O-atom exchange between H2O and CO2 mediated by a bis(dithiolene)tungsten complex

Article abstract of DOI:10.1021/ic300906j

Inspired by the CO₂-reductatse activity of tungsten-dependent formate dehydrogenases (W-FDHs), a reduced W-FDH model, [W^IV(OH) (S₂C₂Ph₂)₂]^-, was prepared in situ through hydrolysis of [W^IV(OPh)(S₂C ₂Ph₂)₂]^- (1) and its reactivity with CO₂ was investigated. The reaction between [W^IV(OH) (S₂C₂Ph₂)₂]^- and CO ₂ at room temperature leads to the formation of [W ^IV(O)(S₂C₂Ph₂)₂] ^2- (2), which slowly oxidizes to [W^V(O)(S ₂C₂Ph₂)₂]^- (3). Isotopic labeling experiments reveal that the O atom in CO₂ incorporates into 3. This implies that there is carbonic anhydrase like activity, in which carbonation and decarboxylation are mediated by a bis(dithiolene)tungsten complex.

Full text of DOI:10.1021/ic300906j

Communication
pubs.acs.org/IC
O‑Atom Exchange between H2O and CO2 Mediated by a
Bis(dithiolene)tungsten Complex
Junhyeok Seo and Eunsuk Kim*
Department of Chemistry, Brown University, Providence, Rhode Island 02912, United States
S
* Supporting Information
Mo-FDHH, in which MoIV was in a square-pyramidal geometry
with two equatorial pyranopterins and a SeCys axial ligand.9a
However, the same crystallographic data have been reevaluated
more recently,9d revealing that the SeCys in the reduced state is
no longer coordinated to the metal center. Instead, the −SH
(or −OH) has been proposed for the axial ligand9d (Figure 1).
ABSTRACT: Inspired by the CO2-reductatse activity of
tungsten-dependent formate dehydrogenases (W-FDHs),
a reduced W-FDH model, [WIV(OH)(S2C2Ph2)2]−, was
prepared in situ through hydrolysis of [WIV(OPh)-
(S2C2Ph2)2]− (1) and its reactivity with CO2 was
investigated. The reaction between [WIV(OH)-
(S2C2Ph2)2]− and CO2 at room temperature leads to the
formation of [WIV(O)(S2C2Ph2)2]2− (2), which slowly
oxidizes to [WV(O)(S2C2Ph2)2]− (3). Isotopic labeling
experiments reveal that the O atom in CO2 incorporates
into 3. This implies that there is carbonic anhydrase like
activity, in which carbonation and decarboxylation are
mediated by a bis(dithiolene)tungsten complex.
Figure 1. Active-site structures in the reduced Mo-FDH (E. coli).9d
In light of the reexamined FDH structure9d and the efficient
electrocatalytic activity of W-FDH,8 here we report our studies
on the CO2 reactivity of a synthetic model complex of W-FDH.
Synthetic models for FDHs are limited,10 although a
significant number of bis(dithiolene)molybdenum and -tung-
sten complexes are known.11 Our own attempts to directly
synthesize discrete HO- or HS-bound bis(dithiolene)tungsten
analogues have not been successful. However, Holm and co-
workers have previously reported12 that a group of phenolate-
bound tungsten or molybdenum bis(dithiolene) complexes
undergo hydrolysis in the presence of excess H2O to yield
WIVO complexes via the formation of {WIVOH} species.
These reports inspired us to explore the CO2 reactivity of
[WIV(OH)(S2C2Ph2)2]−, which can be prepared in situ through
hydrolysis of [WIV(OPh)(S2C2Ph2)2]− (1) in the presence of
H2O.
We first sought a hydrolysis condition that can be achieved
with the minimum amount of H2O possible in order to favor
the binding of CO2 over H2O to [WIV(OH)(S2C2Ph2)2]−. We
found that a relatively small amount (10 equiv) of H2O was
sufficient enough to induce hydrolysis of 112b in acetonitrile to
yield [WIV(O)(S2C2Ph2)2]2− (2). However, we noticed that 2
was not stable over 12 h in our reaction conditions, and it
slowly converted into another well-known13 compound,
[WV(O)(S2C2Ph2)2]− (3; Scheme 1), which is likely associated
with the reduction of H+ to hydrogen (vide infra).
arbon dioxide (CO2) has been implicated as one of the
1
Cmain causes for global warming, thus, the transformation
of CO2 into environmentally friendly and valuable chemicals is
of great interest. Among the various chemical transformations,
the reduction of CO2 is particularly significant because the
reduced forms of CO2 such as formic acid, methanol, or
methane can be directly used as fuels, which establishes CO2 as
an abundant and inexpensive source for alternative energy.2,3
Conversion of CO2 to formic acid can be achieved (electro)-
chemically. The development of efficient catalysts for such a
transformation comprises an active area of research.2−5 There
have been continuing efforts in developing coordination
complexes that can catalyze the reduction of CO2 with H2,4
insertion of CO2 into an M−H bond,2,5 or electrochemical
reduction of CO2.3 Although these complexes show much
promise, none has shown the comparable efficiency and
selectivity of the CO2-reducing enzymes found in nature.
