Formation and reactivity of chromium(V)-thiolato complexes: A model for the intracellular reactions of carcinogenic chromium(VI) with biological thiols

Article abstract of DOI:10.1021/ja101675w

The nature of the long-lived EPR-active Cr(V) species observed in cells and biological fluids exposed to carcinogenic Cr(VI) has been definitively assigned from detailed kinetic and spectroscopic analyses of a model reaction of Cr(VI) with p-bromobenzenethiol (RSH) in the presence or absence of cyclic 1,2-diols (LH₂) in aprotic or mixed solvents. The first definitive structures for Cr(V) complexes with a monodentate thiolato ligand, [Cr^VO(SR) ₄]^- (g_iso = 1.9960, A_iso = 14.7 × 10^-4 cm^-1), [Cr^VOL(SR) ₂]^- (g_iso = 1.9854, A_iso = (15.8-16.2) × 10^-4 cm^-1) and [Cr^V(O) ₂(SR)₂]^- (g_iso = 1.9828, A _iso = 6.8 × 10^-4 cm^-1) were assigned by EPR spectroscopy and electrospray mass spectrometry. The unusually low A _iso (⁵³Cr) value for the latter species is consistent with its rare four-coordinate, bis-oxido structure. The [Cr^VOL(SR) ₂]^- species are responsible for the transient g _iso ≈ 1.986 EPR signals observed in living cells and animals treated with Cr(VI) (where RSH and LH₂ are biological thiols and 1,2-diols, respectively). For the first time, concentrations of Cr(V) intermediates formed during the reduction of Cr(VI) were determined by quantitative EPR spectroscopy, and a detailed reaction mechanism was proposed on the basis of stochastic simulations of the kinetic curves for Cr(V) species. A key feature of the proposed mechanism is the regeneration of Cr(V) species in the presence of Cr(VI) through the formation of organic free radicals, followed by the rapid reactions of the formed radicals with Cr(VI). The concentration of Cr(V) grows rapidly at the beginning of the reaction, reaches a steady-state level, and then drops sharply once Cr(VI) is spent. Similar mechanisms are likely to operate during the reduction of Cr(VI) in biological environment rich in reactive C-H bonds, including the oxidative DNA damage by Cr(V) intermediates.

Full text of DOI:10.1021/ja101675w

Published on Web 06/07/2010
Formation and Reactivity of Chromium(V)-Thiolato
Complexes: A Model for the Intracellular Reactions of
Carcinogenic Chromium(VI) with Biological Thiols
Aviva Levina, Lianbo Zhang, and Peter A. Lay*
School of Chemistry, The UniVersity of Sydney, Sydney, New South Wales 2006, Australia
Received February 26, 2010; E-mail: peter.lay@sydney.edu.au
Abstract: The nature of the long-lived EPR-active Cr(V) species observed in cells and biological fluids
exposed to carcinogenic Cr(VI) has been definitively assigned from detailed kinetic and spectroscopic
analyses of a model reaction of Cr(VI) with p-bromobenzenethiol (RSH) in the presence or absence of
cyclic 1,2-diols (LH2) in aprotic or mixed solvents. The first definitive structures for Cr(V) complexes with a
monodentate thiolato ligand, [CrVO(SR)4]- (giso ) 1.9960, Aiso ) 14.7 × 10-4 cm-1), [CrVOL(SR)2]- (giso
)
1.9854, Aiso ) (15.8-16.2) × 10-4 cm-1) and [CrV(O)2(SR)2]- (giso ) 1.9828, Aiso ) 6.8 × 10-4 cm-1) were
assigned by EPR spectroscopy and electrospray mass spectrometry. The unusually low Aiso (53Cr) value
for the latter species is consistent with its rare four-coordinate, bis-oxido structure. The [CrVOL(SR)2]- species
are responsible for the transient giso ≈ 1.986 EPR signals observed in living cells and animals treated with
Cr(VI) (where RSH and LH2 are biological thiols and 1,2-diols, respectively). For the first time, concentrations
of Cr(V) intermediates formed during the reduction of Cr(VI) were determined by quantitative EPR
spectroscopy, and a detailed reaction mechanism was proposed on the basis of stochastic simulations of
the kinetic curves for Cr(V) species. A key feature of the proposed mechanism is the regeneration of Cr(V)
species in the presence of Cr(VI) through the formation of organic free radicals, followed by the rapid
reactions of the formed radicals with Cr(VI). The concentration of Cr(V) grows rapidly at the beginning of
the reaction, reaches a steady-state level, and then drops sharply once Cr(VI) is spent. Similar mechanisms
are likely to operate during the reduction of Cr(VI) in biological environment rich in reactive C-H bonds,
including the oxidative DNA damage by Cr(V) intermediates.
Introduction
diols (carbohydrates and their derivatives), leading to long-lived
Cr(V)-diolato complexes (giso ≈ 1.979),5-8 which are thought
Biological thiols, including glutathione (GSH, γ-Glu-Cys-
Gly) and metallothioneins (small Cys-rich proteins) play a
variety of physiological roles, including detoxification of heavy
metal ions1 such as Cr(VI), which is a recognized human
carcinogen2 and a major occupational and environmental
hazard.3 A crucial piece of evidence for the participation of thiols
in Cr(VI) metabolism in cultured mammalian cells4 and in the
blood of Cr(VI)-treated animals5 comes from the observation
of a transient EPR signal at giso ≈ 1.986, which has been
assigned to Cr(V)-thiolato complexes based on the model
studies5,6 and on a decrease in the intensity of this signal in
GSH-depleted cells.4b The transient Cr(V)-thiolato complexes
are able to exchange ligands with ubiquitous biological 1,2-
to be the key reactive species in Cr(VI)-induced genotoxicity
and carcinogenicity.6-9 Recent advances in the understanding
of Cr(VI) reactions with biological thiols include structural
characterization (by X-ray absorption spectroscopy) and detailed
reactivitystudiesofCr(V)-GSH10 andCr(VI)-GSH11 complexes.
Due to the d1 electronic structure of Cr(V) ions, EPR
spectroscopy has long been used as a selective analytical
technique for studies of Cr(V) intermediates formed during the
reduction of Cr(VI) in biological and model systems,7-9,12 but
the difficulty in obtaining quantitative data from EPR spectra8,12,13
complicates detailed kinetic and mechanistic studies on such
systems. A method for the determination of Cr(V) concentration
from EPR spectra, based on external calibration curves
obtained for stable Cr(V) complexes (such as Na-
[CrVO(ehba)2], where ehba ) 2-ethyl-2-hydroxybutanoato)14 has
(1) (a) Stefanidou, M.; Maravelias, C. Curr. Top. Toxicol. 2004, 1, 161–
167. (b) Babu, G. N.; Ranjani, R.; Fareeda, G.; Murthy, S. D. S. J.
Phytol. Res. 2007, 20, 1–6.
(2) International Agency for Research on Cancer (IARC). OVerall
EValuations of Carcinogenicity to Humans. IARC, 2003. http://
www.iarc.fr.
(7) (a) Signorella, S.; Palopoli, C.; Santoro, M.; Garcia, V.; Daier, V.;
Gonzalez, J. C.; Roldan, V.; Frascaroli, M. I.; Rizzotto, M.; Sala, L. F.
Trends Inorg. Chem. 2001, 7, 197–207. (b) Codd, R.; Irwin, J. A.;
Lay, P. A. Curr. Opin. Chem. Biol. 2003, 7, 213–219.
(3) Guertin, J., Jacobs, J. A., Avakian, C. P., Eds. Chromium(VI)
Handbook; CRC Press: Boca Raton, 2005.
(8) Levina, A.; Codd, R.; Lay, P. A. In Biological Magnetic Resonance;
Hanson, G. R., Berliner, L. J., Eds.; Springer Publishers: New York,
2009; Vol. 28, Part 4, pp 551-579.
(4) (a) Levina, A.; Harris, H. H.; Lay, P. A. J. Am. Chem. Soc. 2007,
129, 1065–1075. (b) Borthiry, G. B.; Antholine, W. E.; Myers, J. M.;
Myers, C. R. J. Inorg. Biochem. 2008, 102, 1449–1462.
(5) Sakurai, H.; Takechi, K.; Tsuboi, H.; Yasui, H. J. Inorg. Biochem.
1999, 76, 71–80.
(9) (a) Sugden, K. D.; Stearns, D. M. J. EnViron. Pathol. Toxicol. Oncol.
2000, 19, 215–230. (b) Levina, A.; Codd, R.; Dillon, C. T.; Lay, P. A.
Prog. Inorg. Chem. 2003, 51, 145–250. (c) Codd, R.; Dillon, C. T.;
Levina, A.; Lay, P. A. Coord. Chem. ReV. 2001, 216-217, 537–582.
(6) Levina, A.; Lay, P. A. Coord. Chem. ReV. 2005, 249, 281–298.
9
8720 J. AM. CHEM. SOC. 2010, 132, 8720–8731
10.1021/ja101675w  2010 American Chemical Society

