of the polymers obtained, and is consistent with the detrimental effects of phosphine scavenger. Reaction of RuCl2(PPh3)3 with bis(dicyclohexyl)-1,4-phosphinobutane (dcypb) under N2 gives access to PPh3-free RuCl(dcypb)(mu-Cl) 3Ru(dcypb)(N2) 5 in which the presence of N 2 and dative chloride bridges stabilize two equivalents of "RuCl 2(dcypb)". Under Ar or vacuum atmosphere, decomposition occurs via Ru-promoted dehydrogenation of the dcypb ligand, while reaction with chlorinated solvents rapidly yields paramagnetic RuCl(dcypb)(mu-Cl)3RuCl(dcypb) 6. However, the N2-stabilized species 5 is easily handled under N2 in non-chlorinated solvents, giving an ideal entry point into Ru-dcypb chemistry. The N2 ligand within 5 is readily displaced under H2 or CO atmosphere, yielding RuCl(dcypb)(mu-Cl)3Ru(dcypb)(H2) 6 or RuCl2(dcypb)(CO)2, the latter as a mixture of ccc (8) and tcc (9) isomers. Interestingly, the CO ligands can be displaced, and the symmetrical dimer RuCl(dcypb)(CO)(mu-Cl)2RuCl(dcypb)(CO) 10 is formed by decarbonylation of 8/9 on prolonged storage under N2. The incorporation of the stable silylene ligand 1,3-di-tert-butyl-1,3,2-diazasilol-2-ylidene (SiLN2) in to the ligand scaffold of a metathesis catalyst was undertaken to probe its utility as a phosphine mimic. Reaction of [(dcypb)ClRu(mu-Cl)3Ru(dcypb)(N2)] (5) with 4 equivalents of the silylene yields coordinatively unsaturated RuCl(eta 3-dcypb)(SiLN2) (12). Complex 12 is a rare example of a trans-spanning diphosphine complex, this geometry resulting from an unprecedented attack of the metal on the tetramethylene ligand backbone. (Abstract shortened by UMI.)

This thesis is divided into 5 major sections. The first section covers the reactions of Ti(NR2)4 (R is methyl or ethyl) with the surface of partially dehydroxylated silica. This leads to the formation of well-defined grafted titanium complexes. The stoichiometry of the grafting reactions depends upon the degree of dehydroxylation of the silica. When the silica is pretreated at 500°C, ≡SiOTi(NR2)3 is formed, while on silica which has been dehydroxylated at 200°C, (≡SiO) 2Ti(NR2)2 is the product. The second section examines the spontaneous transformations experienced by the bis(dialkylamido)titanium fragments upon standing or exposure to vacuum. beta-H abstraction results in elimination of one equivalent of amine, leaving an eta 2-imine ligand coordinated to titanium. While the N-methylmethanimine complex is stable, the N-ethylethanimine complex undergoes a further loss of ethylene to yield a stable ethylimidotitanium(IV) surface complex. These transformations are catalyzed by the residual surface hydroxyls. The third section describes the reactivity of the bis(dialkylamido)titanium fragments towards amines. Exposure to ammonia induces beta-H abstraction, producing 1 equivalent of dialkylamine and resulting in the formation of eta 2-imine complexes instead of the expected transamination products. The N-ethylethanimine complex, under prolonged vacuum, loses ethylene to produce ethylimidotitanium(IV). Other amines, including dialkylamines and aprotic amines such as pyridine and NEt3, induce the same imine complex formation. The fourth section deals with the reactions of the imine and imido complexes with Ti(NR2)4. Both reactions result in grafting of an additional 2 wt. % Ti onto the silica, in the absence of surface hydroxyls and the formation of one equivalent of the corresponding amine. Dinuclear surface complexes containing bridging imine, imido, or siloxo ligands are proposed as products. The final section involves the characterization of the surface complexes produced upon exposure of partially dehydroxylated silica to Mo(NR2) 4 (R is methyl or ethyl). When R is methyl, grafting on silica pre-treated at 200 or 500°C results in reactions analogous to Ti(NMe2) 4. However, Mo(NEt2)4 displays a reactivity that is significantly different from its Ti analogue. The maximum Mo loading is double that expected for the number of surface hydroxyl groups available. Furthermore, the amount of diethylamine released is 1.5 times higher than expected (treatment of silica at 500°C) or twice as high (silica pre-treatment at 200°C). Spontaneous beta-H abstraction and elimination of diethylamine is proposed to form an N-ethylethanimine complex of molybdenum(IV), which then reacts further with Mo(NEt2)4 to form a dimolybdenum complex with bridging imine ligands.

