High turnover remote catalytic oxygenation of alkyl groups: How steric exclusion of unbound substrate contributes to high molecular recognition selectivity

Article abstract of DOI:10.1021/ja076039m

H-bonding mediated molecular recognition between substrate and ligand -COOH groups orients the substrate so that remote, catalyzed oxygenation of an alkyl C-H bond by a Mn-oxo active site can occur with very high (>98%) regio- and stereoselectivity. This paper identifies steric exclusion - exclusion of non H-bonded substrate molecules from the active site - as one requirement for high selectivity, along with the entropic advantage of intramolecularity. If unbound substrate molecules were able to reach the active site, they would react unselectively, degrading the observed selectivity. Both of the faces of the catalyst are blocked by two ligand molecules each with a -COOH group. The acid p-^tBuC₆H₄COOH binds to the ligand -COOH recognition site but is not oxidized and merely blocks approach of the substrate therefore acting as an effective inhibitor for ibuprofen oxidation in both free acid and ibuprofen ester form. Dixon plots show that inhibition is competitive for the free acid ibuprofen substrate, no doubt because this substrate can compete with the inhibitor for binding to the recognition site. In contrast, inhibition is uncompetitive for the ibuprofen-ester substrate, consistent with this ester substrate no longer being able to bind to the recognition site. Inhibition can be reversed with MeCOOH, an acid that can competitively bind to the recognition site but, being sterically small, no longer blocks access to the active site.

Full text of DOI:10.1021/ja076039m

Published on Web 01/16/2008
High Turnover Remote Catalytic Oxygenation of Alkyl
Groups: How Steric Exclusion of Unbound Substrate
Contributes to High Molecular Recognition Selectivity
Siddhartha Das, Gary W. Brudvig,* and Robert H. Crabtree*
Department of Chemistry, Yale UniVersity, 225 Prospect Street, New HaVen, Connecticut 06520
Received August 10, 2007; E-mail: gary.brudvig@yale.edu; robert.crabtree@yale.edu
Abstract: H-bonding mediated molecular recognition between substrate and ligand -COOH groups orients
the substrate so that remote, catalyzed oxygenation of an alkyl C-H bond by a Mn-oxo active site can
occur with very high (>98%) regio- and stereoselectivity. This paper identifies steric exclusionsexclusion
of non H-bonded substrate molecules from the active sitesas one requirement for high selectivity, along
with the entropic advantage of intramolecularity. If unbound substrate molecules were able to reach the
active site, they would react unselectively, degrading the observed selectivity. Both of the faces of the
catalyst are blocked by two ligand molecules each with a -COOH group. The acid p-tBuC6H4COOH binds
to the ligand -COOH recognition site but is not oxidized and merely blocks approach of the substrate
therefore acting as an effective inhibitor for ibuprofen oxidation in both free acid and ibuprofen ester form.
Dixon plots show that inhibition is competitive for the free acid ibuprofen substrate, no doubt because this
substrate can compete with the inhibitor for binding to the recognition site. In contrast, inhibition is
uncompetitive for the ibuprofen-ester substrate, consistent with this ester substrate no longer being able to
bind to the recognition site. Inhibition can be reversed with MeCOOH, an acid that can competitively bind
to the recognition site but, being sterically small, no longer blocks access to the active site.
1. Introduction
catalytic turnover could occur. Similarly, the Greico and the
Sames groups reported some non-porphyrin Mn-based synthetic
model complexes8,9 where selectivity was attained by covalent
attachment of the substrate but again only with single turnover
chemistry.
Selective and controlled oxidation of alkyl groups with
synthetic catalysts, especially regio- and stereoselective oxida-
tion of saturated C-H bonds remote from an existing functional
group, is a long-standing challenge.1,2 In contrast, enzymes
routinely achieve highly selective functionalization of saturated
C-H bonds. These include the cytochrome P450 dependent
monooxygenases that are highly specific in both alkane and
alkene oxygenation in numerous biosynthetic processes.3 The
fatty acid desaturases, with their nonheme carboxylate bridged
diiron center, catalyze O2 dependent dehydrogenation of satu-
rated fatty acyl CoA with high stereo- and regiospecificity.4-6
Weak noncovalent molecular recognition forces can lead to
reversible binding of substrates and are, thus, important elements
in our design strategy for multiple turnovers. An enzyme anchors
and orients the substrate via H-bonding, hydrophobic interactions
and π-stacking. These are thought to bring the substrate reactive
site close to the catalyst active site in a near attack conformer
(NAC)10,11 that facilitates reaction by controlling the approach
of the substrate to the catalyst. In typical synthetic catalysts,
reversible attachment of substrates via a molecular recognition
functional group is a promising strategy. With the right
conditions, molecular recognition should be tight enough to hold
the substrate properly but still be weak enough to allow catalytic
turnover.
Soon after the first crystal structure of cytochrome P450 was
reported, the Groves group reported the first porphyrin-based
model system, in which the substrate was covalently attached
to the porphyrin via an ester bond.7 Thus, although regioselec-
tivity was attained in oxygenation of saturated C-H bonds, no
In classic work, Breslow and co-workers reported a series of
porphyrin-based Mn catalysts, where cyclodextrin groups are
attached to the porphyrin ring.12-15 Cholesterol substrates having
(1) Meunier, B. Biomimetic Oxidations Catalyzed by Transition Metal Com-
plexes; Imperial College Press: London, 2000.