Formation and Reactivity of Cr(V)-Thiolato Complexes
A R T I C L E S
been proposed.15 However, due to the sensitivity of EPR spectra
to the position of the sample in the cavity of an EPR
spectrometer, simultaneous measurements on the analyzed
sample and the standard within the cavity are required for
accurate calibration.13 In this work, a simple way of obtaining
quantitative results from EPR spectroscopy (placing two capil-
laries containing a reaction mixture and a standard Cr(V)
solution into the same EPR tube) has been used for the first
detailed kinetic study of the Cr(VI) reaction with a simple model
thiol, p-bromobenzenethiol (RSH), in a variety of solvents,
including mixed aqueous-organic systems. The reactions of the
resultant Cr(V)-thiolato complexes with carbohydrate models
(cis-1,2-cyclopentanediol and cis-1,2-cyclohexanediol)7,12,16
were also studied to mimic the conditions of the formation of
Cr(V)-diolato complexes in biological systems. The choice of
the model thiol was determined by the following considerations:
(i) apart from the thiolato S, there are no other potential donor
atoms for Cr in the RSH molecule, which excludes the formation
of chelated thiolato-Cr(V) species (unlike those for multifunc-
tional ligands such as GSH and its derivatives);10,17 (ii) the
nonionic nature of RSH (as opposed to that of GSH and its
derivatives)10,17 and characteristic 79Br/81Br isotopic distribution
patterns are beneficial for the studies of its Cr complexes by
electrospray mass spectrometry (ESMS);18 (iii) a stable and well-
characterized Cr(VI)-thiolato complex, (Ph4As)[CrVIO3(SR)],11,19
can be used as a starting material for kinetic experiments; and
(iv) RSH is relatively air-stable and nonvolatile and is easier to
handle than other nonionic thiols. The only significant disad-
vantage of the RSH chosen for this research was its low water
solubility. Preliminary studies of Cr(V) complexes formed in
the reactions of Cr(VI) with p-bromobenzenethiol have been
reported,20 but no definitive characterization of these species
has been achieved to date. Detailed information on the mech-
anisms of formation and decomposition of Cr(V)-RSH com-
plexes, obtained in this work, has shed a new light on the nature
of Cr(V) species formed during the reactions of Cr(VI) with
biological thiols such as glutathione in vitro10 as well as in
cultured mammalian cells4 and in living animals.5
Reagents. Acetonitrile (MeCN) and N,N-dimethylformamide
(DMF) (both of HPLC grade, Aldrich) were freshly distilled from
activated molecular sieves (4 Å, Aldrich) before use. Other
commercial reagents of analytical or higher purity (purchased from
Aldrich, Sigma, or Merck) were used without further purification.
Water was purified by the Milli-Q technique. The Cr(VI/V)
complexes, (Ph4As)[CrVI(O)3(SR)] (where RSH is p-bromobenzene-
thiol),11,19 Na[CrVO(ehba)2],14 K[CrVOL2](whereLH2)cis-1,2-cyclo-
pentanediol or cis-1,2-cyclohexanediol),16 and K[CrVO(Aib3-DMF)]
(where Aib ) 2-amino-2-methylpropanoic acid)21 were synthesized
and characterized as described previously. Attempts to isolate
[CrVO(SR)4]-, based on the previously developed methods for the
synthesis of Cr(V)-hydroximato22 and -1,2-diolato16 complexes,
used the reaction of Cr(VI) (50 mM, from (NH4)2Cr2O7) with RSH
(500 mM) in anhydrous DMF (0.50 mL) for 15 min at 295 K
(optimal conditions for the formation of [CrVO(SR)4]-, see Results),
followed by the addition of a mixture of n-hexane (5.0 mL)
and anhydrous acetone (2.0 mL). The resultant dark, oily precipitate
was separated by decanting the solvent and drying the residue under
a stream of N2. This procedure led to decomposition of [CrVO(SR)4]-
with the formation of [CrVI(O)3(SR)]- and small amounts of
[CrV(O)2(SR)2]-, as shown by electrospray mass spectrometry
(ESMS) and EPR spectroscopy (see Results). Previous attempts in
the literature to isolate pure Cr(V)-RSH (RSH ) p-bromoben-
zenethiol) complexes were also unsuccessful.20
Analytical Techniques. For EPR spectroscopy, freshly prepared
and Ar-saturated stock solutions of Cr(VI) and RSH in DMF,
DMSO (dimethylsulfoxide), or MeCN were mixed under an Ar
atmosphere, and the mixture was transferred by a Hamilton syringe
into quartz capillaries (length, 10 cm; internal diameter, 0.50 mm),
which were then sealed from the top with vacuum grease. A
capillary containing the reaction mixture was placed into a quartz
EPR tube (length, 20 cm; internal diameter, 3 mm) together with
a second capillary, containing a stable Cr(V) complex (0.050-2.0
mM Na[CrVO(ehba)2] in DMF solution, where ehba ) 2-ethyl-2-
hydroxybutanoato(2-)),14 which was used as a concentration
calibrant. All the preparations were carried out in dim light since
both Cr(VI) and Cr(V) complexes are photosensitive.16,23,24 Time-
dependent EPR spectra (X-band) were collected on an Elexsys
spectrometer (Bruker), equipped with an internal NMR gaussmeter,
at 295 ( 1 K. Stock solutions of Cr(VI) complexes (20 mM of
(Ph4As)[CrVI(O)3(SR)] or 100 mM of (NH4)2Cr2O7 in DMF) and
the calibration solutions of Na[CrVO(ehba)2] (0.050-2.0 mM) in
DMF were stable for at least 12 h at 295 K when protected from
light (determined by electronic absorption spectroscopy).11,14 The
validity of the calibration technique was confirmed by filling the
two capillaries with DMF solutions of two stable Cr(V) complexes,
Experimental Section
Caution! Cr(VI) compounds are human carcinogens,2 and Cr(V)
complexes are mutagenic and potentially carcinogenic;9 appropriate
precautions should be taken to avoid skin contact and inhalation
of their solutions and dusts.
[CrVO(ehba)2]- and [CrVO(Aib3-DMF)]- 21 at concentration ratios
,
of 0.10-10; the ratios obtained from EPR spectral simulations were
within 10% of the expected values (Figure S1 in Supporting
Information). The giso value of [CrVO(ehba)2]- (1.9777 in DMF
solutions)14d was significantly lower than those of the Cr(V) species
formed in the reaction mixtures (1.9790-1.9960, see Results), so
that the calibrant signal did not interfere significantly with those
of the measured Cr(V) species (typical examples are shown in
Figure S2, Supporting Information). Spectral simulations with
WinSim software25 were used for the determination of Cr(V)
concentrations, as well as of the giso and Aiso (53Cr) values of the
Cr(V) species (where second-order corrections were applied).
Kinetic curves for Cr(V) species, obtained from the EPR spectral
simulations, were modeled using Chemical Kinetics Simulator
(10) Levina, A.; Zhang, L.; Lay, P. A. Inorg. Chem. 2003, 42, 767–784.
(11) Levina, A.; Lay, P. A. Inorg. Chem. 2004, 43, 324–335.
(12) (a) Farrell, R. P.; Lay, P. A. Comments Inorg. Chem. 1992, 13, 133–
175. (b) Barr-David, G.; Charara, M.; Codd, R.; Farrell, R. P.; Irwin,
J. A.; Lay, P. A.; Bramley, R.; Brumby, S.; Ji, J.-Y.; Hanson, G. R.
J. Chem. Soc., Faraday Trans. 1995, 91, 1207–1216.
(13) Mazur, M. Anal. Chim. Acta 2002, 456, 129–146.
(14) (a) Krumpolc, M.; Rocˇek, J. J. Am. Chem. Soc. 1979, 101, 3206–
3209. (b) Krumpolc, M.; Rocˇek, J. Inorg. Chem. 1985, 24, 617–621.
(c) Judd, R. J.; Hambley, T. W.; Lay, P. A. J. Chem. Soc., Dalton
Trans. 1989, 2205–2210. (d) Bramley, R.; Ji, J.-Y.; Judd, R. J.; Lay,
P. A. Inorg. Chem. 1990, 29, 3089–3094.
(15) Chappell, J.; Chiswell, B.; Canning, A. Talanta 1998, 46, 23–38.
(16) Bartholoma¨us, R.; Irwin, J. A.; Shi, L.; Meejoo, S.; Levina, A.; Lay,
P. A. Inorg. Chem. 2010. Submitted.
(17) Zhang, L. EPR and XAFS Studies of Biologically Relevant Chromi-
um(V) Complexes and Manganese(II)-Activated Aminopeptidase P.
Ph.D. Thesis, The University of Sydney: Sydney, Australia, 1998.
(18) (a) Gaskell, S. J. J. Mass Spectrom. 1997, 32, 677–688. (b) Cole, R. B.
J. Mass Spectrom. 2000, 35, 763–772.
(21) Barnard, P. J.; Levina, A.; Lay, P. A. Inorg. Chem. 2005, 44, 1044–
1053.
(22) Gez, S.; Luxenhofer, R.; Levina, A.; Codd, R.; Lay, P. A. Inorg. Chem.
2005, 44, 2934–2943.
(23) Mitewa, M.; Bontchev, P. R. Coord. Chem. ReV. 1985, 61, 241–272.
(24) Bontchev, P. R.; Mitewa, M.; Russev, P.; Petrov, G.; Malinovski, A.;
Kabassanov, K. J. Inorg. Nucl. Chem. 1979, 41, 1451–1456.
(25) Duling, D. R. J. Magn. Reson. 1994, B104, 105–110. The WinSim
software is available via the Internet at http://epr.niehs.nih.gov/.
(19) Mazurek, W.; Fallon, G. D.; Nichols, P. J.; West, B. O. Polyhedron
1990, 9, 777–779.
(20) Mazurek, W.; Nichols, P. J.; West, B. O. Polyhedron 1991, 10, 753–
762.
9
J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010 8721