Layered double hydroxides (LDH) containing NiAl or NiV of varying M(II):M(RI) (2:1, 5:1, 9:1 for NiAl and 2:1, 5:1 for NiV) ratios have been prepared by coprecipitation and characterized by XRD, elemental analysis, and FT-IR. The LDHs electrochemical properties were investigated in half-cells to determine their potential application as positive electrode in Ni-Cd and NiMH batteries. The Ni2Al LDH exchanges the most electrons (up to 1.6) and is the most stable during discharging in the KOH electrolyte. The large quantity of electrons and accrued stability was attributed to the aluminium in the layers acting in a two fold manner: High charge to radius ratio increased the electrostatic interaction between the anions and the metal layers increasing the stability of the LDHs and the increase in Bronsted acidity of the hydroxyl groups was accredited for the high exchange of electrons. Notwithstanding the high exchange of electrons of the NiAl LDHs the powders exhibit lower discharge capacity compared to commercial electrode material (beta-Ni(OH) 2) due to their low density. The NiV LDHs only exchanged up to 1.2 electrons (Ni2V) and were found only stable up to a maximum of 14 days in electrolytic solutions of the cells. In the second part of the study ZnAl LDHs were synthesized and intercalated with phenyl phosphonic acid (PPA) or 1,4-phenylene bis phosphonic acid (B-PPA) in order to create microporous materials. The compounds were characterized by XRD, chemical, IR and solid state NMR analysis in addition to being tested for microporosity. The ZnAl LDHs treated with phenyl phosphonic acid gave rise to materials with a d spacing of 14.7A and when treated to B-PPA resulted in materials with a d spacing of 9.6 A. The results showed that the grafting of both phosphonates to the metal layers had occurred. The d spacing of 14.7 A was interpreted as being due to a bilayer of PPA molecules in head to head configuration in the layers. The d spacing of 9.6 A was consistent with a monolayer of B-PPA molecules cross-linking the metal layers of the material. Both materials exhibited little or no microporosity due to the "packing" of the phosphonate moieties in the layers. The last section of the thesis consisted of attempting to synthesize MCM or FSM type molecular sieves by the treatment of ZnAl LDH with hexadecane sulfonic acid or sodium dodecyl sulfate used as templates under different reaction conditions. The materials were characterized by XRD, chemical analysis and IR analysis. The results were consistent with LDHs intercalated with either a monolayer or a bilayer arrangement of surfactant molecules and therefore did not exhibit MCM or FSM mesoporous frameworks.

We present a local moment mean field theory for atomically disordered magnetic alloys. In this model, the level of approximation regarding atomic environment can be decreased gradually. The method applied to one-dimensional alloys at several levels of approximation. The results for the sample average spin indicate a convergence as the level of approximation is decreased. Other interesting results for 1D alloys are obtained and discussed. The method is also applied to the fcc Fe-Ni alloy series, with only three free parameters, the exchange integrals $\rm J\sb{Fe-Fe},\ J\sb{Fe-Ni}$ and $\rm J\sb{Ni-Ni}.$ Although the most extreme behavior of the purely magnetic properties of the Fe rich alloys is not reproduced, deviations from the behavior of normal ferromagnetic alloys are predicted, and the order of magnitude of the calculated paraprocess susceptibility is in agreement with experimental data. Considering the simplicity of the model, our results indicate that a more realistic local moment model (as opposed to a more complicated model based on itinerant electron ferromagnetism) could reproduce the purely magnetic properties of fcc Fe-Ni. Some suggestions to render the model more realistic for fcc Fe-Ni are provided. Finally, a simple approach to include magnetocrystalline coupling to the model is suggested, which could be used to explain Invar behavior in fcc Fe-Ni alloys.

Abstract Not Available.

A shock tube coupled to a laser-schlieren detection system in conjunction with computer simulation was used to study the formation of small clusters in nearly saturated Freon-11 vapour. When a shock wave passed through Freon-11 vapour a temperature and pressure jump was produced. A new equilibrium cluster distribution at the new temperature and pressure was reached by formation and stabilization of clusters, this being an exothermic reaction. This process is accompanied by a temperature and density change. The laser-schlieren technique records a signal proportional to the density gradient. In the experimental studies the initial pressure of CCl$\sb3$F was varied from 26 mm Hg to 762 mm Hg and the speed of the shock wave was changed from 141.1 m/s to 320.7 m/s. Several carefully chosen gases (Ar, CH$\sb4$, CCl$\sb4$ and SF$\sb6$) were used for comparison studies. The time scale for the observed process was 10-100 $\mu$s. Characteristic of bond formation, the observed process displayed a negative activation energy suggesting a strong competition between cluster formation, dissociation and stabilization. Oscillatory signals were observed in those cases where the initial pressure was high, the postshock temperature was low and the postshock pressure was high. It was found that the oscillatory signal was biperiodic at an early stage. Then this oscillatory signal became a chaotic one. The experimentally reproducible chaotic signals are proof that the processes occurring in our system are nonlinear kinetic processes. A computational model for the rate of formation of clusters (cluster size 5) based on the scheme $\rm A\sb{n-1} + A = A\sbsp{n}{*}$ and $\rm A\sbsp{n}{*} + M = A\sb{n} + M$ was built and constrained by experimental observation, thermodynamics and nonlinear kinetics. With the help of computer simulation it was confirmed that the negative signal observed is caused by the cluster formation process followed by exothermic cluster collisional stabilization. The oscillation is a direct result of thermal feedback affecting the equilibrium cluster distribution.

The intermetallic phase, Na$\sb4$Sn, can be synthesised by a solid state reaction starting from stoichiometric quantities of the elements. It crystallizes in the orthorhombic space group Pnam with 4 asymmetric units per unit cell (a = 9.744(3)A, b = 22.751(10)A, c = 5,5390(15)A). The reactivity of Na$\sb4$Sn with organic and inorganic halides, elements and organic cations was investigated. The interconversion between and $\rm P\sb{21}\sp{3-}$ and P$\sb \sp{2-}$ has been demonstrated. Coordinating solvents, such as THF and pyridine, stabilise the more reduced P$\sb{21}\sp{3-}$ while the less reduced P$\sb \sp{2-}$ is favored when a non-coordinating solvent like EtOH is used. The reaction of P$\sb{21}\sp{3-}$ with Bu$\sb3$SnCl yields the new polyphosphides P$\sb \sp{2-}.$ The structure of this new anion was deduced from $\sp{31}$p studies.