Mass spectrometric screening of chiral catalysts by monitoring the back reaction : palladium-catalysed allylic substitution

The principle of microscopic reversibility serves as the basis of a novel screening method for
the evaluation of chiral catalysts within asymmetric allylic substitution reactions. Monitoring
the back reaction of quasienantiomeric products 142a and 142b by electrospray ionisation
mass spectrometry (ESI-MS) reveals the intrinsic enantioselectivity of palladium catalysts in
the nucleophilic addition step and thus in the overall substitution process.
In this manner, an equimolar mixture of mass-labelled quasienantiomers 142a and 142b was
subjected to typical reaction conditions and the ratio of the resulting cationic allyl-palladium
complexes A and B, which are the only species visible in the mass spectrum, was determined
with high accuracy.
A series of linear diaryl allylation products with commonly used nucleophiles as leaving
groups were initially analysed with respect to their leaving group ability, which correlates
well with the pKa values of the respective nucleophiles. Based on these results, screening
procedures for allylic alkylations and aminations using substrates derived from acetyl acetone
and phthalimide were developed. In both cases, the ratios of allyl intermediates A/B closely
match with the enantiomeric product ratios of corresponding preparative reactions as
determined by HPLC analysis of the products. The methodology is fast, reliable, and enables
the simultaneous screening of catalyst mixtures, as long as the catalysts possess different
molecular masses. Three palladium complexes were tested in a single reaction vessel and the
most powerful derivative was readily identified.
After having established a protocol for the evaluation of chiral palladium catalysts in allylic
substitutions of linear diaryl substrates, the methodology was extended to the more
demanding carbocyclic substrates. An efficient screening for the kinetic resolution of
cyclohexenyl benzoate was developed by the use of cyclic allyl esters 65a and 65b.
Moreover, analysis of the back reaction starting from the allylation products 90a and 90b
allowed for determining the efficiency of the catalysts in the overall substitution process. The
observed data were in good agreement with the corresponding preparative reactions.
In an additional project, new pyridyl-phosphite 208, bis(N-sulfonylamino)phosphine 209, and
phosphinooxazoline ligands 133 were successfully synthesised and evaluated by the ESI-MS
procedures and conventional preparative catalytic reactions.
The back reaction screening method was further extended to the mass spectrometric
evaluation of racemic catalysts. Using a scalemic mixture of quasienantiomers 142a and 142b
enabled to determine the enantioselectivities of chiral catalysts by testing their racemates.
This approach was used to study phosphinooxazoline ligands 143, which possess a
stereogenic phosphorus atom as the only source of chirality, because the synthesis of the
enantiomerically pure compounds is not straightforward.
A series of derivatives was synthesised and evaluated by ESI-MS. Conventional preparative
reactions of the separated enantiomers were performed as well. Ligands derived from dialkyl,
alkylaryl, and diaryl phosphines induced only low to moderate selectivities in allylic
substitutions and the results are not competitive with the established phox ligands derived
from chiral aminoalcohols. The screening methodology was successfully used to determine
these selectivities without the time-consuming preparation of the enantiomerically pure
compounds.
Finally, the ESI-MS analysis of iridium-catalysed allylic substitution reactions proved the
actual existence of a previously postulated allyl-iridium intermediate. Preliminary results are
promising and provide information which is in accordance with the suggested reaction
mechanism.