A R T I C L E S
Levina et al.
software,26 based on the stochastic approach (a typical number of
reacting species in the simulations was 5 × 105).
The ESMS analyses were performed using a Finnigan LCQ mass
spectrometer; typical experimental settings were as follows: sheath
gas (N2) pressure, 60 psi; spray voltage, 4.0 kV; capillary temper-
ature, 473 K; cone voltage, 5 V; tube lens offset, 20 V; and m/z
range, 100-2000 (both in negative- and positive-ion modes; no
signals of Cr species were detected in the latter). Analyzed solutions
(5 µL, 0.50 mM Cr in DMF) were injected into a flow of MeOH
(flow rate, 0.30 mL min-1). Acquired spectra were the averages of
10 scans (scan time, 10 ms). Simulations of the mass spectra were
performed using IsoPro software.27 Concentrations of unreacted
RSH in the reaction mixtures after the complete reduction of Cr(VI)
to Cr(III) were determined spectrophotometrically with Ellman’s
reagent,28 as described previously.29 Electronic absorption (UV-vis)
spectroscopy was performed on a Hewlett-Packard HP 8452 A
diode-array spectrophotometer (spectral range, 200-800 nm;
resolution, 2.0 nm; acquisition time, 0.20 s). All the spectroscopic
and kinetic measurements were performed at 295 ( 1 K.
Results
Reactions of Cr(VI) with RSH: Identification of Cr(V)
Intermediates. On the basis of the results of previous studies
of Cr(VI) reduction by glutathione in aqueous solutions,10 the
reaction of Cr(VI) with RSH was initially studied under the
conditions that facilitated the detection and characterization of
Cr(V) intermediates ([Cr(VI)] ) 50 mM and [RSH] ) 500 mM
in DMF solutions at 295 K). The source of Cr(VI) in these
experiments was (NH4)2Cr2O7, due to its higher solubility in
DMF compared with those of the Na+ or K+ chromates or of
(Ph4As)[CrO3(SR)]. The color of the reaction mixture changed
within several minutes from orange to dark green. The reaction
mixtures were diluted 100-fold with DMF at various time points,
and electronic spectra of dilute solutions were recorded within
∼30 s after the dilution (Figure S3a in Supporting Information).
When the reaction mixture was diluted at an early stage (∼30
s after mixing of the reagents), the resultant spectrum showed
a characteristic absorbance band of [CrVIO3(SR)]- (λmax ) 414
nm), which is different from that for a solution of [Cr2VIO7]2-
in DMF in the absence of RSH (λmax ) 380 nm) (Figure S3a
[Supporting Information]).11 This change was followed by the
appearance of a broad absorbance band with λmax ≈ 720 nm,
which reached its maximal intensity at ∼15 min after mixing
of the reagents, and then decayed to the background level within
the next ∼70 min (Figure S3a). In other experiments, the
reaction of Cr(VI) (50 mM) with RSH (500 mM) in DMF
solution was allowed to proceed for 15 min at 295 K (to reach
Figure 1. Typical EPR spectra (X-band, 295 K) of Cr(V) species formed
during the reactions of Cr(VI) ((NH4)2Cr2O7, a-d) or Cr(V) (K[CrVO(cpd)2],
where cpd ) cis-1,2-cyclopentanediol,16 (e) complexes with p-bromoben-
zenethiol (RSH) in DMF solutions. All the reactions were performed at
295 K, and the spectra were collected in the absence of a Cr(V) standard
(see the Experimental Section). Signals due to the 53Cr hyperfine splitting8,12
are designated with asterisks. Reaction conditions: (a) Cr(VI) (50 mM)
reacted with RSH (500 mM) in DMF for ∼3 min, undiluted reaction
mixture; (b) Cr(VI) (50 mM) reacted with RSH (500 mM) in DMF for 15
min, then the reaction mixture was diluted 100-fold with DMF, and the
spectrum was collected at 5 min after the dilution; (c) same as (b), but 15
min after the dilution; (d) Cr(VI) (50 mM) reacted with RSH (500 mM) in
DMF for 15 min, then the Cr(V) complexes were precipitated by a mixture
of hexane and acetone (see the Experimental Section) and redissolved in
DMF at [Cr] ∼ 10 mM, the spectrum was recorded at ∼3 min after the
dissolution in DMF; and (e) Cr(V) (0.50 mM) reacted with RSH (10 mM)
in DMF for ∼3 min, undiluted reaction mixture.
the maximum concentration for the intermediate with λmax
≈
720 nm), then the reaction mixtures were diluted 100-fold with
DMF, and time-dependent absorbance changes (295 K) in dilute
solutions were followed by electronic absorption spectroscopy
(Figure S3b in Supporting Information). In these experiments,
the intensity of absorbance at ∼720 nm decayed to the
background level within ∼60 min, and a shoulder at ∼414 nm
appeared, which was indicative of the presence of [CrVI(O)3-
(SR)]-.11
Undiluted reaction mixtures ([Cr(VI)]0 ) 50 mM and [RSH]0
EPR signal of a Cr(V) intermediate (Figure 1a; giso ) 1.9960;
Aiso(53Cr) ) 14.7 × 10-4 cm-1). A marked decrease in the
linewidths of the 53Cr satellite signals with an increase in
magnetic field was observed for this species (inset in Figure
1a). All these signals (designated with asterisks in Figure 1a)
were of the same integral intensity (within 10% experimental
error, as determined by EPR spectral simulations), so these
signals were due to the 53Cr satellites of the main Cr(V) species
(giso ) 1.9960), and not due to the other minor Cr(V) species.
) 500 mM in DMF solution at 295 K) showed a single isotropic
(26) (a) Chemical Kinetics Simulator, Version 1.01; IBM Almaden Research
Center, New York, 1996, http://www.almaden.ibm.com/st/computational_
science/ck/?cks; (b) Levina, A.; Lay, P. A.; Dixon, N. E. Inorg. Chem. 2000,
39, 385–395.
(27) Senko, M. IsoPro 3.0; Freeware, 1998.
(28) Ellman, G. L. Arch. Biochem. Biophys. 1959, 82, 70–77.
(29) Levina, A.; Bailey, A. M.; Champion, G.; Lay, P. A. J. Am. Chem.
Soc. 2000, 122, 6208–6216.
9
8722 J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010

Formation and Reactivity of Cr(V)-Thiolato Complexes
A R T I C L E S
Table 1. Mass Spectrometric and EPR Spectroscopic Data for Cr
Intermediates Detected During the Reactions of Cr(VI) or Cr(V)
with p-Bromobenzenethiol and a Cyclic Diola
d
speciesb
-m/zc
giso
Aiso × 10-4 cm-1
[CrV(O)2(SR)2]-
[CrVO(SR)4]-
460.1
820.0
543.7
1.9828
6.8
14.7
15.8e
1.9960
[CrVO(cpd)(SR)2]-
1.9852 (60%)e
1.9856 (40%)e
[CrVO(chd)(SR)2]-
557.9
1.9853 (50%)e
16.2e
1.9855 (50%)e
[CrVI(O)3(SR)]-
[CrIVO(SR)3]- f
[CrIII(SR)4]- f
287.1
631.1
804.0
-
-
-
-
-
-
a Reaction conditions corresponded to those described in the captions
to Figures p-bromo-
and 2. b Designations of ligands: RSH
benzenethiol; cpdH2 cis-1,2-cyclopentanediol; and chdH2
1
)
)
)
cis-1,2-cyclohexanediol. c Values for the highest peak in the isotopic
distribution (Figures 2 and S4). d Hyperfine splitting due to the 53Cr
nuclei (S ) 3/2; natural abundance, 9.55%).12 e The EPR signals were
simulated for a mixture of two geometric isomers;7,16 the Aiso values
were similar for both isomers and were not resolved. f These species
were only observed under ESMS conditions; the Cr(IV) and Cr(III)
species in solutions are likely to bind solvent molecules as extra
ligands.18
The intensity of the giso ) 1.9960 signal increased within ∼15
min and then decreased to zero within ∼80 min after mixing
of the reagents, in accordance with the changes in the absorbance
intensity at ∼720 nm (Figure S3a [Supporting Information]).
Dilution of the reaction mixture (100-fold with DMF) led to
the appearance and growth of a second EPR signal (giso
)
1.9828; Figure 1b,c). On the basis of the electronic absorption
and EPR spectroscopic data (Figures 1 and S3 [Supporting
Information]), isolation of the main Cr(V) intermediate (giso
)
1.9960) from the concentrated reaction mixture was attempted
by a method previously used for the isolation of Cr(V)-
hydroximato and -1,2-diolato complexes16,22 (see the Experi-
mental Section). The resultant black powder was dissolved in
DMF ([Cr] ∼ 10 mM), and the EPR spectrum of the solution
showed only the giso ) 1.9828 signal (Figure 1d), which had
an unusually low Aiso(53Cr) value12 for a Cr(V) complex (6.8 ×
10-4 cm-1; inset in Figure 1d). The EPR spectroscopic
parameters of the Cr(V) species are summarized in Table 1.
The conditions used for the ESMS experiments corresponded
to those used for the electronic absorption and EPR spectro-
scopic measurements (Figures 1 and S3 [Supporting Informa-
tion]; [Cr(VI)]0 ) 50 mM and [RSH]0 ) 500 mM in DMF
solutions at 295 K), including dilutions (100-fold with DMF)
of the reaction mixtures at various time points, followed by
immediate collection of the spectra, as well as the studies of
time-dependent changes in dilute reaction mixtures. Typical
ESMS data are shown in Figure 2 and Figure S4, Supporting
Information, and the -m/z values of major Cr species are
summarized in Table 1. At ∼30 s after mixing of the reagents,
the dominant ESMS signal was due to [CrVI(O)3(SR)]- (m/z )
-287.3; Figure S4a [Supporting Information] and Table 1).11
As the reaction progressed, the signal of [CrVO(SR)4]- (m/z )
-820.0; Table 1) became more abundant, reaching its maximal
intensity at ∼15 min after initiation of the reaction (Figure 2a).
The spectrum in Figure 2a was collected within ∼30 s after
dilution of the reaction mixture. When the dilute reaction
mixtures were left at 295 K for a longer time, the signals of
[CrVIO3(SR)]- and [CrV(O)2(SR)2]- (m/z ) -460.1; Table 1)
grew on the expense of that of [CrVO(SR)4]- (see Figure S4b,
time after dilution 10 min; and Figure 2b, time after dilution
40 min), in parallel with the relative growth of the giso ) 1.9828
Figure 2. Typical ESMS data (see the Experimental Section for instru-
mental parameters) for the reactions of Cr(VI) ((NH4)2Cr2O7, a,b) or Cr(V)
(K[CrVO(cpd)2], where cpd ) cis-1,2-cyclopentanediol);16 (c) complexes
with p-bromobenzenethiol (RSH) in DMF solutions at 295 K (see also
Figure S4 in Supporting Information). A summary of the m/z values for
major Cr species is given in Table 1. Experimental (black lines) and
simulated27 (red columns) isotopic distributions of the Cr(V) signals are
shown in the insets. Reaction conditions: (a) Cr(VI) (50 mM) reacted with
RSH (500 mM) in DMF for 15 min, then the reaction mixture was diluted
100-fold with DMF, and the spectrum was collected at ∼30 s after the
dilution; (b) same as (a), but 40 min after the dilution; (c) Cr(V) (0.50
mM) reacted with RSH (10 mM) in DMF for ∼3 min, undiluted reaction
mixture.
signal in EPR spectra under the same conditions (Figure 1b,c).
The growth and decay of the [CrVO(SR)4]- signal in ESMS
were accompanied by corresponding changes of the [CrIVO-
(SR)3]- signal (m/z ) -633.1; Table 1 and Figures 2a and S4a,b
[Supporting Information]). This Cr(IV) species is likely to form
from reduction of [CrVO(SR)4]- under the conditions of
negative-ion ESMS,18 as shown by their roughly constant
abundance ratio (25-35% Cr(IV) relative to Cr(V)) at a low
cone voltage (-5 V; Figures 2a and S4a,b [Supporting Informa-
tion]), and by a marked increase in the relative abundance of
the Cr(IV) species at a high cone voltage (-100 V; Figure S4c
[Supporting Information]). When the reaction was allowed to
proceed for ∼60 min prior to the dilution, most of the signals
due to the Cr(VI/V/IV) species had disappeared, and a weak
signal appeared due to a Cr(III) species ([CrIII(SR)4]-; m/z )
-804.0; Table 1 and Figure S4d [Supporting Information]).
Finally, the black solid isolated from the concentrated reaction
9
J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010 8723