(2) Shilov, A. E.; Shul’pin, G. B. ActiVation and Catalytic Reactions of
Saturated Hydrocarbons in Presence of Metal Complexes; Kluwer Aca-
demic Publishers: Boston, 2000.
(3) Ortiz de Montellano, P. R. Cytochrome P450: Structure Mechanism and
Biochemistry, 3rd ed.; Kluwer Academic/Plenum Publishers: New York,
2004.
(8) Dangel, B. D.; Johnson, J. A.; Sames, D. J. Am. Chem. Soc. 2001, 123,
8149-8150.
(9) Kaufman, M. D.; Grieco, P. A.; Bougie, D. W. J. Am. Chem. Soc. 1993,
(4) Behrouzian, B.; Buist, P. H. Prostaglandins Leukot. Essent. Fatty Acids
2003, 68, 107-112.
115, 11648-11649.
(10) Garcia-Viloca, M.; Gao, J.; Karplus, M.; Truhlar, D. G. Science 2004, 303,
(5) Behrouzian, B.; Savile, C. K.; Dawson, B.; Buist, P. H.; Shanklin, J. J.
Am. Chem. Soc. 2002, 124, 3277-3283.
186-195.
(11) Gao, J. L.; Ma, S. H.; Major, D. T.; Nam, K.; Pu, J. Z.; Truhlar, D. G.
Chem. ReV. 2006, 106, 3188-3209.
(6) Fox, B. G.; Lyle, K. S.; Rogge, C. E. Acc. Chem. Res. 2004, 37, 421-429.
(7) Groves, J. T.; Nemo, T. E.; Myers, R. S. J. Am. Chem. Soc. 1979, 101,
1032-1033.
(12) Breslow, R.; Zhang, X. J.; Huang, Y. J. Am. Chem. Soc. 1997, 119, 4535-
4536.
9
1628
J. AM. CHEM. SOC. 2008, 130, 1628-1637
10.1021/ja076039m CCC: $40.75 © 2008 American Chemical Society

Remote Catalytic Oxygenation of Alkyl Groups
A R T I C L E S
Scheme 1.
Scheme 2.
hydrophobic tails bound to the hydrophobic cyclodextrin cavity
gave multiturnover oxidation at specific C-H bonds. In another
classic example, Groves et al. reported a “membrane spanning”
iron-porphyrin catalyst.16 Steroid groups on the porphyrin
enabled the catalyst to intercalate in a lipid bilayer to place the
Fe active site at midmembrane. Remarkable regioselectivity was
seen in epoxidation of several sterol substrates that were oriented
by the bilayer. Suslick and co-workers saw high shape selectivity
in the epoxidation of alkenes17-20 by bis-pocket Mn porphyrin
and dendrimer Mn porphyrin catalysts. None of these systems
use attractive, labile molecular recognition interactions, however.
In porphyrin-based synthetic catalysts, efficiency tends to be
low,12 so we have turned to a simpler and more highly catalytic
system based on a dimanganese-bis-µ-oxo terpyridine com-
plex21,22 with oxone (HSO5-) as the primary oxidant that prior
studies suggest does not operate via freely diffusing radicals.23
Our modified terpyridine ligand (LMR) (Scheme 1), containing
a molecular recognition group capable of binding carboxylic
acid substrates, gives the molecular recognition (MR) catalyst,
(13) Breslow, R.; Huang, Y.; Zhang, X. J.; Yang, J. Proc. Natl. Acad. Sci. U.S.A.
1997, 94, 11156-11158.
(14) Yang, J.; Breslow, R. Angew. Chem., Int. Ed. 2000, 39, 2692-2694.
(15) Yang, J.; Gabriele, B.; Belvedere, S.; Huang, Y.; Breslow, R. J. Org. Chem.
2002, 67, 5057-5067.
C
MR(IV,IV), [H2O(LMR)Mn(µ-O)2Mn(LMR)H2O](NO3)4. With
(16) Groves, J. T.; Neumann, R. J. Am. Chem. Soc. 1987, 109, 5045-5047.
(17) Cook, B. R.; Reinert, T. J.; Suslick, K. S. J. Am. Chem. Soc. 1986, 108,
7281-7286.
oxone as the primary oxidant, this catalyst gives highly regio-
and stereoselective oxygenation of saturated C-H bonds with
multi-hundred turnovers. Control experiments with a similar but
nonrecognition (NR) catalyst, CNR(III,IV), [H2O(LNR)Mn(µ-
(18) Suslick, K. S.; Cook, B. R. J. Chem. Soc., Chem. Commun. 1987, 200-
202.
(19) Bhyrappa, P.; Young, J. K.; Moore, J. S.; Suslick, K. S. J. Am. Chem. Soc.
1996, 118, 5708-5711.
(20) Bhyrappa, P.; Young, J. K.; Moore, J. S.; Suslick, K. S. J. Mol. Catal. A:
Chem. 1996, 113, 109-116.
