Vanadium-based, extended catalytic lifetime catechol dioxygenases: Evidence for a common catalyst

Article abstract of DOI:10.1021/ja051594e

In 1999, a catechol dioxygenase derived from a V-polyoxometalate was reported which was able to perform a record > 100 000 total turnovers of 3,5-di-tert-butylcatechol oxygenation using O₂ as the oxidant (Weiner, H.; Finke, R. G. J. Am. Chem. Soc. 1999, 121, 9831). An important goal is to better understand this and other vanadium-based catechol dioxygenases. Scrutiny of 11 literature reports of vanadium-based catechol dioxygenases yielded the insight that they all proceed with closely similar selectivities. This, in turn, led to a "common catalyst hypothesis" for the broad range of vanadium based catechol dioxygenase precatalysts presently known. The following three classes of V-based compounds, 10 complexes total, have been explored to test the common catalyst hypothesis: (i) six vanadium-based polyoxometalate precatalysts, (n-Bu₄N)₄H ₅PV₁₄O₄₂, (n-Bu₄N) ₇SiW₉V₃O₄₀, (n-Bu₄N) ₅[(CH₃CN)_xFe^II·SiW ₉V₃O₄₀], (n-Bu₄N)₉P ₂W₁₅V₃O₆₂, (n-Bu₄N) ₅Na₂[(CH₃CN)_xFe^II· P₂W₁₅V₃O₆₂], and (n-Bu ₄N)₄H₂-γ-SiW₁₀V ₂O₄₀; (ii) three vanadium catecholate complexes, [V ^VO(DBSQ)(DTBC)]₂, [Et₃NH]₂[V ^IVO(DBTC)₂]·2CH₃OH, and [Na(CH ₃-OH)₂]₂[V^V(DTBC)₃] ₂·4CH₃OH (where DBSQ = 3,5-di-tert-butylsemiquinone anion and DTBC = 3,5-di-tert-butylcatecholate dianion), and (iii) simple VO(acac)₂. Product selectivity studies, catalytic lifetime tests, electron paramagnetic resonance spectroscopy (EPR), negative ion mode electrospray ionization-mass spectrometry (negative ion ESI-MS), and kinetic studies provided compelling evidence for a common catalyst or catalyst resting state, namely, Pierpont's structurally characterized vanadyl semiquinone catecholate dimer complex, [VO(DBSQ)(DTBC)]₂, formed from V-leaching from the precatalysts. The results provide a considerable simplification and unification of a previously disparate literature of V-based catechol dioxygenases.

Full text of DOI:10.1021/ja051594e

Published on Web 06/01/2005
Vanadium-Based, Extended Catalytic Lifetime Catechol
Dioxygenases: Evidence for a Common Catalyst
Cindy-Xing Yin and Richard G. Finke*
Contribution from Department of Chemistry, Colorado State UniVersity,
Ft. Collins, Colorado 80523
Received March 12, 2005; E-mail: rfinke@lamar.colostate.edu
Abstract: In 1999, a catechol dioxygenase derived from a V-polyoxometalate was reported which was
able to perform a record >100 000 total turnovers of 3,5-di-tert-butylcatechol oxygenation using O2 as the
oxidant (Weiner, H.; Finke, R. G. J. Am. Chem. Soc. 1999, 121, 9831). An important goal is to better
understand this and other vanadium-based catechol dioxygenases. Scrutiny of 11 literature reports of
vanadium-based catechol dioxygenases yielded the insight that they all proceed with closely similar
selectivities. This, in turn, led to a “common catalyst hypothesis” for the broad range of vanadium based
catechol dioxygenase precatalysts presently known. The following three classes of V-based compounds,
10 complexes total, have been explored to test the common catalyst hypothesis: (i) six vanadium-based
polyoxometalate precatalysts, (n-Bu4N)4H5PV14O42, (n-Bu4N)7SiW9V3O40, (n-Bu4N)5[(CH3CN)xFeII‚SiW9V3O40],
(n-Bu4N)9P2W15V3O62, (n-Bu4N)5Na2[(CH3CN)xFeII‚P2W15V3O62], and (n-Bu4N)4H2-γ-SiW10V2O40; (ii) three
vanadium catecholate complexes, [VVO(DBSQ)(DTBC)]2, [Et3NH]2[VIVO(DBTC)2]‚2CH3OH, and [Na(CH3-
OH)2]2[VV(DTBC)3]2‚4CH3OH (where DBSQ ) 3,5-di-tert-butylsemiquinone anion and DTBC ) 3,5-di-tert-
butylcatecholate dianion), and (iii) simple VO(acac)2. Product selectivity studies, catalytic lifetime tests,
electron paramagnetic resonance spectroscopy (EPR), negative ion mode electrospray ionization-mass
spectrometry (negative ion ESI-MS), and kinetic studies provided compelling evidence for a common catalyst
or catalyst resting state, namely, Pierpont’s structurally characterized vanadyl semiquinone catecholate
dimer complex, [VO(DBSQ)(DTBC)]2, formed from V-leaching from the precatalysts. The results provide a
considerable simplification and unification of a previously disparate literature of V-based catechol
dioxygenases.
Scheme 1. Definitions and Fe3+ versus Fe2+ Intradiol Catechol
Dioxygenases (cleavage of the C-C bond between the two
hydroxyl groups) versus Extradiol Catechol Dioxygenases
(cleavage of one of the C-C bonds adjacent to the two hydroxyl
groups)
Introduction
Oxygen activation for selective oxidation reactions continues
to be a central and challenging subject in catalysis.1-4 In nature,
the discovery of dioxygenase enzymes, that is, enzymes
catalyzing the incorporation of both oxygen atoms of O2 into
substrates with no added protons or electrons, dates back to the
1950s with the discovery of oxygen incorporation into catechol
by pyrocatechase of Pseudomonas sp.5 Since that time, mild,
selective, and facile man-made dioxygenases have been the Holy
Grail of biomimetic studies of oxidation catalysis.3
One important class of dioxygenase enzymes, catechol
dioxygenases, catalyze the degradation of aromatic compounds.6
Catechol dioxygenases have been classified into two groups,
intradiol and extradiol dioxygenases (Scheme 1). Catechol
dioxygenase enzymes containing non-heme FeII/III or MnII have
been discovered in nature.7-10 Biomimetic model systems con-
taining Fe(II/III),8,10-19 V(IV/V) (Table 1), Ru(II),20 Rh(III),21,22
as well as several other metals23 are known to catalyze catechol
dioxygenation reactions.
(1) Hayaishi, O. Molecular Mechanisms of Oxygen ActiVation; Academic
Press: New York, 1974.
(2) Oyama, S. T.; Desikan, A. N.; Hightower, J. W. In Catalytic SelectiVe
Oxidation; Oyama, S. T., Hightower, J. W., Eds.; ACS Symposium Series
523; American Chemical Society: Washington, DC, 1993; pp 1-14.
(3) Hill, C. L.; Weinstock, I. A. Nature 1997, 388, 332-333.
(4) Sheldon, R. A. In Biomimetic Oxidations Catalyzed by Transition Metal
Complexes; Meunier, B., Ed.; Imperial College Press: London, 2000; pp
613-662.
(5) (a) Hayaishi, O.; Hashimoto, K. J. Biochem. (Tokyo) 1950, 37, 371-374.
(b) Hayaishi, O.; Katagiri, M.; Rothberg, S. J. Am. Chem. Soc. 1955, 77,
5450-5451.
(6) Nozaki, M. In Molecular Mechanisms of Oxygen ActiVation; Academic
Press: New York, 1974; pp 135-165.
(7) Nozaki, M. Top. Curr. Chem. 1979, 78, 145-186.
(8) Que, L., Jr.; Ho, R. Y. N. Chem. ReV. 1996, 96, 2607-2624.
(9) Que, L., Jr.; Reynolds, M. F. Met. Ions Biol. Syst. 2000, 37, 505-525.
(10) Costas, M.; Mehn, M. P.; Jensen, M. P.; Que, L., Jr. Chem. ReV. 2004,
104, 939-986.
9
10.1021/ja051594e CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 9003-9013
9003

A R T I C L E S
Yin and Finke
the previous record of ∼500 TTOs.28 Practically, it means that
a few milligrams of V-based precatalyst is able to convert 10 g
of DTBC into grams of isolable, pure products in e312 h.24
Scheme 2. Oxidative Cleavage of 3,5-DTBC by Vanadium-Based,
Nonenzymatic Catechol Dioxygenase Catalysts
Our prior work also shows that the O2-uptake kinetics, beginning
7-
with V-based polyoxometalates, such as SiW9V3O40
or
P2W15V3O629-, could be curve-fit with the two-step mechanism,
A f B, then A + B f 2B, where the latter step is the kinetic
definition of autocatalysis in which a product (B) is also a
reactant.29 Kinetics studies of four of the 28 precatalysts
examined revealed that the true catalysts were not the starting
vanadium complexes; rather, the V-based precatalysts reacted
with a DTBC-derived product, B, to generate the true catalyst.
This precatalyst conversion demands that understanding the
nature of the true catalyst and eventually improving the broad
selectivity seen with such V-based dioxygenase catalysts
(Scheme 2) are important goals within the broader context of
dioxygenase chemistry.
Vanadium-based nonenzymatic dioxygenase catalysts react
with 3,5-di-tert-butylcatechol (hereafter 3,5-DTBC) to produce
mixtures of three classes of products (Scheme 2):24 intradiol
cleavage products, 3,5-di-tert-butyl-1-oxacyclohepta-3,5-diene-
2,7-dione (2) and 3,5-di-tert-butyl-5-(carboxymethyl)-2-furanone
(5); extradiol cleavage products, 4,6-di-tert-butyl-2H-pyran-2-
one (3) and spiro[1,4-benzodioxin-2(3H),2′(2H)-pyran]-3-
one,4′,6,6′,8-tetrakis(1,1-dimethylethyl) (4); and the autoxidation
product, 3,5-di-tert-butyl-1,2-benzoquinone (6). The product
numbers in Scheme 2 will be used subsequently throughout this
paper.
