Olefin cis-dihydroxylation with bio-inspired iron catalysts. evidence for an FeII/FeIV catalytic cycle

Article abstract of DOI:10.1021/ja1021014

Iron(II) complexes of a series of N-acylated dipyridin-2-ylmethylamine ligands (R-DPAH) have been investigated as catalysts for the cis-dihydroxylation of olefins to model the action of Rieske dioxygenases that catalyze arene cis-dihydroxylation. The Rieske dioxygenases have a mononuclear iron active site coordinated to a 2-histidine-1-carboxylate facial triad motif. The R-DPAH ligands are designed to provide a facial N,N,O-ligand set that mimics the enzyme active site. The iron(II) complexes of the R-DPAH ligands activate H ₂O₂ to effect the oxidation of olefin substrates into cis-diol products. As much as 90% of the H₂O₂ oxidant is converted into cis-diol, but a large excess of olefin is required to achieve the high conversion efficiency. Reactivity and mechanistic comparisons with the previously characterized Fe(TPA)/H₂O₂ catalyst/oxidant combination (TPA = tris(pyridin-2-ylmethyl)amine) lead us to postulate an Fe^II/Fe^IV redox cycle for the Fe(R-DPAH) catalysts in which an Fe^IV(OH)₂ oxidant carries out the cis-hydroxylation of olefins. This hypothesis is supported by three sets of observations: (a) the absence of a lag phase in the conversion of the H ₂O₂ oxidant into a cis-diol product, thereby excluding the prior oxidation of the Fe(II) catalyst to an Fe(III) derivative as established for the Fe(TPA) catalyst; (b) the incorporation of H₂¹⁸O into the cis-diol product, thereby requiring O-O bond cleavage to occur prior to cis-diol formation; and (c) the formation of cis-diol as the major product of cyclohexene oxidation, rather than the epoxide or allylic alcohol products more commonly observed in metal-catalyzed oxidations of cyclohexene, implicating an oxidant less prone to oxo transfer or H-atom abstraction.

Full text of DOI:10.1021/ja1021014

Published on Web 11/24/2010
Olefin Cis-Dihydroxylation with Bio-Inspired Iron Catalysts.
Evidence for an FeII/FeIV Catalytic Cycle
Paul D. Oldenburg, Yan Feng, Iweta Pryjomska-Ray, Daniel Ness, and
Lawrence Que, Jr.*
Department of Chemistry and Center for Metals in Biocatalysis, UniVersity of Minnesota,
Minneapolis, Minnesota 55455, United States
Received March 11, 2010; E-mail: larryque@umn.edu
Abstract: Iron(II) complexes of a series of N-acylated dipyridin-2-ylmethylamine ligands (R-DPAH) have
been investigated as catalysts for the cis-dihydroxylation of olefins to model the action of Rieske
dioxygenases that catalyze arene cis-dihydroxylation. The Rieske dioxygenases have a mononuclear iron
active site coordinated to a 2-histidine-1-carboxylate facial triad motif. The R-DPAH ligands are designed
to provide a facial N,N,O-ligand set that mimics the enzyme active site. The iron(II) complexes of the R-DPAH
ligands activate H2O2 to effect the oxidation of olefin substrates into cis-diol products. As much as 90% of
the H2O2 oxidant is converted into cis-diol, but a large excess of olefin is required to achieve the high
conversion efficiency. Reactivity and mechanistic comparisons with the previously characterized Fe(TPA)/
H2O2 catalyst/oxidant combination (TPA ) tris(pyridin-2-ylmethyl)amine) lead us to postulate an FeII/FeIV
redox cycle for the Fe(R-DPAH) catalysts in which an FeIV(OH)2 oxidant carries out the cis-hydroxylation
of olefins. This hypothesis is supported by three sets of observations: (a) the absence of a lag phase in the
conversion of the H2O2 oxidant into a cis-diol product, thereby excluding the prior oxidation of the Fe(II)
catalyst to an Fe(III) derivative as established for the Fe(TPA) catalyst; (b) the incorporation of H218O into
the cis-diol product, thereby requiring O-O bond cleavage to occur prior to cis-diol formation; and (c) the
formation of cis-diol as the major product of cyclohexene oxidation, rather than the epoxide or allylic alcohol
products more commonly observed in metal-catalyzed oxidations of cyclohexene, implicating an oxidant
less prone to oxo transfer or H-atom abstraction.
Many iron-based enzymes catalyze the stereospecific oxida-
tion of CdC bonds.1,2 The most extensively studied of these,
the cytochromes P450, have active sites consisting of a heme
that is attached to the protein backbone via an iron-coordinating
cysteinate residue.3,4 More recently, nonheme iron enzymes have
become better characterized and have been shown to promote
novel oxidative chemistry.2,5 Of particular interest is an enzyme
family called the Rieske dioxygenases, which catalyze the cis-
dihydroxylation of CdC bonds in the biodegradation of arenes
in the environment.6,7 These enzymes have a mononuclear iron
Figure 1. NDO active site.
center that is ligated by a 2-His-1-carboxylate facial triad,8,9
a
iron enzymes;10,11 as an example, the active site of naphthalene
1,2-dioxygenase (NDO) is illustrated in Figure 1.8 There are
two potential cis-coordination sites on the iron octahedron and
these sites are used for the side-on binding of O2, as demon-
strated by X-ray crystallography.12 On the basis of biochemical
studies, this O2 adduct is proposed to be an iron(III)-peroxo
intermediate13,14 that either attacks a substrate directly or
recurring motif among oxygen activating mononuclear nonheme
(1) Bertini, I.; Gray, H. B.; Stiefel, E. I.; Valentine, J. S., Eds. Biological
Inorganic Chemistry. Structure and ReactiVity; University Science
Books: Sausalito, CA, 2007.
(2) Costas, M.; Mehn, M. P.; Jensen, M. P.; Que, L., Jr. Chem. ReV. 2004,
104, 939–986.
(3) Groves, J. T. In Cytochrome P450: Structure, Mechanism, and
Biochemistry, 3rd ed.; Ortiz de Montellano, P. R., Ed.; Kluwer
Academic/Plenum Publishers: New York, 2005; pp 1-43.
(4) Nam, W.; Ryu, Y. O.; Song, W. J. J. Biol. Inorg. Chem. 2004, 9,
654–660.
(9) Ferraro, D. J.; Gakhar, L.; Ramaswamy, S. Biochem. Biophys. Res.
Commun. 2005, 338, 175–190.
(5) Abu-Omar, M. M.; Loaiza, A.; Hontzeas, N. Chem. ReV. 2005, 105,
2227–2252.
(10) Hegg, E. L.; Que, L., Jr. Eur. J. Biochem. 1997, 250, 625–629.
(11) Koehntop, K. D.; Emerson, J. P.; Que, L., Jr. J. Biol. Inorg. Chem.
2005, 10, 87–93.
(6) Gibson, D. T. Microbial Degradation of Organic Compounds; Marcel
Dekker: New York, 1984.
(12) Karlsson, A.; Parales, J. V.; Parales, R. E.; Gibson, D. T.; Eklund,
H.; Ramaswamy, S. Science 2003, 299, 1039–1042.
(7) Gibson, D. T.; Parales, R. E. Curr. Opin. Biotechnol. 2000, 11, 236–
243.
(13) Wolfe, M. D.; Altier, D. J.; Stubna, A.; Popescu, C. V.; Mu¨nck, E.;
Lipscomb, J. D. Biochemistry 2002, 41, 9611–9626.
(8) Kauppi, B.; Lee, K.; Carredano, E.; Parales, R. E.; Gibson, D. T.;
Eklund, H.; Ramaswamy, S. Structure 1998, 6, 571–586.
(14) Wolfe, M. D.; Lipscomb, J. D. J. Biol. Chem. 2003, 278, 829–835.
9
10.1021/ja1021014  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 17713–17723 17713

A R T I C L E S
Oldenburg et al.
Scheme 1. Ligands Used for Iron-Based Oxidation Catalysis
undergoes O-O bond heterolysis to form a HOsFeVdO species
that carries out the cis-dihydroxylation.13,15
oxidant. The demonstration of iron-catalyzed cis-hydroxyl-
ation using H2O2 as an oxidant is a particularly significant
development, as iron centers typically react with H2O2 to
generate hydroxyl radicals that give rise to nonspecific
oxidation products. The most effective iron complexes thus
far are ones that use tetradentate N4 ligands and have two
available cis-oriented coordination sites that are proposed to
facilitate O-O bond cleavage (Scheme 1, complexes 10 and
11).18,20 These complexes catalyze both the epoxidation and
cis-dihydroxylation of olefins by H2O2. At present however,
the reaction conditions that elicit the best yields of cis-diols
require (a) syringe-pump introduction of the H2O2 oxidant
to minimize HO• formation and (b) the use of a large excess
of the olefin substrate in order to trap the highly reactive
metal-based oxidant responsible for the highly stereoselective
cis-dihydroxylation. These systems are nevertheless worth
investigation to provide a basis for the subsequent develop-
ment of catalysts suitable for applications in synthetic and
pharmaceutical chemistry.