Formate dehydrogenase (FDHs) are enzymes that catalyze
−
the two-electron oxidation of formate (HCO2 ) to CO2
coupled with the reduction of NAD(P)+ to NAD(P)H.6 In
some prokaryotes, the FDHs contain molybdenum (Mo) or
tungsten (W) cofactors, by which an efficient CO2-reductase
activity (reduction of CO2 to formate) has been also observed.7
Recently, Hirst and co-workers have shown that a purified W-
FDH can activate an electrode to reduce CO2 to formate more
efficiently than any of the known synthetic catalysts.8
IR spectroscopy provides us with rich information that
supports the generation of 3 as the final hydrolysis product as
well as the released free phenol (Figure 2). The IR spectrum of
The X-ray structures of several Mo- or W-containing FDHs
are known.9 In the oxidized state, the active site contains a
MoVI or WVI ion in a trigonal-prism geometry with two
pyranopterin molecules, one selenocysteine (SeCys), and a
sixth −SH (or OH) ligand.9a−c The coordination geometry of
the reduced FDHs was first determined from Escherichia coli
Received: May 3, 2012
Published: July 26, 2012
© 2012 American Chemical Society
7951
dx.doi.org/10.1021/ic300906j | Inorg. Chem. 2012, 51, 7951−7953

Inorganic Chemistry
Communication
Because the formation of [WIV(OH)(S2C2Ph2)2]− was
implied as an intermediate during our hydrolysis reaction, we
studied the desired CO2 reactivity of 1 in the presence of H2O.
An acetonitrile solution of 1 containing 10 equiv of H2O was
exposed to 1 atmospheric pressure of CO2 at room temperature
for 12 h.16 To our surprise, 3 was again identified as the final
reaction product from the 1/H2O/CO2 reaction, with no
indication of generating either the reduced CO2 products (e.g.,
Scheme 1
17
−
formate, CO, oxalate, etc.) or a hydrated product (HCO3 ).
However, when we carry out the same reaction using 10 equiv
of H218O, we observe that only ∼32% of 3 contains the
18
W O moiety (889 and 869 cm−1; Figure 3), suggesting that
Figure 2. IR spectra (Nujol) of (a) the crude reaction products of 1
and 10 equiv of H2O (black dotted line) and those of 1 and 10 equiv
of H218O (blue solid line), (b) the isolated reaction product 3 (black
dotted line), 3 in the presence of 2 equiv of externally added phenol
(green solid line), and reisolated 3 after removal of phenol by
precipitation (gray solid line).
Figure 3. IR spectra (Nujol) of the crude reaction products of 1/
H218O (10 equiv)/CO2 (red solid line) and those of 1/H218O (10
equiv) (blue dotted line).
H2O is no longer the major source of oxygen in 3. Because CO2
is the only other possible O-atom donor in the reaction
mixture, the reaction was further evaluated by using C18O2.
When the reaction was carried out using 18O-enriched CO2
the crude reaction products of 1/H2O exhibits two strong
WVO frequencies at 937 and 915 cm−1, both of which shift to
889 and 869 cm−1 when the reaction is carried out using H218O
(Figure 2a). No such isotopic shifts were observed when
independently synthesized 2 or 3 was treated with 10 equiv of
H218O.14 This indicates that the O atoms in the reaction
products originate from H2O through hydrolysis of 1, during
which the intermediate [WIV(OH)(S2C2Ph2)2]− must have
been generated. The 937 cm−1 peak from the crude reaction
products is assigned to the stretching frequency of WVO in 3,
consistent with the literature value.13 The presence of the
additional peak at 915 cm−1 is presumably due to the WVO
moiety hydrogen-bonded to free phenol that is released from 1.
Mass spectrometry further confirms the presence of free phenol
in the reaction mixture. When 3 was isolated by precipitation
with diethyl ether, only the 937 cm−1 band was observed in the
IR spectrum as expected. However, when the isolated 3 is
redissolved in acetonitrile, followed by the addition of phenol,
the 915 cm−1 peak reappears (Figure 2b).15 The appearance of
the 915 cm−1 band is not observed when phenolate (PhO−) is
added to a solution of 3, which supports the assignment of the
915 cm−1 peak for the hydrogen-bonded WVO moiety. This
cycle can be repeated multiple times without causing
decomposition of 3.
