The reaction of [Ru2(μ-O2CR)4L2]+ with π-acid ligands, such as phosphines, leads to the "disassembly" of the dimer and Ru-Ru bond cleavage. Monophosphines lead to a disproportionation reaction producing an oxidized Ru2(III,III) μ-oxo complex of the form Ru2(μ-O)(μ-O2CR)2(η2-O2CR)2(PR3)2, and a reduced mononuclear Ru(II) complex of the form trans-Ru(η2-O2CR)2(PR3)2. The reaction with diphosphines (dp) appears to lead to a reductive disassembly forming octahedral complexes of the type [cis-Ru(II)(η2-dp)2(η2-O2CR)]+. Reaction with the oxalate dianion, C2O42- (ox), leads to a surprising oxidative disassembly and mononuclear Ru(III) complexes having the formula [trans-Ru(III)(η2-C2O4)2(solvent)2]-. The investigation of these types of reactions forms the largest project currently being pursued in our lab and we are exploring three aspects of this reaction chemistry:
a) Reaction of the [Ru2(μ-O2CR)4L2]+ Core with Chiral Diphosphines: We have investigated numerous reactions of the [Ru2(μ-O2CR)4L2]+ core with diphenylphosphines (dpp) of the form Ph2P(CH2)xPPh2 (x = 1-3). These disassembly reactions offer a far superior route to [cis-Ru(η2-dpp)2(η2-O2CR)]+ complexes than the "conventional" method of reacting the carboxylate directly with cis-Ru(η2-dpp)2Cl2, in that they result in much higher yields and can incorporate much bulkier R groups. For example, we were able to synthesize [cis-Ru(η2-dpp)2(η2-O2CMc)]+ compounds (Mc = ferrocenyl or ruthenocenyl) in > 70% yields, whereas with the direct method, no significant amount of product was obtained.
To date, all of the [cis-Ru(η2-dpp)2(η2-O2CR)]+ complexes we have made and crystallographically characterized were from reactions with the achiral diphosphines mentioned above. These have all yielded racemic mixtures of Δ and Λ octahedral products. Current studies involving the disassembly reactions of [Ru2(μ-O2CCH3)4(H2O)2]+ with the chiral diphosphines S,S- and R,R-Chiraphos (S,S- and R,R-Ph2PCH(CH3)CH(CH3)PPh2) have yielded very interesting results. The disassembly reaction with S,S-Chiraphos yields exclusively the Λ product, [Λ-cis-Ru(η2-S,S-Chiraphos)2(η2-O2CCH3)]+, while reaction with R,R-Chiraphos yields exclusively the Δ product, [Δ-cis-Ru(η2-S,S-Chiraphos)2(η2-O2CCH3)]+. Evidence for these "chiral induction" (induction of chirality at a metal center by a chiral ligand) products comes from the X-ray structures of the two complexes, which verify the absolute configuration, as well as their circular dichroism spectra which are exact "mirror images" of each other. The NMR spectra also confirm the presence of only one isomer in each case. Transfer of chirality to a metal center using a chiral ligand has been investigated by von Zelewsky with the best known example being the use of a chiral tetradentate N-donor heterocycle known as Chiragen, which induces the formation of the Δ-isomer, Δ-Ru(Chiragen)Cl2, when reacted with RuCl2(CH3CN)4. Chiral induction reactions employing chiral aminophosphines and BINAP (bis(diarylphosphino)-1,1-binaphthyl) have also been studied, but there appear to be no examples of chiral induction via the disassembly of an achiral dimer.
We are currently generalizing these reactions to see if other chiral diphosphines also result in "chiral induction" at the metal center, either completely or partially. For example, does R-Prophos (R-Ph2PCH2CH(CH3)PPh2), in which there is only one chiral carbon (instead of two as in Chiraphos), lead to the formation of exclusively the Δ form (as with the R,R-Chiraphos), an enantiomeric excess of the Δ-form or no discrimination at all? Will BINAP work in our case? It is more sterically rigid than Chiraphos and we have found that the disassembly process is not as efficient with less "flexible" ligands. Two other chiral ligand types worth investigating are the Duphos family (bis(diethylphospholano)-alkanes or benzenes) and the DIPAMP family (bis[(2-methoxyphenyl)phenylphosphino]ethane) in which the phosphorous centers are chiral. We will also extend this work to include reactions with chiral diphosphinoferrocenyl ligands such as the Josiphos family (eg. (R)-1-[(1S)-2-(diphenylphosphino)-ferrocenyl]ethyldicyclohexylphosphine), as well as the disassembly of diruthenium dimers containing chiral carboxylates, such as the R- or S forms of 2-chloropropionate and 2-methylbutyrate.