Olefin cis-dihydroxylation is an important chemical trans-
formation in both natural product and drug syntheses and
typically involves the use of OsO4.16,17 The established key role
of the mononuclear iron center in the Rieske dioxygenases has
thus spurred efforts to design nonheme iron catalysts that mimic
the action of these enzymes and develop a more environmental-
ly benign alternative to the osmium-based dihydroxylations.
Indeed, examples of such synthetic iron18-20 as well as
manganese catalysts21,22 have been reported using H2O2 as an
(15) Bassan, A.; Blomberg, M. R. A.; Siegbahn, P. E. M. J. Biol. Inorg.
Chem. 2004, 9, 439–452.
(16) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. ReV.
1994, 94, 2483–2547.
(17) Ojima, I. Catalytic Asymmetric Synthesis, 2nd ed.; Wiley-VCH: New
York, 2000.
(18) Chen, K.; Costas, M.; Kim, J.; Tipton, A. K.; Que, L., Jr. J. Am. Chem.
Soc. 2002, 124, 3026–3035.
(19) Costas, M.; Que, L., Jr. Angew. Chem., Int. Ed. 2002, 41, 2179–2181.
(20) Oldenburg, P. D.; Que, L., Jr. Catal. Today 2006, 117, 15–21.
(21) De Vos, D. E.; de Wildeman, S.; Sels, B. F.; Grobet, P. J.; Jacobs,
P. A. Angew. Chem., Int. Ed. 1999, 38, 980–983.
(22) de Boer, J. W.; Brinksma, J.; Browne, W. R.; Meetsma, A.; Alsters,
P. L.; Hage, R.; Feringa, B. L. J. Am. Chem. Soc. 2005, 127, 7990–
7991.
One major difference between the above complexes and the
Rieske dioxygenase enzymes is the presence of a carboxylate
ligand in the iron coordination environment of the latter. This
9
17714 J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010

Olefin Cis-Dihydroxylation with Bio-Inspired Fe Catalysts
A R T I C L E S
prompted us and others23-27 to explore polydentate ligands with
a N,N,O donor set to assess how the introduction of an oxygen
ligand can affect the performance of the iron catalyst (Scheme
1, species 1-5 and 7-9). In our initial efforts, we designed
Ph-DPAH,25 a facial tridentate ligand providing a donor set with
two pyridines and an amide carbonyl, and showed that its
iron(II) complex 1 catalyzed the cis-dihydroxylation of 1-octene,
resulting in an unprecedentedly high 76% conversion of the
H2O2 oxidant to product. This paper reports on further explora-
tions of the catalytic chemistry of 1 and of related complexes
(Scheme 1), as well as detailed mechanistic studies that reveal
a novel oxidative mechanism involving an FeII/FeIV cycle.
Vacuo, yielding a brown oil. This oil was redissolved into CH2Cl2
and washed with 3 × 50 mL fractions of 1 M HCl. The collected
aqueous phases were rendered basic (pH ≈ 8) by treatment with
sat. NaHCO3 solution. Product was then extracted into CH2Cl2 and
dried by Na2SO4. Solvent was removed in Vacuo, yielding a light-
brown oil. Purified H3C-DPAH was obtained as a white powder
after recrystallization from hot MeOH in 30% yield. 1H NMR
(CDCl3, δ ppm from TMS): 8.55 (d, 2H), 7.91 (d, 1H), 7.64 (td,
2H), 7.43 (d, 2H), 7.16 (d, 2H), 6.22, (d, 1H), 2.15 (s, 3H). 13C
NMR (CDCl3): 169.61, 158.90, 149.27, 136.88, 122.51, 122.25,
59.08, 23.44. IR (in CD3CN soln): ν(CdO) 1677 cm-1
.
Synthesis of (Di-(2-pyridyl)methyl)trifluoroacetamide (F3C-
DPAH). To a solution containing 1.15 g (6.2 mmol) of di-(2-
pyridyl)methylamine in 3 mL of pyridine was added a solution of
0.52 mL (6.8 mmol) trifluoroacetic acid in 8 mL pyridine. This
mixture was heated to 80 °C, at which point a solution of 2.2 mL
(8.5 mmol) triphenyl phosphite in 5 mL of pyridine was added
over a two-hour period. This mixture was stirred for 24 h at 80 °C,
after which the solution was cooled to room temperature and solvent
was removed in Vacuo, yielding a brown oil. Purified F3C-DPAH
was obtained from a silica gel column eluted with a hexanes/ethyl
acetate gradient in 58% yield. 1H NMR (CDCl3, δ ppm from TMS):
9.07 (d, 1H), 8.57 (d, 2H), 7.67 (td, 2H), 7.40 (d, 2H), 7.21 (ddd,
2H), 6.17, (d, 1H). 13C NMR (CDCl3): 156.71, 149.18, 137.15,
122.97, 121.88, 117.69, 58.77. IR (in CD3CN soln): ν(CdO) 1728
Experimental Section
Materials and Synthesis. All reagents were purchased from
Aldrich and used as received unless otherwise noted. All olefin
substrates were passed over basic alumina immediately prior to use.
CH3CN was distilled from CaH2. H218O2 (90% 18O-enriched, 2 wt
% solution in H216O) and H218O (95% 18O enriched) were obtained
from ICON Isotopes. The syntheses of the di-(2-pyridyl)methyl-
amine synthon28 and [FeII(OTf)2 ·2NCCH3]29 have been reported
previously.
Syntheses of (Di-(2-pyridyl)methyl)-4-methoxybenzamide (4-
MeO-C6H4-DPAH) and (Di-(2-pyridyl)methyl)-4-(trifluoro-
methyl)benzamide (4-F3C-C6H4-DPAH). To a solution containing
either 7.8 mmol of 4-methoxybenzoyl chloride or 7.8 mmol of
4-(trifluoromethyl)benzoyl chloride and 1.3 mL of triethylamine
in 20 mL of THF was slowly added 1.46 g (7.8 mmol) of di-(2-
pyridyl)methylamine at 0 °C. With stirring, this mixture was
warmed to 60 °C and heated for 20 min, after which the solution
was cooled back to 0 °C and filtered to remove Et3N·HCl. THF
was removed to afford a cream-colored solid, which was redissolved
in CH2Cl2 and washed with NaOH. The organic layer was then
collected, and the solvent was removed in Vacuo. Purified products
were obtained as white powders after recrystallization from hot
MeOH. 4-MeO-C6H4-DPAH (70% yield): 1H NMR (CDCl3, δ ppm
from TMS): 8.68 (d, 1H), 8.58 (dm, 2H), 7.94 (dm, 2H), 7.66 (td,
2H), 7.5 (d, 4H), 7.17, (ddd, 2H), 6.39 (d, 1H), 3.86 (s, 3H). 13C
NMR (CDCl3): 166.13, 162.28, 159.11, 149.23, 136.91, 129.13,
126.59, 127.52, 122.25, 113.71, 59.32, 55.43. IR (in CD3CN soln):
cm-1
.