18
(85% C18O2) and H2O, ∼50% of W O complexes were
observed in the reaction products, confirming that the O atom
in CO2 incorporates into 3.16
Complex 1 does not react with CO2 in dry solvent under the
same experimental conditions, suggesting that the formation of
the {WIVOH} moiety is crucial to initiating reactivity with CO2.
This behavior along with the absence of any carbon-containing
reaction products derived from CO2 lead us to conclude that
the reaction of 1/CO2/H2O does not exhibit the anticipated
FDH activity. Instead, [WIV(OH)(S2C2Ph2)2]− reacts with CO2
to form a presumable bicarbonato or carbonato intermediate,
which immediately undergoes decarboxylation to generate
[WIV(O)(S2C2Ph2)2]− (2) and CO2 (Scheme 2). The
formation of (bi)carbonato species from CO2 with a metal
hydroxide is common for many transition-metal complexes,18
and examples of tungsten(IV) carbonato complexes exist.19 In
our reaction, however, the (bi)carbonato species does not
appear to be stable probably because of the oxophilic character
Scheme 2
To gain insight into the unexpected oxidation of 2 to 3
during hydrolysis, we studied the reactivity of independently
synthesized 2 with a proton source. When 1 equiv of HBF4 was
added to an acetonitrile solution of 2 at room temperature, the
absorption bands at 350, 449, and 527 nm from 2 decayed with
a concomitant formation of 3 (λmax = 720 nm) in the UV−vis
spectrum.16 In addition, the H2 evolution from the reaction was
qualitatively detected from the headspace gas of the reaction
flask with a commercial H2 dosimeter.16
7952
dx.doi.org/10.1021/ic300906j | Inorg. Chem. 2012, 51, 7951−7953

Inorganic Chemistry
Communication
Notes
of tungsten(IV) in a bis(dithiolene) ligand environment. In
fact, with the only exception of FDHs, all other known
bis(pyranopterin)-bound tungsten/molybdenum enzymes cata-
lyze the O-atom-abstraction chemistry.6 In order to evaluate the
prospect of the proposed (bi)carbonato intermediate during
the reaction (Scheme 2), 1 was reacted with 1 equiv of
Et4NHCO3. Upon the addition of bicarbonate to 1, the oxo
compound 2 formed immediately, as monitored by UV−vis
spectroscopy.16 The result supports a notion that the O atom of
CO2 is inserted into 2 through a (bi)carbonate-bound tungsten
intermediate.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We are grateful for support provided by the ACS-PRF (50840-
DNI3) and Brown University.
■
REFERENCES
■
(1) Crowley, T. J. Science 2000, 289, 270−277.
(2) Sakakura, T.; Choi, J. C.; Yasuda, H. Chem. Rev. 2007, 107,
2365−2387.
The FDHs utilize a bis(dithiolene)-bound MIV (M = Mo or
W) ion to reduce CO2 to formate. The use of higher-valent
metal ions with noninnocent ligands by the enzyme is a strategy
very different from that of most synthetic systems in which the
metal ions in much lower valences, 0, 1+, and 2+, are utilized to
activate CO2.20 Our current study shows that the presence of a
nucleophilic ligand such as OH− is crucial to initiating CO2
reactivity with a bis(dithiolene)tungsten(IV) species to likely
form a tungsten (bi)carbonate intermediate. Indeed, a recent
calculation study21 on the mechanism of Mo-FDH suggests that
a Mo(IV) thiocarbonato intermediate forms during the
oxidation cycle of formate to CO2. With respect to the reverse
reaction by W-FDH, one can expect the conversion of W(IV)
thiocarbonate species to W(VI) formate (Scheme 3), similar to
(3) (a) Benson, E. E.; Kubiak, C. P.; Sathrum, A. J.; Smieja, J. M.
Chem. Soc. Rev. 2009, 38, 89−99. (b) Rakowski DuBois, M.; DuBois,
D. L. Acc. Chem. Res. 2009, 42, 1974−1982.
(4) Jessop, P. G.; Joo, F. Coord. Chem. Rev. 2004, 248, 2425−2442.
(5) Darensbourg, D. J. Inorg. Chem. 2010, 49, 10765−10780.
(6) (a) Hille, R. Chem. Rev. 1996, 96, 2757−2816. (b) Moura, J. J.;
Brondino, C. D.; Trincao, J.; Romao, M. J. J. Biol. Inorg. Chem. 2004, 9,
791−799. (c) Andreesen, J. R.; Makdessi, K. Ann. N.Y. Acad. Sci. 2008,
1125, 215−229.
(7) (a) de Bok, F. A.; Hagedoorn, P. L.; Silva, P. J.; Hagen, W. R.;
Schiltz, E.; Fritsche, K.; Stams, A. J. Eur. J. Biochem. 2003, 270, 2476−
2485. (b) Axley, M. J.; Grahame, D. A. J. Biol. Chem. 1991, 266,
13731−13736. (c) Yamamoto, I.; Saiki, T.; Liu, S. M.; Ljungdahl, L. G.