O)2Mn(LNR)H2O](ClO4)3 (LNR
: 4′-phenyl-2,2′:6′,2′′-terpyri-
dine), which lacks the key -COOH group, do not show any
selectivity.22 The (IV,IV) or (III,IV) designations refer to the
oxidation states of the Mn ions in the catalyst precursor. The
(21) Rivera, G.; Crabtree, R. H. J. Mol. Catal. A: Chem. 2004, 222, 59-73.
(22) Das, S.; Incarvito, C. D.; Crabtree, R. H.; Brudvig, G. W. Science 2006,
312, 1941-1943.
(23) Wessel, J.; Crabtree, R. H. J. Mol. Catal. A: Chem. 1996, 113, 13-22.
9
J. AM. CHEM. SOC. VOL. 130, NO. 5, 2008 1629

A R T I C L E S
Das et al.
Figure 1. Molecular model (Chem 3D) of catalyst CMR(IV,IV) docked to a cartoon substrate. Due to the ∼32° angle between the KTA -COOH group and
the plane containing the imide group, the substrate needs a bent -CH2- group after the -COOH group for the rest of the substrate to come into proximity
with the active site (ideally 4-5 Å between target C-atom and Mn).
Figure 2. Molecular model (Chem 3D) of S1 docked to the catalyst CMR(IV,IV) via -COOH‚‚‚HOOC- type H-bonding. The distance between the target
C-atom and the manganese ion is ∼4.5 Å.
Figure 3. Molecular models (Chem 3D) of the catalyst CMR(IV,IV) docked to (a) cis-S2 and (b) trans-S2 via H-bonding. The distance between the target
C-atom and the manganese ion is 4.7 and 4.6 Å for cis and trans isomers, respectively. In both cases, the isomers are drawn in a way so that C, H, and O
atoms are approximately colinear to facilitate NAC formation.
difference in oxidation state reflects differences in the ease of
isolation between the MR and NR catalysts: every indication
suggests they both take part in exactly the same catalytic cycle.
Selectivity can also be degraded by exposing the MR system
to excess MeCOOH, which successfully competes for the
recognition site.22
9
1630 J. AM. CHEM. SOC. VOL. 130, NO. 5, 2008

Remote Catalytic Oxygenation of Alkyl Groups
A R T I C L E S
Table 1. Product Distribution from Ibuprofen (S1) Oxidation (by 1H
bond followed by a fast “rebound” of the resulting metal bound
OH group to form the C-OH bond of the product, does not
also apply to this system.
NMR Spectroscopy)d
The terpy ligand is not only more available than porphyrins
but also more oxidation resistant.35,36 Porphyrin-based catalysts
without halogen or nitro substituents are oxidized after only
ca. 10 turnovers even for a reactive benzylic substrate,12,13 while
our catalysts give as many as 500 turnovers without any special
protective substitution.
Incorporation of the MR group is relatively straightforward.
In ligand LMR (Scheme 1), we have attached a p-phenylene
linker and a Kemp’s triacid (KTA) “U”-turn unit to the terpy
ligand to provide a rigid37 -COOH group projecting toward
the active site. Ligand LMR, thus, lacks any oxidation-prone
component.
3. Synthesis of Catalysts
The ligands LMR and LNR and the corresponding catalysts
CMR(IV,IV) and CNR(III,IV) were synthesized according to
standard methods as described in our prior report and sum-
marized in Schemes S1 and S2 of the Supporting Information.22
Catalyst CMR(III,IV) was synthesized from Mn(LMR)Cl2 using
a limited amount of oxone.22
a Subtrate/catalyst/oxidant ) 100:0.1:500. bSolvent/CD3CN. cSolvent/
CH3NO2. dTotal Turnovers ) mol of products per mol of catalyst; oxidant:
TBAO; subtrate/catalyst/oxidant ) 100:1:500, solvent: CH3CN, if not
mentioned otherwise.
4. Standardization of Catalytic Conditions
In this paper, we show how the substrate that is bound to the
recognition group blocks the reactive site and prevents unbound
substrate from reacting unselectively. This steric exclusion effect
is a contributing feature in obtaining molecular recognition-
induced selectivity. For this work, we have been able to isolate
the mixed-valent Mn(III,IV) di-µ-oxo dimanganese complex
(CMR(III,IV)) derived from LMR. The complex, CMR(III,IV),
found to be more soluble than CMR(IV,IV), gives essentially
the same selectivity and activity suggesting that the active
species is the same. With CMR(III,IV), we have been able to
select a suitable reversible inhibitor, p-tert-butylbenzoic acid
(I1), that is not itself oxidized but almost entirely inhibits the
catalyst by binding to the molecular recognition site and thus
blocking the active site.
With a -COOH‚‚‚HOOC- H-bonding motif for molecular
recognition, we needed an oxidation resistant solvent that is polar
enough to dissolve the catalyst and promote catalysis but not
so polar as to interfere with H-bonding. Acetonitrile with 0.5%
water proved suitable. Acetonitrile is not completely oxidation
resistant, however. When CH3CN was replaced with CD3CN,
the turnover was found to improve significantly without
affecting the selectivity, presumably due to C-D bonds being
more oxidation resistant than C-H bonds. We have also used
CH3NO2. However the catalyst was insufficiently soluble in
benzonitrile or nitrobenzene. Catalytic runs were performed with
1 eq of substrate, 0.001-0.005 eq of catalyst and 5 eq of oxone
in minimum amounts of CH3CN or CH3NO2 as solvents.
