When (Et$\sb4$N) $\sb3$(Bi$\{$Fe(CO)$\sb4\}\sb4$) is treated with main group metal halides complicated disproportionation reactions occur. The reaction with BiCl$\sb3$ in MeCN yields (Et$\sb4$N) (BiFe$\sb3$(CO)$\sb{10}$), (Et$\sb4$N) $\sb2$(Bi$\sb2$Fe$\sb4$(CO)$\sb{13}$), and (Et$\sb4$N) $\sb2$(Bi$\sb4$Fe$\sb4$(CO)$\sb{13}$) while treatment with SnEt$\sb2$Cl$\sb2$ or CBr$\sb4$ gives the oxidation product (Et$\sb4$N) $\sb2$(Bi$\sb4$Fe$\sb4$(CO)$\sb{13}$) in high yield. Oxidation of (Et$\sb4$N) $\sb3$(Bi$\{$Fe(CO)$\sb4\}\sb4$) with 2 equivalents of (Cu(MeCN)$\sb4$) (BF$\sb4$) or MeI affords (Et$\sb4$N) (BiFe$\sb3$(CO)$\sb{10}$). (Et$\sb4$N) $\sb2$(Bi$\sb2$Fe$\sb4$(CO)$\sb{13}$) is oxidized by (Cu(MeCN)$\sb4$) (BF$\sb4$) forming Bi$\sb2$Fe$\sb3$(CO)$\sb9$ and reacts with CO (850 psi) to produce (Et$\sb4$N) $\sb2$(Bi$\sb4$Fe$\sb4$(CO)$\sb{13}$). Bi$\sb2$Fe$\sb3$(CO)$\sb9$ is reduced readily with Na/Hg forming one- and two-electron-reduction products. Two-electron-reduction can also be achieved by treating Bi$\sb2$Fe$\sb3$(CO)$\sb9$ with 2 equivalents of cobaltocene in CH$\sb2$Cl$\sb2$. The elemental analyses and spectroscopic data for the two-electron-reduction product support the formulation as (Cp$\sb2$Co) $\sb2$(Bi$\sb2$Fe$\sb3$(CO)$\sb9$). The (Bi$\sb2$Fe$\sb3$(CO)$\sb9$) $\sp{2-}$ anion can be reconverted to Bi$\sb2$Fe$\sb3$(CO)$\sb9$ in 90% spectroscopic yield when treated with (Cu(MeCN)$\sb4$) (BF$\sb4$). The reaction of (Et$\sb4$N) $\sb2$(Fe$\sb2$(CO)$\sb8$) with BiCl$\sb3$ or SbCl$\sb3$ forms compounds proposed to be (Et$\sb4$N) (EFe$\sb3$(CO)$\sb{12}$) (E = Bi, Sb) based on elemental analyses and spectroscopic data. The treatment of (Et$\sb4$N) (BiFe$\sb3$(CO)$\sb{12}$) or (Et$\sb4$N) (SbFe$\sb3$(CO)$\sb{12}$) with Cr(CO)$\sb5$(THF) produces (Et$\sb4$N) (EFe$\sb3$Cr(CO)$\sb{17}$) (E = Bi, Sb), respectively, while methylation of (Et$\sb4$N) (BiFe$\sb3$(CO)$\sb{12}$) afford Bi$\sb2$Fe$\sb2$(CO)$\sb8$Me$\sb2$. Oxidation of (Et$\sb4$N) (BiFe$\sb3$(CO)$\sb{12}$) with (Cu(MeCN)$\sb4$) (BF$\sb4$) yields Bi$\sb2$Fe$\sb3$(CO)$\sb9$ whereas the same reaction using of (Et$\sb4$N) (SbFe$\sb3$(CO)$\sb{12}$) gives Sb$\sb2$Fe$\sb6$(CO)$\sb{22}$. Refluxing (Et$\sb4$N) (BiFe$\sb3$(CO)$\sb{12}$) or (Et$\sb4$N) (SbFe$\sb3$(CO)$\sb{12}$) in acetonitrile produces (Et$\sb4$N) $\sb2$(Bi$\sb2$Fe$\sb4$(CO)$\sb{13}$) and (Et$\sb4$N) $\sb2$(Sb$\sb2$Fe$\sb5$(CO)$\sb{17}$), respectively. (Et$\sb4$N) (EFe$\sb3$Cr(CO)$\sb{17}$) (E = Bi, Sb), Bi$\sb2$Fe$\sb2$(CO)$\sb8$Me$\sb2$, (Et$\sb4$N) $\sb2$(Sb$\sb2$Fe$\sb5$(CO)$\sb{17}$), and Sb$\sb2$Fe$\sb6$(CO)$\sb{22}$ have been crystallographically characterized. (Et$\sb4$N) (EFe$\sb3$Cr(CO)$\sb{17}$) (E = Bi, Sb) displays a central main group atom bonded to one