Tatsuno and co-workers reported catechol oxygenation ca-
7-
talysis using the heteropolyvanadates, PV14O429-, MnV13O38
,
and NiV13O38 .
7- 30 Isolation of the reaction intermediate afforded
a blue compound, tentatively assigned as the dimer complex
“[V(DBSQ)(DBCatH)]2” by elemental analysis, IR, EPR results,
and a cryoscopic measurement of the approximate molecular
weight (DBCatH ≡ monoprotonated di-tert-butylcatecholate
anion).30
The present contribution also builds off an important earlier
observation by Professor C. G. Pierpont at the University of
Colorado who, in his review of metal compounds containing
catecholate and semiquinone ligands, noted that the “product
distribution is rather similar” for catechol oxidations beginning
with different sources of vanadium (see p 347 elsewhere).31
Significantly, he also proposed that a mixed catecholate and
semiquinone vanadium oxo compound, [VO(DBSQ)(DTBC)]2,
could be a soluble component of catechol oxidation by
vanadium-based catalysts. Hence, the questions raised and
addressed herein include the following: is there a common
catalyst for the broad range of vanadium precursors employed
for V-based catechol dioxygenases? Alternatively, do vanadium-
based dioxygenases inherently lack selectivity for some as of
yet unexplained reason(s)?
Herein, we test the specific hypothesis that there is, in fact,
a common catalyst for the rather different vanadium-based cate-
chol dioxygenase systems described in the literature. We report
reaction selectivity, catalytic lifetime, EPR, MS, and kinetic
evidence which support this “common vanadium catechol dioxy-
genases catalyst” hypothesis. Our evidence supports Pierpont’s
suggestion that [VO(DBSQ)(DTBC)]2 is that common catalyst.
A subsequent paper in this series presents kinetic studies which
reveal how virtually any V-based precursors can yield the same
V-based DTBC dioxygenase catalyst, via the novel concept of
autoxidation-product-initiated dioxygenase catalysis.32
Previously, a survey of 28 vanadium-containing precatalysts
revealed that the vanadium polyoxometalate plus FeII precatalyst,
(n-Bu4N)5[(CH3CN)xFeII‚SiW9V3O40], was able to accomplish
125 000 Total Turnovers (hereafter TTOs) of DTBC dioxygen-
ation in ∼300 h.24 (For leading reviews on polyoxometalates,
see elsewhere.25-27) This level of combined catalytic activity
and lifetime is an improvement of 2 orders of magnitude over
(11) Funabiki, T. In Catalysis by Metal Complexes; Funabiki, T., Ed.; Kluwer
Academic Publishers: Dordrecht, The Netherlands, 1997; Vol. 19, pp 105-
155.
(12) Kru¨ger, H.-J. In Biomimetic Oxidations Catalyzed by Transition Metal
Complexes; Meunier, B., Ed.; Imperial College Press: London, 2000; pp
363-413.
(13) Yamahara, R.; Ogo, S.; Masuda, H.; Watanabe, Y. J. Inorg. Biochem. 2002,
88, 284-294 and references therein.
(14) Jang, H. G.; Cox, D. D.; Que, L., Jr. J. Am. Chem. Soc. 1991, 113, 9200-
9204.
(15) Dei, A.; Gatteschi, D.; Pardi, L. Inorg. Chem. 1993, 32, 1389-1395.
(16) Ito, M.; Que, L., Jr. Angew. Chem., Int. Ed. Engl. 1997, 36, 1342-1344.
(17) Jo, D.-H.; Que, L., Jr. Angew. Chem., Int. Ed. 2000, 39, 4284-4287.
(18) Pascaly, M.; Duda, M.; Schweppe, F.; Zurlinden, K.; Mu¨ller, F. K.; Krebs,
B. J. Chem. Soc., Dalton Trans. 2001, 828-837.
(19) Fe-based catechol dioxygenase model systems have been intensively studied
during the past decade.13 In most of the studies, 3,5-di-tert-butylcatechol
is employed as the substrate. The best intradiol and extradiol models
reported provide almost exclusively intradiol or extradiol products,
respectively. Premier Fe model systems to date are the catechol intradiol
dioxygenase [FeIII(TPA)DTBC]BPh4 (TPA ) tris(2-pyridylmethyl)amine,
tetradentate ligand), which adds oxygen to bound DTBC to yield the
intradiol cleavage products (product 2 (60%) and 5 (38%), Scheme 2) in
an overall 98% yield,14 and the catechol extradiol dioxygenase [FeIII(L)-
DBC]Cl (L ) TACN or Me3TACN, TACN ) 1,4,7-triazacyclononane,
tridentate ligand), which yields a near-quantitative yield of the extradiol
cleavage product 3 (Scheme 2, plus an isomer of product 3, 3,5-di-tert-
butyl-2H-pyran-2-one) with or without added Lewis bases.15-17
(20) Matsumoto, M.; Kuroda, K. J. Am. Chem. Soc. 1982, 104, 1433-1434.
(21) Bianchini, C.; Frediani, P.; Laschi, F.; Meli, A.; Vizza, F.; Zanello, P. Inorg.
Chem. 1990, 29, 3402-3409.
Experimental Section
Reagents. 3,5-DTBC (purchased from Aldrich, 99%, mp 96-99 °C)
was recrystallized three times using n-pentane under argon (mp 99-
100 °C) and stored in a Vacuum Atmosphere drybox (O2 level e5
(22) Vlcek, A., Jr. Chemtracts: Inorg. Chem. 1991, 3, 275-280.
(23) Nishinaga, A. In Catalysis by Metal Complexes; Kluwer Academic
Publishers: Dordrecht, The Netherlands, 1997; Vol. 19, pp 157-194.
(24) Weiner, H.; Finke, R. G. J. Am. Chem. Soc. 1999, 121, 9831-9842.
(25) Pope, M. T.; Mu¨ller, A. Polyoxometalates: From Platonic Solids to Anti-
Retroviral Activity. In Topics in Molecular Organization and Engineering;
Kluwer Academic Publishers: Dordrecht, The Netherlands, 1994; Vol. 10.
(26) Hill, C. L. Chem. ReV. 1998, 98, 1-390.
(28) Tatsuno, Y.; Tatsuda, M.; Otsuka, S. J. Chem. Soc., Chem. Commun. 1982,
1100-1101.
(29) Watzky, M. A.; Finke, R. G. J. Am. Chem. Soc. 1997, 119, 10382-10400
and ref 34 therein.
(30) Tatsuno, Y.; Nakamura, C.; Saito, T. J. Mol. Catal. 1987, 42, 57-66.
(31) Pierpont, C. G.; Lange, C. W. Prog. Inorg. Chem. 1994, 41, 331-442.
(32) Yin, C.-X.; Sasaki, Y.; Finke, R. G. Inorg. Chem., submitted.
(27) Pope, M. T.; Mu¨ller, A. Polyoxometalate Chemistry From Topology Via
Self-Assembly to Applications; Kluwer Academic Publishers: Dordrecht,
The Netherlands, 2001.
9
9004 J. AM. CHEM. SOC. VOL. 127, NO. 25, 2005

Vanadium-Based Catechol Dioxygenases
A R T I C L E S
ppm). (NB: it is important to recrystallize the 3,5-DTBC substrate more
than one time to remove impurities such as 3,5-di-tert-butyl-1,2-
benzoquinone.) The polyoxometalate precursors, (n-Bu4N)4H5PV14O42,33
(n-Bu4N)7SiW9V3O40,34 (n-Bu4N)9P2W15V3O62,34,35 (n-Bu4N)5FeII‚
SiW9V3O40,24 and (n-Bu4N)5Na2FeII‚P2W15V3O62,24 were synthesized
according to the most recent literature procedures. (n-Bu4N)4H2-γ-
SiW10V2O40, a gift from Prof. N. Mizuno, was prepared as described36
and characterized by 183W NMR. VO(acac)2 (Aldrich, 95%) and VCl3
(Aldrich) were stored in the drybox and used as received. NaOCH3
(Fisher Scientific) was stored under N2 and used as received. [CoIII(3,5-
DBSQ)(CN)4]2- crystals37 were a gift from Prof. M. Wicholas. HPLC
grade solvents (1,2-dichloroethane, acetonitrile, diethyl ether, and ethyl
acetate) and anhydrous grade methanol were purchased from Aldrich
and stored in the drybox; each of the above solvents was dried by
standing for at least 48 h over ∼5 vol % 3 or 4 Å molecular sieves
that had been preactivated by heating at >170 °C under vacuum (e1
Torr) for at least 12 h, then cooled under dry N2 in the drybox.
Anhydrous grade toluene was purchased from Aldrich, stored in the
drybox, and used without further drying. Et3N (Mallinckrodt) was
distilled over BaO under Ar and stored in the drybox. Anhydrous
certified ACS grade diethyl ether was purchased from Fisher Scientific
and used as received. Argon gas was purchased from General Air
(99.985%) and used as received.
2.0037). Single-crystal X-ray diffraction crystallography was performed
on a Bruker SMART 1K CCD X-ray diffractometer. CHN elemental
analyses were performed by Atlantic Microlab, Inc. (Norcross, Georgia)
or Galbraith Laboratories (Knoxville, Tennessee).
Preparation of Polyoxometalate Precursors. The polyoxo-
metalate precursors, (n-Bu4N)4H5PV14O42,33 (n-Bu4N)7SiW9V3O40,34 (n-
Bu4N)9P2W15V3O62,34,35 (n-Bu4N)5FeII‚SiW9V3O40,24 and (n-Bu4N)5Na2-
FeII‚P2W15V3O62,24 were prepared according to the most recent literature
procedures. Further details about the syntheses and characterizations
are provided in the Supporting Information.