To maintain -COOH‚‚‚HOOC- H-bonding, all the carboxy-
lic acid groups should remain protonated. The primary oxidant,
tetrabutylammonium oxone, has a pH of ∼1.5 in H2O consistent
with this requirement. Deprotonation destroys the selectivity.22
Schemes 1 and 2 show the structures of all the substrates,
inhibitors, and catalysts used in this study.
2. Discussion of Our Molecular Recognition Catalyst
CMR(IV,IV)
5. Results and Discussion
There is no reason to think that the usual proposed
mechanism,24-34 H atom abstraction from the substrate C-H
5.1. Molecular Modeling. Molecular models of catalyst CMR
-
(IV,IV) were constructed by importing crystal structure param-
eters of ligand LMR22 and the Mn(µ-O)2Mn core35 followed by
energy minimization (MM2, CAChe 5). The crystal structure
of LMR showed that the -COOH group is bent at an angle of
∼32° with respect to the plane of the imide group. In a
p-substituted benzoic acid substrate, the remote C-H bonds
would, therefore, be too far from the active site for reaction;
these substrates proved to be useful inhibitors, however. This
led us to incorporate a -CH2- unit next to the -COOH group
(24) Lundberg, M.; Blomberg, M. R. A.; Siegbahn, P. E. M. Inorg. Chem. 2004,
43, 264-274.
(25) Jin, S. X.; Bryson, T. A.; Dawson, J. H. J. Biol. Inorg. Chem. 2004, 9,
644-653.
(26) Newcomb, M.; Hollenberg, P. F.; Coon, M. J. Arch. Biochem. Biophys.
2003, 409, 72-79.
(27) Newcomb, M.; Zhang, R.; Chandrasena, R. E. P.; Halgrimson, J. A.; Horner,
J. H.; Makris, T. M.; Sligar, S. G. J. Am. Chem. Soc. 2006, 128, 4580-
4581.
(28) Shaik, S.; de Visser, S. P.; Kumar, D. J. Biol. Inorg. Chem. 2004, 9, 661-
668.
(29) Hirao, H.; Kumar, D.; Thiel, W.; Shaik, S. J. Am. Chem. Soc. 2005, 127,
13007-13018.
(30) Hirao, H.; Kumar, D.; Que, L.; Shaik, S. J. Am. Chem. Soc. 2006, 128,
8590-8606.
(34) Groves, J. T.; Haushalter, R. C.; Nakamura, M.; Nemo, T. E.; Evans, B. J.
J. Am. Chem. Soc. 1981, 103, 2884-2886.
(31) Shaik, S.; Kumar, D.; de Visser, S. P.; Altun, A.; Thiel, W. Chem. ReV.
2005, 105, 2279-2328.
(35) Limburg, J.; Vrettos, J. S.; Liable-Sands, L. M.; Rheingold, A. L.; Crabtree,
R. H.; Brudvig, G. W. Science 1999, 283, 1524-1527.
(32) Ingold, K. U.; MacFaul, P. A. In Biomimetic Oxidations Catalyzed by
Transition Metal Complexes; Meunier, B., Ed.; Imperial College Press:
London, 2000.
(36) Limburg, J.; Vrettos, J. S.; Chen, H. Y.; de Paula, J. C.; Crabtree, R. H.;
Brudvig, G. W. J. Am. Chem. Soc. 2001, 123, 423-430.
(37) Rebek, J. Science 1987, 235, 1478-1484.
(33) Groves, J. T. J. Chem. Educ. 1985, 62, 928-931.
9
J. AM. CHEM. SOC. VOL. 130, NO. 5, 2008 1631

A R T I C L E S
Das et al.
Figure 4. Both the cis and trans isomers of S2 give the same carbon-centered radical, which after rebound of the OH group from Mn-OH gives trans-P2.
cis-S2 after H-atom abstraction gives a C-centered radical, which to avoid steric hindrance rotates 180°.
Table 2. Product Distribution from S2 (cis-S2 + trans-S2) (by 1H NMR Spectroscopy)c
b
c
a Solvent: CD3CN. Solvent: CH3NO2. Total Turnovers ) mol of products per mol of catalyst; oxidant: TBAO; subtrate/catalyst/oxidant ) 100:0.1:
500, solvent: CH3CN, if not mentioned otherwise.
to give the substrates a bent structure that was expected to bring
the remote C-H bonds close to the active site. From crystal
structures of cytochrome P450-substrate complexes, it was
proposed that the carbon atom to be oxygenated needs to be
within an ∼4-5 Å distance from the metal.3 To impart some
rigidity to the substrate, we incorporated a benzene ring or a
cyclohexyl group. Molecular modeling suggested that the C-6
or C-7 C-H bonds would then come close to the active site
(Figure 1).
Ibuprofen [2-(4-isobutyl-phenyl)-propionic acid] (S1) and
4-methyl-cyclohexyl acetic acid (S2) were found to be the most
suitable substrates. These molecules have relatively rigid rings,
either planar C6H4 or chair C6H10, and have at least two alternate
sites of attack. We expected selective oxidation at the remote
benzylic and tertiary C-H bonds in S1 and S2, respectively,
group to give ketone P1 with the molecular recognition catalyst
(Table 1). Otherwise, a competitive oxidation pathway also
occurs leading to decarboxylation of the -COOH group to give
product P1′ (Table 1).38 Control catalyst CNR(III,IV), lacking
molecular recognition, gave both P1 and P1′ (3:1), but with the
molecular recognition catalyst CMR(IV,IV), essentially only P1
is formed (∼98%) (Table 1).
