Artificial Enzymes - Artificial Enzymes - page 30

added an additional binding ester group to the substrate, in 134 [195]. This made three
binding interactions with the catalyst 131 , and the hydroxylation product was now in-
deed the C-9 hydroxy compound 135 . Our computer models made it clear why this was
successful [195]. Some of this work has been reviewed [196].
The third binding interaction introduced in the hydroxylation of 134 made the sub-
strate present its face, not its edge, to the Mn=O of catalyst 131 , but it seemed likely that
double binding of the substrate could also achieve this if the binding groups were at-
tached to C-3 and C-6 rather than C-3 and C-17 as in 134 . Thus we synthesized a new
substrate, 136 , and examined its hydroxylation by catalyst 131 [197]. We saw that in-
deed there was hydroxylation at C-9 to form 137 , but there was also some hydroxylation
at C-15 to form 138 . With the lack of a binding interaction at C-17, apparently the
substrate had slipped a little to the left, moving C-15 into position to be attacked.
To move the substrate back in place, by shortening the distance from a cyclodextrin to
the Mn=O group, models suggested that the cyclodextrin be attached to the meta rather
than para position of the phenyl rings in the catalyst. Thus we synthesized catalyst 139 ,
with such a meta attachment, and indeed it did cleanly convert substrate 136 into the
C-9 hydroxy product 137 , with no hydroxylation at C-15 [197]. However, as there was no
fluorine in the phenyl linkers, the catalyst was destroyed after only 2.5 turnovers.
This was solved by replacing the phenyl linkers with trifluoropyridine rings in cat-
alyst 140 [197]. The compound was easily made, and it performed the C-9 hydroxyla-
tion of substrate 136 to form product 137 with 90 turnovers.