Syntheses of [FeII(4-MeO-C6H4-DPAH)2](OTf)2 (2), [FeII(4-
F3C-C6H4-DPAH)2](OTf)2 (3), [FeII(H3C-DPAH)2](OTf)2 (4),
and [FeII(F3C-DPAH)2](OTf)2 (5). In a N2-containing glovebox, a
mixture of the R-DPAH ligand (1.0 mmol) and FeII(OTf)2 ·2NCCH3
(0.5 mmol) was stirred for 5 h in 10 mL of CH3CN. A yellow-
green powder formed. Solvent was removed in Vacuo, and CH3CN
was added until the resulting solid was completely dissolved. For
2 and 3, vapor diffusion of Et2O into this solution at -20 °C resulted
in the formation of purified 2 after 5 days as bright green, block-
shaped single crystals in a 62% yield and purified 3 after 7 days as
bright yellow-green, plate-shaped single crystals in a 53% yield,
both suitable for X-ray crystallographic analysis. See Supporting
Information for details regarding X-ray crystallographic analysis
and Tables S1 and S2 for crystal data and structure refinement for
2 and 3. For 4 and 5, layering Et2O over the CH3CN solution resulted in
the formation of purified product after 1 day at -20 °C as a yellow-green
powder in 58% yield for 4 and 65% yield for 5. Characterization of 2:
ESI/MS: m/z 843 ([Fe(MeOC6H4-DPAH)2(OTf)]+), 524 ([Fe(MeOC6H4-
DPAH)(OTf)]+), 347 ([Fe(MeOC6H4-DPAH)2]2+), 320 ([(MeOC6H4-
DPAH)+H]+). Anal. Calcd (found) for C40H34F6FeN6O10S2: C, 48.40
(48.26); H, 3.45 (3.43); N, 8.47 (8.53); F, 11.48 (11.54). IR (in CD3CN
soln): ν(CdO) 1618 cm-1. Characterization of 3: ESI/MS: m/z
562 ([Fe(F3CC6H4-DPAH)(OTf)]+), 385 ([Fe(F3CC6H4-DPAH)2]2+),
358 ([(F3CC6H4-DPAH)+H]+). Anal. Calcd (found) for
C39H27F12FeN6O8S2: C, 44.96 (44.50); H, 2.64 (2.59); N, 7.86 (7.85);
1
ν(CdO) 1657 cm-1. 4-F3C-C6H4-DPAH (65% yield): H NMR
(CDCl3, δ ppm from TMS): 8.94 (d, 1H), 8.59 (d, 2H), 8.08 (d,
2H), 7.71 (m, 4H), 7.51 (d, 2H), 7.21, (ddd, 2H), 6.42 (d, 1H). 13
C
NMR (CDCl3): 165.20, 158.31, 149.07, 137.35, 136.97, 132.81,
127.65, 125.44, 122.61, 122.16, 59.16. IR (in CD3CN soln):
ν(CdO) 1668 cm-1
.
Synthesis of (Di-(2-pyridyl)methyl)acetamide (H3C-DPAH).
To a solution containing 0.88 g (4.8 mmol) of di-(2-pyridyl)-
methylamine in 2 mL of pyridine was added a solution of 0.3 mL
(5.2 mmol) of acetic acid in 5 mL of pyridine. This mixture was
heated to 80 °C, at which point a solution of 1.7 mL (6.5 mmol) of
triphenyl phosphite in 3 mL of pyridine was added over a 2 h period.
This mixture was stirred for 24 h at 80 °C, after which the solution
was cooled to room temperature and the solvent was removed in
F, 21.33 (21.32). IR (in CD3CN sol’n): ν(CdO) 1632 cm-1
.
Characterization of 4: ESI/MS: m/z 659 ([Fe(H3C-DPAH)2(OTf)]+),
432 ([Fe(H3C-DPAH)(OTf)]+), 255 ([Fe(H3C-DPAH)2]2+), 228
([(H3C-DPAH)+H]+). Anal. Calcd (found) for C28H26F6FeN6O8S2:
C, 41.60 (41.72); H, 3.24 (3.14); N, 10.39 (10.50); F, 14.10 (14.35).
IR (in CD3CN soln): ν(CdO) 1630 cm-1. Characterization of 5:
ESI/MS: m/z 617 ([Fe(F3C-DPAH)2-H]+), 486 ([Fe(F3C-DPA-
H)(OTf)]+), 377 ([Fe(F3C-DPAH)(NCCH3)-H]+), 304 ([(F3C-
DPAH)+Na]+), 282 ([(F3C-DPAH)+H]+). Anal. Calcd (found) for
C28H20F12FeN6O8S2 ·H2O: C, 36.70 (36.52); H, 2.20 (2.15); N, 9.17
(23) Beck, A.; Weibert, B.; Burzlaff, N. I. Eur. J. Inorg. Chem. 2001, 521–
527.
(24) Beck, A.; Barth, A.; Hu1bner, E.; Burzlaff, N. Inorg. Chem. 2003,
42, 7182–7188.
(25) Oldenburg, P. D.; Shteinman, A. A.; Que, L., Jr. J. Am. Chem. Soc.
2005, 127, 15672–15673.
(9.03); F, 24.88 (24.75). IR (in CD3CN soln): ν(CdO) 1633 cm-1
.
Instrumentation. 1H and 13C NMR spectra were recorded on a
Varian Unity 300 or 500 MHz spectrometer at ambient temperature.
Chemical shifts (ppm) were referenced to the residual protic solvent
peaks. FTIR spectra were obtained with a Thermo Nicolet Avatar
370 FT-IR instrument. X-ray crystallographic analyses were
completed by mounting the crystal on a Bruker-AXS platform
diffractometer with a CCD area detector and sealed-tube 3-KW
(26) Oldenburg, P. D.; Ke, C.-Y.; Tipton, A. A.; Shteinman, A. A.; Que,
L., Jr. Angew. Chem., Int. Ed. 2006, 45, 7975–7978.
(27) Bruijnincx, P. C. A.; Buurmans, I. L. C.; Gosiewska, S.; Moelands,
M. A. H.; Lutz, M.; Spek, A. L.; Koten, G. v.; Klein Gebbink, R. J. M.
Chem.sEur. J. 2008, 14, 1228–1237.
(28) Renz, M.; Hemmert, C.; Meunier, B. Eur. J. Org. Chem. 1998, 1271–
1273.
(29) Hagen, K. S. Inorg. Chem. 2000, 39, 5867–5869.
9
J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010 17715

A R T I C L E S
Oldenburg et al.
X-ray generators. High-resolution electrospray mass spectral (ESI-
MS) experiments were performed on a Bruker (Billerica, MA)
BioTOF II time-of-flight spectrometer. Product analyses from
catalysis experiments were performed on a Perkin-Elmer AutoSys-
tem gas chromatograph (AT-1701 column, 30 m) with a flame
ionization detector. Gas chromatography/mass spectral analyses
were performed on an HP 6890 GC (HP-5 column, 30 m) with an
Agilent 5973 mass analyzer. A 4% NH3/CH4 mix was used as the
ionization gas for chemical ionization analyses.
Reaction Conditions for Catalytic Oxidations. In a typical
reaction, 10 equiv of H2O2 (diluted from 35% H2O2 solution with
CH3CN resulting in a 70 mM solution) were delivered by syringe
pump over a period of 5 min at 25 °C in air to a vigorously
stirred 3.7 mL solution of 1.3 µmol of iron catalyst and 1000
equiv of olefin substrate in CH3CN. The final concentrations
were 0.35 mM iron catalyst, 3.5 mM H2O2, and 0.35 M olefin.
The solution was stirred for an additional 60 min after syringe
pump addition, after which organic products were esterified by
1 mL of acetic anhydride together with 0.1 mL of 1-methylimi-
dazole and extracted with CHCl3. An internal standard (naph-
thalene) was added, and the solution was washed with 1 M
H2SO4, sat. NaHCO3, and H2O. The organic layer was dried with
Na2SO4 and subjected to GC analysis. The products were
identified by comparison of their GC retention times and GC/
MS with those of authentic compounds.
Isotope Labeling Studies. Similar conditions as those described
above were used for isotope labeling studies except for the following
details. In experiments involving H218O, 1000 equiv of H218O were
added to the commercially available 30% H2O2 solution and
delivered by syringe pump to the catalyst/substrate solution with
the final H218O concentration equaling 0.35 M. In experiments
involving H218O2, 10 equiv of H218O2 (diluted by CH3CN from the
commercially available 2% H218O2/H2O solution, which contains a
1:100 molar ratio of H218O to H216O) were used instead of H2O2,
with the final H218O2 concentration equaling 3.5 mM. The diol
esterification procedure was the same as that diagrammed above.
The data reported either summarize a single reaction or are the
average of 2-3 reactions, and the % 18O values reported were
calculated based on the 18O-enrichment of the reagents containing
the isotope.
Figure 2. ORTEP plots of complexes 2 and 3 with 50% probability
ellipsoids. Hydrogen atoms, counterions, and noncoordinating solvent
molecules have been omitted for clarity.
Table 1. Bond Lengths (Å) Observed for the Iron Centers in 1-3
and 7-9, As Well As Two Forms of Naphthalene 1,2-Dioxygenase
(NDO)
Fe-N2a
Fe-N3a
Fe-O1
Fe-O2
ref
1
2
3
7
2.171
2.167(2)
2.174(2)
2.181
2.187(2)
2.178(2)
2.043
2.043(2)
2.047(2)
25
This work
This work
23, 24
2.169-2.340 2.154-2.353 2.003-2.080
2.122(2)
2.124
8
2.120(2)
2.137
2.1
2.228(2)
2.145
2.2
27
9b
2.150 26
NDOc 2.0
NDOb 2.1
2.6
2.4
8
14
2.0
2.3
a For NDO, N2
) His208 and N3 )
His213. b Structure was
Results and Discussion
obtained for related [FeII(L1)(Cl)] complex. c 1NDO.pdb. d 1O7G.pdb.