Preparation and Characterization of Vanadium-Catecholate
Complexes. Three vanadium catecholate complexes, [VO(DBSQ)-
(DTBC)]2,38 [Et3NH]2[VO(DTBC)2]‚2CH3OH,39 and [Na(CH3OH)2]2-
[V(DTBC)3]2‚4CH3OH,40,41 were chosen from the literature as the best
available examples of relatively simple, well-characterized vanadium-
catecholate complexes that provide three alternative hypotheses as to
common components in V-based catechol dioxygenase reactions. The
details of their synthesis and characterization are provided in the
Supporting Information.
Synthesis and Characterization of Deprotonated Di-tert-butyl-
catecholate Salt, Na2(3,5-DTBC). The details of these experiments
are recorded in the Supporting Information, including Figures S8 and
S9.
Instrumentation. Air-sensitive samples were prepared in a drybox
prior to analyses. GC analyses were performed on an HP (Hewlett-
Packard) 5890 Series II gas chromatograph equipped with a FID
detector and a SPB-1 capillary column (30 m, 0.25 mm i.d.) with the
following temperature program: initial temperature, 200 °C (initial time,
2 min); heating rate, 2 °C/min; final temperature, 240 °C (final time,
3 min); injector temperature, 250 °C; FID detector temperature, 250
°C. GC-MS analyses were performed under the same temperature
program on an Agilent 5973N/GC 6890 instrument equipped with a
mass selective detector (70 eV) and an Agilent HP-5MS column (30
m). Negative ion electrospray ionization mass spectrometry (negative
ion ESI-MS) analyses were performed on a Thermo Finnigan LCQ
Advantage Duo MS directly coupled with a syringe pump (5 µL/min
feeding speed and 15 µL/min at purging; spray voltage -4.5 kV,
capillary voltage -(38-42) V, capillary temperature 180 °C) or on a
Fisons VG Quattro-SQ spectrometer by directly injecting an acetonitrile
solution (spray voltage -2.9 kV, sample cone voltage -25 V, source
temperature 80 °C). NMR spectra were obtained in CDCl3 or CD3CN
Selectivity Experiments. The DTBC dioxygenase selectivities of
the 10 vanadium model compounds which follow were examined with
the same substrate-to-vanadium ratio to test whether they produce a
similar product distribution: (n-Bu4N)4H5PV14O42, (n-Bu4N)7SiW9V3O40,
(n-Bu4N)5[FeII‚SiW9V3O40], (n-Bu4N)9P2W15V3O62, (n-Bu4N)5Na2[FeII‚
P2W15V3O62], (n-Bu4N)4H2-γ-SiW10V2O40, [VVO(DBSQ)(DTBC)]2,
[Et3NH]2[VIVO(DBTC)2]‚2CH3OH, [Na(CH3OH)2]2[VV(DTBC)3]2‚
4CH3OH, and VO(acac)2. Oxygenation experiments were carried out
on a volume-calibrated oxygen-uptake line, as detailed in the Supporting
Information elsewhere.24 The standard procedure used for these
experiments is as follows: 400 ( 5 mg (ca. 1.8 mmol) of three-times-
recrystallized 3,5-DTBC was weighed in the drybox into a 50 mL
round-bottom reaction flask equipped with a septum, sidearm, and an
3
egg-shaped 3/4 in. × /8 in. Teflon-coated magnetic stir bar. Using a 10
mL glass syringe, approximately 8 mL of predried HPLC grade 1,2-
dichloroethane was transferred into the flask. Then, the flask was sealed
with a Teflon stopcock and taken out of the drybox. The flask was
connected to the oxygen-uptake line through an O-ring joint, and the
reaction solution was frozen in a dry ice/ethanol bath (-76 °C). Two
pump-and-fill cycles with O2 as the refill gas were performed. Next,
the dry ice bath was replaced with a temperature-controlled oil bath.
The temperature of the flask was increased to 40 ( 0.7 °C and allowed
to equilibrate with stirring for 25 min. In the drybox, 0.5-2.0 mg of a
predetermined vanadium precatalyst (mole ratio of substrate to the
vanadium in the precatalysts of ca. 1000:1) was weighed into a glass
1
(Cambridge Isotope Lab). H, 31P, and 51V NMR were recorded in 5
mm o.d. tubes on a Varian Inova (JS-300) NMR spectrometer. 1H NMR
was referenced to the residual impurity in the deuterated solvent; 31P
NMR was referenced to 85% H3PO4 in H2O using the external
substitution method, and 51V NMR was referenced to neat VOCl3
(Aldrich) using the external substitution method. Spectral parameters
for 31P NMR include: tip angle ) 60° (pulse width 10 µs); acquisition
time, 1.6 s; sweep width, 10 000 Hz. Spectral parameters for 51V NMR
include: 51V tip angle ) 90° (pulse width 3.1 µs); acquisition time,
0.096 s; sweep width, 83 682.0 Hz. Infrared spectra were obtained on
a Nicolet 5DX spectrometer as KBr pellets (Aldrich, spectrophotometric
grade) or in a CaF2 cell (A ) 0.1 mm). UV-visible spectra were
obtained on an HP 8452A diode spectrophotometer in glass UV cells
sealed by ground-glass stopcocks. Electron paramagnetic resonance
(EPR) spectra were recorded on a Bruker EMX 200U EPR spectrom-
eter. Quartz EPR tubes of 4 mm o.d. were used, and DPPH (2,2-
diphenyl-1-picrylhydrazyl) was used as the reference compound (g )
(38) Cass, M. E.; Green, D. L.; Buchanan, R. M.; Pierpont, C. G. J. Am. Chem.
Soc. 1983, 105, 2680-2686.
(39) Cooper, S. R.; Koh, Y. B.; Raymond, K. N. J. Am. Chem. Soc. 1982, 104,
5092-5102.
(40) Luneva, N. P.; Mironova, S. A.; Lysenko, K. A.; Antipin, M. Y. Russ. J.
Coord. Chem. 1997, 23, 844-849.
(41) (a) Cass, M. E.; Gordon, N. R.; Pierpont, C. G. Inorg. Chem. 1986, 25,
3962-3967. (b) Cass, M. E. Ph.D. Thesis, University of Colorado, Boulder,
CO, 1984; p 67. (c) Private communication with Professor Cort G. Pierpont.
In our preparation of Na[V(DTBC)3]‚4CH3OH using a V(II) solution
(prepared by Zn/Hg amalgam reduction of a NH4VO3 solution), crystals
identified by UV-visible, EPR, and X-ray crystallography as ZnII(Cat-N-
SQ)(BQ-N-SQ) were obtained instead (Chaudhuri, P.; Hess, M.; Hilden-
brand, K.; Bill, E.; Weyhermuller, T.; Wieghardt, K. Inorg. Chem. 1999,
38, 2781-2790), where (Cat-N-SQ)2- and (BQ-N-SQ)0 are two redox
isomers of (Cat-N-BQ)-, where these ligands are the iminosemiquinone
(SQ) and iminobenzoquinone (BQ) versions of the 3,5-di-tert-butylbenzo-
quinone 1-(2-hydroxy-3,5-di-tert-butylphenyl)imine monoanion. That this
known, dark-green crystalline compound was not the intended vanadium
complex Na[V(DTBC)3]‚4CH3OH was determined by its room-temperature
EPR (a multiline spectrum, shown as Figure S4 of the Supporting
Information, and its UV-visible maximum in CHCl3 is at 736 nm instead
of the reported41a 650 nm of Na[V(DTBC)3]‚4CH3OH).
(33) (a) Preuss, F.; Schug, H. Z. Naturforsch., B: Anorg. Chem., Org. Chem.
1976, 31B, 1585-1591. (b) Kato, R.; Kobayashi, A.; Sasaki, Y. J. Am.
Chem. Soc. 1980, 102, 6571-6572. (c) Kato, R.; Kobayashi, A.; Sasaki,
Y. Inorg. Chem. 1982, 21, 240-246.
(34) Finke, R. G.; Rapko, B.; Saxton, R. J.; Domaille, P. J. J. Am. Chem. Soc.
1986, 108, 2947-2960.
(35) Hornstein, B. J.; Finke, R. G. Inorg. Chem. 2002, 41, 2720-2730.
(36) Nakagawa, Y.; Uehara, K.; Mizuno, N. Inorg. Chem. 2005, 44, 14-16.
(37) Arzberger, S.; Soper, J.; Anderson, O. P.; La Cour, A.; Wicholas, M. Inorg.
Chem. 1999, 38, 757-761.
9
J. AM. CHEM. SOC. VOL. 127, NO. 25, 2005 9005

A R T I C L E S
Yin and Finke
vial and dissolved in ca. 0.2 mL of 1,2-dichloroethane (0.2 mL toluene
was used for [Na(CH3OH)2]2[V(DTBC)3]2‚4CH3OH because of its
greater solubility in toluene). The resultant precatalyst solution was
drawn into a 1 mL gastight syringe and brought out of the drybox with
its needle protected from air via its insertion into a septum-capped vial.
This precatalyst solution was then injected through the sidearm of the
reaction flask and t ) 0 was set. Pressure readings from the manometer
were used to follow the reaction ((1 Torr or ca. (1% precision over
a pressure loss of ca. 80-100 Torr). The reaction was stopped at 23.5
h unless noted otherwise, and the reacted mixture was diluted 21-fold
using 1,2-C2H4Cl2 prior to GC analysis.
31P NMR Following Oxygen-Uptake Using Precatalyst (n-
Bu4N)9P2W15V3O62. Results are shown in Figure S10; and the
experimental details are included in the Supporting Information.
