Electrochemistry and spectroscopy of sulfate and thiosulfate complexes of iron porphyrins

Article abstract of DOI:10.1016/S0020-1693(01)00665-X

The electrochemical and spectroscopic properties of the complex formed by the addition of thiosulfate to ferric porphyrins were examined. The NMR spectrum of the thiosulfate-ferric porphyrin complex was consistent with a high-spin ferric complex, while the EPR spectrum at liquid nitrogen temperatures indicated that the complex under these conditions was low-spin. Such behavior has been previously observed for other ferric porphyrin complexes. The visible spectra were characterized by a shift in the Soret band to higher energies, with smaller changes in the longer wavelength region. The complex was reasonably stable in DMF, but slowly reduced over several hours to Fe^II(TPP) and S₄O₆^2-. The voltammetric behavior of the thiosulfate complex in DMF consists of two waves, the first of which was irreversible. The ferric/ferrous reduction in the presence of thiosulfate was shifted negatively about 400 mV, compared to the Fe(TPP)(Cl) reduction. The visible, NMR and EPR spectra were most consistent with a Fe-S bonded ferric porphyrin-thiosulfate complex, Fe(P)(S-SO₃)^-. The kinetics of the reduction of ferric porphyrin by thiosulfate in DMSO indicated an autocatalytic mechanism, where the first step is the formation of the catalyst. The identity of the catalyst could not be determined because it must be present at low concentrations, but it is formed from the reaction of the ferric complex with thiosulfate. Coordination of thiosulfate to the porphyrin was not necessary for the reduction to occur, and the reduction of Fe(TPP)(Cl) by thiosulfate was accelerated by the addition of sulfate. Under these conditions, sulfate had replaced thiosulfate as the axial ligand for the ferric porphyrin. In the presence of sulfate, the reduction occurred in a single kinetic pseudo-first order step.

Full text of DOI:10.1016/S0020-1693(01)00665-X

www.elsevier.com/locate/ica
Inorganica Chimica Acta 328 (2002) 13–22
Electrochemistry and spectroscopy of sulfate and thiosulfate
complexes of iron porphyrins
Philip W. Crawford 1, Michael D. Ryan *
Department of Chemistry, Wehr Chemistry Building, Marquette Uni6ersity, PO Box 1881, Milwaukee, WI 53201-1881, USA
Received 4 May 2001; accepted 26 July 2001
Abstract
The electrochemical and spectroscopic properties of the complex formed by the addition of thiosulfate to ferric porphyrins were
examined. The NMR spectrum of the thiosulfate–ferric porphyrin complex was consistent with a high-spin ferric complex, while
the EPR spectrum at liquid nitrogen temperatures indicated that the complex under these conditions was low-spin. Such behavior
has been previously observed for other ferric porphyrin complexes. The visible spectra were characterized by a shift in the Soret
band to higher energies, with smaller changes in the longer wavelength region. The complex was reasonably stable in DMF, but
slowly reduced over several hours to FeII(TPP) and S4O62−. The voltammetric behavior of the thiosulfate complex in DMF
consists of two waves, the first of which was irreversible. The ferric/ferrous reduction in the presence of thiosulfate was shifted
negatively about 400 mV, compared to the Fe(TPP)(Cl) reduction. The visible, NMR and EPR spectra were most consistent with
a Fe–S bonded ferric porphyrin–thiosulfate complex, Fe(P)(SꢀSO3)−. The kinetics of the reduction of ferric porphyrin by
thiosulfate in DMSO indicated an autocatalytic mechanism, where the first step is the formation of the catalyst. The identity of
the catalyst could not be determined because it must be present at low concentrations, but it is formed from the reaction of the
ferric complex with thiosulfate. Coordination of thiosulfate to the porphyrin was not necessary for the reduction to occur, and
the reduction of Fe(TPP)(Cl) by thiosulfate was accelerated by the addition of sulfate. Under these conditions, sulfate had
replaced thiosulfate as the axial ligand for the ferric porphyrin. In the presence of sulfate, the reduction occurred in a single kinetic
pseudo-first order step. © 2002 Elsevier Science B.V. All rights reserved.
Keywords: Sulfate complexes; Thiosulfate complexes; Iron porphyrins; Electrochemistry; Spectroelectrochemistry; Kinetics
dehydrogenase [3]). The importance of these transfor-
mations in nature has lead us to investigate the reac-
1. Introduction
tions and electrochemistry of iron porphyrins in the
Enzymes that metabolize small molecular compounds
presence of thiosulfate and sulfate. These ligands fit in
of nitrogen and sulfur play an important role in the
with our general interest in the electrochemistry of
biosynthesis of amino acids by plants or as a bacterial
complexes with electroactive ligands. Previously we
source of energy. In particular, interest has grown
have reported on the electrochemistry of iron por-
substantially in microbial sulfur metabolism, as exem-
phyrins with nitrite [4], hydroxylamine [5], and nitric
plified by the book edited by Peck and LeGall [1].
oxide [6,7].
Many of the enzymes that are in this system are not yet
Ligands which are easily oxidized are known to
well characterized, but several are known to contain
readily reduce ferric porphyrins, probably after com-
heme prosthetic groups (e.g. thiosulfate dehydrogenase
plexation. For example, piperidine [8–10], cyanide
from T. acidiphilus [2], sulfite reductase, and sulfite
[9,11,12], alkyl thiols [9,13,14], amines [15], te-
trathiomolybdate (tungstate) [16], and hydroxide have
all been shown to induce the reduction of iron(III). In
general, with the exception of hydroxide, the reduction
appears to involve coordination of the ligand species
(L) to the Fe(III) center followed by intramolecular
* Corresponding author. Tel.: +1-414-288 7065; fax: +1-414-288
7066.
E-mail address: michael.ryan@marquette.edu (M.D. Ryan).
1 Present address: Department of Chemistry, Southeast Missouri
State University, Cape Girardeau, MO 63701, USA.
0020-1693/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S0020-1693(01)00665-X