A R T I C L E S
Levina et al.
mixture (see the Experimental Section), after dissolution in DMF
([Cr] ∼ 1 mM) showed a main signal due to [CrVI(O)3(SR)]-
and a minor one due to [CrV(O)2(SR)2]- (no detectable levels
of [CrVO(SR)4]-; Figure S4e [Supporting Information]), in
agreement with the presence of only one Cr(V) signal in the
corresponding EPR spectrum (giso ) 1.9828, Figure 1d). Typical
experimental and calculated isotopic distributions in the ESMS
signals of Cr(V) complexes (arising from the roughly equal
abundance of 79Br and 81Br isotopes in the ligand)27 are shown
in the insets within Figure 2.
In summary, two Cr(V) intermediates have been identified
(by a combination of electronic absorption and EPR spec-
troscopies and ESMS) during the reduction of Cr(VI) by excess
RSH in DMF solutions: [CrVO(SR)4]-, (m/z ) -820.0; giso
)
1.9960; Aiso(53Cr) ) 14.7 × 10-4 cm-1; λmax ∼720 nm), which
is predominantly formed at high concentrations of the reagents;
and [CrV(O)2(SR)2]- (m/z ) -460.1; giso ) 1.9828; Aiso(53Cr)
) 6.8 × 10-4 cm-1; no distinct absorbance peaks at 400-800
nm), which is formed on dilution of the reaction solutions with
excess DMF or at an attempted isolation of [CrVO(SR)4]-. These
spectral features were then used in the identification and
quantification of intermediates in the kinetic studies.
Figure 3. Typical kinetic curves for the formation of [CrVO(SR)4]- (giso
) 1.9960) species in deuterated and undeuterated solvents during the
reactions of [CrVI(O)(SR)3]- (10 mM) with RSH (250 mM) at 295 K, as
observed by quantitative EPR spectroscopy with Na[CrVO(ehba)2] as an
internal standard (see the Experimental Section for details). A full set of
kinetic curves for all the reaction conditions listed in Table 2 is given in
the kinetic curves for the reactions in mixed solvents (DMF/H2O or DMF/
MeOH) shown in Figure S5, Supporting Information.
Kinetic Experiments and Modeling. Detailed kinetic studies
of the Cr(VI) reduction by RSH (focused on the formation and
decay of Cr(V) species) were carried out by quantitative EPR
spectroscopy at 295 K, using Na[CrVO(ehba)2] (0.050-2.0 mM
solutions in DMF) as an internal standard (see the Experimental
Section). Two parallel experimental series were carried out using
either (Ph4As)[CrVI(O)3(SR)]11,19 or (NH4)2Cr2O7 as the sources
of Cr(VI), with [Cr(VI)]0 ) 1.0-10 mM and [RSH]0 ) 20-750
mM, in DMF, DMSO, MeCN, or mixed solvents (DMF/H2O
or DMF/MeOH; a full list of kinetic experiments is given in
Table S1, Supporting Information). In the experiments using
[Cr2VIO7]2- as a source of Cr(VI), the initial concentration of
RSH was increased by the value of [Cr(VI)]0, to allow for the
formation of [CrVI(O)3(SR)]-.11 Under the studied reaction
conditions, the choice of the Cr(VI) source did not significantly
affect the reaction kinetics. Kinetic data for all the reaction
conditions listed in Table S1 (including the use of alternative
sources of Cr(VI)) are shown in Figure S5, Supporting Informa-
tion.
The most prominent feature of the reactions of Cr(VI) (10
mM) with excess RSH (g50 mM) in DMF or DMSO solutions
was the rapid formation of the giso ) 1.9960 species ([Cr-
VO(SR)4]-, Table 1) followed by a period of a steady-state
concentration and then by a rapid decay (solid black and red
lines in Figure 3, and red line in Figure 4a). This unusual kinetic
behavior was confirmed by electronic spectroscopy (using the
specific absorbance of [CrVO(SR)4]- at 720 nm, Figure S3
[Supporting Information]); a comparison of typical kinetic
curves obtained by the both methods is shown in Figure S6,
Supporting Information. Previously, similar kinetic behavior was
observed (but not explained) by Rocˇek and co-workers30 for
Cr(V) species (detected by their specific absorbance at ∼750
nm) that are formed during the reactions of Cr(VI) with
2-hydroxycarboxylic acids in acidic aqueous solutions. In the
current work, since [CrV(O)2(SR)2]- (unlike [CrVO(SR)4]-) did
not show any characteristic absorbance peaks (see above), there
was no alternative to the use of EPR spectroscopy for detailed
studies of the formation and decay of both of the Cr(V)
intermediates. In addition, unlike most of the previous kinetic
studies of Cr(VI) reduction,30,31 electronic absorption spectros-
copy could not be used to follow the changes in concentrations
of Cr(VI) species, since their absorbance at ∼400 nm was
masked by that of the excess ligand and Cr(V) species (Figure
S3, Supporting Information).
In contrast to the reactions in DMF or DMSO solutions, the
kinetics of the formation and decay of Cr(V) species in the
reactions of Cr(VI) with a large excess of RSH in MeCN
solutions were close to those for typical sequential pseudo-first-
order reactions6,31 (solid green line in Figure 3). The use of
deuterated MeCN did not significantly alter the reaction kinetics
compared with that in the undeuterated solvent, in contrast to
the major differences in the reaction kinetics in deuterated vs
undeuterated DMF or DMSO (dashed lines in Figure 3).
Addition of protic solvents such as H2O (5-20% vol/vol) or
MeOH (10-65% vol/vol) to the reaction solutions (10 mM
Cr(VI) and 250 mM RSH) in DMF caused the Cr(V) intermedi-
ate ([CrVO(SR)4]-, giso ) 1.9960) to form and decay more
rapidly, without major changes in the maximal amounts of
resultant Cr(V) (rows 21-27 in Table S1 and the corresponding
data in Figure S5, Supporting Information). The giso ) 1.9960
value for [CrVO(SR)4]- was solvent independent. Higher
concentrations of H2O could not be used because of precipitation
of RSH, and the reactions in pure MeOH were too fast to
monitor the formation and decay of Cr(V) intermediates by EPR
spectroscopy. The minimal amount of H2O added to DMF in
the mixed-solvent studies (5% vol/vol or ∼3 M) was at least 2
orders of magnitude higher than the amounts of H2O expected
to form in the redox reaction (30 mM from 10 mM Cr(VI),
Scheme 1) or to be present in dried and freshly distilled DMF
(<10 mM).32
The most simple scheme that adequately describes the
observed kinetic phenomena (obtained by the stochastic simula-
(30) (a) Krumpolc, M.; Rocˇek, J. J. Am. Chem. Soc. 1977, 99, 137–143.
(b) Mahapatro, S. N.; Krumpolc, M.; Rocˇek, J. J. Am. Chem. Soc.
1980, 102, 3799–3806.
(31) Lay, P. A.; Levina, A. Inorg. Chem. 1996, 35, 7709–7717.
(32) Izutsu, K. Electrochemistry in Nonaqueous Solutions; Wiley-VCH:
Weinhem, Germany, 2002; pp 85-106 and 287-300.
9
8724 J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010

Formation and Reactivity of Cr(V)-Thiolato Complexes
A R T I C L E S
Figure 4. Typical experimental and calculated26 kinetic curves for the reaction of [CrVIO3(SR)]- (10 mM) with RSH (a, 100 mM or b, 20 mM) in DMF
solutions at 295 K. The experimental data were obtained by quantitative EPR spectroscopy with Na[CrVO(ehba)2] as an internal standard (see the Experimental
Section). The kinetic model and designations of the reacting species correspond to Table 2. Expanded views (a, Right) and (b, Right) of experimental (filled
circles) and calculated (lines) kinetic curves for the Cr(V) species (the species giving rise to the giso ) 1.9828 signal were not detected at [RSH]0 ) 100
mM). The arrows in (a) show the time point when a rapid decay of Cr(V) begins, corresponding to the disappearance of Cr(VI).
tion method)26 is shown in Table 2 (eqs 1-8), and the values
of the rate constants that were dependent on the reaction
conditions (k1, k4, and k6) are listed in Table S1 (Supporting
Information) (the other rate constants are given in Table 2).
Typical examples of calculated kinetic curves for the main
reacting species, as well as experimental kinetic curves for Cr(V)
species present during the reactions of Cr(VI) (10 mM) with
high or low concentrations of RSH (100 or 20 mM, respec-
tively), are shown in Figure 4a and b. A comparison of
experimental and calculated kinetic curves (for Cr(V) species
only) for all the reaction conditions used (see Table S1
[Supporting Information]) is shown in Figure S5, and a
comparison of calculated kinetic curves for Cr(V) species under
various reaction conditions is given in Figure S7 (Supporting
Information). Many alternative kinetic schemes were considered
and rejected, since they failed to reproduce the observed
dependences of the kinetic curves for Cr(V) intermediates on
the reaction conditions (typical examples are shown in Table
S2 and Figure S8, Supporting Information).
assumed to be the only reactive form of Cr(VI).11 According
to the proposed kinetic scheme (Table 2), the first Cr(V)
intermediate formed by one-electron reduction of Cr(VI)
(designated as Cr(V)a, eq 1 in Table 2) was unstable and was
not detected by EPR spectroscopy (its calculated concentrations
never exceeded 2% of the total Cr(V) concentration). Further
reactions of Cr(V)a with excess RSH (eqs 2 and 3 in Table 2)
produced two relatively stable Cr(V) species, designated as
Cr(V)b and Cr(V)c that correspond to [CrV(O)2(SR)2]- and
[CrVO(SR)4]-, respectively, (see the Discussion section for
details). These species reacted with solvents, such as DMF or
DMSO, to produce free radicals (designated as Solv•; eqs 4 and
7 in Table 2), which then rapidly reacted with Cr(VI) to
regenerate Cr(V) species (eq 5 and 7 in Table 2). The ability of
DMF to form radicals on the reactions with oxidants, including
Cr(V), has been reported previously.21,33 These reactions led
to relatively constant levels of Cr(V) species in the reaction
mixtures in the presence of excess Cr(VI), and to their rapid
decay when Cr(VI) was spent (see the representative calculated
kinetic curves in Figure 4a). The same relation between the
kinetic curves for Cr(VI) and Cr(V) species formed in the
reactions of Cr(VI) with 2-hydroxycarboxylic acids in acidic
Since all the reactions were performed in the presence of
excess RSH over Cr(VI) (see Table S1 [Supporting Informa-
tion]) and no kinetic differences were observed when either
[CrVI(O)3(SR)]- or [Cr2VIO7]2- was used as a source of Cr(VI)
(Figure S5 [Supporting Information]), [CrVI(O)3(SR)]- was
(33) Muzart, J. Tetrahedron 2009, 65, 8313–8323.
9
J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010 8725