Structural Analysis. A series of R-DPAH ligands (R )
4-MeO-C6H4, 4-F3C-C6H4, CH3, and CF3; Scheme 1) can readily
be obtained through the acylation of di(2-pyridyl)methylamine.
These ligands are variations of the previously reported Ph-DPAH
ligand, with each providing a N,N,O donor set. The particular
R groups were picked to modulate the basicity of the carbonyl
oxygen, which in turn may affect the Lewis acidity of the
iron center and its interaction with H2O2. As with [FeII(Ph-
DPAH)2](OTf)2 (1), complexes 2-5 were synthesized via
combination of [FeII(OTf)2(NCMe)2] and 2 equiv of ligand.
Single crystals suitable for X-ray analysis were obtained for 2
and 3, and their structures are shown in Figure 2. Upon iron
coordination, the R-DPAH ligands provide a facial array of two
pyridines and a carbonyl oxygen in a good approximation of
the facial 2-His-1-carboxylate ligand environment found in a
number of oxygen activating mononuclear nonheme iron
enzymes.10,11 The structures of 2 and 3 closely resemble that
of 1,25 with a center of inversion present at each iron center.
Diffraction quality crystals could not be obtained for 4 and 5,
but elemental analyses showed a 1:2 metal-ligand stoichiometry
as found for 1-3.
ligand sets developed respectively by Burzlaff (7)23,24 and
Klein Gebbink (8).27 Complex 9 differs from 7 and 8 in
having an approximately tetragonal N,N,O,O ligand set
derived from two pyridines and a bidentate carboxylate.26
The Fe-Npyridine and Fe-Npyrazole bond distances all fall
between 2.1 and 2.2 Å, typical of high-spin iron(II) com-
plexes. The Fe-Oamide distances of 2.043-2.047 Å observed
for 1-3 are appreciably shorter than the Fe-Oester distance
of 8 but comparable to the Fe-Ocarboxylate distances found for
the complexes of Burzlaff (7). This difference reflects a strong
interaction between the R-DPAH amide carbonyl oxygen
atom with the metal center, comparable to that of a
monodentate carboxylate ligand.
Spectroscopic studies suggest that the complexes remain
for the most part intact upon dissolution in CH3CN but
undergo ligand scrambling. ESI-MS analysis of CH3CN
solutions of the complexes show that the [FeL2]2+ species
represent the dominant peaks in the spectra. For exam-
ple for 1, ions are observed at m/z 783 for {[Fe(Ph-
DPAH)2](OTf)}+ and m/z 317 for {[Fe(Ph-DPAH)2]}2+, but
less intense peaks corresponding to [Fe(Ph-DPAH)(OTf)]+
(m/z 494) and the protonated free ligand (m/z 290) can also
be detected (Figure 3 top). Similar features are observed for
Table 1 compares the bond lengths observed in 1-3 with
those of related complexes 7,23,24 8,27 and 9,26 as well as
those reported for naphthalene dioxygenase (NDO).8,12
Complexes 7 and 8 are iron(II) complexes with facial N,N,O
9
17716 J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010

Olefin Cis-Dihydroxylation with Bio-Inspired Fe Catalysts
A R T I C L E S
to achieve the desired catalytic outcomes, the substrate olefins
must be present in large excess to trap the iron oxidant as soon
as it is formed, and the H2O2 must be introduced by syringe
pump to keep the H2O2 concentration low and avoid H2O2
disproportionation and generation of hydroxyl radicals. As such,
these reaction conditions are not very practical for synthetic
chemists. However, these are the only iron catalysts thus far
for the conversion of olefins into cis-diol products that use H2O2
as the oxidant. Thus, the mechanistic insights obtained from
this study may be useful for the development of practical
catalytic iron chemistry to effect such transformations.
Complexes 1-4 represent the first examples of iron com-
plexes that use H2O2 as the oxidant and carry out olefin cis-
dihydroxylation of both electron-rich and electron-poor olefins.
Indeed for the oxidation of styrene, 76-94% of the H2O2 oxidant
was converted to the corresponding cis-diol product, and there
was e1% epoxide product. With 1-octene, cis-2-heptene, and
dimethyl fumarate as substrates, conversions of the H2O2 oxidant
for 3 and 4 were comparable to those observed with 1. Lower
conversions (40-56%) were generally found for 2, where an
electron-donating methoxy substituent was introduced on the
phenyl ring; however, with styrene as the substrate, 2 exhibited
the best catalytic activity, resulting in 94% oxidant conversion
to diol. Complex 5, with the very electron-withdrawing CF3
group in place of the phenyl, was not a good catalyst and gave
low oxidant conversions (4-21%). The trends in the catalytic
activities of 1-5 suggest that the R-DPAH ligand framework
can tolerate a range of R groups, except for the electron-
withdrawing trifluoromethyl group adjacent to the amide
functionality, which diminishes the performance of 5. We also
investigated the catalytic behavior of related complex 6,30 which
has one meridional tridentate ligand and three labile binding
sites (Scheme 1). Like 1-4, 6 greatly favored cis-dihydroxyl-
ation over epoxidation but was in general a less efficient catalyst
than 1-4. Among all of the iron complexes examined with cis-
2-heptene and dimethyl fumarate as substrates, >93% of the
diol products (and usually 99%) were observed with retention
of configuration. These results emphasize that this oxidation is
a cis-only dihydroxylation and preclude the possibility that
epoxidation occurs initially followed by epoxide ring-opening.
Additional experiments were carried out with 1 and 4 to
examine their behavior with two other substrates of interest.
With cyclohexene as substrate, 10 equiv of H2O2 afforded 5.6
equiv of cis-cyclohexane-1,2-diol, 0.5 equiv of cyclohexenol,
0.4 equiv of cyclohexenone, and no cyclohexene oxide. The
high conversion to the cis-diol product is quite remarkable, as
cyclohexene oxidation by nonheme iron complexes is most often
associated with facile formation of allylic oxidation products.31-35
On the other hand, no cis-dihydroxylation of naphthalene was
observed, a different outcome from what was reported for
complex 10.36
Figure 3. ESI-MS analysis of 1 (0.35 mM) (top) and a mixture of 1 and
2 (bottom) in CH3CN (L ) Ph-DPAH and L′ ) 4-MeO-C6H4-DPAH).
the other complexes (Figure S1), suggesting the existence
of the ligand dissociation equilibrium shown below.
[FeL2]2+ h [FeL(solv)x]2+ + L
In support of this equilibrium, an equimolar mixture of 1 and
2 in CH3CN affords an ESI-MS spectrum that shows a statistical
mixture of [FeL2]2+ ions with the mixed ligand complex
[FeLL′]2+ being the most abundant species in the mixture
(Figure 3 bottom). This spectrum is observed immediately upon
mixing 1 and 2 in CH3CN and injecting the solution into the
mass spectrometer, showing that ligand scrambling occurs
essentially upon dissolution.
FTIR studies shed light on the nature of the species in CH3CN
solution. The Ph-DPAH ligand exhibits an amide νCdO at 1662
cm-1. This feature downshifts by 37 cm-1 in the solution of
the corresponding iron complex 1, and there is no evidence for
the νCdO of the free ligand in this solution. Similar results are
obtained for 2-5, with respective downshifts of 39, 36, 47, and
95 cm-1 observed for these complexes in solution. These
downshifts indicate a strong interaction between the ligand
amide carbonyl oxygen atoms and the metal center that is also
reflected by the relatively short Fe-O distances observed in
the crystal structures of 1-3. Only in the case of 5 can some
uncoordinated amide be discerned. Taken together, these
observations indicate that complexes 1-5 in solution retain in
large part the 1:2 metal-to-ligand composition found in the
crystal structures of 1-3. Furthermore, the ligand dissociation
equilibrium postulated above on the basis of the ESI-MS
data must lie predominantly to the left in favor of the
[FeL2]2+complex.
(30) Kryatov, S. V.; Taktak, S.; Korendovych, I. V.; Rybak-Akimova, E. V.;
Kaizer, J.; Torelli, S.; Shan, X.; Mandal, S.; MacMurdo, V.; Mairata
i Payeras, A.; Que, L., Jr. Inorg. Chem. 2005, 44, 85–99.
(31) Guajardo, R. J.; Hudson, S. E.; Brown, S. J.; Mascharak, P. K. J. Am.
Chem. Soc. 1993, 115, 7971–7977.
(32) Kojima, T.; Leising, R. A.; Yan, S.; Que, L., Jr. J. Am. Chem. Soc.
1993, 115, 11328–11335.
(33) Gosiewska, S.; Lutz, M.; Spek, A. L.; Klein Gebbink, R. J. M. Inorg.