5.3. Regio- and Stereoselective Oxygenation of Cyclohex-
ane Acetic Acid (S2). 4-Methylcyclohexane acetic acid, sub-
strate S2, was employed to investigate alkyl CH hydroxylation.
Our sample was a 5:4 mixture of cis and trans isomers (cis-S2
and trans-S2). Both isomers react (cis > trans), but only a single
product, trans-P2, was formed (Table 2). Control catalysis with
C
NR(III,IV) gave both isomers and many other unidentified
products. The methyl ester of the product from the molecular
recognition catalyst was isolated by preparative thin layer
with the molecular recognition catalysts CMR(IV,IV) and CMR
-
1
(III,IV), as shown in Figures 2 and 3.
chromatography and characterized as trans-P2 based on H
5.2. Regioselective Oxygenation of Ibuprofen (S1). For
ibuprofen, selective oxidation occurs at the remote benzylic CH2
(38) Ohe, T.; Mashino, T.; Hirobe, M. Tetrahedron Lett. 1995, 36, 7681-7684.
9
1632 J. AM. CHEM. SOC. VOL. 130, NO. 5, 2008

Remote Catalytic Oxygenation of Alkyl Groups
A R T I C L E S
Figure 5. Possible products seen from cis-stilbene (S3) oxidation with metalloporphyrins following several mechanisms.
Table 3. Mechanistic Investigations with CMR(IV, IV)b
b
a Based on GC-MS, quantitative analysis was not performed. Yields are w.r.t. substrates; oxidant: TBAO.
1
1
1
NMR, H NOE, H COSY, H homonuclear decoupling, and
13C NMR data (Scheme 2). This suggests that both substrate
isomers give a common intermediate radical39 that has time to
undergo bond rotation and accept the OH group exclusively
from one side of the molecule in the “rebound” step (Figure
4).
5.4.2. Epoxidation of cis-Stilbene. cis-Stilbene (S3) is one
of the substrates commonly used to investigate the radical vs
cationic mechanism of oxygenation.23,40-42 cis-Stilbene can give
a variety of products (Figure 5).23 With CMR(IV,IV), the major
products were cis-stilbene oxide, benzophenone, and 1,2-
diphenylethanone, but no trans-stilbene oxide (Table 3). These
products can be explained by fast “rebound” or by a carbocation
intermediate (Figure 5). The absence of any trans-stilbene oxide
indicates the absence of any freely diffusing, long-lived carbon-
radical intermediate (Table 3).
5.4. Mechanism of C-H Bond Oxygenation. 5.4.1. Effect
of Air. In C-H bond oxygenation, any long-lived carbon-
centered radical intermediate typically gives radical chain
products with air.23 Our catalytic runs under air or N2 gave the
same product distribution (Table 3), so any radical intermediate
is “caged” by the molecular recognition and is short-lived.
(40) Castellino, A. J.; Bruice, T. C. J. Am. Chem. Soc. 1988, 110, 158-162.
(41) Ostovic, D.; Bruice, T. C. Acc. Chem. Res. 1992, 25, 314-320.
(42) Collman, J. P.; Brauman, J. I.; Meunier, B.; Raybuck, S. A.; Kodadek, T.
Proc. Natl. Acad. Sci. U.S.A. 1984, 81, 3245-3248.
(39) Rauk, A.; Sorensen, T. S.; Maerker, C.; Carneiro, J.; Sieber, S.; Schleyer,
P. V. J. Am. Chem. Soc. 1996, 118, 3761-3762.
9
J. AM. CHEM. SOC. VOL. 130, NO. 5, 2008 1633

A R T I C L E S
Das et al.
Figure 6. Molecular models (Chem 3D) of catalyst CMR(III,IV) docked to S4 and S5. (a) Bound S4 blocks access of unbound S4. (b) The same in a space
filling representation. (c) Bound S5 does not block the active site. (d) The same in a space filling representation.
Table 4. Investigation of Molecular Recognition Mechanisma
a Yields are w.r.t. substrates; oxidant: TBAO (5 equiv of w.r.t. substrates).
5.4.3. Effect of Halogenating Agents. CBrCl3 and N-
bromosuccinimide (NBS) give rapid (ca. 108 L‚mol-1 s-1 43
Br
Synthetic catalysts generally do not have such a cavity, but
in our case, the approach of unbound substrates must be blocked
by the bound substrate in order to achieve the high regioselec-
tivity we see. Where this is not the case, selectivity can be
greatly reduced.19
)
atom transfer to alkyl radicals. With ibuprofen and CMR(IV,-
IV), CBrCl3 and NBS (0.125 M, 5 equiv vs substrate) did not
yield any halogenated product (Table 3) suggesting that the half-
life of the radical formed must be e10-7 s. This indicates the
With molecular recognition catalyst CMR(III,IV) and 4-
methylcyclohexane carboxylic acid (S4), no oxidation was seen.