J. Biol. Chem. 1983, 258, 1826−1832.
(8) Reda, T.; Plugge, C. M.; Abram, N. J.; Hirst, J. Proc. Natl. Acad.
Sci. U.S.A. 2008, 105, 10654−10658.
(9) (a) Boyington, J. C.; Gladyshev, V. N.; Khangulov, S. V.;
Stadtman, T. C.; Sun, P. D. Science 1997, 275, 1305−1308.
(b) Jormakka, M.; Tornroth, S.; Byrne, B.; Iwata, S. Science 2002,
295, 1863−1868. (c) Raaijmakers, H.; Macieira, S.; Dias, J. M.;
Teixeira, S.; Bursakov, S.; Huber, R.; Moura, J. J.; Moura, I.; Romao,
M. J. Structure 2002, 10, 1261−1272. (d) Raaijmakers, H. C.; Romao,
M. J. J. Biol. Inorg. Chem. 2006, 11, 849−854.
Scheme 3
(10) (a) Groysman, S.; Holm, R. H. Inorg. Chem. 2007, 46, 4090−
4102. (b) Sarkar, S.; Das, S. K. Proc.Indian Acad. Sci., Chem. Sci.
1992, 104, 533−534.
(11) (a) Enemark, J. H.; Cooney, J. J.; Wang, J. J.; Holm, R. H. Chem.
Rev. 2004, 104, 1175−1200. (b) Groysman, S.; Holm, R. H.
Biochemistry 2009, 48, 2310−2320.
the typical O-atom-abstraction chemistry known for this class
of enzyme.6 In our current synthetic model system, however,
we were not able to imitate the reduction step. This may be due
to the lack of a proton delivery channel near the metal active
site, i.e., selenocysteine and histidine, that is strictly conserved
in the FDH proteins.
In summary, we have studied the CO2 reactivity of a
structural analogue of W-FDH. A bis(dithiolene)tungsten
complex 1 itself does not react with CO2 at ambient
temperature. However, an in situ generated hydrolysis product,
[WIV(OH)(S2C2Ph2)2]−, reacts with CO2 and displays a
carbonic anhydrase like activity. A tungsten (bi)carbonate
species is presumed to form during the reaction. Future studies
will focus on the factors critical to inducing an O-atom
abstraction from a putative tungsten (bi)carbonate intermediate
over the oxide (O2−) abstraction chemistry observed in this
study.
(12) (a) Lim, B. S.; Holm, R. H. J. Am. Chem. Soc. 2001, 123, 1920−
1930. (b) Sung, K. M.; Holm, R. H. J. Am. Chem. Soc. 2001, 123,
1931−1943.
(13) Goddard, C. A.; Holm, R. H. Inorg. Chem. 1999, 38, 5389−
5398.
(14) Syntheses of 2 and 3 are known.13
(15) The complete conversion of the 937 to 915 cm−1 peak requires
at least 3 equiv of phenol.16 Upon substitution of phenol with p-
nitrophenol, the 915 cm−1 peak further shifts to 910 cm−1.16
(16) See the Supporting Information.
(17) The 13C NMR and IR spectra of the reaction mixtures of 1/
H2O/13CO2 did not display any signals for the CO2-redrived products.
The chemical detections for CO using FeII(TPP) and
Cp*RuCl(PCy3) were also negative.
(18) Palmer, D. A.; Vaneldik, R. Chem. Rev. 1983, 83, 651−731.
(19) (a) Ito, T.; Sugimoto, S.; Ohki, T.; Nakano, T.; Osakada, K. J.
Organomet. Chem. 1992, 428, 69−83. (b) Alvarez, R.; Carmona, E.;
Galindo, A.; Gutierrez, E.; Marin, J. M.; Monge, A.; Poveda, M. L.;
Ruiz, C.; Savariault, J. M. Organometallics 1989, 8, 2430−2439.
(c) Green, M. L. H.; Parkin, G.; Ohare, D.; Wong, L. L.; Derome, A. E.
J. Organomet. Chem. 1986, 317, 61−68.
(20) Gibson, D. H. Chem. Rev. 1996, 96, 2063−2095.
(21) Mota, C. S.; Rivas, M. G.; Brondino, C. D.; Moura, I.; Moura, J.
J.; Gonzalez, P. J.; Cerqueira, N. M. J. Biol. Inorg. Chem. 2011, 16,
1255−1268.
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental details and UV−vis and IR spectra. This material
is available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: Eunsuk_Kim@brown.edu.
■
7953
dx.doi.org/10.1021/ic300906j | Inorg. Chem. 2012, 51, 7951−7953