A heterogeneous, selective oxidation catalyst based on Mn triazacyclononane grafted under reaction conditions

Article abstract of DOI:10.1039/b920391e

A unique method has been developed to synthesize an active heterogeneous oxidation catalyst by the in situ grafting of a 1,4,7-trimethyl-1,4,7- triazacyclononane manganese complex on carboxylic acid-functionalized supports serving dual roles as surface tether and necessary co-catalyst, massively increasing total turnovers as compared to the homogeneous analog.

Full text of DOI:10.1039/b920391e

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
A heterogeneous, selective oxidation catalyst based on Mn
triazacyclononane grafted under reaction conditionsw
Nicholas J. Schoenfeldt, Andrew W. Korinda and Justin M. Notestein*
Received (in Berkeley, CA, USA) 8th October 2009, Accepted 19th January 2010
First published as an Advance Article on the web 5th February 2010
DOI: 10.1039/b920391e
A unique method has been developed to synthesize an active
heterogeneous oxidation catalyst by the in situ grafting of a
1,4,7-trimethyl-1,4,7-triazacyclononane manganese complex on
carboxylic acid-functionalized supports serving dual roles as
surface tether and necessary co-catalyst, massively increasing
total turnovers as compared to the homogeneous analog.
attachment of P, N, or S ligands, or strong electrostatic
adsorption of an otherwise proficient homogeneous catalyst,
and often results in a diminishment of reactivity or selectivity.5
Also, immobilizing active homogeneous catalysts can create
‘Trojan horse’ effects where the solid is a reservoir for active
soluble species.6 Our approach instead combines complex 1
with co-catalytic CA-functionalized SiO2 gel (S–CA), fumed
SiO2 (FS–CA) or Al2O3 gel (A–CA). Under reaction
conditions, 1 grafts itself to the surface creating a solid catalyst
(20) that is more efficient for cyclooctene oxidation with
aqueous H2O2 when in the grafted form.
The selective oxidation of olefins to cis-diols or epoxides remains
of great industrial importance.1 Manganese 1,4,7-trimethyl-1,4,7-
triazacyclononane (tmtacn) complexes, e.g. 1, were first reported
in the late 1980’s and were quickly realized to be powerful low
temperature oxidation catalysts.2 Mn(IV,IV) complex 1 efficiently
catalyzes alkene oxidation with H2O2 in the presence of carboxylic
acid (CA) co-catalysts that suppress the inherent catalase activity
of 1 and enable tuning of the catalyst selectivity towards epoxide
or cis-diol.3 The exact role of the co-catalyst and the precise
nuclearity of the active species in homogeneous systems remain
under investigation, but it has been suggested that the system
1–H2O2–CA forms a reduced bis(m-carboxylato)–Mn2III complex
2 that is the resting state for alkene oxidation.3b,c Although
explored less than for other molecular oxidation catalysts, a route
to heterogenized Mn triazacyclononane catalysts supported on
SiO2 has been developed via modification of the ligand,4 but its
mechanism of reaction also remains incompletely understood.
Seizing on the requirement of CA co-catalysts for efficient
H2O2 utilization in many homogeneous systems, we have
developed a novel protocol for immobilization. Catalysts
‘heterogenized’ on solid supports often rely on covalent
Complex 1 was synthesized from tmtacn, MnCl2ꢀ4H2O,
and KPF6.7 Complex
2 was synthesized from tmtacn,
Mn(OAc)3ꢀ2H2O, CH3COONa and KPF6.7c CA-functionalized
supports, S–CA, FS–CA and A–CA, were synthesized by
grafting and subsequent deprotection of 2-(carbomethoxy)-
ethyl-trimethoxysilane.8 The silane precursor concentration
was varied to synthesize supports with molar loadings rangiꢁn2g
from 0.06–0.66 mmolCA(gSiO ) ) .
ꢁ1 or 0.07–0.80 CA(nmsurface
2
Supported structures and their loadings were verified by 13C
CP/MAS NMR spectroscopy and thermogravimetric analysis.
Alkene oxidation was performed in sealed batch or semi-batch
reactors in acetonitrile (MeCN) at 0 1C with controlled
amounts of S–CA, FS–CA, A–CA, or valeric acid, o-dichloro-
benzene internal GC standard, and 30% aqueous H2O2.