EPR Following Oxygen-Uptake Experiments. The procedure
employed here is exactly as reported above in the “Selectivity
Experiments”, except that (i) the precatalyst amount was chosen so
that the amount of vanadium from the precatalyst was ca. 6-8 µmol,
(ii) toluene was employed as the solvent in place of 1,2-dichloroethane,
with the toluene being frozen using liquid nitrogen instead of dry ice/
acetone and (iii) the precatalyst was premixed with DTBC in the drybox
instead of being injected into a stirred DTBC/O2 solution. The oxygen
pressure loss was monitored with a manometer. An aliquot of the
reaction solution was drawn under an oxygen flow at ca. 0.5-2 h after
the oxygen pressure started to decrease. The reaction solution was
sampled again after no further pressure change was observed for >2
h.
Quantitative EPR Analysis. The standard solutions were prepared
by dissolving an air-stable, semiquinone-containing crystal, [CoIII(3,5-
DBSQ)(CN)4]2- 37
, into toluene (1.0 M) and serial diluting to concentra-
tions of 0.10 and 0.010 M. The reaction solution (catalyzed by the
precatalyst VO(acac)2) was harvested ca. 2 h after oxygen-uptake
ceased, as described above (in the section titled EPR Following Oxygen-
Uptake Experiments). The double-integrated area of [VO(DBSQ)-
(DTBC)]2 was compared to that of [CoIII(3,5-DBSQ)(CN)4]2- to
estimate the concentration of [VO(DBSQ)(DTBC)]2.
Negative Ion ESI-MS Analysis. These experiments were performed
on a mass spectrometer (Thermo Finnigan LCQ) directly coupled to a
syringe pump. The solvent in the syringe serves as the matrix for the
negative ion ESI-MS. Solid samples were diluted in the drybox using
predried acetonitrile to ca. 10 µM (ca. 0.01-0.1 mg in 1 mL) and placed
in a 2 mL septum screw-capped vial. The sample was then transferred
into a rinsed 500 µL syringe outside the drybox; the syringe needle
was inserted into the MS sampler hose as fast as possible to avoid
oxygen contamination. The injection rate was controlled by a syringe
pump. The MS peaks were observed ca. <1 min after the injection.
The influence of water in the ESI-MS matrix was examined with a set
of control experiments using [VO(DBSQ)(DTBC)]2 as the sample and
varying the ratio of CH3CN:H2O from 1:0, 9:1, 1:1, 1:9, to 0:1 in the
solvent. Although this instrument is not optimal for pure water samples
(as shown by the lower signal level as the water ratio goes up), the
same peaks were observed as in neat acetonitrile.
Kinetic Competence Experiments. The kinetic competence experi-
ments were set up the same way as the selectivity experiments described
above, except that a higher concentration (0.46-0.48 mM) of the
precursor was used to increase the O2-uptake rate.
Attempted Catalyst Isolations. Isolation and characterization of
the actual catalyst, if possible, would of course provide direct evidence
of the identity of the true catalyst, although the obvious caveat here is
that good catalysts are often metastable transients that are not readily
isolable (see “Halpern’s Rules” as detailed elsewhere50). Nevertheless,
it was important to attempt to isolate an active catalyst (or even a
deactivated form of the active catalyst) from the vanadium-containing
Catalytic Lifetime Experiments. These experiments were per-
formed to determine if the V complexes, (n-Bu4N)4H5PV14O42, (n-Bu4N)7-
SiW9V3O40, (n-Bu4N)5[FeII‚SiW9V3O40], (n-Bu4N)9P2W15V3O62, (n-Bu4N)5-
Na2[FeII‚P2W15V3O62], [VVO(DBSQ)(DTBC)]2, [Et3NH]2[VIVO(DBTC)2]‚
2CH3OH, [Na(CH3OH)2]2[VV(DTBC)3]2‚4CH3OH, and VO(acac)2, dif-
fer in their catalytic lifetimes. Experiments were performed with the
apparatus described in the literature;24 a catalytic lifetime experiment
is briefly described below, in which about half the amount of substrate
(∼29 mmol), yielding e50 000 TTOs, is employed in comparison to
the largest scale previously employed (∼63 mmol substrate which
yielded 100 000-150 000 TTOs24). The precatalyst amounts were in
the range of 0.4-0.7 µmol for all the catalytic lifetime experiments.
The e50 000 TTO catalytic lifetime experiment proceeded as
follows: ca. 6.50 g (29.2 mmol) of three-times-recrystallized 3,5-DTBC
was weighed into a 250 mL round-bottom flask with a sidearm and an
1
egg-shaped 1 in. × /2 in. Teflon-coated magnetic stir bar. About 125
mL 1,2-C2H4Cl2, measured in a graduated cylinder, was added to the
reaction flask. Then, 1.1-4.0 mg of a chosen precatalyst was weighed
into a vial (the mole ratio of substrate to the catalyst is ca. 45000:1),
dissolved in ca. 0.2 mL of 1,2-C2H4Cl2, and quantitatively transferred
into the reaction flask via a pipet. The flask was sealed, brought out of
the drybox, and frozen in a dry ice/ethanol bath (-76 °C) for 1 h.
Next, the flask was connected to a condenser under an argon flow.
Two pump-and-fill cycles with O2 as the refill gas were performed,
followed by heating the solution to 65 ( 1 °C via a temperature-
controlled oil bath (this took ca. 30 min). Once a constant temperature
was achieved, t ) 0 was noted. The flask was connected to an oxygen
tank through a bubbler, keeping a slow positive flow of O2 (ca. 0-5
psig) atop the solution. An aliquot of ca. 0.15 mL was drawn from the
solution every 24-48 h with a disposable polyethylene 1 mL syringe
equipped with a predried stainless steel needle. The aliquot was diluted
21-fold using 1,2-C2H4Cl2, and then subjected to GC analysis.
The 100 000-150 000 TTO catalytic lifetime experiments were
performed via the same procedures as above, except ca. 14 g of 3,5-
DTBC was employed (substrate-to-precursor mole ratio within 100 000-
150 000).
VO(acac)2 Aging Studies. The effects of aging the VO(acac)2 were
tested due to our inability to reproduce the low TTOs observed (<6000)
in the TTO catalytic lifetime experiment reported in our previous
studies.24 A putatively fresh sample was purchased from Aldrich (95%)
and stored in the drybox upon receiving. One aged sample of VO(acac)2
was prepared by placing the fresh VO(acac)2 in a vial, taking the vial
out of the drybox, and then exposing the sample to atmospheric oxygen
for 2 months. The other sample was prepared by heating ground
VO(acac)2 (from a bottle which had been stored outside the drybox
for over 2 years) in a 60 °C thermostatic oven in air for 6 days. During
this time, the color darkened from green to green-black. The above
two aged samples together with a fresh sample were examined by EPR
in CH2Cl2 (the two aged samples did not totally dissolve, presumably
due to the presence of insoluble decomposition products). Double-
integrated EPR peak areas of the three compounds were compared,
but no difference was found within experimental error. The first aged
sample was also tested in a 100 000-150 000 TTO catalytic lifetime
experiment and yields the same catalytic lifetime within experimental
error as the fresh sample (Table S3 of the Supporting Information).
(42) Tatsuno, Y.; Tatsuda, M.; Otsuka, S.; Tani, K. Chem. Lett. 1984, 1209-
1212.
(43) Casellato, U.; Tamburini, S.; Vigato, P. A.; Vidali, M.; Fenton, D. E. Inorg.
Chim. Acta 1984, 84, 101-104.
(44) Roman, E.; Tapia, F.; Barrera, M.; Garland, M. T.; Le Marouille, J. Y.;
Giannotti, C. J. Organomet. Chem. 1985, 297, C8-C12.
(45) Galeffi, B.; Postel, M.; Grand, A.; Rey, P. Inorg. Chim. Acta 1987, 129,
1-5.
(46) Galeffi, B.; Postel, M.; Grand, A.; Rey, P. Inorg. Chim. Acta 1989, 160,
87-91.
(47) Nishida, Y.; Kikuchi, H. Z. Naturforsch., B: Chem. Sci. 1989, 44, 245-
247.
(48) Russo, U.; Zarli, B.; Zanonato, P.; Vidali, M. Polyhedron 1991, 10, 1353-
1361.
(49) Zhang, B. Y.; Zhang, Y.; Chen, B. W.; Wang, K. Chin. Chem. Lett. 1997,
8, 547-550.
(50) Hagen, C. M.; Vieille-Petit, L.; Laurenczy, G.; Su¨ss-Fink, G.; Finke, R.
G. Organometallics 2005, 24, 1819-1831; see footnote 45 therein.
9
9006 J. AM. CHEM. SOC. VOL. 127, NO. 25, 2005

Vanadium-Based Catechol Dioxygenases
A R T I C L E S
Table 1. A Compilation of the Product Distributions and Mass Balance of 11 Literature Reports of V-Based Oxidative Cleavage of
3,5-DTBC
Product Distributions (%)
Mass Balance and TTOs
b
c
d
precatalyst(s)a
2
3
4
6
mass balance
TTOmax
TTOconv
TTOyield
ref
1
2
3
VO(acac)2, VO(salen),
VCl(salen), VCl(saldpt)
[V(salen)(DbcatH)2]‚1/2CH2Cl2,
[V(saldpt)(DBcatH)]‚CH2Cl2
[VO(acac)(OCH3)]2, [VO(tmh)(OCH3)]2,
(VOaap)2, (VOdmba)2, (VOdba)2,
39-43
6-15
22-28
e83%
1000
1000
e750
28
42
11
24
77%
100
100
77
42
43
e
46-48
7-10
22-25
e82%
100
e82
(VO(acac)OCH3)2
4
[LVO(acac)], [LVO(phen)]+PF6
,
<10f
70f
<80%
44
-
[LVO(bipy)]+PF6- (O2 or H2O2 as oxidant)
5
6
VO(acac)(TCCat)
45
34-36
6
24
15-21
75%
e82%
100
100
75
e82
45
30
7-
PV14O429-, MnV13O38
NiV13O387-, VO(acac)2
[VO(acac)(DTBCat)]
,
11-20
3-13
7
8
45
15-39
6
24
5-19
75%
58%
46
47
cis-V2W4O194-, 1,4,9-[PV3W9O40]6-
V10O286-, VO(acac)2
VO(HL)2OH,
,
100
100
e58
e82
9
45-48
5-10
6-15
23-25
e82%
100
48
(VO)2L2SO4‚4H2O
10
11
TEA[VO(TEA)]
80g
40-57
e82%
e94%
10
4500
8.2
4230
49
24
SiW9V3O407-, FeII‚SiW9V3O40
P2W15V3O629-, FeII‚P2W15V3O62
,
10-18
9-25
4500
5-
7-
,
PV14O429-, VO(acac)2, etc. (28 total)
a Inactive vanadium-containing precatalysts omitted from the above table, but listed here for the sake of completeness are [V(salen)(cat)]‚0.1CH2Cl2 and
[V(salen)(4-Bcat)]‚H2O;42 Na[VO(DBcat)2] and Na2[VO(OCH3)(DBcat)2];30 VO(acac)(cat) and VO(acac)(3-Bcat);46 [VW5O193-];47 VO(NTA);49 and
[SiW11VO405-].24 b TTOmax ) molessubstrate/molesprecatalyst c TTOconv ) molessubstrate, consumed/molesprecatalyst d TTOyield ) molessum of products/molesprecatalyst e The
. . .