b) Reaction of the [Ru2(μ-O2CR)4L2]+ Core with N-N and P-N Ligands: We have found that the reaction of [Ru2(μ-O2CCH3)4(H2O)2]+ with 2,2'-dipyridyl leads to a complete reductive disassembly to [Ru(2,2'-dipy)3]2+ with no retention of carboxylate (acetate in this case); however, when steric hindrance is introduced next to the heterocyclic nitrogens, such as in 6,6'-dimethyl-2,2'-diyridyl, the acetate is retained and [Ru(2,2'-dipy)2(η2-O2CCH3)]+ is formed.
We are currently investigating this "steric directing" reaction further in two ways, first, by varying the bulk and position of the substituents on the 2,2'-dipyridyl. For example, we will try 5,5'- and 4,4'-dimethyl-2,2'-dipyridyl, as well as 5,5'- and 4,4'-di-(tert-butyl)-2,2'-dipyridyl, and see how this effects the nature of the disassembly products as well as the more "rigid" 1,10-phenanthroline ligand along with the sterically hindered 2,9-dimethyl-1,10-phenanthroline (neocuproine). In addition, we will investigate "hybrid" P-N donor ligands, such as (2-aminophenyl)diphenylphosphine, PPh2C6H4NH2, which can be easily prepared by the reduction of the corresponding phosphine oxide with methylpolysiloxane, and iminophosphine ligands, such as the N-(2-(diphenylphosphino)-benzylidene)amines, which can be prepared in high yield from the condensation reaction of 2-(diphenylphosphino)benzaldehyde with the appropriate primary amine, to see if and how they disassemble the [Ru2(μ-O2CR)4]+ core.
c) Further Derivatization of the Oxalate Disassembled Product, trans-[Ru(ox)2(H2O)2]- : The disassembly reaction of [Ru2(μ-O2CCH3)4(H2O)2]+ in mildly acidic solution with the oxalate dianion (ox) leads to a mononuclear complex, trans-[Ru(ox)2(H2O)2]- . We were the first to isolate such a species as an [NH2(CH3)2]+ salt in 2000 using a novel dimethylamine diffusion into the acidic solution to precipitate crystals of [NH2(CH3)2][Ru(ox)2(H2O)2]. No previous X-ray characterization had been done on any ruthenium bis-oxalato species and very few of these complexes have been reported on.
We are currently generating a series of Ru-bis-oxalato derivatives, trans-[Ru(ox)2(L2)]- , by substituting the axial water ligands for various other donor ligands (L), such as dimethylsulfoxide (dmso), dimethlyformamide (dmf), pyridine (py), and triphenylphosphine (PPh3). The chemistry of ruthenium oxalates has hardly been explored. Most of the syntheses are to be carried out by reacting the [NH2(CH3)2][trans-Ru(ox)2(H2O)2] starting material with a 2-4 molar equivalent of one of the above ligands in aqueous solution at room temperature or with heating. Early indications have shown that all of these reactions lead to a colour change, and we are optimistic in that products will be isolatable. Care will be taken in the aqueous reactions with pyridine, since substantial amounts of the pyridinium cation, pyH+, will form at pH < 6.5. This will provide for some interesting chemistry as the dimethylammonium cation, [NH2(CH3)2]+, could be exchanged for pyH+ as the pH is lowered. The reactions in the neat donor solvents may also provide for additional substitution of one or both of the oxalate dianions, which will need to be monitored. The dependence of the electronic spectra and the redox potentials on the nature of the axially substituted ligand will be of particular interest. Extension of the work to generate polymers of the form {M[trans-Ru(ox)2(L-L)]}x, where M is a cation and L-L a bidentate bridging ligand, such as pyrazine, will also be explored.
It should be noted that the exact overall mechanism of "disassembly" and/or "disproportionation" of the [Ru2(μ -O2CR)4L2]+ core has not been fully elucidated, although we and others have made progress on those disassemblies mediated by mono- and di-phosphines. The oxidative disassembly involving oxalate is still a mystery, and are studying it further by varying solvent, reaction stoichiometry, pH, etc.