Air was not excluded. Full synthesis, characterization, and
catalysis details are found in the ESIw.
Combining S–CA or FS–CA with 1 yielded substantially more
total turnover numbers (TON = molproduct mol1ꢁ1) to cis-
cyclooctanediol and cis-cyclooctane epoxide products than 1
alone or 1 with un-functionalized SiO2 or Al2O3 (Table 1). In
the absence of effective CA co-catalysts, Mn tmtacn complexes
are known to rapidly decompose H2O2.2a,3a,9 Thus, total TON
are a function of the initial concentrations and the relative rate
constants for oxidation vs. H2O2 decomposition, assuming
the catalytic cycles progress through the same intermediate
(see ESIw). S–CA co-catalysts led to B10ꢂ higher total TON
and relative oxidation rate constants than its closest homo-
geneous analog, valeric acid, when similar amounts of co-catalyst
were present, either alone or with unmodified SiO2. Formation of
the catalyst in situ creates a much more active catalyst than
deliberate grafting and isolation of the air-stable catalyst
(see ESIw) suggesting that the surface species is a true catalyst
intermediate rather than another isolable pre-catalyst form. Slow
addition of H2O2 in semi-batch reactions, which is typically
employed for this class of catalyst, yields 450 total TON towards
cis-diol and epoxide with only 2 eq. S–CA. Slightly higher
catalyst loadings enable B100% cyclooctene conversion at the
expense of TON (see ESIw). In the absence of alkenes, S–CA
Northwestern University, Department of Chemical and Biological
Engineering, 2145 Sheridan Rd., Tech. E136, Evanston, IL, USA.
E-mail: j-notestein@northwestern.edu; Fax: +1 847 491 3728;
Tel: +1 847 491 5357
w Electronic supplementary information (ESI) available: Synthesis,
catalysis, and characterization procedures; solution and diffuse reflectance
UV-visible spectra, NMR spectra, and thermogravimetric analyses of
solid materials. See DOI: 10.1039/b920391e
ꢃc
This journal is The Royal Society of Chemistry 2010
1640 | Chem. Commun., 2010, 46, 1640–1642

View Article Online
Table 1 Alkene oxidation with 1 and various co-catalysts under equivalent conditionsa
% Epoxide : % cis-diol : % otherc
ꢁ1
Co-catalyst
R/molCA molMn
Reactant
Total TON
None
SiO2
Al2O3
S–CA
0
0
0
1
10
2
1
10
1
10
10
500
10
10
10
10
10
10
10
10
cis-Cyclooctene
cis-Cyclooctene
cis-Cyclooctene
cis-Cyclooctene
cis-Cyclooctene
cis-Cyclooctene
cis-Cyclooctene
cis-Cyclooctene
cis-Cyclooctene
cis-Cyclooctene
cis-Cyclooctene
cis-Cyclooctene
Cyclohexene
o5
24
87 : 13 : 0
76 : 24 : 0
75 : 25 : 0
66 : 34 : 0
61 : 39 : 0
55 : 45 : 0
58 : 42 : 0
57 : 43 : 0
89 : 11 : 0
86 : 14 : 0
51 : 49 : 0
39 : 61 : 0
63 : 6 : 31c
91 : 3 : 6
75 : 15 : 10
90 : 3 : 7
92 : 3 : 5
83 : 16 : 1
79 : 11d : 10
41 : 55d : 4
o5
147
465
455
198
o5
17
S–CA
S–CAb
FS–CA
A–CA
Valeric acid
Valeric acid/SiO2
Valeric acid
Valeric acid
S–CA
S–CA
S–CA
S–CA
S–CA
13
53
644
835
585
238
609
544
824
248
213
Norbornene
Cyclohexenone
1-Octene
Styrene
a-Methylstyrene
trans-b-Methylstyrene
cis-Stilbene
S–CA
S–CA
S–CA
a
Unless noted, all reactions were performed in a batch reactor at 0 1C with 3 h reaction time, 2.0 mL MeCN, 0.13 mM 1, 0.17 M cis-cyclooctene,
b
c
and 130 mL 30 wt% H2O2. Semi-batch. See ESIw for conditions. Allylic oxidation products: 9% cyclohexenol, 9% radical-derived epoxide,
3% trans-diol, 4% cyclohexenone, 7% other. Mixture of cis- and trans-diol.
d
(3 eq.) reduced the catalase activity of 1 by 20% at 0 1C or 45% at
25 1C (see ESIw). Thus, the supported CA was more effective both
at decreasing catalase activity and increasing oxidation activity.