Chim. Acta 2007, 360, 405–417.
Catalytic Activities. Table 2 compares the catalytic activities
of various complexes in the oxidation of a variety of olefins
using H2O2 as the oxidant. Also included in the comparison
are the results for 10 and 11, the prototypical complexes in our
studies of bio-inspired oxidation catalysis. For these experiments
(34) Martinho, M.; Banse, F.; Bartoli, J.-F.; Mattioli, T. A.; Battioni, P.;
Horner, O.; Bourcier, S.; Girerd, J.-J. Inorg. Chem. 2005, 44, 9592–
9596.
(35) Mukherjee, A.; Martinho, M.; Bominaar, E. L.; Mu¨nck, E.; Que, L.,
Jr. Angew. Chem., Int. Ed. 2009, 48, 1780–1783.
9
J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010 17717

A R T I C L E S
Oldenburg et al.
Table 2. Oxidation of Olefins Catalyzed by Various Iron Complexes with H2O2 Oxidanta
1-octene
styreneb
cis-2-heptene
epox [RC]c
tert-butyl acrylate
dimethyl fumarate
complex
epox
diol
epox
diol
diol [RC]c
epox
diol
epox
diol [RC]c
1
2
3
4
5
6
8
0.1(1)
7.6(3)
5.6(4)
7.7(4)
7.7(2)
1.3(3)
5.3(3)
2.7
0.67(2)
5.3(3)
5.5(2)
0.1(1)
<0.1
<0.1
<0.1
<0.1
8.0(5)
9.4(3)
7.6(4)
8.9(7)
2.1(2)
0.7(1) [57]
0.7(1) [45]
0.5(1) [60]
0.5(1) [44]
0.8(1) [43]
4.9(4) [99]
4.1(3) [99]
5.1(3) [99]
5.0(1) [99]
0.6(1) [99]
—
—
—
—
—
—
5.5(2)
4.5(1)
2.9(4)
5.2(1)
0.4(1)
2.8(3)
—
—
—
—
—
5.3(4) [99]
4.0(5) [99]
5.0(4) [99]
5.0(4) [99]
0.8(1) [99]
<0.1
<0.1
<0.1
0.18(2)
<0.1
1.6
0.11(1)
2.2(1)
0.3(2)
2.3
1.7
3.1 [84]
3.2 [92]
9d
10e
11e
0.37(1) [51]
1.9(1) [80]
0.4(1) [35]
0.47(5) [96]
3.0(3) [96]
4.1(4) [93]
—
0.58(4)
2.6f
—
—
—
0.42(4) [99]
5.2g [99]
1.7(1)
0.6(1)
2.4(1)
5.2(1)
<0.1f
—
6.2g
a Reaction conditions: 10 equiv H2O2 was added by syringe pump over a 5 min period (to minimize H2O2 disproportionation) to a solution of catalyst
(0.35 mM) and substrate (0.35 M) in CH3CN (3.7 mL). This solution was stirred for an additional 60 min prior to workup. See Experimental Section for
further details. Yields expressed as turnover numbers, TON (µmol of product/µmol of catalyst). b Results were obtained under an Ar atmosphere by
degassing solutions prior to oxidant introduction so that only a minor amount of the benzaldehyde autoxidation product was observed. See Table S3 for
the quantification of this product. c %RC: 100 × (A - B)/(A + B), where A ) yield of cis-diol with retention of configuration and B ) yield of epimer.
d Results from ref 26. e Results from ref 18. f Result from ref 38. g Results from ref 39.
Figure 4. Diol yields resulting from the oxidations of 1-octene (O), styrene (S), tert-butyl acrylate (A), and dimethyl fumarate (F), and equimolar mixtures
of substrates (O+S, S+A, and A+F) catalyzed by 1 with 10 equiv of H2O2 as oxidant.
The longevity of catalyst 1 has been previously investigated
by increasing the amount of H2O2 introduced,25 revealing
significant deterioration of catalyst activity after the addition
of more than 10 equiv of H2O2 relative to catalyst. Examination
of the ESI-MS data for 1 after a typical catalytic reaction with
subsequent treatment with Na2EDTA to remove iron from the
solution shows the presence of peaks arising from species other
than the free ligand peaks that indicate ligand oxidation (Figure
S2). These peaks correspond to N-(dipyridin-2-ylmethylene)ben-
zamide, benzamide, and dipyridin-2-ylmethanone. These prod-
ucts indicate that the Ph-DPAH ligand was oxidized into its
imine derivative, which then hydrolyzed into the observed
ketone and amide. A similar oxidation of the related N-(bis(2-
pyridyl)methyl)pyridine-2-carboxamide ligand has been re-
ported.37 Neither a 1:1:1 mixture of iron(II), benzamide, and
dipyridin-2-ylmethanone nor a 1:2 mixture of iron(II) and
dipyridin-2-ylmethanone was found to be catalytically active.
Competitive Oxidations. To gather information regarding the
interaction of substrate with the oxidant formed in the course
of catalysis, substrate competition studies were carried out with
pairs of olefin substrates. With 1 as the catalyst, equimolar
amounts of two different substrates were oxidized under normal
catalytic and workup conditions, and an interesting trend was
observed (Figure 4). In individual experiments, 1 catalyzes
1-octene and styrene cis-dihydroxylation resulting in 7.6
turnovers of octane-1,2-diol and 8.0 turnovers of phenylethane-
1,2-diol. However, when both substrates compete for the active
oxidant generated from 1, the dihydroxylation of styrene was
clearly favored, with 6.4 turnovers of phenylethane-1,2-diol and
merely 0.9 turnover of octane-1,2-diol. When styrene and tert-
butyl acrylate compete for the active oxidant, acrylate was
favored, with 4.4 turnovers of the diol from acrylate and 1.9
turnovers of styrene diol. Finally, for competition between tert-
butyl acrylate and dimethyl fumarate, the fumarate was clearly
favored by the active oxidant with less than 0.5 turnover
resulting in the formation of acrylate diol. Clearly the more
electron-deficient olefin is favored in these oxidations. A similar
reactivity trend has been previously reported for 11.39 The results
(36) Feng, Y.; Ke, C.-y.; Xue, G.; Que, L., Jr. Chem. Commun. 2009, 50–
52.
(37) Zhu, S.; Brennessel, W. W.; Harrison, R. G.; Que, L., Jr. Inorg. Chim.
Acta 2002, 337, 32–38.
9
17718 J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010

Olefin Cis-Dihydroxylation with Bio-Inspired Fe Catalysts
A R T I C L E S
for both 1 and 11 suggest the active oxidant to be more
nucleophilic in nature,40,41 but the exact nature of this oxidant
is not well understood.
A Hammett correlation analysis was also attempted for a
series of styrene substrates. Equimolar amounts of styrene and
p-substituted styrenes were utilized as substrates using 1 as the
catalyst under otherwise normal catalytic conditions, and relative
yields were analyzed by NMR analysis of the diol products.
The substituents of the styrenes changed the diol yields by no
more than a factor of 2 with no obvious correlation with σp
observed (Figure S3).
Isotopic Labeling Studies. 18O labeling experiments have
proven useful in previous studies for deducing whether peroxide
O-O bond cleavage occurs prior to the attack of substrate by
establishing the source of the oxygen atoms incorporated into
the product.18,20,42 As commercially available H218O2 usually
comes as a 2 wt % solution in H216O, every equivalent of H218O2
added in our labeled studies is accompanied by 100 equiv of
H216O. To corroborate these results, complementary labeling
experiments with 10 equiv of H216O2/1000 of equiv H218O are
also carried out. Previous experiments carried out for 10 and
11 with cyclooctene as substrate revealed two distinct labeling
patterns.18 For 10 and related complexes of tetradentate ligands
with strong enough ligand fields to support a low-spin iron(II)
center in the [FeII(L)(NCCH3)2]2+ complexes, the cis-diol
product incorporated one oxygen atom from H2O2 and the other
from H2O (Table 3), results that led us to postulate a water-
assisted mechanism for activating the peroxo O-O bond for
cleavage. In contrast, for complexes like 11, which have weaker
field ligands due to the presence of R-substituents on the pyridine
ligands, both diol oxygens derived exclusively from H2O2,
thereby requiring a nonwater-assisted pathway for H2O2 activation.
