990
V. Nandipati et al. / Tetrahedron Letters 55 (2014) 985–991
carbonyl of the urea or amide pyrrolidine substituent. kmax for
hosts 2, 5, and 6 is 420 nm, whereas that of 3, 4, 7, and 8 is
425 nm (host 1 kmax is 424 nm). Thus, we believe the red shift in
kmax for hosts 3, 4, 7, and 8 (and 1) compared to 2, 5, and 6 is attrib-
uted to this intramolecular zinc coordination which is not exhib-
ited in hosts 2, 5, and 6.
Upon guest binding, the kmax for host’s 2–8 complexes are typ-
ically 429–431 nm. Thus, there seems to be a general trend where-
by these types of porphyrin hosts which have introverted
functional groups exist in a conformation with the metallo center
intramolecularly complexed if this interaction is not sterically pre-
cluded (which it may be in 5 and 6, or the sulfonamide group may
be a poor ligand for zinc). Other porphyrin examples with this fea-
ture have been reported14 (this is a general phenomena in porphy-
rin-based metallo proteins as well). This represents another
example of conformationally induced guest binding.15
We believe hosts 3 and 4 bind mandelate through carboxylate
coordination to zinc and a hydrogen bond interaction between
the urea and the hydroxy group. Host 3 may bind R-mandelate
with greater affinity than S-mandelate because the R-isomer fits
the porphyrin binding cavity better. If R-mandelate binds in the
manner depicted in Figure 6, the phenyl group would not experi-
ence a steric interaction. If S-mandelate binds in a similar fashion
however, the phenyl group could experience a steric interaction
with the porphyrin surface. To avoid a steric interaction, we believe
S-mandelate binds to 3 in a different, un-defined conformation. A
1H NMR titration of 3 with R- and S-mandelate supports this
(Fig. 7).
Host 1 did not show any selectivity for other chiral guests that
we studied, just as is observed here for 3 and 4. This begs the ques-
tions: why is there no selectivity for other guests such as N-acetyl-
alanine and N-acetylphenylalanine stereoisomers? And why do not
hosts 5 and 6 show any stereoselectivity in guest binding, even
with mandelate stereoisomers? Since 5 and 6 did not show selec-
tivity in binding mandelate isomers, we predicted hosts 7 and 8
would not either. As Table 1 shows, hosts 7 and 8 show no selectiv-
ity in binding mandelate isomers; we expect no selectivity for 7
and 8 with other guests of Table 1 and thus they have not been
studied with other guests at this point. We believe the other chiral
guests examined show no stereopreference in guest binding be-
positioned to hydrogen bond with a guest than the N–H of the
amide or sulfonamide substituents of hosts 5–8. Small changes in
structure can have huge effects on molecular recognition, particu-
larly in rigid systems—one atom’s diameter, the direction of an
amide bond, etc. can make all the difference in the stability of a
host–guest complex.16
Figure 7 shows the 1H NMR titration of 2 with S-mandelate. Pro-
ton Hb shifts downfield with increasing concentration of S-mandel-
ate, but the shift is only apparent when large amounts of guest are
added (ꢀ5 equiv addition at a time), which implies weak binding of
S-mandelate to 2, which is in line with the results of Table 1. Thus,
the urea proton of 2 seems to contribute little to binding of guests
by itself.
For the titration of 3 with R-mandelate, the signals for the urea
protons (labeled b and g) rapidly move downfield with small incre-
ments of guest addition, which suggest the binding constant is
large compared to that of 2 with mandelate. The rapid change in
chemical shift of Hb suggests a cooperative interaction between
Hb, Hg and the metallo center in guest binding. The urea phenyl
proton signals (labeled h–j) are initially ꢀ7 ppm but move upfield
and two sets of protons become non-chemical shift equivalent in
the process. When 3 is titrated with S-mandelate, the signal for
urea Hb moves upfield! The signal for the urea phenyl protons also
moves upfield and splits into three sets of non-chemical shift
equivalent protons. This suggests that 3 forms conformationally
different complexes with R- and S-mandelate.
In summary, we have developed a class of easy-to-synthesize
and modular chiral porphyrin-based hosts that are unique because
recognition motifs are directed over the porphyrin surface in a
well-defined fashion where they work in tune with the metal cen-
ter for cooperative guest binding. Preliminary results suggest that
we can predict a priori which stereoisomer of mandelate stereoiso-
mers will bind these hosts. Future work will be conducted to deter-
mine if these hosts might in general stereoselectively bind other
chiral
a-hydroxy carboxylate compounds. We are working to im-
prove on the chiral selectivity. Others working in this field have
noted the importance of geometrical confinement of the guest to
achieve good enantioselectivity in chiral recognition.17 Variations
of these types of porphyrins with optimally positioned recognition
motifs are in the works.
cause either the alkyl group on the a-position is too small to have
a spatial preference in these complexes (such as in N-acetylala-
Acknowledgments
nine) or as in the case of N-acetylphenylalanine isomers, the CH2
spacer of the
a
-side chain allows the phenyl ring the ability to
We gratefully acknowledge support from the Robert A. Welch
Foundation (T-0014), NSF-REU (Grant No. 0851966), the Chemistry
Department at Texas A&M-Commerce and the Office of Graduate
Studies & Research at Texas A&M-Commerce for funding this work.
avoid a steric interaction with the porphyrin surface (Fig. 6 illus-
trates this tentative explanation for the lack of stereoselectivity
with other guests.). Thus, if the amino acid derivatives form com-
plexes similar to mandelate with hosts 3 and 4 such that the car-
boxylate coordinates zinc and the N-acyl group hydrogen bonds
to the host, it is not surprising that N-acetylalanine stereoisomers
show no selectivity in binding since the methyl group does not
have much greater steric demands than a hydrogen. For phenylal-
anine, although the benzyl side chain is relatively large, the guest
could adopt a conformation in which the CH2 group is directed at
the porphyrin surface rather than the phenyl group, which would
minimize steric interactions and hence eliminate selectivity in
binding.
Supplementary data
Supplementary data (HD mass spectral data for hosts 2–8. 1H
NMR and 13C NMR spectra of compounds (room temperature and
45 °C)) associated with this article can be found, in the online ver-
References and notes
Of course, the host/amino acid complexes could exhibit an en-
tirely different conformation than those of
a-hydroxycarboxylate
guests (mandelate) which would be reasonable if the N-acyl group
of amino acids does not match the hydrogen bonding motifs of the
host. Lastly, hosts 5–8 may show no selectivity with these guests
because the sulfonamide or amide substituents may not be opti-
mally positioned to interact with a guest that simultaneously com-
plexes zinc—the urea substituent of hosts 3 and 4 has an N–H
group 2 atoms further from the pyrrolidine ring that may be better