The enhanced reactivity with solid co-catalysts is likely due
to the high local surface densities of CA, equivalent to an
average CA separation of 0.85 nm, which is more than
adequate to multiply-coordinate Mn even at low total CA
due to the surface chelate effect.10,11 Weak Brønsted acidity of
SiO2 may also assist in activating the H2O2, as for Ti–SiO2.12
Strong support effects on catalyst productivity and selectivity
were observed, demonstrating a role for the oxide support beyond
that of inert spectator. Co-catalyst A–CA did not increase total
TON above that of 1 alone, in contrast to the large increases in
productivity for functionalized SiO2. Functionalized, non-porous
fumed SiO2 (FS–CA) exhibited slightly higher productivity than
S–CA at low co-catalyst ratios. FS–CA and A–CA have 0.45
with decreasing ring size, and the higher yields with styrene
and n-octene than one might expect for monosubstituted alkenes.
cis-Stilbene, which is both bulky and relatively electron deficient,
has a low yield. Multiply substituted alkenes generally give lower
selectivity to epoxide. Only 3% selectivity to trans-cyclohexane-
diol was observed, indicating low epoxide hydrolysis activity.
The effects of CA surface loading and co-catalyst ratio
R on catalyst productivity are given in Fig. 1. For constant
amounts of Mn, total TON increased sharply up to R E
1–1.5 molCA molMnꢁ1, and rose more slowly at greater ratios.
For constant CA, H2O2 utilization E was maximized near
equimolar CA and Mn (inset, F = 0.5). Low Mn loadings are
kinetically limited, and H2O2 decomposition dominates with
excess Mn. For S–CA, changing CA surface loading did not
have a significant effect on total TON, suggesting that the CA
groups may cluster more than is implied by their average surface
density. The effects of surface density remain under investigation.
Elemental analysis indicates that >75% of the Mn is complexed
with R = 2.0, UV-visible spectra indicate loss of >95% of 1 from
solution with R = 1.5, and even at R = 1.0, the solution above
the S–CA solids displayed no oxidation productivity after being
removed from the solids, even with additional H2O2 or S–CA
added to the decanted solution (see ESIw). These results indicate
that the curve shape in Fig. 1 is due to an increasing fraction of
productive surface Mn complexes with increasing S–CA content,
rather than simple trapping of an active complex. This ability to
assign two-electron oxidation activity primarily to the surface
species is in contrast with the free-radical, metal-initiated
oxidations observed for the limited prior examples of metals
combined with CA-functionalized supports.13
and >0.8 CA(nmsurface)
ꢁ2, respectively, comparable to S–CA, so
TON are not likely influenced by geometric ability to form the
active structure 20. Differences in TON and selectivity among
supports are hypothesized to be due to the density and pKa of
residual surface hydroxyls, which is under investigation.
For no co-catalyst, low levels of valeric acid, or un-
functionalized SiO2 or Al2O3, none of which interact strongly
with 1, intrinsically low selectivity to cis-cyclooctanediol was
observed. All functionalized supports gave relatively constant
cis-diol selectivitiesꢁ1near 40% for the co-catalyst ratios
R = molCA molMn investigated, suggesting similar active
species for all SiO2-supported catalysts.
High oxidation productivity was observed for diverse alkenes.
Total TON to epoxide and diol, that is, the relative rate of
oxidation vs. H2O2 decomposition, increases with increasing
electron density at the double bond and with decreasing
steric congestion. Cyclohexenone has a low yield of oxidation
products, as expected from the electron-withdrawing ketone. The
dependence on sterics is demonstrated by higher yields with
a-methylstyrene than with b-methylstyrene, increasing yields
The structure of the Mn species in the catalyst cycle remains
a subject of debate,3 however the stoichiometry at maximum
TON is consistent with either a dimeric surface structure with
two to three bridging and surface-linking CA groups (e.g. 20)
or a monomeric surface structure with one surface-linking
CA group per monomer. However, electronic spectra of the
ꢃc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 1640–1642 | 1641

View Article Online
the analogy can be made between the active sites in this work and
those of many oxidation enzymes, in which a ‘metal cofactor’
(e.g. 1) is embedded in a coordinating scaffold (e.g. S–CA) to
create a new, more productive catalyst (e.g. 20) whose reactivity is
sensitive to subtle changes to the scaffold structure.
