62nd Annual Report on Research 2017

The key objective for this project is the development of a general, straightforward methodology for the synthesis of terminal disulfide complexes. The disulfide functional group is a subject of interest because it has been implicated, for example, as the catalytically essential moiety along reactive edges of MoS₂. Lacking, however, are satisfactory general methods for preparing well-defined M(η²-S₂) small molecules that might support catalytic activity under homogeneous conditions in which informative spectroscopic study would be feasible.

Our approach to this objective has been to assess the feasibility of using bis(trialkylsilyl)disulfide molecules as delivery agents for the disulfide ligand (Scheme 1). The lynchpin to the idea is that the relatively weak Si‒S can be exploited to deliver the S₂^2- group to an appropriate L_nMⁿ complex while a much more thermodynamically favorable R₃SiCl or (R₃Si)₂O byproduct can result from ligand exchange at metal. Among choices for bis(trialkylsilyl)disulfide molecules, our attention has been given to ⁱPr₃SiSSSiⁱPr₃, in part because of the availability of ⁱPr₃SiSH. We have found, however, that a survey of the reactivity of ⁱPr₃SiSSSiⁱPr₃ with a variety of transition metal chloride complexes leads to no observable reaction. These outcomes are attributed to kinetic inertness provided by the steric profile given by the multiple ⁱPr groups.

As an alternative to the reaction type illustrated in Scheme 1, we have studied the possibility of dissecting the reaction into parts, as illustrated in Scheme 2. The platinum complex illustrated has been selected for focused study because a variant of it has been shown to support a terminal disulfide ligand. We have recently demonstrated that displacement of Cl^- from [Pt(dppn)Cl₂] (dppn = 1,8-bis(diphenylphosphinonaphthalene)) by Na^+-SSiⁱPr₃ does indeed occur to afford [(dppn)Pt(SSiⁱPr₃)₂] ((b) in Scheme 2). This product molecule has been identified by X-ray crystallography (Figure 1). Further, we find that the ^-SSiⁱPr₃ ligands in [(dppn)Pt(SSiⁱPr₃)₂] are, as intended, subject to further reactivity. Initial efforts to unmask the ⁺SiⁱPr₃ protecting groups and oxidize the resulting sulfide anion to disulfide made use of commercially available [ⁿBu₄N]⁺F^- solutions. However, while reactions have certainly observed between [(dppn)Pt(SSiⁱPr₃)₂] and [ⁿBu₄N]⁺F^-_, the reactivity has not been clean and has not yielded the intended terminal disulfide ((c) in Scheme 2). This complication is, in hindsight, attributed to the fact that commercially sold solutions of [ⁿBu₄N]⁺F^- in THF are typically ~5% H₂O.

This supposition is supported by a recent observation that use of Na⁺[OSiMe₃] as unmasking agent, although more bulky, is more clean in its reactivity and yields the di-platinum compound illustrated in Figure 2 and as (d) in Scheme 2. Mixed sulfide-chloride bridging ligands are inferred on the basis of net charge; Cl^- is presumed to have originated from the use of CH₂Cl₂ in the recrystallization of this product or from residual NaCl from the preceding step.

In continuing work on development of a protocol for use of R₃SiSSSiR₃ as delivery agents for inorganic S₂^2-, we are exploring several variants of the dppn ligand that differ in the steric profile that they project toward the other side of the metal ion. These variants include the 3,5-dimethylphenyl and the 2,4,6-trimethylphenyl derivatives. It is currently unclear how sensitive the [(phosphine)₂Pt(η²-S₂)] is toward ambient atmosphere and how much steric bulk is essential to stabilize it against a tendency to form (μ₂-S)₂ or (η²,η²,μ₂-S₂) species. If the reaction conditions and supporting ligand attributes can be identified which lead to isolation of [(phosphine)₂Pt(η²-S₂)], our studies will be extended toward gold, iridium, and mercury complexes because these elements appear most likely to support the intended ligand type.

A second theme of research outlined in the original project proposal was an examination of ways by which sulfur might be incorporated into polyoxometallate anions. The general strategy is to modify well-defined polyoxometallates with silane reagents (Scheme 3), which have been demonstrated to be effective for oxo-to-silyloxide and oxo-to-sulfido transformations in mono-metallic systems. Initial efforts focused upon [W₁₂SiO₄₀]^4-, but no decisive evidence has been developed indicating that exchange of terminal oxo ligand in [W₁₂SiO₄₀]^4- is occurring. In hindsight, what may be complicating the assessment of possible products in Scheme 3 is the occurrence of multiple isomers of [W₁₂SiO₄₀]^4-, which makes difficult the crystallization of samples well-suited for X-ray diffraction. In view of this apparent challenge, we are directing attention to [Mo₆O₁₉]^2-, which is not subject to isomerism, and we are endeavoring to accomplish simple silylation of terminal oxo ligand with [Et₃Si]⁺ according to Scheme 4. This reaction type is the first step in oxo-for-sulfido exchange, and such intermediates have been isolated in monometallic systems. Our current belief is that a reactive silylium cation, such as [Et₃Si]⁺[B(C₆H₃-3,5-(CF₃)₂)₄], should silylate a terminal oxo ligand in [Mo₆O₁₉]^2- and sidestep any complicating undesired redox chemistry. The [Et₃Si]⁺ cation has no capacity to reduce the [Mo₆O₁₉]^2- cluster and instead can only leave it as fully oxidized, diamagnetic, and mono-negative, which will facilitate product identification by ¹H NMR spectroscopy and mass spectrometry.

This research has motivated both the PI, James P. Donahue, and the primary student involved, Ms. Malathy Selvachandran, to first of all be mindful of the importance of orienting chemical research to applied ends and to find the ways to articulate a justification for fundamental chemistry in terms of an objective that is useful. The beginning of every presentation on this project has required the student to explain why it is that terminal disulfide complexes are important and why it is that there is value in devising more efficient ways toward them. Further, this research has brought focus to the need to be broadly read in the generally relevant area and to be innovative in finding alternative pathways to a synthetic target if one’s first choice does not work. For the PI, this project has been significant in that it is the first time he has undertaken and thought about a synthetic methodology problem as opposed to a targeted synthesis problem.