A R T I C L E S
Levina et al.
Scheme 1. Proposed Mechanism for the Cr(VI) + RSH Reaction
(Equation Numbers and Designations of the Reacting Species
Correspond to Those in Table 2)a
corresponding reactions in Table S1 [Supporting Information])
are consistent with the production of Solv• radicals during the
reactions of C-H bonds of the solvent molecules with Cr(V).
The corresponding reactions in MeCN solutions were at least
an order magnitude slower than those in DMF or DMSO
solutions (see the k4 values in rows 17-20, Table 2), which
was consistent with much lower reactivity MeCN toward
oxidants compared with DMF or DMSO. For instance, elec-
trochemical oxidation of MeCN occurs at +3.0 V vs +1.5 V
for both DMF and DMSO (using the ferricenium/ferrocene
couple as a reference).32 Changes in the reaction kinetics caused
by the addition of H2O or MeOH to DMF solutions were
modeled by the increases in the rate constants k1 and k4 (Table
S1 [Supporting Information]). The calculated values of k4 also
increased at very high concentrations of RSH (g500 mM, Table
S1 [Supporting Information]), which pointed to a change in the
reaction mechanism under these conditions. The reaction kinetics
at low RSH concentrations (e20 mM at [Cr(VI)]0 ) 10 mM)
could not be described adequately by the scheme in Table 2,
probably because of the participation of more than one form of
Cr(VI) (e. g., [CrVI(O)3SR]- and [HCrVIO4]-).11 Detailed studies
of the reaction kinetics at very high or very low RSH
concentrations were beyond the scope of the current work, since
these conditions were not relevant to the biological context of
the reactions of interest.4,6
The involvement of solvents in the redox reactions (eqs 4, 6
and 7 in Table 2) led to the solvent-dependence of the reaction
stoichiometry, with the amount of RSH required for the
complete reduction of Cr(VI) to Cr(III) being generally lower
(by about one molar equivalent) in DMF solutions compared
with MeCN solutions (Figure S9 in Supporting Information).
The observed reaction stoichiometry (2-5 mols of RSH required
for the reduction of one mol of [CrVI(O)3(SR)]- to Cr(III)
products) was also dependent on the initial concentration of RSH
(due to the various proportions of the two main Cr(V)
intermediates being formed), as confirmed by kinetic modeling
(Figure S9 [Supporting Information]).
The determination of Cr(V) concentrations during the reac-
tions of Cr(VI) with RSH (by quantitative EPR spectroscopy)
showed that under no conditions did the total concentration of
Cr(V) intermediates exceed 20% (mol.) of the initial concentra-
tion of Cr(VI) (Figure S5). This finding was supported by the
results of kinetic modeling (e.g., Figure 4), which taken together,
suggest that Cr(V) species always coexist in the reaction
mixtures with larger amounts of Cr(VI) and Cr(III) species.
Therefore, these reactions could not be used for quantitative
generation of Cr(V) thiolato complexes in solutions for structural
studies by X-ray absorption spectroscopy (in contrast to our
previous studies of reactive Cr(VI/V/IV) complexes).10,11,34 Note
that ESMS data (Figure 2a) suggested a higher yield of
[CrVO(SR)4]-, possibly due to the higher sensitivity of the ion
detector to this species compared with that for [CrVI(O)3(SR)4]-.
An attempted isolation of the [CrVO(SR)4]- complex (see the
Experimental Section) led to its decomposition (Figures 1d and
S4e [Supporting Information]).
a Some alternative variants of the mechanism, rejected on the basis of
kinetic analysis, are shown in Table S1 and Scheme S1, Supporting
Information.
aqueous media was observed by Rocˇek and co-workers,30 who
calculated these curves from electronic absorption spectra. The
rest of the proposed kinetic scheme consisted of the reduction
of Cr(V)c to a Cr(III) product by excess RSH, and quenching
of the Solv• radicals (eqs 6 and 8 in Table 2).
Reactions of Cr(V) Thiolato Complexes in Dilute Solutions
and Ligand-Exchange Reactions with 1,2-Diols. In order to
determine the possible fate of Cr(V) thiolato intermediates in
biological media, the reactivities of Cr(V)-thiolato complexes
were studied under the conditions that were used previously
The observed deuterium kinetic isotope effect (KIE ) k4H/
k4D) values for the reactions of Cr(V)c with DMF (1.7) or
DMSO (2.3) solvents (see Figure 3 and the k4 values for the
(34) Levina, A.; Codd, R.; Foran, G. J.; Hambley, T. W.; Maschmeyer,
T.; Masters, A. F.; Lay, P. A. Inorg. Chem. 2004, 1046–1055.
9
8726 J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010

Formation and Reactivity of Cr(V)-Thiolato Complexes
A R T I C L E S
Table 2. Proposed Kinetic Model for the Reaction of Cr(VI) with Excess RSH at 295 Ka
reactionb
rate constant
c
k1 ) (1-10) × 10-3 M-1 s-1
k2 g 2 M-1 s-1
Cr(VI) + RSH f Cr(V)a
Cr(V)a + RSH f Cr(V)b
(1)
(2)
k3 ) 20 ( 5 M-2 s-1
Cr(V)b + 2 RSH f Cr(V)c
(3)
(4)
(5)
Cr(V)c + 2 SolvHd f Cr(III) + 2 Solv•
Cr(VI) + Solv• f Cr(V)a
c
k4 ) (0-10) × 10-3 s-1
k5 g 250 M-1 s-1
Cr(V)c + RSH+SolvHd f Cr(III)
Cr(V)b + 2 SolvHd f Cr(III) + 2 Solv•
Solv• f oxidized solvent
(6)
(7)
(8)
(9)
c
k6 ) (3-30) × 10-3 M-1 s-1
k7 ) (1.0 ( 0.2) × 10-3 s-1
k8 g (2.0 ( 0.2) × 10-3 s-1
k9 ) (6 ( 2) × 10-3 M-1 s-1
f h
,
g h
,
;
(1.0 ( 0.3) × 10-3 M-1 s-1
;
or (5 (
Cr(V)b + LH2 f Cr(V)de
f i
,
0.6) × 10-2 M-1 s-1
f g h
, ,
k10 ) (2.0 ( 0.5) × 10-4 M-1 s-1
;
or (1.0 ( 0.07) ×
Cr(V)d + LH2 f Cr(V)ee
(10)
f i
,
10-2 M-1 s-1
a The most simple kinetic model (see Scheme 1 for details) that adequately describes the kinetics of formation and decomposition of Cr(V) species
under the conditions listed in Table S1 (Supporting Information), using either (Ph4As)[CrVI(O)3(SR)] or (NH4)2Cr2O7 as a source of Cr(VI) (see Figure
S5, Supporting Information, for comparison of the experimental and calculated kinetic curves). Representative alternative models are shown in Table S2
and Scheme S1, Supporting Information. b Equation numbers (in parentheses) and designations of the reaction intermediates correspond to Scheme 1. c The
value of the rate constant is solvent-dependent, see Table S1 [Supporting Information] for details. d SolvH ) solvent molecule (e.g., DMF). Concentrations of
the solvents were assumed to be constant in the kinetic analysis. e Ligand-exchange reactions with large excesses of 1,2-diolato ligands, see Scheme 1. f For
LH2 ) cis-1,2-cyclopentanediol (cpdH2). g For LH2 ) cis-1,2-cyclohexanediol (chdH2). h In DMF solution. i In DMF/H2O (4:1 vol/vol) solution.
for identification of the reaction intermediates (Figures 1, 2, S3
and S4 [Supporting Information]). Namely, the [CrVO(SR)4]-
complex was generated by the reaction of Cr(VI) (50 mM, from
(NH4)2Cr2O7) with RSH (500 mM) in DMF solutions (reaction
time 15 min at 295 K, Figure S3a [Supporting Information]),
then the reaction mixtures were diluted 100-fold with various
solvents, and the decomposition of Cr(V) was followed by
quantitative EPR spectroscopy. Concentrations of all of the
reacting species at the time point of dilution were calculated
from the kinetic model (Table 2). A good correspondence
between calculated and experimental values of Cr(V) concentra-
tions under these conditions is shown in Figure S5 (data set 13
[Supporting Information]). The calculated concentration values
for all the reacting species were then divided by the factor of
100 (to account for the dilution) and used to calculate the kinetic
profiles for Cr(V) species in dilute solutions from the same
kinetic model (typical examples are shown in Figure 5 and
Figure S10, Supporting Information). For instance, Figure 5a
shows a good correspondence between the experimental and
calculated kinetic curves for the decay of Cr(V)c ([CrVO(SR)4]-,
giso ) 1.9960) and the growth and decay of Cr(V)b
([CrV(O)2(SR)2]-, giso ) 1.9828) in a dilute solution. By contrast,
direct reactions of low concentrations of Cr(VI) and RSH (0.50
and 5.0 mM, respectively, to approximate the conditions in the
dilute reaction mixture) in DMF solutions led to negligibly slow
formation of Cr(V) intermediates, as confirmed by kinetic
modeling using the parameters contained in Tables S1 and 2
(Figure S10a). Experimentally, no giso ) 1.9960 species and
only marginal concentrations of the giso ) 1.9828 species could
be detected under these conditions (Figure S10a).
Reactions of the preformed [CrVO(SR)4]- species with DMF
solutions containing various concentrations of cyclic 1,2-diols
(cis-1,2-cyclopentanediol, cpdH2, or cis-1,2-cyclohexanediol,
chdH2) were studied as models of the formation of Cr(V)
complexes with biological 1,2-diols (carbohydrates and glyco-
proteins) during the reduction of Cr(VI) with biological thiols
(e.g., glutathione) in diol-rich biological media.7,8,16 As shown
in Figure 5a,b, the rate of decomposition of the giso ) 1.9960
species in dilute ([Cr] ) 0.50 mM) DMF solutions was not
significantly affected by the presence of a large excess of cpdH2
(1.0 M), but the presence of the diol led to complete replacement
of the giso ) 1.9828 signal with that of a new Cr(V) species
(Cr(V)d in Figure 5b, giso ) 1.9854, a Cr(V)-thiolato/diolato
complex, see below) and then to a gradual formation of the giso
) 1.9790 species (Cr(V)e in Figure 5b, described previously
as a Cr(V)-bis-diolato complex, [CrVO(cpd)2]-).16 Similar
changes were observed when the concentrated reaction mixture
was diluted with DMF containing chdH2 (1.0 M), but in this
case, the giso ) 1.9828 species were still detectable at the
beginning of the reaction (Figure S10b [Supporting Informa-
tion]). This kinetic behavior could be modeled satisfactorily by
the addition of two ligand-exchange reactions to the kinetic
scheme (eqs 9 and 10 in Table 2; the ligand concentrations were
1.0 M in all cases and the reactions were treated as pseudo-
first-order processes, with the reverse reactions considered
negligible).26 Note that according to the kinetic model (Table
2, eq 9) the diol ligands reacted initially with Cr(V)b
([CrV(O)2(SR)2]-) and not with Cr(V)c ([CrVO(SR)4]-). After
prolonged reaction times (g100 min after the dilution at 295
K), when the giso ) 1.9960 species were no longer detectable
9
J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010 8727