The authors thank Evonik Degussa for Aeroperl fumed
SiO2 and Dow Chemical and the Dow Methane Challenge, the
Camille and Henry Dreyfus New Faculty Awards Program,
and Northwestern University for financial support.
Notes and references
1 (a) R. A. Sheldon and J. K. Kochi, Metal Catalyzed Oxidations of
Organic Compounds, Academic Press, New York, 1981;
(b) E. M. McGarrigle and D. G. Gilheany, Chem. Rev., 2005,
105, 1563–1602; (c) D. Mandelli, K. B. Voitiski, U. Schuchardt and
G. B. Shul’pin, Chem. Nat. Compd., 2002, 38, 243–245.
2 (a) K. Wieghardt, U. Bossek, B. Nuber, J. Weiss, J. Bonvoisin,
M. Corbella, S. E. Vitols and J. J. Girerd, J. Am. Chem. Soc., 1988,
110, 7398–7411; (b) R. Hage, J. E. Iburg, J. Kerschner, J. H. Koek,
E. L. M. Lempers, R. J. Martens, U. S. Racheria, S. W. Russell,
T. Swarthoff, M. R. P. van Vliet, J. B. Warnaar, L. V. D. Wolf and
B. Krijnen, Nature, 1994, 369, 637–639; (c) V. C. Quee-Smith,
L. DelPizzo, S. H. Jureller, J. L. Kerschner and R. Hage, Inorg.
Chem., 1996, 35, 6461–6465; (d) D. De Vos and T. Bein, Chem.
Commun., 1996, 917–918.
3 (a) D. E. De Vos, B. F. Sels, M. Reynaers, Y. V. Subba Rao and
P. A. Jacobs, Tetrahedron Lett., 1998, 39, 3221–3224; (b) J. W. De
Boer, P. L. Alsters, A. Meetsma, R. Hage, W. R. Browne and
B. L. Feringa, Dalton Trans., 2008, 6283–6295; (c) H. Killic,
W. Adam and P. L. Alsters, J. Org. Chem., 2009, 74, 1135–1140.
4 (a) D. E. De Vos, S. D. Wildeman, B. F. Sels, P. J. Grobet and
P. A. Jacobs, Angew. Chem., Int. Ed., 1999, 38, 980–983;
(b) D. E. De Vos and P. A. Jacobs, Catal. Today, 2000, 57,
105–114; (c) Y. V. S. Rao, D. E. De Vos, T. Bein and
P. A. Jacobs, Chem. Commun., 1997, 355–356.
5 Some relevant reviews are: (a) A. Choplin and F. Quignard, Coord.
Chem. Rev., 1998, 180, 1679–1702; (b) M. Dusi, T. Mallat and
A. Baiker, Catal. Rev. Sci. Eng., 2000, 42, 213–278; (c) D. E. De
Vos, M. Dams, B. F. Sels and P. A. Jacobs, Chem. Rev., 2002, 102,
3615–3640; (d) M. H. Valkenberg and W. F. Holderich, Catal. Rev.
Sci. Eng., 2002, 44, 321–374; (e) A. Corma, Catal. Rev. Sci. Eng.,
2004, 46, 369–417; (f) R. Augustine, S. Tanielyan, S. Anderson,
Y. Gao, P. Goel, N. Mahata, I. Nair, C. Reyes, H. Yang and
A. Zsigmond, Methodologies in Asymmetric Catalysis, ACS
Symp. Ser., 2004, 880, 15–28; (g) J. M. Notestein and A. Katz,
Chem.–Eur. J., 2006, 12, 3954–3965.
Fig. 1 Effect of CA to Mn molar ratio and CA surface density of S–CA
on total TON to epoxide and cis-diol at a constant 2.7 ꢄ 0.3 ꢂ 10ꢁ3 mmol
1 under batch reactor conditions. (Inset) Effect of mole fraction Mn vs. CA
on H2O2 utilization at constant 2.9 ꢄ 0.4 ꢂ 10ꢁ3 mmol surface CA under
batch reactor conditions. Lines are includedꢁa1s a guide to the eye. D 0.06,
B 0.14, J 0.25, and & 0.51 mmolCA(gSiO
) .