product distribution was not reported for the denoted precatalyst. f We speculate the reason for these different selectivities is that the bulky ligand (L) blocks
the oxygen binding to the vanadium-catecholate compound. g This yield should be the sum of 2 and 6 due to its “red” color; the yield of 5 is 3%. Abbreviations:
salen ) ethylenebis(salicylideneaminato); saldpt ) N,N′-(3,3′-dipropylamine)bis(salicylideneaminato); tmh ) 2,2,6,6-tetramethylheptandione; H2aap )
o-hydroxyacetophenone;H2dmba)1,5-bis(p-methoxyphenyl)-1,3,5-pentanedione;H2dba)1,5-diphenyl-1,3,5-pentanetrione;L(entry4))[η-CpCo{P(O)(OC2H5)2}3]-;
bipy ) 2,2′-bipyridine; TCCat ) tetrachlorocatecholate; DBcat/DTBCat ) 3,5-di-tert-butylcatecholate; H2L (entry 9) ) 2,2′-dihydroxy-3,3′-diacetyl-5,5′-
dichlorodiphenylmethane; TEA ) triethanolamine; cat ) pyrocatecholate; Bcat ) tert-butylcatechol; and NTA ) nitrilotriacetic acid.
less than 1%).17 These results show that, as one would expect,
polyoxometalate precatalysts, especially in light of earlier indications
that this might be possible.24,30 However, in the end, we were not able
there is not an inherent insensitivity in the DTBC oxygenation
to isolate analytically pure catalysts from these particular studies, so
product selectivities, at least when considering different metals.
that we turned to the use of in situ spectroscopic studies (vide infra).
The hypothesis for the remainder of this work then became
Details of the attempted isolation studies are in the Supporting
the “common catalyst hypothesis”, namely, that there is a
Information, along with figures showing IR, negative ion ESI-MS, or
common catalyst for the rather different vanadium-based
catechol dioxygenase systems described in the literature, Table
1. The results in Table 1 are, of course, themselves consistent
EPR of the resultant materials (Figures S11-S14).
Results and Discussion
with and supportive of this hypothesis.
A Literature Survey of V-Based 3,5-Di-tert-butylcatechol
Oxygenations. The published selectivity and lifetime data for
the 11 reports of vanadium-based DTBC oxygenations are
shown in Table 1. The tabulated data yield the following
insights: muconic acid anhydride (2) is the major product (40-
50%), followed in decreasing order by 3,5-di-tert-butylquinone
(6, 20-30%), pyranone (3), the spiro product (4), and the
furanone product (5); the last three products have a combined
yield of 10-20%. Clearly, the product selectiVities are rather
similar for all of the V-based DTBC precatalysts examined in
the literature to date and which are actiVe, as Professor Pierpont
had hinted.31 A second, relevant insight comes from the literature
for Fe catechol dioxygenase models;13 the product distributions
therein are often quite different compared to those in Table 1.
For example, Fe(III) model complexes with tripodal N4 donor
ligands and catecholate ligands catalyze the intradiol cleavage
of 10-100 equiv of 3,5-DTBC to product 2 in 30-84% yield
(the yield of the autoxidation product 6 is less than 6%);18
[FeIII(tacn)Cl(DTBC)] (where tacn ) 1,4,7-triazacyclononane)
reacts with O2 to produce extradiol cleavage product 3 in up to
78% yield with the addition of 1 equiv of AgBF4 and 20 equiv
of 4-methylpyridine (the yield of an isomer of product 3, 3,5-
di-tert-butyl-2H-pyran-2-one, is 20%; the yield of product 6 is
Choice of Three Simple, Well-Characterized Vanadium-
Containing DTBC Complexes for Further Investigation.
Given the data in Table 1 and the resultant common catalyst
hypothesis, the next task was to identify vanadium-DTBC and/
or -DBSQ complexes known in the literature that were logical
starting points as either (i) precatalysts closer in composition
and structure to the hypothesized true catalyst(s), or conceivably
(ii) actual catalysts themselves. Three vanadium-DTBC or
DTBC/DBSQ complexes quickly rose to the top of the list
from a scrutiny of the literature: Pierpont and co-worker’s38
[VVO(DBSQ)(DTBC)]2, Raymond and co-worker’s39 [Et3NH]2-
[VIVO(DTBC)2]‚2CH3OH, and Luvena and co-worker’s40
[Na(CH3OH)2]2[VV(DTBC)3]2‚4CH3OH (Chart 1). In addition,
it proved to be productive to examine the simple VO(acac)2 as
an example of a well-known V-precatalyst without DTBC or
DBSQ ligands. The VO(acac)2 was purchased commercially
(Aldrich, 95%), whereas the other three complexes were
prepared according to the literature as detailed in the Supporting
Information.
All three vanadium catecholate complexes were characterized
by UV-visible spectroscopy, negative ion electrospray ioniza-
tion mass spectrometry (ESI-MS), and elemental analysis (plus
9
J. AM. CHEM. SOC. VOL. 127, NO. 25, 2005 9007

A R T I C L E S
Yin and Finke
Chart 1. Structures of the Three Vanadium Catecholate
Complexes Studied Herein (tert-butyl groups are shown as R, only
one isomer of [VIVO(DTBC)2]2- is shown, and coordinated
methanol is not shown in the middle structure for the sake of
clarity). Also Shown Is the Structure of
also the same, and interestingly, no induction periods are
observed for [VO(DBSQ)(DTBC)]2 or VO(acac)2 (the absence
of an induction period for VO(acac)2 is as reported in the
literature24). Note that the lack of an induction period is kinetic
evidence that we are closer to the catalyst in these cases, if not
at a true catalyst in the [VO(DBSQ)(DTBC)]2 case, vide infra.
(ii) The vanadium-containing polyoxometalates are all the
same within experimental error in their selectivities and O2-
uptake values.51 Interestingly, there are no discernible induction
periods for two of the polyoxometalates, (n-Bu4N)4H5PV14O42
and (n-Bu4N)4H2-γ-SiW10V2O40. This, plus knowledge of the
properties of these particular polyoxometalates,33,36 suggests the
release of VVO2+ or possibly VIVO2+ from the polyoxometalate
structure,52 as supported by the thermogravimetric studies on
H2(VVO2)PW12O40 and by the structure of Na2(VIVO)[SiW12O40]‚
13H2O.52a
[Na(CH3OH)2]2[V(DTBC)3]2‚4CH3OH Obtained through X-ray
Single-Crystal Diffraction
(iii) Note that Fe is not needed for the oxygenation catalysis
activity, only V is required for facile dioxygenase catalysis from
the 15 entries in Table 2. A control using (n- Bu4N)4SiW12O40
confirms that polyoxometalates lacking V are inactiVe toward
catechol oxygenation.24
(iv) The control using (n-Bu4N)5SiW11VO40 (Table 2, entry
11 (data from our previous study24)) reveals the inactiVity of
this single Vanadium-containing polyoxometalate. This inactivity
is consistent with either its stability and/or the possibility that
more than one vanadium is needed for oxygenation catalysis.
(v) The two active species isolated from (n-Bu4N)5[FeII‚
SiW9V3O40]24 (grayish black, in the presence of DTBC under
O2) and (n-Bu4N)4H5PV14O4230 (blue, in the presence of DTBC
under N2) give the same selectivities within experimental error
as the other species in Table 2. The blue species was tentatively
assigned by Tatsuno and co-workers as “[V(DBSQ)(DBCatH)]2”
(DBCatH, monoprotonated di-tert-butylcatecholate anion).30
(vi) The (no vanadium) control of Na2(DTBC) alone was also
examined under our conditions. Previously, deprotonated 3,5-
DTBC was reported to react with dissolved oxygen in ether.53
Depending on the amount of H2O present, 3,5-di-tert-butyl-5-
(carboxymethyl)-2-furanone (product 5) can be obtained in 20-
60% yield.53 In our hands, the light blue disodium salt of 3,5-
DTBC in diethyl ether plus O2 yields furanone 5 in 13-21%
yield (following acidifying the product mixture with 5% HCl
and then extracting with 1,2-C2H4Cl2). The major product in
the vanadium-catalyzed reactions, muconic acid anhydride
(product 2), is barely detectable by GC (<0.01%). If the above
reaction is carried out in 1,2-C2H4Cl2 (i.e., in the solvent used
for most of our selectivity experiments,) instead of in Et2O, then
the major products are 3,5-di-tert-butyl-1,2-benzoquinone (30%
yield) and an unknown compound (m/z ) 248, yield ∼57%
estimated from the area ratio to 3,5-di-tert-butyl-1,2-benzo-
quinone). All other products (i.e., 2-5) are less than 1%. A
single-crystal X-ray crystallography in the case of [Na(CH3-
OH)2]2[VV(DTBC)3]2‚4CH3OH). The elemental analysis results
of different batches of [VO(DBSQ)(DTBC)]2 show that a
variable amount of methanol solvates are present in the
compounds; hence, the difficulty in obtaining single crystals is
ascribed to crystal deterioration due to solvate loss at room
temperature.38 In summary, analytically pure samples of each
of the three literature vanadium DTBC/DBSQ complexes were
prepared so that their DTBC oxygenation selectivities, kinetics,
and catalytic lifetimes could be studied.