A R T I C L E S
Levina et al.
the reaction system was limited by the low water-solubility of
RSH (Table S1 [Supporting Information]). Decomposition of
the preformed [CrVO(SR)4]- complex after 100-fold dilution
with a DMF/H2O mixture (4:1 vol/vol) was ∼10 times faster
than that after the dilution with pure DMF, and led to a ∼3-
fold decrease in the amount of the resultant [CrV(O)2(SR)2]-
intermediate (compare Figures 5a and S10c [Supporting Infor-
mation]). Kinetics of Cr(V) decomposition in H2O-containing
solutions was satisfactorily described (Figure S10c [Supporting
Information]) by the general kinetic model of the Cr(VI) + RSH
reaction (Table 2), using the rate constants optimized for the
DMF/H2O systems (Table S1 [Supporting Information]). A
similar reaction in a DMF/H2O mixture containing cpdH2 (1.0
M) was characterized (compared with the DMF/cpdH2 system,
Figure 5b) by a ∼10-fold increase in the rates of both
decomposition of [CrVO(SR)4]- (giso ) 1.9960) and formation
of [CrVO(cpd)2]- (giso ) 1.9790), as well as by a ∼5-fold
decrease in the amounts of the giso ) 1.9854 intermediate (Figure
S10d [Supporting Information]). The optimized kinetic model
for this reaction included increases in rate constants k1 and k4
(Table S1), as well as increases in the values of k9 and k10
compared with the same reaction in the absence of H2O (Table
2). In summary, a comparison of kinetic data in Figures 5b and
S10d (Supporting Information) showed that the presence of H2O
decreased the stability of Cr(V)-thiolato and Cr(V)-thiolato/
diolato complexes, and promoted the formation of Cr(V)-bis-
diolato complexes.
To confirm the nature of the giso ) 1.9854 intermediates
(Cr(V)d in Figures 5 and S10 [Supporting Information]) as
Cr(V)-thiolato/diolato complexes, the same species were gener-
ated in higher yields by the reactions of the isolated K[CrVOL2]
complexes (where L ) cpd or chd)16 with RSH in DMF
solutions (Figure 1e and Figure S11 in Supporting Information).
Previously, ligand-exchange reactions of [CrVOL2]- with GSH
were shown to occur in neutral aqueous solutions.16 According
to EPR spectroscopic data, the maximum yields of the mixed-
ligand species were achieved at [Cr(V)]0 ) 0.50 mM and [RSH]
) 10 mM (Figures 1e and S11b), and ESMS spectra under these
conditions were dominated by the [CrVOL(SR)2]- signals (m/z
) -543.7 or -557.9 for L ) cpd or chd, respectively, Figures
2c and S4f [Supporting Information], and Table 1). Similar to
the observations in earlier studies of Cr(V) bis-diolato16 and
2-hydroxycarboxylato35 complexes, EPR signals of the
[CrVOL(SR)2]- species were fitted by a combination of two
signals with close giso values (Table 1), which were attributed
to different geometric isomers (splits in the giso ) 1.9854 signals
were obvious from the second-derivative spectra, see the inset
in Figure 1e). Unlike for the Cr(V)-bis-diolato complexes,7,8,16
superhyperfine splitting of the [CrVOL(SR)2]- EPR signals due
to the 1H nuclei in the diolato ligands were not resolved. Kinetic
studies of the reactions of K[CrVOL2] with RSH by electronic
absorption spectroscopy (Figure S12 in Supporting Information)
confirmed the formation of [CrVO(SR)4]- (λmax ≈ 720 nm,
Figure 6) at high concentrations of RSH (e.g., 50 mM at [Cr]
) 0.24 mM, Figure S12b [Supporting Information]), and also
showed the formation of a spectrally distinct [CrVOL(SR)2]-
species (λmax ≈ 550 and 650 nm, Figure 6) at lower RSH
concentrations (e.g., 5.0 mM at [Cr] ) 0.70 mM, Figure S12a
[Supporting Information]). In the both cases, the ligand-
exchange reactions were complete within 1-2 min (Figure S12
[Supporting Information]), in contrast to the much slower
Figure 5. Typical kinetic curves (a, b) and EPR spectra (c) for the
decomposition of [CrVO(SR)4]- (giso ) 1.9960 species) in dilute DMF
solutions at 295 K in the presence (b, c) or absence (a) of cis-1,2-
cyclopentanediol (cpdH2, 1.0 M). The [CrVO(SR)4]- complex was generated
by the reaction of Cr(VI) (50 mM, from (NH4)2Cr2O7) with RSH (500 mM)
in DMF for 15 min at 295 K, then the reaction mixture was diluted 100-
fold with an appropriate solvent. Experimental kinetic data (dots in a and
b) were obtained by quantitative EPR spectroscopy with Na[CrVO(ehba)2]
as an internal concentration standard (see the Experimental Section). The
EPR spectrum in c was obtained from an average of ten scans in the absence
of the Cr(V) standard (the experimental conditions correspond to those in
b, reaction time after the dilution 120-130 min). Calculated kinetic curves
(lines in a and b) were obtained from the kinetic model in Table 2 (see text
for details). Designations of the reaction intermediates (in italics) correspond
to Table 2 and Scheme 1.
(Figures 5b and S10b [Supporting Information]), time-dependent
EPR spectra of the reaction solutions closely resembled those of
Cr(VI)-treated mammalian cells4,6 or living animals,5 with the giso
) 1.9854 signal being gradually replaced by that at giso ) 1.9790
(e.g., Figure 5c). Previously,6 the time-dependent changes in the
EPR spectra of Cr(VI)-treated mammalian cells were successfully
modeled with a system containing Cr(VI) (0.50 mM), GSH (2.0
mM) and chdH2 (15 mM) in an aqueous buffer solution (pH )
7.4). In a direct reaction of low concentrations of Cr(VI) (0.50
mM) and RSH (5.0 mM) with cpdH2 (1.0 M) in DMF solution,
only slow formation of the giso ) 1.9790 species was observed
(Figure S10a [Supporting Information]).
The influence of H2O on the stability of Cr(V)-thiolato
complexes and their ligand-exchange products with diols was
studied in relation to the likely biological roles of such
complexes,6-9 but the amount of H2O that could be added to
(35) (a) Codd, R.; Lay, P. A. J. Am. Chem. Soc. 1999, 121, 7864–7876.
(b) Codd, R.; Lay, P. A. Chem. Res. Toxicol. 2003, 16, 881–892.
9
8728 J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010