In contrast, for control catalyst CNR(III,IV) lacking a molecular
recognition function, several oxidized products were found (1H
NMR). A molecular model showed that the docked substrate,
lacking the extra -CH2- group of the successful substrate, does
not present an oxidizable C-H group to the active site (Figure
6a). Not only does oxygenation of the docked substrate not
occur, but the docked substrate also prevents the approach of
undocked substrate molecules. In contrast, the sterically smaller
p-toluic acid (S5) yielded 1,4-benzenedicarboxylic acid (P5) with
both molecular recognition catalyst CMR(III,IV) and control
catalyst CNR(III,IV). A molecular model shows that when S5
is docked to CMR(III,IV), its benzylic methyl group cannot
approach the active site, but it also cannot prevent external
substrate molecules from approaching (Figure 6c and Table 4).
Substrates, S1 and S2, while docked to the catalyst CMR(III,-
IV) result in significant steric bulk at the active site. These
molecules, unlike S4, have reactive groups close to the active
“rebound” happens at a rate >108 L‚mol-1 s-1
5.5. Mechanism of Molecular Recognition. 5.5.1. H-
bonding is required. The MR selectivity obtained with CMR
.
-
(IV,IV) disappears with control catalyst CNR(III,IV) lacking
the key -COOH group. With excess acetic acid (4 equiv vs
substrate), the regioselectivity also disappeared (Table 4). Thus,
acetic acid may disrupt the proposed recognition by binding to
the ligand -COOH.
5.5.2. Inhibition of Reaction of Unbound Substrate is
Required. In enzymes, there is only enough room for one
substrate molecule to occupy the active site at any time. Some
enzymes even undergo structural changes after binding of one
substrate molecule, presumably in part to help prevent the entry
of any further substrate molecules. Thus, a bound substrate can
be seen as sterically excluding any other substrate molecules
in the medium.
(43) Dneprovskii, A. S.; Eliseenkov, E. V.; Chulkova, T. G. Russ. J. Org. Chem.
2002, 38, 338-343.
9
1634 J. AM. CHEM. SOC. VOL. 130, NO. 5, 2008

Remote Catalytic Oxygenation of Alkyl Groups
A R T I C L E S
Table 5. Inhibition of Molecular Recognition Catalyst CMR(III,IV)a
a Yields are w.r.t. substrates; oxidant: TBAO (5 equiv of w.r.t. substrates).
Table 6. Inhibition and Reactivation of Molecular Recognition Catalyst CMR(III,IV)e
a Aliquot consisting of half of the total catalytic solution was collected and worked up as the first fraction. bSecond fraction: total catalytic solution minus
first fraction. cEquivalents of acetic acid are w.r.t. the substrate in second fraction. dAfter first inactivation without acetic acid present (see text). eYields are
w.r.t. substrates; oxidant: TBAO (5 equiv of w.r.t. substrates).
site. Neither the target C-H bonds nor the other oxidation-
prone C-H bonds in these molecules are terminal and unhin-
dered.
equiv vs substrate) was added to the catalytic medium the yield
of P6 decreased to a mere 1.4%. No oxidized product at all
was obtained from S4 alone (Tables 4 and 5).
As the substrate used in these catalyses is in large excess
relative to the catalyst concentration (substrate/catalyst ) 1000:
1), we propose that under the catalytic conditions all the catalytic
recognition sites are saturated with docked substrates. The
docked substrate thus covers the active site with a protective
steric “umbrella” and makes it sterically inaccessible to unbound
substrates, thereby preventing any nonregioselective catalysis.
This steric exclusion of free substrates may well contribute to
high regioselectivity in any systems like ours.
Although the substrates S1 and S2, while remaining bound to
the MR group, could undergo bond rotations that would cause
them to move far away from the active site and, thus, leave the
Mn oxo site open to attack unbound substrate, this presumably
does not occur because the observed regioselectivity is very
high. Solvophobic and aromatic stacking forces between
substrate and ligand may be responsible for this effect.
5.6. Inhibition and Reactivation of the Molecular Recogni-
tion Catalyst. 5.6.1. Inhibition. Since the steric exclusion effect
of bound S4 prevents approach of any unbound S4 substrate to
the active site of catalyst CMR(III,IV), other bulky carboxylic
acids should act similarly. Potential substrate MeC6H10COOH,
S4, was found instead to act as an inhibitor with the methyl
ester of ibuprofen (S6) as the substrate. S6, lacking any -COOH
group, cannot bind to the catalyst CMR(III,IV) but can only
approach the active site by diffusion. Thus, S6 cannot compete
with the inhibitor for binding at the molecular recognition site.
In the absence of inhibitor S4, S6 is oxidized to give P6 as the
major product with molecular recognition catalyst CMR(III,-
IV), giving ∼30% yield after 3 h at 20 °C. But when S4 (1
Molecular modeling indicated that bulky 4-tert-butyl benzoic
acid (I1) should be an even better inhibitor. Under catalytic
conditions, the tert-butyl C-H bonds remain unaffected. When
I1 is bound to the catalyst CMR(III,IV), models suggest that
this bulky tert-butyl group occupies and blocks the region close
to the active site of the catalyst (Figure 7). With I1 (1 equiv vs
substrate) present, the yield of oxidation product P6 decreased
even further to 0.9% under the same conditions (Table 5).