14
P.W. Crawford, M.D. Ryan / Inorganica Chimica Acta 328 (2002) 13–22
one-electron transfer, via homolytic cleavage of the
Fe(III)ꢀL bond [10,12]. Radical species have been de-
tected in some cases [9–11].
(Fe(TPP)(Cl)) [28] and [Fe(TPP)]2(SO4) [18] were syn-
thesized by literature procedures.
Bis(tetraphenylphosphonium) thiosulfate [29] and
bis(tetraphenylphosphonium) tetrathionate were pre-
pared via the metathesis reaction between the sodium
salt of the anion and tetraphenylphosphonium chloride
(Ph4PCl). The difference infrared spectrum between
Ph4PCl and (Ph4P)2S2O3 gave bands at 548, 670, 1006
and 1134 cm−1, matching those observed for sodium
thiosulfate and those calculated for the thiosulfate ion
[15,30]. The infrared spectrum for (Ph4P)2S4O6 exhib-
ited bands which were similar to those observed for
sodium tetrathionate [31] (592, 629, 1014, 1230 br, and
1245 cm−1). Bis(triphenylphosphoranylidene) ammo-
nium sulfate, (PNP)2(SO4), was prepared from a litera-
ture procedure [32]. Chemical reduction of ferric
porphyrins were carried out using the zinc amalgam
method [33,34].
Recently, interest has developed in this laboratory
and others in the interactions between iron porphyrins
and sulfur-oxy anions in non-aqueous solvents. The
first report from this laboratory has been published
[17]. Dianions such as sulfate were found to form
dimeric ferric porphyrin complexes in non-coordinating
solvents such as methylene chloride [17–20]. Three
reduction waves were observed in cyclic voltammetry
due to the reduction of the dimer, the sulfate complex,
and the ferrous porphyrin.
[FeIII(P)]2(SO4)+e− X FeIII(P)(SO4)− +FeII(P)
(Wave Ia)
FeIII(P)(SO4)− +e− X FeII(P)+SO42−
FeII(P)+e− X FeI(P)−
(Wave Ib)
(Wave II)
In DMSO, the visible spectroscopy and cyclic
voltammetry indicated that the dimer was dissociated,
and, as a result, Wave Ib was significantly attenuated
[17]. Addition of sulfate to a solution of the dimer or
Fe(P)(Cl) gave rise to two waves: Wave Ib and Wave
II.
This work was continued with an investigation of the
reaction with iron porphyrins and thiosulfate in non-
aqueous solvents. Thiosulfate is known to be a reducing
agent for Fe(III) complexes [21–24]. Chen et al. [25–27]
have recently investigated the oxidation of thiosulfate
in an aqueous solution by water soluble iron and cobalt
porphyrins, as well as the photocatalytic reaction. The
work in this report involves the electrochemistry, spec-
troscopy, and spectroelectrochemistry of thiosulfate
with iron porphyrins. In order to clarify some of the
questions raised by the thiosulfate results, the chemistry
and electrochemistry of the ferric–sulfate complexes in
DMSO and DMF were re-examined.
2.2. Methods
Electrochemical data were obtained with either an
ECO 553 potentiostat with an EG&G Princeton Ap-
plied Research (PARC) 175 waveform generator or an
IBM EC/225 voltammetric analyzer. Voltammetric data
were recorded on a Hewlett–Packard 7045 X-Y
recorder. Scan rates ranged from 5 to 200 mV s−1. A
three-electrode IBM cell was used for the cyclic voltam-
metric experiments, consisting of Sargent–Welch plat-
inum flag electrodes as the working and auxiliary
electrodes, and a Ag/AgNO3 (0.1 M in acetonitrile)
reference electrode (SRE). The SRE was used as the
reference electrode in all electrochemical experiments.
The reference electrode was separated from the por-
phyrin solution by either an asbestos plug or a porous
Vicor tip salt bridge filled with the appropriate solvent
and supporting electrolyte (0.1 M). All solutions were
deoxygenated either by bubbling the solution for 15
min with prepurified dinitrogen or by performing the
experiments under an argon atmosphere in an HE-43-2
Dri-Lab glove box. The inert atmosphere was main-
tained over the solutions during all measurements. Be-
tween scans the electrodes were disconnected to prevent
any back oxidation of ferrous porphyrin to ferric por-
phyrin. An optically transparent thin layer electrode
(OTTLE) cell [35] was used for the spectroelectrochem-
ical studies.
2. Experimental
2.1. Materials
Tetraphenylporphyrin (H2TPP), octaethylporphyrin
iron(III) chloride (Fe(OEP)(Cl)), tetraphenylphospho-
nium chloride, dimethylformamide (DMF), dimethyl
sulfoxide (DMSO), DMF-d7, DMSO-d6, toluene, tolu-
ene-d8, and sodium tetrathionate dihydrate were pur-
chased from Aldrich Chemical Co. in the highest purity
available. All solvents were used without further purifi-
cation. The supporting electrolyte, tetrabutylammo-
nium perchlorate (TBAP), was obtained from G.F.
Smith Chemical Co., and was used as received. Sodium
thiosulfate was obtained from Fisher Scientific Co.
All visible spectra were obtained on either a Perkin–
Elmer 320 UV–Vis spectrophotometer with a PE 3600
data station or a Hewlett–Packard 8452A diode array
spectrophotometer. Room temperature (r.t.) spectra
were obtained with 1 cm, 1 or 2 mm quartz cells. A GE
Omega 300 MHz NMR spectrometer was used to ob-
tain NMR spectra. Infrared samples were prepared as
KBr pellets and the spectra obtained with an Analect
FX-6200 FT infrared spectrometer. All mathematical
Tetraphenylporphyrinato
iron(III)
chloride