Fe$\sb2$(CO)$\sb8$ unit, one Cr(CO)$\sb5$ ligand, and one Fe(CO)$\sb4$ moiety. Bi$\sb2$Fe$\sb2$(CO)$\sb8$Me$\sb2$ contains a Bi$\sb2$Fe$\sb2$ parallelogram. A Me group is bonded to each pyramidal bismuth atom and the iron atoms are pseudooctahedrally coordinated. (Et$\sb4$N) $\sb2$(Sb$\sb2$Fe$\sb5$(CO)$\sb{17}$) has a square-planar Sb$\sb2$Fe$\sb3$ core geometry, with the external Fe(CO)$\sb4$ group bonded to each Sb atom. Sb$\sb2$Fe$\sb6$(CO)$\sb{22}$ is composed of an Fe$\sb2$(CO)$\sb6$ unit bridged by two Sb atoms. The coordination of each antimony atom is completed by bonding to an Fe$\sb2$(Co)$\sb8$ moiety.

Treating $\rm SbCl\sb3,\ SnCl\sb4.5H\sb2O,\ or\ Pb(OAc)\sb2.3H\sb2O$ with NaCo(CO)$\sb4$ forms the open complexes $\rm Sb\{Co(CO)\sb4\}\sb3,\ Sn\{Co(CO)\sb4\}\sb4$, and Pb$\rm \{$Co(CO)$\sb4\}\sb4$, respectively. The unstable antimony complex was determined to be Sb$\rm \{Co(CO)\sb4\}\sb3$ by the similarity of its infrared spectrum with that of known $\rm Bi\{Co(CO)\sb4\}\sb3$. Isostructural $\rm Sn\{Co(CO)\sb4\}\sb4$ and $\rm Pb\{Co(CO)\sb4\}\sb4$ were characterized by single crystal X-ray diffraction. Each contains a Group 14 element tetrahedrally surrounded by four trigonal bipyramidal Co(CO)$\sb4$ groups. Heating solutions of $\rm Bi\{Co(CO)\sb4\}\sb3$ forms BiCo$\sb3$(CO)$\sb9$, which contains a closed cobalt triangle capped by a bismuth atom. A bridging carbonyl lies along each Co-Co bond. The thermal decomposition is reversible, and a kinetic study of the carbonylation of BiCo$\sb3$(CO)$\sb9$ indicates this transformation to be first-order with respect to (BiCo$\sb3$(CO)$\sb9$) and P(CO) in n-hexane (500-900 psi, 50-65 C), with an activation energy of 68+/$-$2 kJ.mol$\sp{-1}$. Reducing $\rm Bi\{Co(CO)\sb4\}\sb3\ forms\ \lbrack Bi\{Co(CO)\sb4\}\sb4\rbrack \sp-$. This complex is isostructural to Sn$\rm \{Co(CO)\sb4\}\sb4\ and\ Pb\{Co(CO)\sb4\}\sb4$, but it is not isoelectronic, as the bismuth is a hypervalent 10-electron center. Reducing BiCo$\sb3$(CO)$\sb9$ produces (Bi$\sb2$Co$\sb4$(CO)$\sb{11}$) $\sp-$ and (Co(CO)$\sb4\rbrack \sp-$. Single crystal X-ray analysis (Cp$\sb2$Co$\sp+$ salt) determined the core structure of the bismuth anion to consist of a Bi$\sb2$Co$\sb2$ tetrahedron with Co(CO)$\sb3$ units capping the two Bi$\sb2$Co triangular faces. X-ray analyses of (PPN) (Sb$\sb2$Co$\sb4$(CO)$\sb{11}$) and (PPN) $\sb2$ (Sb$\sb2$Co$\sb4$(CO)$\sb{11}$) showed these complexes to have nearly identical frameworks, which are isostructural and isoelectronic to their bismuth homologues. These complexes are electron rich and do not conform to conventional bonding formalisms. Extended Huckel calculations agree with observed structural changes that upon reduction of the monoanion, the added electron enters an antibonding orbital primarily localized between two cobalt atoms bridged by a carbonyl. Treatment of BiCo$\sb3$(CO)$\sb9$ with PPh$\sb3$ gives $\rm Bi\{Co(CO)\sb3PPh\sb3\}\sb3.