Formation and Reactivity of Cr(V)-Thiolato Complexes
A R T I C L E S
unstable Cr(IV) intermediates (Cr(IV)a and Cr(IV)b, eqs 4 and
6 in Scheme 1) had to be assumed in order to explain the
conversion of Cr(V) intermediates into Cr(III) products. One
of these species (Cr(IV)a in Scheme 1) corresponds to the
[CrIVO(SR)3]- signal observed in ESMS of the reaction mixtures
(Figures 2 and S4 [Supporting Information] and Table 1; solvent
ligands are usually removed from the complexes under ESMS
conditions).18 Although the appearance of the [CrIVO(SR)3]-
signal in ESMS was probably due to the reduction of
[CrVO(SR)4]- in the gas phase (see Results),18 the observation
of this species under ESMS conditions supports the existence
of small steady-state concentrations of a related Cr(IV) species
(Cr(IV)a in Scheme 1) in the reaction solutions. The concentra-
tions of Cr(IV) intermediates in the reaction mixtures, calculated
from the kinetic model (Table 2), did not exceed 1% (molar
concentration) of the total Cr.
The structures of the Cr(III) products, proposed in Scheme 1
(Cr(III) in eqs 4, 6 and 7) do not correspond to that of the only
Cr(III) species observed by ESMS ([CrIII(SR)4]- in Table 1 and
Figure S4d [Supporting Information]), but its low abundance
(Figure S4d) suggests that this is probably a minor product. The
Cr(III) products of the Cr(VI) + RSH reaction probably consist
mainly of polymeric species (not detected by ESMS), which are
formed from Cr(III) monomers containing hydroxido ligands (such
as the Cr(III) species in Scheme 1).36 The polymeric nature of the
Cr(III) products is supported by the observed slow (hour time scale
at 295 K) formation of a gray-green precipitate during the reactions
of Cr(VI) with excess RSH in DMF solutions.
Figure 6. Typical electronic absorption spectra of Cr(V)-bis-diolato,16 diolato/
thiolato and tetrathiolato complexes (cpdH2 ) cis-1,2-cyclopentanediol; RSH
) p-bromobenzenethiol). The ꢀ values for the [CrVO(cpd)(SR)2]- and
[CrVO(SR)4]- complexes were calculated from the time-dependent spectral
changes during the reactions of Cr(V)-bis-diolato complexes (∼0.5 mM) with
RSH (5-50 mM) in DMF solutions (Figure S12 in Supporting Information).
The total concentrations of Cr(V) in the reaction solutions were determined by
quantitative EPR spectroscopy with Na[CrVO(ehba)2] as an internal concentra-
tion standard (see the Experimental Section).
reactions of Cr(V)-thiolato complexes with diol ligands
(Figures 5 and S10 [Supporting Information]).
The unusual shapes of the kinetic curves for the formation
and decomposition of Cr(V) intermediates (Figures 3 and 4)
prevented conventional determination of the rate constants (in
contrast to the previous kinetic studies of Cr(VI) reactions with
thiols)31 and warranted full modeling of the observed kinetic
curves by the stochastic simulation method (Figures 4 and S5
[Supporting Information]).26 The optimized kinetic model (Table
2) suggests that the stabilization of Cr(V) species in DMF or
DMSO compared with MeCN solutions (Figure 3) occurs
through a chain reaction with the participation of solvent free
radicals, formed during the reactions of Cr(V/IV) intermediates
with the C-H bonds of the solvents (eqs 4 and 5 in Table 2
and Scheme 1). These reactions are likely to be facilitated for
DMF or DMSO by the formation of H-bonds between coordi-
nated solvent molecules and oxido groups of the Cr(V/IV)
intermediates, while a similar reaction is much less likely in
the case of MeCN due to steric considerations (as shown in the
last line of Scheme 1). Although stabilization of Cr(V)
complexes during the reduction of Cr(VI) in moderately reactive
solvents (such as DMF) as opposed to nonreactive ones (such
as MeCN)32 has been noted previously,24 no mechanistic details
of this phenomenon have been described. On the other hand,
the presence of highly reactive solvents such as MeOH
drastically decreased the lifetime of Cr(V) species (Table S1
and Figure S5 [Supporting Information]).
Discussion
Mechanisms of the Cr(VI) Reduction Reaction with a
Model Thiol. Despite two decades of extensive research into
the formation of Cr(V) intermediates during the reduction of
Cr(VI) (usually performed by EPR spectroscopy),7-9,12 mecha-
nistic details of these reactions remain unresolved, largely
because of the difficulties with obtaining accurate values of
Cr(V) concentrations from EPR spectroscopic data.8,12,13 In this
work, quantitative EPR spectroscopy was applied for the first
time for detailed mechanistic analysis of the reaction of Cr(VI)
with a simple thiol ligand (RSH ) p-bromobenzenethiol), used
as a model of Cr(VI) reduction by glutathione and other thiols
in biological systems.6,10 The proposed mechanism of Cr(VI)
reduction to Cr(III) by excess RSH (Scheme 1) was derived
from the following considerations: (i) the known structure19 of
the initial Cr(VI) compound, [CrVI(O)3(SR)]- (Cr(VI) in Scheme
1; could also be formed in situ from [Cr2VIO7]2- and excess
RSH);11 (ii) the molecular formulas of the Cr(V) intermediates
determined by ESMS (Cr(V)b, Cr(V)c, and Cr(V)d in Scheme
1, see also Table 1 and Figure 2); and (iii) the optimized kinetic
scheme (Table 2; the equation numbers correspond to those in
Scheme 1). The main reaction steps included into Table 2 (and
correspondingly into Scheme 1) are explained in detail in the
Results section. A similar mechanism, including the formation
of organic radicals during the reactions of the organic substrate
with a Cr(IV) intermediate, followed by the reaction of these
radicals with Cr(VI) to produce Cr(V) intermediates, has been
proposed for Cr(VI) reactions with alcohols24 and 2-hydroxy-
carboxylic acids30 in acidic aqueous media. The current work
presents the first full modeling of the anomalous kinetic curves
for the Cr(V) intermediates30 that arise from such a mechanism
(Figures 4 and S5 [Supporting Information]).
Some of the modifications of the proposed mechanism
(Scheme 1) that were rejected based on the lack of agreement
with the observed kinetic data, are presented in Scheme S1 (see
also Table S2 and Figure S8, Supporting Information). These
include the following: (i) the reaction of a Cr(IV) intermediate
with [CrVI(O)3SR]- as the main pathway for regeneration of
Cr(V) (instead of the pathway based on solvent radicals, eqs 4
and 5 in Scheme 1); (ii) the formation of a reactive Cr(IV)
Although the optimized kinetic model (Table 2) did not
require an involvement of any Cr(IV) species, the formation of
(36) Stu¨nzi, H.; Marty, W. Inorg. Chem. 1983, 22, 2145–2150.
9
J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010 8729

A R T I C L E S
Levina et al.
intermediate in the reaction of [CrVO(SR)4]- with RSH, rather
than with a solvent molecule (in eq 4, Scheme 1); (iii) the
reaction of [CrVO(SR)4]- with [CrVI(O)3(SR)]- as a route for
the formation of [CrV(O)2(SR)2]- during the decomposition of
[CrVO(SR)4]-; and (iv) the formation of a Cr(IV) species during
the reaction of Cr(V)a with RSH in parallel with the formation
of Cr(V)b (eq 2 in Scheme 1). Given the high sensitivity of the
shapes of the calculated kinetic curves on the reaction mech-
anism (representative examples are shown in Figure S8 [Sup-
porting Information]), it is unlikely that an alternative kinetic
scheme that is consistent with both the observed kinetic data
(Figure S5 [Supporting Information]) and the known chemistry
of Cr(VI/V/IV) complexes6 can replace that in Scheme 1.
Three newly described Cr(V) species ([CrV(O)2(SR)2]-,
[CrVO(SR)4]-, and [CrVOL(SR)2]-, where L is a 1,2-diolato(2-)
ligand), identified by EPR spectroscopy and ESMS (Table 1),
are among the few known Cr(V) complexes with monodentate
ligands.37 Unfortunately, these complexes proved to be too
unstable for structural studies (see Results). No crystal structures
of Cr(V) complexes with thiolato ligands are known to date,37
and the only Cr(V)-thiolato complex structurally characterized
by X-ray absorption spectroscopy was that with glutathione
(where Cr(V) is bound to both thiolato and amido moieties of
the ligand).10 The [CrVO(SR)4]- and [CrVOL(SR)2]- complexes
belong to a large group of mono-oxido, five-coordinate Cr(V)
complexes (probably of a trigonal bipyramidal geometry),12,36
and their giso and Aiso (53Cr) values (Table 1) are consistent with
those for other Cr(V) complexes of this type.8,9,12 By contrast,
[CrV(O)2(SR)2]- is a rare example of a four-coordinate, bis-
oxido Cr(V) complex, and its Aiso (53Cr) value is about half of
those for the five-coordinate, mono-oxido Cr(V) species (Table
1).12 Previously, the Aiso (53Cr) values of typical five-coordinate
Cr(V) complexes were shown to be lower than those for typical
six-coordinate complexes.12 A related four-coordinate, bis-oxido
Cr(V)-amido complex, [CrV(O)2(NtBuAr)2]- (where Ar ) 2,5-
C6H3FMe) has been isolated and characterized by elemental
analyses, NMR spectroscopy and magnetic measurements,38 but
no EPR spectroscopic data for this compound were reported.
Comparison of Cr(VI) Reactions with Glutathione (GSH)
and a Model Thiol. A number of features of the Cr(VI) + RSH
reactions in nonaqueous solvents, observed in this work, were
similar to those found previously for the Cr(VI) + GSH reactions
in aqueous solutions.10 The kinetic profile for the formation and
decomposition of a dark-green Cr(V)-GSH intermediate (followed
by electronic spectroscopy at 750 nm) during the reaction of Cr(VI)
(50 mM) with GSH (500 mM) in aqueous solutions (pH ≈ 7) at
298 K10 was similar to that for the formation and decomposition
of [CrVO(SR)4]- in DMF solutions at the same reagent concentra-
tions (giso ) 1.9960 species, followed by EPR spectroscopy, data
set 13 in Figure S5 [Supporting Information]), except that the
reaction with GSH was ∼20-fold faster.10 In both cases, the kinetic
curves of Cr(V) formation and decomposition were markedly
different from those for the intermediates formed in sequential
pseudo-first-order reactions.6,31 The electronic absorption spectrum
of this Cr(V)-GSH intermediate (recorded after the dilution of
the reaction mixture 100-fold with H2O)10 was also similar to that
of [CrVO(SR)4]- (Figures 6 and S3 [Supporting Information]), but
the visible absorbance band for the former species was less intense
and shifted to lower wavelengths (λmax ≈ 630 vs ∼720 nm). Dilute
solutions of the Cr(VI) + GSH reaction mixtures showed ESMS
signals due to [CrVO(L′H3)2]- (a bis-thiolato-amido chelate
complex) and [CrVO(L′H4)4]- (where L′H5 ) GSH),10 the latter
species being a GSH analogue of [CrVO(SR)4]-. The existence of
an equilibrium between [CrVO(L′H3)2]- and [CrVO(L′H4)4]- in the
Cr(VI) + GSH system was probably responsible for a much
broader EPR signal at giso ≈ 1.9960 compared with the Cr(VI) +
RSH system.10 Dilute solutions of the Cr(VI) + GSH reaction
mixtures showed two main EPR signals (giso ) 1.9960 and 1.9857),
and the relative intensity of the latter signal increased with time as
the total concentration of Cr(V) species decreased (which mimics
the kinetic curves for the giso ) 1.9960 and 1.9828 species in Figure
5a).
Reactions of relatively low concentrations of Cr(VI) (5.0 mM)
and GSH (20 mM) in buffered aqueous solutions (pH ≈ 7) led
to low total yields of Cr(V) species, and the main Cr(V) signals
observed by EPR spectroscopy (particularly at prolonged
reaction times) were those at giso ) 1.9857 (Aiso (53Cr) ) 7.8 ×
10-4 cm-1).10,17 The corresponding Cr(V) species could not be
detected by ESMS due to the high background signals of GSH
ions.10 The low Aiso value (close to that for [CrV(O)2(SR)2]-
but very different from that for the five-coordinate Cr(V) species,
Table 1), and the kinetic behavior (similar to that of the giso
1.9828 species in Figures 4b and 5a) suggest that the giso
)
)
1.9857 species observed in the Cr(VI) + GSH system10,17 are
similar to [CrV(O)2(SR)2]- (giso ) 1.9828, Table 1). The
difference in giso values is likely to be caused by the weak axial
binding of either an additional donor group of the ligand (e.g.,
an amido group in GSH)10 or a solvent molecule (e.g., H2O) to
the [CrV(O)2(SR)2]- species, as suggested by the results of
experiments with various model thiols.17,39
In summary, the two main Cr(V) species (giso ) 1.9960 and
1.9857), formed during the Cr(VI) + GSH reactions in neutral
aqueous solutions,10 are similar in spectral properties and kinetic
behavior to the [CrVO(SR)4]- and [CrV(O)2(SR)2]- complexes
(giso ) 1.9960 and 1.9828, respectively, Table 1), formed in
the Cr(VI) reaction with a simple model thiol, RSH, in DMF
solutions. The same two types of Cr(V)-thiolato species were
observed by EPR spectroscopy in Cr(VI) reactions with other
model thiols and GSH derivatives in neutral aqueous solutions.17
Studies of the Cr(VI) reactions with simple model thiols such
as RSH in nonaqueous media can partially overcome the
practical difficulties in the studies of Cr(V) complexes with GSH
or related thiols in aqueous solutions,10,17 including high reaction
rates (which prevent detailed kinetic studies by EPR spectros-
copy), low concentrations of the Cr(V) species and difficulties
in their detection by ESMS due to the ionic nature of the ligands.
Biological Implications of Cr(V)-Thiolato Complexes. Al-
though most of the kinetic experiments in the Cr(VI) + RSH
systems were performed in nonaqueous solutions, the effect of
H2O on the reaction kinetics (which is crucial for the Cr(VI)
(39) For instance, two types of Cr(V) intermediates were detected for the
reactions of Cr(VI) with N-(2-mercaptoethyl)acetamide (R′SH), a non-
ionic thiol equally soluble in DMF or H2O (see ref 11). Similarly to
the reactions with RSH (see Results), the [CrVO(SR′)4]- species (giso
) 1.9960; Aiso (53Cr) ) 14.6 × 10-4 cm-1; m/z )-540.0) were
generated by the reactions of Cr(VI) (50 mM) with R′SH (500 mM)
in DMF solutions, and then the reaction mixtures were diluted 100-
fold with either DMF or H2O. The [CrV(O)2(SR′)2]- species (m/z )
-320.3) were detected by ESMS in the both solvents, but EPR
spectroscopic parameters of these species were solvent-dependent: giso
(37) Lay, P. A.; Levina, A. In ComprehensiVe Coordination Chemistry II;
McCleverty, J. A., Meyer, T. J., Eds.; Elsevier: Oxford, UK, 2004;
Vol. 4, pp 313-413.
(38) Odom, A. L.; Mindiola, D. J.; Cummins, C. C. Inorg. Chem. 1999,
38, 3290–3295.
) 1.9842 (Aiso (53Cr) ) 7.8 × 10-4 cm-1) in DMF solutions; or giso
)
1.9858 (Aiso not determined due to the low intensity of the signal) in
H2O solutions (A. Levina, unpublished data).
9
8730 J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010