Confirmation of the Similar Selectivities For the Three
Vanadium DTBC/DBSQ Complexes, VO(acac)2, As Well As
for Six V-Polyoxometalate Precatalysts. Selectivity experi-
ments were carried out in 1,2-C2H4Cl2 for the 10 com-
plexes; studies of isolated, active components and control studies
were also performed (entries 11-15, Table 2), for a total of 15
studies. The key results include the following: (i) The selectivi-
ties are the same for the three V-precatalysts, as well as for
VO(acac)2 (entries 1-4, Table 2). The O2-uptake values are
(51) Ca. 14-18% less of product 4 in the cases of the vanadium-based
polyoxometalate precatalysts and VO(acac)2 compared to our literature24
is seen, due presumably to the instability of product 4 in the reaction solution
(higher yields of product 4 can be obtained if sampled immediately after
the O2-uptake ceases).
(52) (a) Cadot, E.; Marchal, C.; Fournier, M.; Te`ze`, A.; Herve`, G. Polyoxo-
metalates: From Platonic Solids to Anti-Retroviral Activity. In Topics in
Molecular Organization and Engineering; Pope, M. T., Mu¨ller, A., Eds.;
Kluwer Academic Publishers: Dordrecht, The Netherlands, 1994; Vol. 10,
pp 315-326. (b) Kozhevnikov, I. V. Chem. ReV. 1998, 98, 171-198 (see
pp 189-191 and references therein). (c) A detailed speciation study on
H+-H2VO4-H2O2 systems: Andersson, I.; Angus-Dunne, S.; Howarth,
O.; Pettersson, L. J. Inorg. Biochem. 2000, 80, 51-58.
(53) Speier, G.; Tyekla´r, Z. J. Mol. Catal. 1990, 57, L17-L19.
9
9008 J. AM. CHEM. SOC. VOL. 127, NO. 25, 2005

Vanadium-Based Catechol Dioxygenases
A R T I C L E S
Table 2. Catalytic Selectivity Experiments, as well as O2-Uptake and Induction-Period Observations, for the Three Classes of
Vanadium-Based Compoundsa
Products Yield (%)
2
3
4
6
induction period?
∆molO2/molDTBC
Four Simple V Complexes
1
2
3
[VO(DBSQ)(DTBC)]2
[Et3NH]2[VO(DTBC)2]‚2CH3OH
[Na(CH3OH)2]2
[V(DTBC)3]2‚4CH3OH
VO(acac)2
40 ( 2
38 ( 1
38 ( 1
12 ( 1
12 ( 1
14 ( 1
-
-
18 ( 1
18 ( 1
18 ( 1
no
yes
yes
0.64
0.63
0.62
1 ( 1
4
42 ( 1
11 ( 1
-
16 ( 1
no
0.63
V-Based Polyoxometalate Precatalysts
5
6
7
8
9
(n-Bu4N)4H5PV14O42
(n-Bu4N)7[SiW9V3O40]
46 ( 2
49 ( 1
40 ( 1
40 ( 2
42 ( 2
43 ( 1
∼1
13 ( 1
17 ( 1
13 ( 1
14 ( 1
13 ( 1
13 ( 1
-
16 ( 1
11 ( 1
12 ( 1
18 ( 1
22 ( 1
18 ( 1
no
0.64
0.61
0.59
0.49
0.55
0.66
2 ( 1
3 ( 1
3 ( 1
4 ( 1
-
yes
yes
yes
yes
no
(n-Bu4N)5[FeII‚SiW9V3O40]
(n-Bu4N)9[P2W15V3O62]b
(n-Bu4N)7[FeII‚P2W15V3O62]b
(n-Bu4N)4H2-γ-SiW10V2O40
(n-Bu4N)5SiW11VO40
10
11
(control; data from literature24)
Isolated Active Components
12
13
isolated active species
39 ( 2
13 ( 1
19 ( 1
13 ( 1
no
no
0.61
0.45
from (n-Bu4N)4H5PV14O42
isolated active species
from (n-Bu4N)5[FeII‚SiW9V3O40]c
40 ( 2
7 ( 1
Controls
14
15
Na2(DTBC)d
30 ( 3
2 ( 1
[CoII(CH3CN)6](BF4)2
e
a Experimental conditions: ca. 400 mg of 3,5-di-tert-butylcatechol, ca. 1000:1 mole ratio of substrate to vanadium in the precatalysts, 40.0 ( 0.7 °C, 9
mL of 1,2-C2H4Cl2 solvent; reaction stopped at 23.5 h; product yields obtained by GC. b These two reactions were stopped at ca. 42 and 24 h, respectively.
c This reaction was stopped at ca. 94 h; the exact number of moles of the precatalyst are not known for this incompletely characterized compound. d This
experiment was run with ca. 410 mg of Na2DTBC (1.55 mmol), 2 equiv of (n-Bu4N)Br (1 g, 3.l mmol, to facilitate the phase-transfer of Na2DTBC into the
organic phase), 8 mL of 1,2-C2H4Cl2 solvent; reaction stopped at 161 h; product mixture was concentrated, acidified with 5% HCl, and then extracted into
1,2-C2H4Cl2 and analyzed by GC. e Substrate to precatalyst mole ratio is ca. 700:1.
control experiment showed that the GC product yields are
identical for the original product solution compared to the
extracted solution acidified with 5% HCl. In short, the major
intradiol cleavage dioxygenase reaction product 2 is obtained
only upon interaction with a metal, in our case vanadium.
(vii) The control experiment using [CoII(CH3CN)6](BF6)2
produces only 2% of benzoquinone product 6, showing that the
vanadium-based systems do give very different products and
selectivities than such slow autoxidation systems.
In summary, the 12 Vanadium precursors in Table 2 all giVe
the same selectiVities within experimental error in DTBC
oxygenation catalysis.
Figure 1. Catalytic lifetime study using nine different vanadium-based
precatalysts. 1 ) [VO(DBSQ)(DTBC)]2; 2 ) [Et3NH]2[VO(DTBC)2]‚
2CH3OH; 3 ) [Na(CH3OH)2]2[VV(DTBC)3]2‚4CH3OH; 4 ) VO(acac)2; 5
) (n-Bu4N)4H5PV14O42; 6 ) (n-Bu4N)7SiW9V3O40; 7 ) (n-Bu4N)5[Fe‚
SiW9V3O40]; 8 ) (n-Bu4N)9P2W15V3O62; 9 ) (n-Bu4N)5Na2[Fe‚P2W15V3O62].
Error bars for the substrate-based TTOs are not available due to the complete
conversion (>99%) of the substrate at the end of each reaction.
Catalytic Lifetimes of the Four Simple V-Complexes and
Five V-Polyoxometalate Precatalysts. (A) e50 000 Total
Turnover (TTO) Experiments.54 A series of catalytic lifetime
experiments (e50 000 TTOs) were carried out with ca. 29 mmol
3,5-DTBC and ca. 0.7 µmol precatalyst to determine if the
catalyst lifetimes vary or are similar. The four simple vanadium
complexes and five vanadium polyoxometalate complexes show
the same catalytic lifetimes within experimental error (Figure
1). The product selectivities are also the same within experi-
mental error for these nine precatalysts examined in the catalytic
lifetime experiments. (Full details are provided in Table S2 of
the Supporting Information.) Rather clearly, the true catalyst
cannot be polyoxometalate-based. In short, the results offer
further support for the hypothesis that a common catalyst is
generated regardless of the V-based precursor employed.
(B) 100 000-150 000 TTO Experiments. Several larger-
scale catalytic lifetime experiments were carried out with mole
ratios of substrate to catalyst (3,5-DTBC to (n-Bu4N)7-
SiW9V3O40 or (n-Bu4N)5FeII‚SiW9V3O40) from 100 000:1 to
150 000:1 to verify the prior record TTOs of ca. 100 000. The
results show the product-based TTO catalytic lifetimes of the
above two precursors of 60 000-80 000 ((5000) are compa-
rable with the previous results employing (n-Bu4N)5FeII‚
SiW9V3O40 of 90 000-100 000 ((10 000) product-based TTOs.24
Details of the results are provided in Table S3 of the Supporting
Information.
Before our previous studies,24 the highest reported catalytic
lifetime of VO(acac)2 was only ∼500 TTOs.28 Our previous
24
result of the modest TTOs (<6000) exhibited by VO(acac)2
was >10-fold better than that in the literature,28 so that the
<6000 TTOs seemed to be reliable at the time. However, given
9
J. AM. CHEM. SOC. VOL. 127, NO. 25, 2005 9009

A R T I C L E S
Yin and Finke
Table 3. EPR Spectra of O2-Uptake Solutions of Three Disparate Kinds of Precatalysts during and after the Reactiona
a The full spectra were scanned at a magnetic field width of 1000 G while the center-field scans were obtained at 50 G. b The spectra of the oxygenation
reaction solution using VO(acac)2 are different from those of (n-Bu4N)4H5PV14O42 or [Et3NH]2[VO(DTBC)2]‚nCH3OH. The reason for the lack of hyperfine
coupling in the center-field-scan spectra is not known, but may result from interference from the additional ligands present in those two cases.
that our current experimental data show that all the vanadium-
based precatalysts exhibit the same, high catalytic lifetimes
within experimental error, a re-examination of the maximum
TTOs exhibited by VO(acac)2 seemed in order. Those experi-
ments reveal that VO(acac)2 exhibits the same 29 000 ((1000)
TTOs of DTBC oxygenation as do all the other V-precatalysts
within experimental error. Even after deliberately exposing the
somewhat O2-sensitive VO(acac)2 to atmospheric oxygen,
neither the double-integrated EPR area nor the catalytic lifetime
result of aged VO(acac)2 differ from those obtained from a fresh
sample (details of the aging procedure are in the Experimental
Section; lifetime results are provided in Table S3, Supporting
Information). Hence, despite the fact that the underlying cause
of the lower TTO performance of VO(acac)2 in previous studies
including ours is not clear,24 the present reproducible result
(three experiments) is that VO(acac)2 behaves as a prototypical
V-DTBC dioxygenase catalyst. In addition and although
VO(acac)2 oxidation is slow (by EPR), we suggest that this
precatalyst be treated as oxygen-sensitive.