$ $\rm Bi\{Co(CO)\sb4\}\sb3, BiCo\sb3(CO)\sb9$, and $\rm Sb\{Co(CO)\sb4\}\sb3$ all react with PhC$\sb2$Ph through an apparent radical mechanism to form Co$\sb2$(CO)$\sb6\{$PhC$\sb2$Ph$\}$ as the only infrared observable product.

The search for a new class of compounds with quasi-low-dimensional frameworks has become important for the development of new transition metal oxides that are of magnetic and electronic importance. The exploratory synthesis of low-dimensional inorganic solids, in particular reduced niobium oxophosphates and oxosilicates containing extended, but confined, M-O-M and M-M bond interactions, has been investigated. These bonding types are common structural features observed in metallic oxides that possess unusual phenomena associated with charge density waves, superconductivity, and magnetic ordering. Ultimately, these new compounds should lead to structural models for experimental and theoretical studies of the behavior of delocalized electrons in a confined lattice. A number of significant solid state compounds with interesting structures and physical properties have been prepared through our exploratory synthesis. Four tetragonal tungsten bronze-related (TTB) mixed-metal oxides have been made; namely BaNb$\sb2\rm O\sb{6-\rm x},$ orthorhombic NbPO$\sb5,$ and ANb$\sb3\rm P\sb3O\sb15$ (A = Cs, and Ba). The niobium oxophosphate compounds, BaNb$\sb7\rm P\sb6O\sb{33}$ and CaNb$\sb2$P$\rm\sb2O\sb{11},$ with mixed-frameworks and fused NbO$\sb6$ octahedra have also been prepared and studied. Finally, a series of niobium-rich oxosilicates with the general formula (Ba$\sb3\rm Nb\sb6Si\sb4O\sb{26})\sb{n}(Ba\sb3Nb\sb8O\sb{21}$) has been identified with quasi-one-dimensional transition metal oxide chains. Various synthetic techniques have been employed to prepare the title compounds, in particular a conventional high-temperature solid-state ceramic method and a molten salt flux method. To characterize these new materials, a variety of analytical techniques have been utilized including both single crystal and powder X-ray diffraction, Weissenberg X-ray photographs, thermal analysis (TGA and DTA), magnetic susceptibility and four-probe conductivity measurements. Additional relevant techniques for characterization of inorganic solids were also used. Our study of structure/property correlations should prove insightful towards a better understanding of the conducting mechanism in technologically important metallic oxides and in the development of new materials with desired applications.