Analogous 18O labeling experiments were carried out with
1-3. Figure 5 graphically illustrates the percentages of 16O18O
and 18O18O isotopomers of the diol products obtained from
H218O2 experiments. Table 3 lists the results for all the ex-
periments and demonstrates agreement between the comple-
mentary H218O and H218O2 experiments. The cis-dihydroxylation
of dimethyl fumarate by 1-3 afforded the cis-diol product with
both its oxygen atoms derived from H2O2. A follow-up labeling
experiment carried out with 1 as the catalyst and a 50:50 mixture
of H216O2/H218O2 resulted in the generation of a diol product
with a 44:7:49 ratio for the 16O16O, 16O18O, and 18O18O iso-
topomers, demonstrating that the two diol oxygens incorporated
into the fumarate substrate must originate from only one
molecule of H2O2. (The observed small fraction of the 16O18O
isotopomer is consistent with the minor amount of H216O18O
present in the 90% enriched H218O2 solution.) When the labeling
experiments were extended to include tert-butyl acrylate,
styrene, and 1-octene as substrates, some label incorporation
from H218O into their respective diol products was observed
(8-29%), and the amount incorporated increased as the olefin
substituents became more electron-rich. Because the latter
substrates were found to react more slowly than dimethyl
fumarate on the basis of the competition experiments presented
Figure 5. Percentages of diol product with two 18O atoms (blue) and with
one 18O atom (red) in olefin cis-hydroxylation reactions catalyzed by
indicated complexes (0.35 mM) with 10 equiv jof H218O2 and 1000 equiv
of H216O in CH3CN solvent at room temperature in the presence of 1000
equiv of olefin. O, S, A, F, and C stand for 1-octene, styrene, tert-butyl
acrylate, dimethyl fumarate, and cyclooctene, respectively. Data for 10 and
11 obtained from ref 18.
Table 3. Percentages of Isotopomers Found in Diol Products with
H218O or H218O2 Used As Sources of 18O Atoms in Olefin
Oxidations Catalyzed by Indicated Iron Catalystsa
10 H216O2/1000 H218
O
10 H218O2/1000 H216
16O16O 16O18O 18O18O
O
Catalyst
Substrate
16O16O
16O18O
1
1-octene
styrene
tert-butyl acrylate
dimethyl fumarate
1-octene
67
81
92
98
72
83
84
97
79
88
95
95
29
18
8
4
1
0
1
4
1
1
3
3
4
2
0
32
20
13
4
30
18
15
2
22
16
12
6
64
79
87
95
65
81
83
95
74
80
86
92
2
2
3
26
17
15
2
18
12
4
styrene
tert-butyl acrylate
dimethyl fumarate
1-octene
styrene
tert-butyl acrylate
dimethyl fumarate
1
a Reaction conditions: 3.5 mM H2O2 and 0.35 M H2O delivered by
syringe pump to a solution of 0.35 mM catalyst and 0.35 M substrate in
CH3CN (3.7 mL) at room temperature. See Experimental Section for
further details.
in Figure 4, the labeling results suggest that there is a
competition between the rate of olefin attack by the iron-derived
oxidant and the rate at which label scrambling occurs between
H2O2 and water at the iron center.
Additional labeling experiments with 1 as a function of
1-octene (Figure 6) and H218O (Figure 7) concentration provided
additional insight. At a constant H218O concentration of 0.175
M, the fraction of the cis-diol product that incorporated an
oxygen atom from water decreased linearly from 33% to 18%,
as the concentration of 1-octene increased 20-fold (Figure 6).
On the other hand, at a constant 1-octene concentration of 0.35
M, label incorporation from water increased with increasing
H218O concentration and reached a plateau at 0.35 M H218O.
This saturation behavior suggests a pre-equilibrium binding of
H2O to the iron center during the catalytic cycle and is supported
by the linearity of the double reciprocal plot of these data (Figure
7, inset). These results demonstrate that 1-octene oxidation and
water scrambling at the iron center occur at comparable rates.
The 1-octene isotope labeling results for 1-3 stand in marked
contrast to the near-zero water incorporation into the diol product
(38) Mas-Balleste´, R.; Que, L., Jr. J. Am. Chem. Soc. 2007, 129, 15964–
15972.
(39) Fujita, M.; Costas, M.; Que, L., Jr. J. Am. Chem. Soc. 2003, 125,
9912–9913.
(40) Wertz, D. L.; Valentine, J. S. Struct. Bonding (Berlin) 2000, 97, 38–
60.
(41) Wada, A.; Ogo, S.; Nagatomo, S.; Kitagawa, T.; Watanabe, Y.;
Jitsukawa, K.; Masuda, H. Inorg. Chem. 2002, 41, 616–618.
(42) Chen, K.; Que, L., Jr. J. Am. Chem. Soc. 2001, 123, 6327–6337.
9
J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010 17719

A R T I C L E S
Oldenburg et al.
Figure 6. Fraction of 18O-labeled diol in the oxidation of 1-octene catalyzed
by 1 with H2O2 in the presence of 0.175 M H218O as a function of the
concentration of 1-octene.
Figure 8. Yield of diol product (9) and percent conversion of H2O2 into
diol product ([) as a function of the amount of H2O2 added for the oxidation
of 1-octene by 1. Note: the data points at 10 equiv of H2O2 are superimposed
on each other.
for 11 is lacking.18,19 However, the corresponding FeIII(6-Me3-
TPA)(OOtBu) complex has been trapped and characterized
spectroscopically to be a high-spin species.46,47 Because 18O
labeling results for 11 do not show any evidence for water
incorporation into the cis-diol product,18 the water-assisted
mechanism postulated for 10 cannot apply in these cases. It is
thus proposed that the peroxide is bound side-on to the iron(III)
center as a means of facilitating O-O bond cleavage, and the
putative high-spin FeIII-η2-OOH intermediate is proposed to
either attack the substrate directly or undergo O-O bond
cleavage to form the HOsFeVdO oxidant analogous to that
proposed for 10.18,39
Figure 7. Percent of 18O-labeled diol obtained in the oxidation of 1-octene
catalyzed by 1 with H2O2 as a function of the concentration of H218O. Inset:
double-reciprocal plot.
The R-DPAH family of catalysts provides a unique op-
portunity to probe the mechanism of the high-spin iron cis-
dihydroxylation-selective catalysts because of the high yields
of cis-diol (relative to the H2O2 oxidant added) obtained from
these reactions. The demonstration that both oxygen atoms of
diol are derived from one molecule of H2O2 makes it reasonable
to propose the initial formation of an Fe/H2O2 adduct. Two
possible iron-peroxo formulations immediately come to mind:
(i) an FeIII-OOH species, like that previously proposed for 11
by extension of the mechanism generally accepted for 10, or
(ii) an FeII-O2H2 adduct. Because of the high conversion of
H2O2 to cis-diol for these complexes, it should be possible to
distinguish experimentally between these two options by
monitoring the percent conversion of H2O2 to the diol product
as a function of added H2O2 equivalents in the oxidation of
1-octene catalyzed by 1 (Figure 8). In the case of option (i),
some lag in the production of the cis-diol product may be
expected due to the requisite initial oxidation of Fe(II) to Fe(III),
but such a preoxidation step is not required for option (ii).
As shown in Figure 8, the catalytic behavior of 1 conforms
more closely to that expected for option (ii). At low H2O2/
catalyst ratios, H2O2 was converted nearly quantitatively to the
diol. For instance, with 0.5 equiv of H2O2 added relative to the
iron catalyst, 0.46 equiv of diol was formed, amounting to 92%
oxidant conversion. Similarly, with 1.0 equiv of H2O2 per iron,
0.94 equiv of diol was formed for a 94% conversion of H2O2.
The percent conversion values decreased with the increasing
number of turnovers, 89%, 82%, and 77% with 2, 4, and 10
equiv of H2O2, respectively, probably due to some catalyst
decomposition. Thus, the first 0.5 equiv of H2O2 was not needed
to oxidize iron(II) to iron(III) prior to H2O2 activation, in contrast
obtained for 6. Complex 6 has a meridional tridentate ligand,
in place of the facial tridentate ligands used for 1-3. This
change in ligand orientation must prevent label scrambling
between H2O2 and water at the iron center in the case of 6.
Mechanistic Considerations. The currently favored mecha-
nism for the family of nonheme iron hydrocarbon oxidation
catalysts we have investigated is based on extensive work on
10 for which a low-spin FeIIIsOOH intermediate can be
observed at -40 °C in CH3CN.18 This intermediate is proposed
to convert to an as yet unobserved HOsFeVdO oxidant that
effects alkane hydroxylation, olefin epoxidation, and olefin
cis-dihydroxylation.18,42 (An S ) /2 EPR signal has recently
1
been reported in the reaction of 10 with peracids at -70 °C
and is proposed to arise from a putative FeVdO oxidant,43 but
additional corroboration of the iron(V) oxidation state assign-
ment is needed.) Because the oxygen atom from added H218O
can be incorporated into all the various oxidation products, it
is proposed that the H2O binds to the low-spin FeIIIsOOH
species and participates in cleaving the OsO bond via a water-
assisted mechanism to afford the HOsFeVdO oxidant where
one oxygen atom derives from H2O. These mechanistic notions
are supported by DFT calculations.44,45
Much less is known of the peroxide activation mechanism
for 11 and related catalysts that favor cis-dihydroxylation of
olefins. These complexes have weaker field ligands and thus
cannot support formation of low-spin iron centers in the catalytic
cycle. By analogy to the mechanism proposed for 10 and related
catalysts, a high-spin FeIII-OOH species has been proposed in
the catalytic cycle, although evidence for such an intermediate
(43) Lyakin, O. Y.; Bryliakov, K. P.; Britovsek, G. J. P.; Talsi, E. P. J. Am.