31P NMR Control Confirming the Expected (n-Bu4N)9-
P2W15V3O62 Precatalyst’s Degradation under Catalytic Con-
ditions. Because a clear prediction of our results so far is that
polyoxometalate structures are at least partially degraded under
DTBC/O2 dioxygenase catalysis, an important control was to
confirm this prediction by a direct method, such as 31P NMR.
As the further information and Figure S10 of the Supporting
Information detail, a paramagnetic species is formed in the “V3
cap” of (n-Bu4N)9P2W15V3O62 upon the addition of just 3,5-
DTBC (i.e., no O2). Following the addition of O2 and catalysis,
the 31P NMR peak shifts and loss of intensity confirm that the
polyoxoanion structure has been at least somewhat degraded
as predicted (Figure S10).
g ) 2.036-2.038. A clear, resolved view at the center signal
(g ) 2.003-2.004) was obtained by scanning at a magnetic
field of 50 G instead of the whole field, 1000 G. The reaction
solution employing VO(acac)2 shows a 10 line signal (g )
2.006-2.007, A51V ) 2.1 G; Table 3) as reported in the
literature.38 We speculate that the EPR detected during the O2-
uptake reactions is a binary or ternary complex between the
catalyst, the substrate, and O2.55
The postreaction, center-field-scan results show that the same
EPR spectrum is observed for the three disparate vanadium
precatalysts shown in Table 3 following 3,5-DTBC oxygenation
in toluene. A nine line spectrum centered at g ) 2.004-2.006
((0.002) with an average hyperfine coupling value (A51V) of
3.04-3.08 ((0.1) G is observed in the EPR spectra of all three
postreaction solutions. Significantly, the literature38 reports a
nine line EPR spectrum for [VO(DBSQ)(DTBC)]2 centered at
g ) 2.004 with A51V ) 2.85 G (this coupling constant was
obtained by simulation38), a spectrum within experimental error
of the EPR we obserVed (Table 3 and Figure 2). This small
coupling constant signifies the weak coordination of the semi-
quinone ligand DBSQ to the single diamagnetic vanadium(V).
This EPR spectrum serves as a signature of [VO(DBSQ)-
(DTBC)]2 since most VIV compounds show an eight line
38
spectrum with hyperfine coupling, A51V ∼ 100 G. Given that
we know that [VO(DBSQ)(DTBC)]2 is as active as any complex
tested, reacts without an induction period (vide infra), displays
the maximum O2-uptake rate of -0.5 mmol O2/h (vide infra),
and also gives a high TTO value, it follows from this EPR and
other evidence that [VO(DBSQ)(DTBC)]2 is at least a resting
state of the true catalyst.
(54) The catalytic lifetime experiments at e50 000 TTOs (with 29 mmol 3,5-
DTBC) are ca. half the largest scale used in the previous work24 (with ca.
63 mmol 3,5-DTBC). All other conditions are the same, 0.5-0.7 µmol
precatalyst, ca. 125 mL of 1,2-C2H4Cl2, 65 °C, and 1 atm O2.
(55) The compound exhibiting the 10 line EPR spectrum is tentatively assigned
to V(DBSQ)3 in the literature38 via a parent ion at m/z ) 712. In our
experience, many different compounds can give this 712 peak (see Table
5 in the main text), so that the identity of the compound responsible for
the 10 line EPR spectrum remains unclear, a point confirmed by Prof.
Pierpont.
EPR Detection of A Common Active Species, [VO(DBSQ)-
(DTBC)]2. EPR of reaction solutions employing three
representative precursors, (n-Bu4N)4H5PV14O42, [Et3NH]2-
[VO(DTBC)2]‚2CH3OH, and VO(acac)2, all show an EPR signal
at center field (g ) 2.003-2.004) and a weak broad signal at
9
9010 J. AM. CHEM. SOC. VOL. 127, NO. 25, 2005

Vanadium-Based Catechol Dioxygenases
A R T I C L E S
conditions. The results, again, support [VO(DBSQ)(DTBC)]2
as a key species intimately connected to the catalytic cycle.
Negative Ion ESI-MS Evidence of a Common Catalyst,
[VO(DBSQ)(DTBC)]2. When crystalline [VO(DBSQ)(DTBC)]2
is dissolved in acetonitrile, the molecular ion peak at m/z )
1015 is observable, together with two main fragment peaks at
m/z ) 507, assigned to [VO(DTBC)2]-, and at m/z ) 711.4,
assigned to [V(DTBC)3]- (Figure 3). When MS-MS was
applied to the molecular ion peak, a peak attributable to
[VO(DTBC)2]- (m/z ) 507) was observed.
The m/z ) 507, 711, and 1015 peaks were simulated using
the Isotope Viewer software (version 1.0; in the Qual Browser
of the Thermo Finnigan MS). The corresponding formulas
reproducing the observed molecular weights and isotopic
abundances are VO(C14H20O2)2 corresponding to [VO(DTBC)2]-,
V(C14H20O2)3 corresponding to [V(DTBC)3]-, and [V2O2-
(C14H20O2)3(C14H21O2)] corresponding to [V2O2(DBSQ)(DBSQ-
H)(DTBC)2]- (where C14H20O2 corresponds to 3,5-di-tert-
butylcatecholate dianion or 3,5-di-tert-butylsemiquinone anion,
and C14H21O2 corresponds to protonated 3,5-di-tert-butylsemi-
quinone (3,5-DBSQ-H) radical). The observed MS spectra and
simulations are shown in Table 4; the match between the
observed and simulated spectra is excellent.
Figure 2. Center-field EPR spectrum of the postreaction solution of 3,5-
DTBC and VO(acac)2 in toluene (a). This resonance is identical within
experimental error to those of postreaction solutions using precatalysts (n-
Bu4N)4H5PV14O42 and [Et3NH]2[VO(DTBC)2]‚nCH3OH, as shown in Table
3, as well as the authentic [VO(DBSQ)(DTBC)]2 in toluene (b).38
As a semiquantitative method, ESI-MS is capable of identify-
ing the main species in solution.56-58 Significantly, two common
major peaks (VO(DTBC)2- of m/z 507, and V(DTBC)3- of m/z
711.5) are detected in the mass spectra for two of the three
vanadium catecholate compounds and the two isolated active
components from the vanadium polyoxometalates (Table 5).
The possibility of either hydrolysis or oxygenation during
the process of electrospray ionization needs to be considered.
Oxygenation was minimized by sealing the sample in the drybox
and exposing it to air only briefly just before injection. The
effect of water was tested with several controls detailed in the
Experimental Section; these experiments show the two common
Figure 3. Negative ion ESI-MS spectrum of crystalline [VO(DBSQ)-
(DTBC)]2. The parent ion peak is at m/z ) 1015; the additional peak at
m/z ) 1037 is due to an exchange of two H+ with Mg2+ of MgSO4 from
the distillation step. The base peak, at m/z ) 507, is due to [VO(DTBC)2]-.
Importantly, if additional substrate was added to the post-
reaction solution of 3,5-DTBC and VO(acac)2 under O2, the
oxygen-uptake continued and the EPR changed back to the EPR
observed during reaction. Additionally, the identical EPR
diagnostic of [VO(DBSQ)(DTBC)]2 was observed again once
all the substrate was consumed (Figure S15 of the Supporting
Information). These additional EPR experiments show that the
vanadium species we observe in the three postreaction solutions
(Table 3) with its characteristic nine line EPR spectrum is
intimately connected to the catalytic cycle of DTBC oxygen-
ation. Since that nine line spectrum is seen for authentic
[VO(DBSQ)(DTBC)]2, it follows that [VO(DBSQ)(DTBC)]2 is
an intimate part of the catalytic cycle.
-
peaks (VO(DTBC)2 and V(DTBC)3-) are not generated
through a reaction with water. Overall, the mass spectral data
further support the hypothesis that the active species in the
reaction solution is [VO(DBSQ)(DTBC)]2 since its diagnostic
-
fragment peak VO(DTBC)2 at m/z ) 507, as seen from
authentic material [VO(DBSQ)(DTBC)]2, is observed in the
postreaction solution from disparate vanadium precatalysts,
including active components isolated from two different V-
containing polyoxometalate precatalysts.
Kinetic Competence of [VO(DBSQ)(DTBC)]2 But Not of
[Na(CH3OH)2]2[V(DTBC)3]2‚4CH3OH. In oxygen-uptake ex-
periments (Figure 4), [VO(DBSQ)(DTBC)]2 was shown to be
an active oxygenation catalyst with no induction period.
Moreover, its rate is as fast as that of any catechol oxygenation
catalysis we have observed for our conditions herein,59 d[O2]/
dtmax ∼ -0.5 mmol/h. This establishes [VO(DBSQ)(DTBC)]2
as a kinetically competent catalyst or catalyst resting state.
Quantitation of the [VO(DBSQ)(DTBC)]2 EPR Signal.