Formation and Reactivity of Cr(V)-Thiolato Complexes
A R T I C L E S
reactions with thiols in biological media) has also been determined
(Table S1 and Figure S5, Supporting Information). Kinetic analysis
(Tables 2 and S1 [Supporting Information]) showed that H2O
accelerates the redox reactions of both [CrVI(O)3(SR)]- and
[CrVO(SR)4]- (eqs 1 and 4 in Scheme 1), which leads to faster
formation and decomposition of Cr(V) species in the presence of
H2O (Figure S5, Supporting Information). Such observations are
consistent with the faster reactions observed in the aqueous Cr(VI)
+ GSH systems.10 When Cr(VI) is reduced by thiols in fully
aqueous media, thiols and their oxidation products can act as the
sources of C-H bonds for the formation of organic radicals (eqs
4 and 7 in Scheme 1). This feature is probably responsible for the
seemingly paradoxical stabilization of Cr(V)-GSH complexes in
neutral aqueous solutions in the presence of high concentrations
of GSH.40 However, in biological systems, other biomolecules such
as membrane lipids41 or the sugar backbone of DNA42 are more
likely to serve as sources of C-H bonds (similar to the structures
of the intermediates shown in the last line of Scheme 1), which
may lead to the observed stabilization of Cr(V) species in Cr(VI)-
treated cultured cells compared with cell-free aqueous solutions.4
Another aspect of Cr(V) chemistry in biological systems is
the formation of thermodynamically more stable Cr(V) 1,2-
diolato complexes (with carbohydrate and glycoprotein ligands)
from the initial Cr(V)-thiolato intermediates.6-8 The current
work has demonstrated the unusually slow rates of the ligand-
exchange reactions of Cr(V)-thiolato complexes (particularly
of [CrVO(SR)4]-, Figure 5b) with 1,2-diols in nonaqueous
solutions, while the opposite reactions (those of Cr(V)-diolato
complexes with thiols) were relatively fast (minutes time scale
at 295 K, as shown by electronic spectroscopy, Figure S12
[Supporting Information]). Due to their d1 electronic structure,
Cr(V) complexes are regarded as kinetically labile, so that their
ligand-exchange reactions are expected to be complete within
seconds at 295 K.12,37 However, crucially for biological systems,
the ligand-exchange reactions of Cr(V)-thiolato complexes with
1,2-diols are drastically accelerated in the presence of H2O
(Figure S10d [Supporting Information]). Therefore, the inef-
ficient replacement of thiolato ligands with diols in nonaqueous
media is likely to be caused by the thermodynamic and kinetic
barriers for the deprotonation of diols in the absence of H2O.
The other difference between the model system and the in ViVo
chemistry is that the Cr(III) products are most likely to be
adducts bound to proteins and DNA43,44 rather than the insoluble
polymers, but this does not affect the redox chemistry involving
the higher oxidation states.
the giso ≈ 1.986 and 1.979 signals, observed in Cr(VI)-treated
biological systems,4-6 can be assigned to Cr(V) species of the
[CrVOL(SR)2]- and [CrVOL2]- types, respectively (where LH2
is a biological 1,2-diol and RSH is a biological thiol). An
alternative structure of the giso ≈ 1.986 species could be
[CrV(O)2(SR)2]- (such as the giso ) 1.9857 species observed
for the Cr(VI) + GSH reaction in aqueous media, see above),
but these species exhibit very narrow EPR signals (∼0.4 × 10-4
cm-1),10 while both the EPR signals observed in Cr(VI)-treated
biological media usually have medium widths (∼(1-1.5) × 10-4
cm-1; such as those in Figure 5c).4-6 In addition, the kinetics
of the ligand-exchange reactions in the Cr(V)-thiolato/diolato
systems make the formation of detectable amounts of
[CrV(O)2(SR)2]- in the presence of H2O and excess diol unlikely
(Figure S10e [Supporting Information]). A definitive distinction
between [CrV(O)2(SR)2]- and [CrVOL(SR)2]- on the basis of
the Aiso values (Table 1) was impossible for biological systems
due to the low abundance of the 53Cr isotope.8,12 Thus, a reaction
of Cr(VI) with a simple model thiol, used in this work, has
proven extremely useful in the assignment of the nature of Cr(V)
species formed in biological systems. These species are likely
to act as reactive intermediates in both the Cr(VI)-induced
carcinogenesis and the insulin-potentiating activities of Cr(III).44,47
Acknowledgment. Financial support of this work was provided
by Australian Research Council (ARC) Large and Discovery Grants
and Australian Professorial Fellowships to P.A.L. and Wellcome
Trust and ARC LIEF Grants for the ESMS and EPR equipment.
Supporting Information Available: (i) Typical examples of
experimental and simulated EPR spectra, including those used
to check the validity of the calibration method; (ii) typical time-
dependent changes in electronic absorption spectra of the Cr(VI)
+ RSH reaction mixtures; (iii) typical EPR spectroscopic data
for the Cr(VI) + RSH + diol reaction mixtures (in addition to
those included in Figure 1); (iv) typical ESMS data for the
reaction mixtures (in addition to those included in Figure 2);
(v) a list of conditions used in kinetic experiments and the
corresponding rate constants (k1, k4 and k6, see Table 2); (vi)
experimental and calculated kinetic curves for Cr(V) species
(obtained by EPR spectroscopy) for all the kinetic experiments;
(vii) comparison of calculated kinetic curves for the formation
of Cr(V) species under various experimental conditions; (viii)
comparison of kinetic data for Cr(V) complexes obtained by
EPR and electronic absorption spectroscopies; (ix) typical
alternative kinetic schemes (to that shown in Table 2 and
Scheme 1) and the corresponding calculated kinetic curves; (x)
experimental and calculated stoichiometry of the Cr(VI) + RSH
reactions under different conditions; (xi) typical kinetic curves
of Cr(V) decomposition in dilute reaction mixtures in the
presence or absence of 1,2-diols (in addition to those shown in
Figure 5); and (xii) typical changes in electronic absorption
spectra during the reactions of Cr(V) diolato complexes with
RSH. This material is available free of charge via the Internet
at http://pubs.acs.org.
The ligand-exchange reactions of Cr(V)-thiolato complexes
with cyclic cis-1,2-diols (models of carbohydrates, Figures 5b,c
and S10 [Supporting Information]) led to EPR spectra that are
similar to those of living cells and animals, and respiratory
mucus treated with Cr(VI).4-6,45,46 On the basis of these data,
(40) (a) O’Brien, P.; Ozolins, Z. Inorg. Chim. Acta 1989, 161, 261–266.
(b) O’Brien, P.; Pratt, J.; Swanson, F. J.; Thornton, P.; Wang, G. Inorg.
Chim. Acta 1990, 169, 265–269.
JA101675W
(41) (a) Schaich, K. M. Lipids 1992, 27, 209–218. (b) Malle, E.; Marsche,
G.; Arnhold, J.; Davies, M. J. Biochim. Biophys. Acta 2006, 1761,
392–415.
(46) Codd, R.; Lay, P. A.; Tsibakhashvili, N. Ya.; Kalabegishvili, T. L.; Murusidze,
I. G.; Holman, H.-Y. N. J. Inorg. Biochem. 2006, 100, 1827–1833.
(47) (a) Levina, A.; Mulyani, I.; Lay, P. A. In The Nutritional Biochemistry
of Chromium(III); Vincent, J. B., Ed.; Elsevier: Amsterdam, 2007;
pp 225-256. (b) Mulyani, I.; Levina, A.; Lay, P. A. Angew. Chem.,
Int. Ed. 2004, 43, 4504–4507. (c) Nguyen, A.; Mulyani, I.; Levina,
A.; Lay, P. A. Inorg. Chem. 2008, 47, 4299–4309.
(42) Levina, A.; Barr-David, G.; Codd, R.; Lay, P. A.; Dixon, N. E.;
Hammershøi, A.; Hendry, P. Chem. Res. Toxicol. 1999, 12, 371–381.
(43) Macfie, A.; Hagan, E.; Zhitkovich, A. Chem. Res. Toxicol. 2010, 23, 341–
347.
(44) Levina, A.; Lay, P. A. Chem. Res. Toxicol. 2008, 21, 563–571.
(45) Codd, R.; Lay, P. A. J. Am. Chem. Soc. 2001, 123, 11799–11800.
9
J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010 8731