Chem. Soc. 2009, 131, 10798–10799.
(44) Bassan, A.; Blomberg, M. R. A.; Siegbahn, P. E. M.; Que, L., Jr.
J. Am. Chem. Soc. 2002, 124, 11056–11063.
(46) Zang, Y.; Kim, J.; Dong, Y.; Wilkinson, E. C.; Appelman, E. H.; Que,
L., Jr. J. Am. Chem. Soc. 1997, 119, 4197–4205.
(45) Bassan, A.; Blomberg, M. R. A.; Siegbahn, P. E. M.; Que, L., Jr.
Chem.sEur. J. 2005, 11, 692–705.
(47) Lehnert, N.; Ho, R. Y. N.; Que, L., Jr.; Solomon, E. I. J. Am. Chem.
Soc. 2001, 123, 12802–12816.
9
17720 J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010

Olefin Cis-Dihydroxylation with Bio-Inspired Fe Catalysts
A R T I C L E S
Scheme 2. Mechanistic Overview
to what was observed for 10.18 These results thus unequivocally
implicate an initially formed FeII-O2H2 adduct as the catalyti-
cally relevant iron-peroxo species.
0.1 equiv of diol was obtained upon addition of 10 equiv of
H2O2, suggesting that the four-pyridine ligand combination
postulated for option (a) is not able to catalyze olefin cis-
dihydroxylation under the conditions that complexes 1-4 can.
To assess the viability of option (b), we tested a 1:1 ratio of
Fe(OTf)2 and Ph-DPAH and found that this combination was
active in catalyzing 1-octene cis-dihydroxylation. Addition of
1 equiv of H2O2 afforded 0.7(1) TON of the cis-diol, while
addition of 4 equiv of H2O2 resulted in 2.9(2) TON, for about
a 73% efficiency in the conversion of H2O2 into cis-diol, just
slightly below that observed for 1 (76%,Table 2 and Figure 8).
However, addition of 10 equiv of H2O2 afforded only 4.1(2)
TON of the cis-diol, suggesting significant catalyst deterioration
after 3 turnovers. Delaying this outcome may be the role played
by the additional Ph-DPAH ligand in 1. Future work will explore
what factors can extend this protective effect to allow more
turnovers to be achieved.
Irrespective of which option affords the initial η2-O2H2 adduct,
the FeII(η2-O2H2)(solv) intermediate can either play the role of
oxidant and attack the olefin directly (Scheme 2, left) or undergo
O-O bond cleavage to form an FeIV(OH)2(solv) species that
reacts with the olefin substrate (Scheme 2, right). In the left
branch of Scheme 2, direct attack of the substrate by the FeII(η2-
O2H2)(solv) adduct would result in O-O bond cleavage and
formation of the first C-O bond to yield a ꢀ-hydroxyethyl
radical (S•) that remains coordinated to an FeIII(OH)(solv)
moiety. A rapid oxygen rebound must occur to form the second
C-O bond and afford the cis-diol product with no loss of
stereochemistry. In the right branch of Scheme 2, the substrate
combines with the dihydroxoiron(IV) moiety to afford the
desired product. In either pathway, O-O bond cleavage is
proposed to be the rate-determining step.
Scheme 2 describes the evolution of this FeII-O2H2 species
into what we propose to be the active oxidant for olefin cis-
dihydroxylation derived from the R-DPAH catalysts. The most
straightforward way to account for the observation that both
diol oxygen atoms derive from one molecule of H2O2 is by
formation of an FeII(η2-O2H2) adduct. Formation of such an
adduct has been considered and found to be energetically
reasonable in computational studies of Fenton chemistry48,49 and
of an iron complex of a tetradentate bispidine ligand that is
capable of olefin oxidation catalysis.50 For the R-DPAH family
of catalysts, such an FeII(η2-O2H2) adduct may be formed either
by (a) dissociation of the two carbonyl ligands upon addition
of H2O2 or (b) loss of one of the R-DPAH ligands. In addition,
there must be an additional metal coordination site to bind water
to account for the observed label scrambling. For option (a),
this would require formation of a seven-coordinate FeII(N2)2(η2-
O2H2)(solv) intermediate, which may be too sterically crowded.
Option (b), on the other hand, is particularly attractive as it can
provide three coordination sites on the face of the six-coordinate
iron center to allow binding of both a side-on bound H2O2 and
a water molecule. Evidence that the proposed ligand dissociation
equilibrium is established rapidly has been presented in Figure
3. Upon addition, the H2O2 would then most likely interact with
the 1:1 [Fe(L)(solv)x]2+ species and shift the dissociation
equilibrium to the right to form the six-coordinate FeII(L)(η2-
O2H2)(solv) intermediate.
To assess the viability of option (a), we tested whether a 1:2
ratio of Fe(OTf)2 and dipyridin-2-ylmethanone in CH3CN could
catalyze the cis-dihydroxylation of 1-octene. However, less than
These two options may be distinguished on the basis of the
H218O-labeling results, because the point at which the labeled
water can become incorporated into the cis-diol product differs.
With the addition of excess water, the CH3CN ligand on the
FeII(η2-O2H2)(NCCH3) adduct can be displaced by water at least
(48) Ensing, B.; Buda, F.; Blo¨chl, P.; Baerends, E. J. Angew. Chem., Int.
Ed. 2001, 40, 2893–2895.
(49) Buda, F.; Ensing, B.; Gribnau, M. C. M.; Baerends, E. J. Chem.sEur.
J. 2001, 7, 2775–2783.
(50) Comba, P.; Rajaraman, G.; Rohwer, H. Inorg. Chem. 2007, 46, 3826–
3838.
9
J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010 17721

A R T I C L E S
Oldenburg et al.
the product from a meridional FeIV(OH)2(H218O) unit, compared
to only one for the facial FeIV(OH)2(H218O) unit.
Scheme 3
The experimental results described above lead us to the
conclusion that the oxidant produced in the reaction of H2O2
with 1-4 is a cis-FeIV(OH)2 species, distinct from the isomeric
FeIV(O)(OH2) species proposed by Baerends in his calculations
on the Fenton reaction49 and from the cis-FeV(O)(OH) oxidant
we have proposed for the reaction of H2O2 with 10.18 The latter
two high-valent iron-oxo species would be expected to carry
out oxo transfer to olefins and form epoxides, but very little
epoxide, if any, is observed in the reactions catalyzed by 1-4.
The products observed in the reaction of cyclohexene with
4/H2O2 are quite illustrative in supporting this point, where the
dominant product is the cis-diol, representing 90% of the
products formed. Cyclohexenol and cyclohexenone make up
the remaining 10%, and no epoxide was observed. This outcome
is distinctly different from oxidations catalyzed by other
nonheme iron complexes. For example, in the case of FeII(cy-
clam)/H2O2, only cyclohexene oxide was observed.51 On the
other hand, allylic oxidation products were the dominant, if not
exclusive, products observed for several other complexes,32-35
demonstrating the susceptibility of the allylic C-H bonds of
cyclohexene to oxidative attack. The very low extent of allylic
oxidation in the case of the R-DPAH complexes supports the
involvement of an oxidant with a much lower hydrogen-atom
abstraction capability than is associated with nonheme ox-
oiron(IV) complexes, which have been shown to be quite facile
at carrying out hydrogen-atom abstractions, particularly of allylic
C-H bonds.34,52-54 A recent comparison of oxoiron(IV) and
hydroxoiron(IV) species supported by the same tetradentate
ligand showed the former to have a 100-fold higher H-atom
abstracting ability than the latter.55 A similar comparison of
Mn(IV) complexes also revealed the same trend.56-58 With a
lower hydrogen-atom affinity than its oxoiron(IV) counterpart,
the proposed dihydroxoiron(IV) oxidant is consequently well
suited to carry out the highly selective cis-dihydroxylation of
the cyclohexene double bond. Lastly, unlike 10, 1-4 do not
catalyze the cis-dihydroxylation of the C1-C2 double bond of
naphthalene, consistent with the formation of a less powerful
oxidant than the FeV(O)(OH) oxidant associated with 10.
