Electronic circular dichroism of monomethyl [16O,17O,18O]-phosphate and [16O,17O,18O]-thiophosphate revisited.

Phosphoryl-transfer reactions have long been of interest due to their importance in maintaining numerous cellular functions. A phosphoryl-transfer reaction results in two possible stereochemical outcomes: either retention or inversion of configuration at the transferred phosphorus atom. When the product is phosphate, isotopically-labeled [16O,17O,18O]-phosphate derivatives can be used to distinguish these outcomes; one oxygen must be replaced by sulfur or esterified to achieve isotopic chirality. Conventionally, stereochemical analysis of isotopically chiral phosphate has been based on 31P NMR spectroscopy and involves complex chemical or enzymatic transformations. An attractive alternative would be direct determination of the enantiomeric excess using chiroptical spectroscopy. (S)-Methyl-[16O,17O,18O]-phosphate (MePi*), 7 and enantiomeric [16O,17O,18O]-thiophosphate (TPi*), 10, were previously reported to exhibit weak electronic circular dichroism (ECD), although with 10 the result was considered to be uncertain. We have now re-examined the possibility that excesses of 7 and 10 enantiomers can be detected by ECD spectrometry, using both experimental and theoretical approaches. 7 and both the (R) and (S) enantiomers of 10 (10a,10b) were synthesized by the 'Oxford route' and characterized by 1H, 31P and 17O NMR, and by MS analysis. Weak ECD could be found for 7, with suboptimal S/N. No significant ECD could be detected for the 10 enantiomers. Time-dependent DFT (TDDFT) calculations of the electronic excitation energies and rotational strengths of the same three enantiomers were carried out using the functional B3LYP and the basis set 6-311G**. The isotopically-perturbed geometries were predicted using the anharmonic vibrational frequency calculational code in GAUSSIAN 03. In the case of 10, calculations were also carried out for the hexahydrated complex to investigate the influence of the aqueous solvent. The predicted excitation wavelengths are greater than the observed wavelengths, a not unusual result of TDDFT calculations. The predicted anisotropy ratios are 2.9 x 10(-5) for 7, -5.3 x 10(-6) for 10a/b, and 1.7 x 10(-6) for 10a/b.(H2O)6. For 7 the predicted anisotropy ratio approximates that observed in this work, 4.5 x 10(-5) at 208 nm. For 10a/b, the upper limits of the experimental anisotropy ratios (<5 x 10(-6) at 225 nm, pH 9; <5 x 10(-6) at 236 nm, pH 12) are comparable to the predicted magnitude of the value for 10a/b. The lower predicted value for 10a/b.(H2O)6 suggests that the aqueous environment affects the ECD significantly. Altogether, the TDDFT calculations together with a stereochemical analysis based on NMR and the MS data support the conclusion that the experimental ECD results for MePi* and TPi* may be reliable in order of magnitude.