Several arsenate and silicate compounds with interesting structural features have been synthesized by the molten salt flux method. These new materials were characterized by single crystal and powder X-ray diffraction, infrared spectroscopy, magnetic susceptibility, and elemental analysis. The compounds generally have an open-framework structural network, although some also exhibit layer-like or quasi-low-dimensional features. In the arsenate family of compounds, two copper arsenates, BaCuAs$\sb2$O$\sb7$ and $\rm Na\sb7Cu\sb4(AsO\sb4)\sb5,$ as well as two isomorphous manganese arsenates, NaMn$\sb4$(AsO$\sb4)\sb3$ and KMn$\sb4$(AsO$\sb4)\sb3,$ were synthesized and characterized. In the silicate system, two isostructural copper silicate chlorides, $\rm Na\sb2Rb\sb2Cu\sb4Si\sb{12}O\sb{29}Cl\sb2$ and $\rm Na\sb2Cs\sb2Cu\sb4Si\sb{12}O\sb{29}Cl\sb2,$ were successfully identified. Our study of structure/property correlations should prove insightful toward a better understanding of the conducting mechanism in technologically important metallic/magnetic oxides and in the development of new materials with desired applications.

Investigation into the synthesis of reduced transition metal phosphates has led to the formation of two novel pseudo-ternary vanadium (III) phosphate compounds, each of which exhibits empty channels within the lattice. $\beta$-BaV$\sb2$(P$\sb2$O$\sb7)\sb2$ crystallizes in the P-1 (No. 2) space group with a = 6.269(1)A, b = 7.864(3)A, c = 6.1592(9)A, $\alpha$ = 101.34(2)$\sp\circ$, $\beta$ = 105.84(1)$\sp\circ$, and $\gamma$ = 96.51(2)$\sp\circ$ and exhibits three channels, two empty (one two-dimensional and one one-dimensional) and one occupied by the barium cation (three-dimensional). This compound is believed to be a high temperature phase. Ba$\sb3$V$\sb6$(PO$\sb4)\sb8$ crystallizes in the P2/c (No. 14) space group with a = 8.969(6)A, b = 9.907(5)A, c = 14.470(4)A, and $\beta$ = 104.26(3)$\sp\circ$. This structure also exhibits an empty, pentagonal-shaped channel visible along the a axis. In addition, the two barium sites form unusual corner sharing polyhedra through oxygen 16. Edge sharing of V(2)O$\sb6$ and P(2)O$\sb4$ polyhedra results in a structure distortion to avoid a short V-P distance. These structures appear stable through a wide temperature range.

The metal vapor synthesis of two novel bis(trifluoromethyl)gold-$\mu$-halide dimers was accomplished by the codeposition of gold atoms with freons (CF$\sb3$X; X = Br, I) at $-$196$\sp\circ$C. The reactor, the reaction procedures and conditions, and the isolation and purification techniques are described here. Solution IR, $\sp{19}$F NMR, mass spectrometry, and x-ray crystallography were used to characterize the two compounds. Both compounds showed moderate vapor pressures and were treated with thallium hexafluoroacetylacetonate, phosphorus trifluoride, and tris(trifluoromethyl)phosphine in an attempt to synthesize monomeric compounds with higher vapor pressures. The dimers were decomposed by PF$\sb3$ and did not react with P(CF$\sb3)\sb3$; neither reaction yielded soluble gold compounds detectable by mass spectrometry. Reaction with Tl(hfacac) resulted in an unstable oily product which gave a mass spectrum characteristic of bis(trifluoromethyl)gold hexafluoroacetylacetonate. This compound decomposed readily and resisted isolation. All of the isolated gold compounds were light sensitive. The bis(trifluoromethyl)gold-$\mu$-halide dimers were resistant to air oxidation.