We expected that a change of inhibitor concentration would
change the extent of inhibition. Indeed, a decrease of the
inhibitor (I1) concentration gave an increased yield of P6. A
number of runs were performed varying the concentration of I1
and measuring the yields at low conversions where yield can
be considered to be directly proportional to the velocity. A Dixon
plot, obtained by plotting the reciprocal of rate (estimated from
the yield at early reaction time) versus the inhibitor concentra-
tion, gave a straight line as found in enzyme inhibitor studies
(Figure 8). By using three different concentrations of substrate
(S6) for each concentration of the inhibitor, three different curves
were obtained in the Dixon plot (Figure 8a). These were almost
parallel, indicating that we have uncompetitive inhibition.44 We
verified that the system obeys saturation (Michaelis-Menten)
kinetics, required for the Dixon approach to be valid, by noting
that an increase in substrate concentration in the range of interest
and in the absence of inhibitor does not cause any change in
(44) Segel, I. H. Enzyme Kinetics. BehaVior and Analysis of Rapid Equilibrium
and Steady-state Enzyme Systems; Wiley-Interscience Publication: New
York, 1975.
9
J. AM. CHEM. SOC. VOL. 130, NO. 5, 2008 1635

A R T I C L E S
Das et al.
The reciprocal of rate vs inhibitor concentration also gave a
linear Dixon plot, but now having straight lines radiating from
a point (Figure 8b) on the negative side of the y-axis, indicating
a competitive inhibition in this case.44 This is consistent with
our hypothesis that both substrates and inhibitors bind to the
molecular recognition site.
5.6.2. Reversal of Inhibition. Reactivation of the inhibited
catalyst was found with acetic acid further confirming the
reversibility of the binding of CMR(III,IV) with I1. With S6 (1
equiv), CMR(III,IV) (0.005 equiv), and I1 (0.3 equiv) as the
inhibitor in CH3CN, half of the catalytic solution was worked
up after 3 h at 0 °C. Only 1.5% conversion of S6 to P6 occurred,
indicating successful inhibition with I1 (Table 6). To the other
half of the catalytic solution, excess acetic acid (2.25 equiv)
was added, an amount too small to cause any significant change
in the solvent polarity but enough to saturate all the catalytic
molecular recognition sites by H-bonding. This addition of CH3-
COOH was expected to disrupt the binding of the inhibitor to
the catalyst, reversing the inhibition. Indeed, after a further 3 h
at 0 °C, P6 product was recovered with 34% conversion (Table
6) showing successful reversal of inhibition by I1 with CH3-
COOH. We are not aware of any similar reactivation in a
synthetic catalyst.
Figure 7. (a) Molecular model and (b) space filling model of I1 bound to
CMR(III,IV). The oxidation resistant tert-butyl group of I1 offers a significant
steric bulk to block the active site.
5.7. Consequences of Steric Exclusion. The need for steric
exclusion to block unselective reaction may explain why it has
previously proved21 difficult to obtain high selectivities via
molecular recognition. A particular feature of our CMR system
is that it has two molecular recognition units, which can act
together to block both faces of the catalyst at once. In order to
do this, both -COOH sites of the ligand must be essentially
fully occupied and steric repulsion between the two substrates
or inhibitors must be sufficient to cause the two groups to
occupy opposite faces of the catalyst.
Conclusion
The results reported here help confirm our earlier proposal
that molecular recognition of the -COOH group in our
substrates by the ligand -COOH group orients the substrate.22
This allows remote catalytic oxygenation of an alkyl C-H bond
by a Mn-oxo active site to occur with very high (>98%) regio-
and stereoselectivity with oxone as the primary oxidant. Our
work identifies steric exclusionsexclusion of free substrate
molecules from the active sitesas an important criterion for
selectivity. If free substrates were able to reach the active site,
they would react unselectively, degrading the selectivity. Both
faces of the catalyst are protected by having two ligand -COOH
groups. p-tBuC6H4COOH (I1) was an effective inhibitor for our
standard substrate, ibuprofen (S1), by binding to the ligand
-COOH recognition site and blocking the approach of a
substrate. Dixon plots show that inhibition is competitive for
ibuprofen, suggesting that the substrate -COOH group com-
petes with the inhibitor -COOH group for binding to the
recognition site. In contrast, inhibition is uncompetitive for the
alternate substrate, ibuprofen-ester (S6), consistent with this
substrate no longer being able to bind to the recognition site. A
much less bulky acid, MeCOOH, can reverse the inhibition
caused by p-tBuC6H4COOH (I1) by displacing it from the ligand
-COOH recognition site and revealing the Mn-oxo active site.
Figure 8. Dixon plots showing the inhibition of oxidation of (a) S6 and
(b) S1 with molecular recognition catalyst CMR(III,IV) and I1 as the
inhibitor.
reaction rate indicating that all the measurements were done in
the saturation regime.
Oxidation of ibuprofen (S1) is also inhibited by I1, although
to a lesser extent than the case for for ibuprofen methyl ester,
S6. The likely cause of the difference is competition for the
molecular recognition site between the -COOH groups of
substrate S1 and inhibitor I1, which is not possible for ester S6.