2
surface species formed under reaction conditions are more
consistent with the former. In the absence of oxidant, solution
and diffuse reflectance UV-visible spectra show 1 disappearing
from MeCN–H2O solution (see ESIw) and adsorbing without
change onto S–CA (Fig. 2c and d). In contrast, with H2O2
present, 1 disappears more rapidly from solution and the solid-
supported species shows a characteristic shoulder at B320 nm,
virtually identical to 2 in solution (Fig. 2a and b). Diffuse
reflectance UV-visible spectra of A–CA with adsorbed 1 are
structureless indicating the absence of a well-defined Mn
complex (see ESIw), in line with the lack of catalyst activity.
In conclusion, carboxylic acids grafted onto SiO2 (e.g. S–CA)
are an especially effective co-catalyst for selective oxidation with
tmtacn manganese(IV) dimer 1. The oxidant-dependant reaction
between 1 and the supported CA synthesizes a heterogeneous
catalyst (20) under reaction conditions, and this heterogeneous
catalyst has >10ꢂ productivity for cyclooctene oxidation vs.
H2O2 decomposition than for the same ratios of analogous
homogeneous co-catalyst valeric acid. Many non-heme oxidation
catalysts require carboxylic acid co-catalysts or additives for
selective oxidation of various substrates,14 and this communica-
tion demonstrates a generalizable route to the synthesis of
molecularly well-defined, solid-supported versions of this type
of catalyst that avoids synthetic modification of the ligand, which
can complicate catalyst development and understanding. Finally,
6 R. A. Sheldon, M. Wallau, I. W. C. E. Arends and U. Schuchardt,
Acc. Chem. Res., 1998, 31, 485–493.
7 (a) G. H. Searle and R. J. Geue, Aust. J. Chem., 1984, 37, 959–970;
(b) S. A. Madison and D. J. Batal, Lever Brothers Co., 1994,
US Pat., 5 284 944; (c) J. H. Koek, S. W. Russell, L. van der Wolf,
R. Hage, J. B. Warnaar, A. L. Spek, J. Kerschner and L. DelPizzo,
J. Chem. Soc., Dalton Trans., 1996, 353–362.
8 J. C. McKeen, Y. S. Yan and M. E. Davis, Chem. Mater., 2008, 20,
5122–5124.
9 (a) A. J. Wu, J. E. Penner-Hahn and V. L. Pecoraro, Chem. Rev.,
2004, 104, 903–938; (b) A. E. M. Boelrijk, S. V. Khangulov and
G. C. Dismoukes, Inorg. Chem., 2000, 39, 3009–3019.
ꢁ1
10 A typical surface coverage of 0.60 mmolCA(gSiO
)
is equivalent to
one group per 0.85 nm in any direction, or 2.7 m2mol cmꢁ3 = 2.7 M.
In the batch reactors, 500 equivalents of valeric acid is 65 mM.
11 A. Mulder, T. Auletta, A. Sartori, S. Del Ciotto, A. Casnati,
R. Ungaro, J. Huskens and D. N. Reinhoudt, J. Am. Chem. Soc.,
2004, 126, 6627–6636.
12 J. M. Notestein, A. Solovyov, L. R. Andrini, F. G. Requejo,
A. Katz and E. Iglesia, J. Am. Chem. Soc., 2007, 129, 15585–15595.
13 (a) T. Maschmeyer, R. D. Oldroyd, G. Sankar, J. M. Thomas,
I. J. Shannon, J. A. Klepetko, A. F. Masters, J. K. Beattie and
C. R. A. Catlow, Angew. Chem., Int. Ed. Engl., 1997, 36,
1639–1642; (b) A. J. Butterworth, J. H. Clark, P. H. Walton and
S. J. Barlow, Chem. Commun., 1996, 1859–1860.
Fig. 2 Normalized UV-visible spectra of (a) 2 in MeCN, (b) 20 via
self-catalyzed grafting of 1 on S–CA from MeCN–H2O2, (c) 1 in
MeCN, and (d) 1 adsorbed on S–CA from MeCN–H2O.
14 L. Que and W. B. Tolman, Nature, 2008, 455, 333–340.
ꢃc
This journal is The Royal Society of Chemistry 2010
1642 | Chem. Commun., 2010, 46, 1640–1642