Part I. An unusual appended-tail porphyrin, UroTPPH$\sb2$, which possesses a site of unsaturation in the tail (lending rigidity) and a terminal imidazole moiety joined at the C-4 rather than N-1, has been examined. Zinc(II) and iron(III) derivatives of the photosensitive (UroTPP)$\sp{2-}$ have been prepared and characterized to explore the potential for (and mode of chelation) of the imidazole-containing tail. The metalloporphyrin compounds are of special interest, with both neutral imidazole and charged imidazolate tail forms being possible. Characterization of these compounds have included physical measurements utilizing NMR, IR, EPR, $\sp{57}$Fe Mossbauer, and MCD spectroscopies and magnetic susceptibility studies. This study is one of the first in-depth studies of chloro and trifluoroacetate derivatives of appended-tail iron(III) porphyrin systems. After deprotonation of the imidazole moiety of the appended-tail, both high- and low-spin forms of the iron(III) derivatives have been observed. The predominant form of these high- and low-spin iron(III) derivatives in the solid state is probably best described as a dimer form, while a high-spin form predominates in solution at room temperature. These (Fe$\sp{\rm III}$(UroTPP)X) compounds with X = Cl$\sp{-}$ and CF$\sb3$CO$\sb{2-}$ have been investigated as potential precursors to bimetallic $\mu$-X synthetic analogs of derivatized active-site structures of cytochrome c oxidase. Part II. Imidazolate-bridged, mixed-metal binuclear porphyrin compounds have been reinvestigated to model the proposed imidazolate-bridged (cytochrome $a\sb3\sp{3+}$(imid)Cu$\sp{2+}$) active-site structure of resting cytochrome c oxidase. The model compounds have been derived from (M$\sp{\rm II}$(TPP)) (M = Co or Mn) and (M$\sp{\rm II}$(imidH)$\sb2$DAP) $\sp{2+}$ (M = Zn or Cu) to form binuclear compounds with (M(imid)M$\sp{\prime\prime}$) centers. A reevaluation of the manganese compounds has been necessary to clarify the oxidation state of the manganese porphyrin center and to reevaluate the magnetic properties reported for the (Mn(imid)Cu) compound. Also, the previously presumed deoxygenated forms of the binuclear cobalt(II) compounds have been actually found to be oxygenated forms of cobalt(III). New deoxygenated forms of the cobalt(II) compounds have been synthesized and characterized in this work. Characterization of the compounds have included physical measurements utilizing IR, uv-visible, and EPR spectroscopies, and magnetic susceptibility studies.

The reaction of $\rm NaAsO\sb2, TeO\sb2$, and $\rm SeO\sb2$ with methanolic solutions of KOH and $\rm Fe(CO)\sb5$ produces the isoelectronic clusters ($\rm HAs\{Fe(CO)\sb4\}\sb3\rbrack\sp{2-}, \lbrack Te\{Fe(CO)\sb4\}\sb3\rbrack\sp{2-}$, and $\rm \lbrack Se\{Fe(CO)\sb4\}\sb3\rbrack\sp{2-}$. The selenium cluster, however, is unstable and rapidly loses CO to form the closo cluster $\lbrack\rm SeFe\sb3(CO)\sb9\rbrack\sp{2-}. \lbrack Te\{Fe(CO)\sb4\}\sb3\rbrack\sp{2-}$ can also be transformed into the related closo cluster ($\rm TeFe\sb3(CO)\sb9\rbrack\sp{2-}$ by either pyrolysis or photolysis. In contrast, pyrolysis or photolysis of $\rm\lbrack HAs\{Fe(CO)\sb4\}\sb3\rbrack\sp{2-}$ yields the higher nuclearity species $\rm \lbrack As\sb2Fe\sb5(CO)\sb{17}\rbrack\sp{2-}$. The stepwise protonation of $\rm\lbrack EFe\sb3(CO)\sb9\rbrack\sp{2-}$ (E = Se, Te) allows for the preparation of either $\rm \lbrack HEFe\sb3(CO)\sb9\rbrack\sp-$ or $\rm H\sb2EFe\sb3(CO)\sb9$ in good yield. Low temperature $\rm\sp1H$ NMR studies of the protonation process suggest that the basic sites on the cluster are the iron-iron bonds rather than the main-group element. Spectral and structural characterization of these related clusters allows for the effect of cluster charge on structure to be probed. Reaction of $\rm\lbrack TeFe\sb3(CO)\sb9\rbrack\sp{2-}$ with solid CuCl produces the unusual Lewis acid-base adduct $\rm\lbrack TeFe\sb3(CO)\sb9(CuCl)\rbrack\sp{2-}$. This cluster can also be viewed as the initial step in the copper oxidation of the metal cluster, which produces the well known cluster $\rm Fe\sb3(CO)\sb9Te\sb2$. The reaction of elemental tellurium with $\rm Na\sb2Fe(CO)\sb4$ yields the unstable cluster $\rm\lbrack Fe\sb2(CO)\sb6(Te\sb2)\sb2\rbrack\sp{2-}$. Reaction of this unstable cluster with alkylating agents such as iodomethane and diiodomethane produces $\rm Fe\sb2(CO)\sb6(TeMe)\sb2$ and $\rm Fe\sb2(CO)\sb6(Te\sb2CH\sb2)$ respectively, along with elemental tellurium. Oxidation of $\rm\lbrack Fe\sb2(CO)\sb6(Te\sb2)\sb2\rbrack\sp{2-}$ with $\rm\lbrack Cu(MeCN)\sb4\rbrack BF\sb4$ gives the known cluster $\rm Fe\sb2(CO)\sb6Te\sb2$ and tellurium metal. The reaction of $\rm TeCl\sb4$ with iron carbonylate anions produces a variety of products: $\rm Fe\sb2(CO)\sb6Te\sb2, Fe\sb3(CO)\sb9Te\sb2$, or $\rm\lbrack TeFe\sb3(CO)\sb9\rbrack\sp{2-}$, depending on the conditions employed. The reactive cluster $\rm Fe\sb2(CO)\sb6Te\sb2$ was structurally characterized as a 2:1 adduct with $\rm\lbrack Et\sb4N\rbrack Cl$. The reaction of the dinuclear group VI anions $\rm\lbrack M\sb2(CO)\sb{10}\rbrack\sp{2-}$ (M = Cr, Mo, W) with $\rm Ph\sb2BiCl$ produces the metallated organobismuth clusters $\rm\lbrack Ph\sb2Bi\{M(CO)\sb5\}\sb2\rbrack\sp-$. The analogous reaction with $\rm\lbrack Fe\sb2(CO)\sb8\rbrack\sp{2-}$ generates $\rm\lbrack Ph\sb2Bi\{Fe(CO)\sb4\}\sb2\rbrack\sp-$. The reaction of the previously synthesized cluster $\rm\lbrack Ph\sb2BiFe(CO)\sb4\rbrack\sp-$ with $\rm Fe\sb2(CO)\sb9$ or $\rm Cr(CO)\sb5THF$ produces $\rm\lbrack Ph\sb2Bi\{Fe(CO)\sb4\}\sb2\rbrack\sp-$ or the heterometallic $\rm\lbrack Ph\sb2Bi\{Fe(CO)\sb4\}\{Cr(CO)\sb5\}\rbrack\sp-$ respectively. Comparison of the structural parameters seen in these clusters and other related organobismuth species provides insight into the nature of the bonding at bismuth in compounds of this type. Additionally, in contrast to previous reports, the reactivity of $\rm\lbrack Ph\sb2BiFe(CO)\sb4\rbrack\sp-$ observed in these studies suggests that the bismuth atom is the site of nucleophillic attack.