Given the high sensitivity of EPR, it was important to determine
whether [VO(DBSQ)(DTBC)]2 is present more than at a trace
level. Hence, the amount of the vanadyl semiquinone dimer
complex in the postreaction solution was compared by double
integration to the air-stable free-radical in crystalline [CoIII(3,5-
(56) Ralph, S. F.; Iannitti, P.; Kanitz, R.; Sheil, M. M. Eur. Mass Spectrom.
1996, 2, 173-179. This key, early paper provides an example showing
that ESI-MS peak intensities correlate well with known compositions in
solution, at least for the particular system examined.
DBSQ)(CN)4]2- 37
, a sample generously provided by Professor
M. Wicholas. This comparison indicates that the [VO(DBSQ)-
(DTBC)]2 semiquinone complex is present at ca. 5 mol % of
the precatalyst (VO(acac)2) concentration (ca. 0.035 mM), an
estimate that is necessarily an approximation since [VO(DBSQ)-
(DTBC)]2 is oxygen-sensitive and slowly degrades under aerobic
(57) Raymond, C. C.; Dick, D. L.; Dorhout, P. K. Inorg. Chem. 1997, 36, 2678-
2681.
(58) (a) Deery, M. J.; Fernandez, T.; Howarth, O. W.; Jennings, K. R. J. Chem.
Soc., Dalton Trans. 1998, 2177-2183. (b) Fyles, T. M.; Zeng, B. Supramol.
Chem. 1998, 10, 143-153. (c) Pasilis, S. P.; Pemberton, J. E. Inorg. Chem.
2003, 42, 6793-6800.
9
J. AM. CHEM. SOC. VOL. 127, NO. 25, 2005 9011

A R T I C L E S
Yin and Finke
Table 4. Observed versus Computer-Simulated Parent Peaks and Isotopic Abundance Patterns for [VO(DTBC)2]-, [V(DTBC)3]-, and
[V2O2(DBSQ)(DBSQ-H)(DTBC)2]-
a Observed peaks: 507.3, 508.3, and 509.3; simulated peaks: 507.2, 508.2, and 509.2. b Observed peaks: 711.4, 712.4, and 713.3; simulated peaks:
711.4, 712.4, and 713.4. c Observed peaks: 1015.1, 1016.1, 1017.1, and 1018.3; simulated peaks: 1015.5, 1016.5, 1017.5, and 1018.3.
Table 5. Negative Ion Electrospray Ionization Mass Spectrometry Identified Peaks of Various Precursors
Relative Abundance
ESI-MS peaks, m/z
451.4
491.4
507.3
711.4
998.4
1015.1
simple V-catecholate complexes
[VO(DBSQ)(DTBC)]2
[Et3NH]2[VO(DTBC)2]‚2CH3OH
[Na(CH3OH)2]2[V(DTBC)3]2‚4CH3OH
<5
<5
100
25
100
100
<5
30
<5
<10
86
<10
<5a
Isolated Active Components
isolated active species from
(n-Bu4N)5[FeII‚SiW9V3O40]
isolated active species from
(n-Bu4N)4H5PV14O42
93
25
100
17
<5
30
100
a [Na(CH3OH)2]2[V(DTBC)3]2‚4CH3OH only shows a trace amount of [VO(DTBC)2]- in its negative ion ESI-MS, as expected since it lacks an oxo
ligand.
-
The remaining question is whether the other common MS
peak (m/z ) 711.5, V(DTBC)3-) is one of the catalytic species.
In kinetic trials, an induction period of about 10 min is observed
when [Na(CH3OH)2]2[V(DTBC)3]2‚4CH3OH is used as the
precatalyst. Therefore, V(DTBC)3 can be ruled out as a
dominant catalyst.
In summary, [VO(DBSQ)(DTBC)]2 is a kinetically competent
catalyst or catalyst resting state. It is proposed as an active
9
9012 J. AM. CHEM. SOC. VOL. 127, NO. 25, 2005

Vanadium-Based Catechol Dioxygenases
A R T I C L E S
challenge. Put in different terms, facile ligand substitution and
32
leaching, especially in the presence of H2O2 and powerfully
chelating ligands, such as DTBC dianion, is hereby established
as the dominant hypothesis to be disproved for future V-based
oxygenation catalysis.
In a subsequent paper, we provide further evidence that
[VO(DBSQ)(DTBC)]2 is a catalyst resting state that reversibly
fragments to its reactive monomer, [VO(DBSQ)(DTBC)], as
well as evidence for a novel autoxidation-product-initiated
mechanism producing H2O2 for the release of V from V-based
precatalysts to form [VO(DBSQ)(DTBC)]2, that is, for a novel
autoxidation-initiated dioxygenase.32
Acknowledgment. We thank Professor Cort G. Pierpont for
multiple insightful discussions relevant to this work, and
Professor Mark Wicholas for his generous gift of air-stable
cobalt semiquione crystals employed in the quantitative EPR
analysis. We thank Mr. Don Dick for his expert assistance with
the GC-MS and negative ion ESI-MS experiments, Mr. Don
Heyse for his expert assistance with the EPR experiments, and
Dr. Susie Miller for her expert assistance with the single-crystal
X-ray crystallography. This research is supported by NSF fund
9531110.
Figure 4. Kinetic competence or incompetence of the three vanadium
catecholate complexes. Experimental conditions: 400 mg (1.8 mmol) of
DTBC, 0.46-0.48 mM precatalyst/catalyst, ca. 9 mL of 1,2-C2H4Cl2, 40
((0.7) °C, and 1 atm O2.
catalyst component, in (at least) the 11 tabulated vanadium
catechol dioxygenases, based on the observation of its signature
EPR spectrum in the postreaction solution of three different
classes of precatalysts and the observation of its diagnostic half-
fragment (VO(DTBC)2-) in the negative ion ESI-MS spectra
of postreaction solutions beginning with disparate precatalysts,
including vanadium catecholate complexes and polyoxometa-
lates. The EPR and negative ion ESI-MS signatures reported
herein can now be used to expand or refute the common catalyst
finding herein with any other vanadium-based DTBC dioxy-
genase system.
Supporting Information Available: Preparation of poly-
oxometalate precursors; preparation and characterization of
[VO(DBSQ)(DTBC)]2, [Et3NH]2[VO(DTBC)2]‚nCH3OH, and
[Na(CH3OH)2]2[V(DTBC)3]2‚4CH3OH; preparation and char-
acterization of deprotonated 3,5-di-tert-butylcatecholate salt,
Na2(3,5-DTBC), and [Et3NH][V(DTBC)3]‚1.5CH3OH; 31P NMR
following oxygen-uptake using (n-Bu4N)9P2W15V3O62; catalyst
isolation attempts; spectroscopic information (UV-visible,
Conclusions
In conclusion, we have (i) shown that a survey of 11 literature
reports of vanadium-based catechol dioxygenases reveal that
they all give very similar product ratios, an insight which, in
turn, implies that a common catalyst is present; (ii) shown that
10 vanadium compounds that we have tested to date behave
remarkably similar in their DTBC oxygenation catalysis selec-
tivities and TTOs, again implying the existence of a common
active catalyst; and (iii) provided EPR, negative ion ESI-MS,
and kinetic data which all point to a common resting form of
the catalyst, namely, Pierpont’s novel [VO(DBSQ)(DTBC)]2
complex.
This study also provides an excellent example of the
phenomenon of “leaching” of a soluble form of the true catalyst
from both heterogeneous60 and homogeneous precatalysts.
Combined with previous work,60 our results indicate, for
example, that the task of preparing single-site, supported Vx
catalysts61,62 (e.g., x ) 1,2), which stay firmly supported and
do not leach, is an important but probably very difficult
1
negative ion ESI-MS, or H NMR) of [VO(DBSQ)(DTBC)]2,
[Et3NH]2[VO(DTBC)2]‚nCH3OH, and [Na(CH3OH)2]2[V-
(DTBC)3]2‚4CH3OH; EPR spectrum of Zn(Cat-N-SQ)(BQ-N-
SQ); crystal data for [Na(CH3OH)2]2[V(DTBC)3]2‚4CH3OH; IR
spectrum of Na2(3,5-DTBC) and 3,5-DTBC; 31P NMR following
the O2-uptake of (n-Bu4N)9P2W15V3O62 with DTBC; IR spec-
trum of (n-Bu4N)5FeII‚SiW9V3O40 and the isolated active species
from (n-Bu4N)5FeII‚SiW9V3O40; negative ion ESI-MS spectrum
of isolated catalyst from (n-Bu4N)5FeII‚SiW9V3O40; solid EPR
of the material isolated from (n-Bu4N)4H5PV14O42; center field,
nine line EPR spectrum of a postreaction solution of 3,5-DTBC
and VO(acac)2 in toluene, which had additional fresh DTBC
and O2 added; tabulated literature overview of vanadium
catecholate organometallic complexes relevant to dioxygenase
catalysis; details for e50 000 TTO catalytic lifetime experiments
(data for Figure 1 in the main text) and 100 000-150 000 TTO
catalytic lifetime experiments. This material is available free
of charge via the Internet at http://pubs.acs.org.
(59) The maximum rate, measured from the post-induction period (or initial
rate in the case of [VO(DBSQ)(DTBC)]2), is approaching the O2 gas-to-
solution mass transfer limit under our experimental conditions of 400 mg
(1.8 mmol) of DTBC, 0.46-0.48 mM precatalyst/catalyst, ca. 9 mL of
1,2-C2H4Cl2, 40 ((0.7) °C, 1 atm O2, and stirring rate of ca. 700-800
rpm.
JA051594E
(61) Rulkens, R.; Tilley, D. T. J. Am. Chem. Soc. 1998, 120, 9959-9960.
(62) Khodakov, A.; Olthof, B.; Bell, A. T.; Iglesia, E. J. Catal. 1999, 181, 205-
(60) Arends, I. W. C. E.; Sheldon, R. A. Appl. Catal., A 2001, 212, 175-187.
216.
9
J. AM. CHEM. SOC. VOL. 127, NO. 25, 2005 9013