DFT calculations provide computational precedents for the
proposed formation of a dihydroxoiron(IV) species from
the reaction of an iron(II) precursor with H2O2. In a study of
the active species generated in the Fenton reaction, Baerends
and co-workers49 found that [FeIV(OH)2(OH2)4]2+ was the initial
high-valent species formed in the reaction of [FeII(OH2)6]2+ and
H2O2, which then isomerized to [FeIV(O)(OH2)5]2+ by proton
transfer. Comba and co-workers50 investigated various pathways
leading to the formation of high-valent iron intermediates in
some of the time to generate the FeII(η2-O2H2)(OH2) species.
In the left branch of Scheme 2, incorporation of a water-derived
oxygen atom into the diol product could occur only in the
formation of the second C-O bond, via scrambling between
the -OH and -OH2 groups by intramolecular proton transfer
just prior to the rebound step. If proton transfer proceeded much
more quickly than the formation of the second C-O bond, then
50% of the diol product would have one oxygen atom derived
from H2O and the other from H2O2. If, however, the rebound is
much faster than proton transfer, then both oxygens would derive
exclusively from H2O2. The fact that 20-30% label incorpora-
tion from water is observed suggests that the rates of rebound
and proton transfer are comparable. However there is one
experimental observation that is inconsistent with this mecha-
nism. Experimental data show that there is an inverse substrate
concentration dependence on the extent of label incorporation
from water (Figure 6). As label scrambling can occur only after
the first C-O bond has already been formed, the extent of label
incorporation cannot be affected by the substrate concentration.
On the other hand, the right branch of Scheme 2 is consistent
with the observed substrate concentration dependence. In this
pathway, the rate-determining cleavage of the η2-peroxo
O-O bond forms an FeIV(OH)2 oxidant. With 18O-labeled
water occupying the solvent binding site, the resulting
FeIV(OH)2(H218O) unit can undergo label scrambling prior to
attacking the olefin substrate. Intramolecular proton transfer from
the bound water to a hydroxo ligand represents a simple
mechanism for facile label scrambling, but other plausible
mechanisms cannot be excluded. If the rate of scrambling is
much faster than the rate of olefin attack, then 33% of diol
product should have both oxygen atoms derived from H2O2 and
67% should have one oxygen atom each from H2O and H2O2.
The fact that the extent of 18O incorporation from labeled water
is significantly lower than the predicted 67% suggests a
competition between intramolecular label scrambling and
intermolecular olefin attack. The extent of label incorporation
from H218O can then be affected by changing the rate of the
latter step. This expectation is indeed borne out by the
experimental results where label incorporation was observed
to decrease with higher olefin concentration (Figure 6) or the
use of more reactive olefins (Figure 5).
The R-DPAH complexes described in this paper are the first
iron catalysts highly selective for cis-dihydroxylation for which
label incorporation from H218O has been observed. We suggest
that the observed label scrambling is due to the availability of
an additional binding site for water and facilitated by the facial
arrangement of the hydroxo and aqua ligands of the
FeIV(OH)2(H218O) unit (Scheme 2). The latter point is supported
by the observation of no label incorporation from H218O into
the cis-diol product from 6 (Figure 5), which is a complex of
the tridentate but meridional ligand Bn-BQA (Scheme 1).
Extending the mechanism proposed for the R-DPAH complexes
to 6 would afford a meridional FeIV(OH)2(H218O) unit (Scheme
3). On the assumption that cis-diol formation entails the transfer
of two adjacent hydroxo groups to the olefin, then two
consecutive proton transfers must occur to incorporate 18O into
(51) Nam, W.; Ho, R. Y. N.; Valentine, J. S. J. Am. Chem. Soc. 1991,
113, 7052–7054.
(52) Kaizer, J.; Klinker, E. J.; Oh, N. Y.; Rohde, J.-U.; Song, W. J.; Stubna,
A.; Kim, J.; Mu¨nck, E.; Nam, W.; Que, L., Jr. J. Am. Chem. Soc.
2004, 126, 472–473.
(53) Sastri, C. V.; Lee, J.; Oh, K.; Lee, Y. J.; Lee, J.; Jackson, T. A.; Ray,
K.; Hirao, H.; Shin, W.; Halfen, J. A.; Kim, J.; Que, L., Jr.; Shaik, S.;
Nam, W. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 19181–19186.
(54) Klinker, E. J.; Shaik, S.; Hirao, H.; Que, L., Jr. Angew. Chem., Int.
Ed. 2009, 48, 1291–1295.
(55) Fiedler, A. T.; Que, L., Jr. Inorg. Chem. 2009, 48, 11038–11047.
(56) Yin, G.; Danby, A. M.; Kitko, D.; Carter, J. D.; Scheper, W. M.; Busch,
D. H. J. Am. Chem. Soc. 2007, 129, 1512–1513.
(57) Yin, G.; Danby, A. M.; Kitko, D.; Carter, J. D.; Scheper, W. M.; Busch,
D. H. J. Am. Chem. Soc. 2008, 130, 16245–16253.
(58) Chattopadhyay, S.; Geiger, R. A.; Yin, G.; Busch, D. H.; Jackson,
T. A. Inorg. Chem. 2010, 49, 7530–7535.
9
17722 J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010

Olefin Cis-Dihydroxylation with Bio-Inspired Fe Catalysts
A R T I C L E S
the reaction of H2O2 with an iron(II) complex supported by a
tetradentate bispidine ligand and found the formation of the
corresponding dihydroxoiron(IV) species to be energetically
quite facile, which they suggested may be the olefin cis-
dihydroxylation agent in their catalytic system. Although no
DFT calculations have been carried out on the FeII(R-DPAH)
catalysts, these other studies support the mechanistic ideas we
have developed in Scheme 2 on the basis of our experimental
results.
In summary, we have explored the oxidation chemistry of
iron complexes coordinated with bio-inspired facially oriented
N,N,O ligands and found them to be effective catalysts for olefin
cis-dihydroxylation with limiting H2O2 oxidant. These com-
plexes show a significant improvement in oxidative efficiency
and cis-dihydroxylation selectivity over the previously developed
iron-based olefin cis-dihydroxylation catalysts that use H2O2 as
the oxidant.20 A mechanism is presented in Scheme 2 that can
rationalize three important features of this catalytic system: (a)
the strong preference for olefin cis-dihydroxylation over epoxi-
dation with very high stereoretention, (b) the predominant
incorporation of both oxygen atoms of H2O2 into the cis-diol
product, and (c) the involvement of a water scrambling pathway
competitive with olefin oxidation. These features are most easily
accounted for by invoking the formation of an fac-(L)FeII(η2-
H2O2)(solv) adduct that forms a fac-(L)FeIV(OH)2(solv) oxidant,
which allows label scrambling to occur in the presence of H218O.
This is the first example of an iron-based cis-dihydroxylation
catalyst for which experimental evidence for an FeII/FeIV redox
cycle has been obtained. The insights derived from this work
differ significantly from those reported in a very recent paper
in which an oxoiron(V) species is implicated as the oxidant
generated by the reaction of a cis-dichloroiron(III) complex of
a macrocyclic tetraaza ligand with oxone; in this system,
catalytic olefin cis-dihydroxylation can be achieved in high yield
with a limiting substrate and a 2-fold excess of the oxone
oxidant.59 Taken together, these two catalytic systems demon-
strate the versatility of high-valent iron in carrying out olefin
cis-dihydroxylation. Our ongoing efforts are aimed at extending
the lifetime of the R-DPAH catalysts and identifying reaction
conditions that allow the Fe/H2O2 cis-dihydroxylation chemistry
described here to be used in synthetic applications.
Acknowledgment. This work was supported by the United
States Department of Energy (DE-FG02-03ER15455). We thank
Dr. Victor G. Young, Jr. of the X-ray Crystallographic Laboratory
at the University of Minnesota for invaluable assistance and Prof.
William Tolman for the use of his Thermo Nicolet Avatar 370 FT-
IR instrument.
Supporting Information Available: Tables S1 and S2 con-
taining crystallographic data for 2 and 3, Table S3 listing
additional oxidation results, Figures S1 and S2 showing ESI-
MS data of 2-5 and decomposed ligand, and Figure S3 showing
resultsfromanattemptedHammettcorrelationinthecis-dihydroxyl-
ation of styrenes. This material is available free of charge via
the Internet at http://pubs.acs.org.
JA1021014
(59) Chow, T. W.-S.; Wong, E. L.-M.; Guo, Z.; Liu, Y.; Huang, J.-S.; Che,
C.-M. J. Am. Chem. Soc. 2010, 132, 13229–13239.
9
J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010 17723