Transition Metal Catalyzed Selective Oxidation of Sugars and Polyols

Oxidation reactions of sugars to their corresponding lactones and
the selective oxidation of a secondary alcohol of sugar substrates
to keto-sugars using transition metal catalysis were investigated.
Research results demonstrate that Shvo’s hydrogen-transfer
catalyst, [(η4-C4Ph4CO)(CO)2Ru]2, selectively oxidized unprotected
sugars to δ-lactones under very mild conditions. This is the first
method to synthesize δ-D-galactonolactone, the first oxidation
product of the 3rd most abundant sugar. δ-D-Galactonolactone was
fully characterized by conformational analysis through NMR
experiments and computational methods. We propose that in aprotic,
anhydrous solvents the δγ lactone rearrangement mechanism proceeds
via a bicyclic transition state. The oxidation of alicyclic vicinal
diols to α-hydroxy ketones was investigated as a model system for
the selective oxidation of a secondary hydroxyl group in sugar
substrates to keto-sugars. The challenge was to develop a method
that allows selective oxidation, yet prevents over-oxidation to the
dione or dicarboxylic acids products. Based on a mathematical model
of consecutive bimolecular reactions, the rate constants for the
initial oxidation step k1(diolα-OH ketone) and second oxidation
step k2(α-OH ketonedione) must be k1 > 10 k2 in order to obtain
a synthetically useful method for α-hydroxy ketone synthesis. The
metal-ligand bifunctional hydrogen transfer reactions of Shvo’s
catalyst and Noyori’s η6-arene N-tosyl-1,2-diaminoethane
ruthenium(II) complexes were investigated with alcohol model
systems and it was concluded that hydrogen transfer reactions are
insufficient in the oxidation of vicinal diols due to a
unfavourable position of the equilibrium. Under oxidizing
conditions, the 16-electron ruthenium complexes of the Noyori
systems compete with a β-elimination process and thus new
degradation resistant ligands were synthesized. The apparent slower
oxidation of Noyori’s η6-arene N-tosyl-1,2-diaminoethane
ruthenium(II) complexes under oxidizing conditions, i.e. in acetone
or cyclohexanone solvent, was investigated through NMR and IR
experiments finding no evidence of a kinetic inhibition by the
solvent. We propose that the slower reactions depend on a
relatively high energy barrier for the reaction of the 16-electron
complex with a hydrogen donor and that a kinetic model must account
for two effectively different oxidation and reduction catalysts. In
an alcohol/ketone equal-concentration experiment with Noyori-type
ruthenium(II) complexes a linear relationship was found between the
initial rate of alcohol consumption/production and ∆G°. Reactions
involving peroxides as the oxidant were investigated in order to
avoid equilibrium processes. The oxidation of alcohol model systems
with several Mo and W catalysts and peroxide sources indicated a
deactivation of the Mo and W catalysts by formation of water or
hydroxide complexes. The product distribution of the oxidation of
trans-1,2-cyclohexanediol with the MoO2(acac)2/Na2CO3 x 1.5 H2O2
method was in agreement with the theoretical model, yet only had a
k1 = 1.5 k2 rate ratio resulting in a maximum α-hydroxy
cyclohexanone content of 45 %. The NiBr2 mediated oxidation of
alcohols and benzoylperoxide was investigated. The selective
oxidation of vicinal diols was unsuccessful with this method.
However, the oxidation reaction of mono-alcohol substrates was
greatly improved using water in the reaction. The reaction
mechanism was investigated and we propose that the actual oxidant
in the NiBr2 mediated benzoylperoxide method is [Br+], which is
generated by the hydrolysis product of benzoylhypobromite,
hypobromic acid (HOBr). Ishii’s stoichiometric NaHSO3/NaBrO3
reagent selectively oxidized vicinal diols to α-OH ketone without
overoxidation. This oxidation reaction was investigated with
regards to substrate, concentration and pH. The reactivity and
selectivity were studied and it was found that the oxidation
mechanism is based on a multitude of comproportionation and
disproportionation equilibria at low pH. A small but steadily
replenished HOBr concentration is the source of the actual oxidant
and effectively acts as a redox buffer. While the NaHSO3/NaBrO3
reagent demonstrated excellent selectivities with the alcohol model
systems, the oxidation of sugar substrates failed due to
side-reactions occurring at the required low pH.