Preparations of (perfluoroaryloxy)bismuth oxide complexes have been undertaken in order to study their reactivity patterns as well as their potential use as precursors to bismuth-containing superconductors via sol gel or CVD processes. The reaction of the dimeric complex (Bi(OC$\sb6$F$\sb5)\sb3$(toluene)$\rbrack\sb2,$ prepared by reaction of BiPh$\sb3$ with HOC$\sb6$F$\sb5$ in refluxing toluene, with NaOC$\sb6$F$\sb5$ in THF yields the hexanuclear complex, $\rm Bi\sb6(O)\sb3(OC\sb6F\sb5)\sb{12}.$ Three sodium-bismuth alkoxide complexes were found. The reaction of BiCl$\sb3$ with three equivalents of NaOC$\sb6$F$\sb5$ in THF yields the tetranuclear sodium salt NaBi$\sb4(\rm O)\sb2(OC\sb6F\sb5)\sb9(THF)\sb2.$ The polymeric chain of $\rm \lbrack NaBi(OPh\sb{f})\sb4\cdot THF\rbrack\infty$ was synthesized from the reaction of (Bi(OC$\sb6$F$\sb5)\sb3$(toluene)$\rbrack\sb2$ with one equivalent of NaOC$\sb6$F$\sb5$ in THF. An octahedron of mixed metals, $\rm Na\sb4Bi\sb2(O)(OPh\sb{f})\sb8\cdot(THF)\sb4$ was synthesized from (Bi(OC$\sb6$F$ \sb5)\sb3$(toluene)) $\sb2$ and two equivalents of NaOC$\sb6$F$\sb5$ in THF. Single crystal structure determination and variable temperature NMR studies were done on theses compounds.

The reactions of transition and main group metals, including Ca, Sc, Ti, V, Cr, Mo, Zn, Ga and Ge, with molecular hydrogen have been investigated using FT-IR matrix isolation spectroscopy. The reactions of gallium with methane in Ar, Kr and pure methane matrices have also been studied. All the above metals (except Ge) encounter an energy barrier and require photoexcitation to insert into the H-H bond to form MH$\sb2.$ Bond angles for the dihydrides (MH$\sb2)$ of the above metals have been estimated through measured relative intensities of symmetric and asymmetric stretching modes. We have found that CaH$\sb2$ and ZnH$\sb2$ are linear (or near linear) molecules, while ScH$\sb2,$ TiH$\sb2,$ VH$\sb2,$ CrH$\sb2,$ MoH$\sb2,$ GaH$\sb2$ and GeH$\sb2$ are bent. The stretching force constants for MH$\sb2$ have been calculated with a diatomic approximation and are found to increase across the row (from Ca to Zn). A correlation between the stretching force constants and the size of the metals and the involvement of 3d electrons in bonding appears to exist. The bent nature of the transition metal dihydrides is also thought to be a consequence of involvement of 3d electrons in bonding. Monohydrides of V, Mo, Ga, and trihydrides of Sc, Cr, Mo have also been isolated as results of photoreactions of the respective metals with H$\sb2.$ Formation of polyhydrides (MH$\sb{\rm n},$ n = 4, 6, etc.) was also observed in the reactions of atomic Ti, and Mo with H$\sb2.$ Reactions of metal dimers and trimers with hydrogen have also been observed. Species with terminal hydrogens resulting from the insertion of the dimers of Ca, Sc, Ti, Cr, Ga and Ge into the H-H bond have also been identified as the reaction products of photolysis. Spontaneous reactions of Ga$\sb2$ with H$\sb2$ yield Ga$\sb2$H$\sb2$ with two bridging hydrogens. The trimer of gallium is thought to be isolated in Ar matrices and found to react with H$\sb2$ to form Ga$\sb3$H$\sb2$ with two terminal hydrogens upon absorption of UV light. A ligand-free molecular coordinated dihydrogen complex with the calcium dimer, $\rm Ca\sb2(H\sb2)\sb2,$ was also observed in Kr matrices. Photolysis with UV light produces methylgallium hydride, CH$\sb3$GaH, in the reactions of gallium with methane in Ar, Kr and pure methane matrices.