9
1636 J. AM. CHEM. SOC. VOL. 130, NO. 5, 2008

Remote Catalytic Oxygenation of Alkyl Groups
A R T I C L E S
6.4. Catalytic Conditions. 6.4.1. Catalyzes with S1, S2, S4, and
S5. 5 × 10-4 mmol of catalyst was dissolved in 50 µL of water and 2
mL of CH3CN. This solution was added to 0.5 mmol of substrate. The
solution was cooled to 0 °C. A solution of 2.5 mmol of TBAO in 8
mL of CH3CN was prepared and was added to the cooled solution of
catalyst and substrate. The resulting solution was stirred for 2 h at
20 °C, 8 h at 0 °C, and 2 days at -20 °C. The reaction was quenched
by addition of 60 mL of a saturated aqueous solution of NaHSO3. The
solution was acidified with 0.1 N HCl. The unreacted substrate and
products were extracted in ether, evaporated under vacuum, and dried
in vacuo over P2O5 overnight. The resulting unreacted substrates and
6. Experimental Section
6.1. Materials. All solvents and reagents were obtained from
commercial sources and were used without further purification. Oxone,
used in all the catalytic runs, is the tetrabutylammonium salt of oxone.
Tetrabutylammonium oxone (TBAO) and K[Oxone] were standardized
using iodometric titration.
6.2. Physical Measurements. 1H and 13C NMR spectra were
recorded in deuterated solvents (Cambridge Isotope Laboratories) at
25 °C on a Bruker 400 or Bruker 500 spectrometer at the Yale
Department of Chemistry Instrument Center. Electrospray ionization-
mass spectroscopy (ESI-MS) was done on a Waters ZQ LC-MS
instrument. Gas chromatography-mass spectroscopy (GC-MS) was
done on an Agilent 5973 GC-MS instrument.
1
product mixtures were analyzed with H NMR, 13C NMR, and ESI-
MS.
6.3. Syntheses. LMR, [(LMR)MnCl2], [(LNR)MnCl2], CMR(IV,IV),
6.4.2. Oxygenation of S6 in Presence of I1. To a 2 mL acetonitrile
solution of S6 (0.5 mmol) and I1 (concentration varies in different runs),
a solution of CMR(III,IV) (2.5 × 10-3 mmol) dissolved in a minimum
amount of CH3CN-H2O (25 µL of H2O and 1.0 mL of CH3CN) was
added and stirred in an ice bath to cool it down to 0 °C. A solution of
TBAO (2.5 mmol) in 21 mL of acetonitrile was added to it and stirred
for 3 h at 0 °C. The workup was done as above. The unreacted substrate
C
NR(III,IV) were synthesized following the literature method.22
6.3.1. [H2O(LMR)Mn(µ-O)2Mn(LMR)H2O](NO3)3 (CMR(III,IV)).
[(LMR)MnCl2] (0.200 g, 0.3 mmol) was dissolved in a minimum amount
of hot (almost boiling) H2O-CH3CN (1:1); the solution was cooled to
0 °C by stirring in an ice bath. 5 mL of ice-cold aq. K-Oxone (0.080
g, 0.2 mmol) were added dropwise very slowly over a time of 10 min.
10 mL of ice-cold aq. KNO3 (4 g, 40 mmol, large excess) were added.
The volume of the solution was minimized by stirring the solution under
vacuo. The brown precipitate that formed was filtered and washed with
ice-cold water (3 × 3 mL) and then with copious diethyl ether to give
pure CMR(III,IV). Very slow addition of oxone and use of a smaller
amount are important to obtain CMR(III,IV) instead of CMR(IV,IV).
Elemental analysis: (C66H64Mn2N11O21) (C2H5OH5C2); Calcd. C, 54.89;
H, 4.87; N, 10.07. Found: C, 54.96; H, 4.87; N, 10.07. A frozen solution
of CMR(III,IV) in CH3CN gave a 16-line EPR spectrum (10 K),
characteristic of a MnIII(µ-O)2MnIV unit.
6.3.2. Syntheses of Methyl Esters of Several Carboxylic Acids.
The carboxylic acid was dissolved in a minimum amount of methanol-
benzene solution (MeOH/PhH ) 2:7) and cooled to 0 °C. An ice-cold
solution of trimethylsilyldiazomethane in diethyl ether was added
dropwise. The reaction was followed from the disappearance of the
reddish color of trimethylsilyldiazomethane accompanied with the
evolution of a gas. Once the reddish color persisted, the addition of
trimethylsilyldiazomethane was stopped. The mixture was stirred at
0 °C for another 10 min and evaporated under vacuum, giving an oily
product.
1
and product mixture was analyzed with H NMR.
6.5. Identification of Products. Identification of products obtained
from oxidation of S1 and S2 with catalyst CMR(IV,IV) has been
discussed in our prior report as well as briefly in the Supporting
Information.22
6.5.1. Oxidation of S5. The product P5 was identified by comparing
1
the H NMR spectrum to that of the commercially available sample.
Acknowledgment. This work was supported by the National
Science Foundation (CHE-0614403). We thank Professors A.
D. Hamilton, K. B. Wiberg, and J. W. Faller for advice and
assistance and Professor L. Que, Jr. for useful suggestions.
Supporting Information Available: Experimental procedures
and spectroscopic data. This material is available free of charge
via the Internet at http://pubs.acs.org.
JA076039M
9
J. AM. CHEM. SOC. VOL. 130, NO. 5, 2008 1637