P.W. Crawford, M.D. Ryan / Inorganica Chimica Acta 328 (2002) 13–22
15
operations in the kinetics section of the work were
The addition of SO42− to solutions of the sulfate
dimer or Fe(TPP)(Cl) in both coordinating and non-co-
ordinating solvents readily led to the formation of the
Fe(TPP)(SO4)− complex. The sulfate complex was
characterized by a Soret band at 417–426 nm (depend-
ing upon the solvent) [17], and additional visible bands
around 565 and 615 nm. Even in DMSO, complete
coordination of the ferric complex was observed with
about 1–2 mM sulfate. At lower concentrations, some
dissociation was observed.
carried out using MATLAB. The PLS Toolbox 2.0
–
(Eigenvector Research, Inc.) algorithms for MATLAB
were used for the chemometric analysis (factor analysis
and multivariate curve resolution).
Product isolation studies for the reaction between
Fe(P)(Cl), where P=TPP or OEP, and (Ph4P)2S2O3
were performed by mixing equimolar amounts of each
reactant in minimal amounts of solvent, DMSO or
DMF, and allowing the reactions to proceed under
argon for varying lengths of time (5–17 h). Large
excesses of toluene or toluene–benzene were then
added and the reaction mixtures left for varying lengths
of time (12 h–5 days) in order to precipitate any
non-porphyrin products. The crystals were then washed
with toluene and dried under argon. Unless noted
otherwise, all solutions containing Fe(P)(Cl) and
(Ph4P)2S2O3 were prepared under an inert atmosphere.
The visible spectra of FeIII(P)(X), where P=TPP or
OEP and X=S2O32− were examined. Because thiosul-
fate is not stable in halogenated hydrocarbons [36], the
majority of the thiosulfate experiments were carried out
in solvents such as DMF and DMSO. All spectroscopy
studies in halogenated solvents were performed soon
after mixing before any reaction with the solvent was
important.
The most significant change in the visible spectrum of
DMF solutions of Fe(TPP)(Cl) in the presence of thio-
sulfate occurred in the Soret region. The band at 416
nm shifted to a broad band at 402 nm in the presence
of thiosulfate (Fig. 1 and Table 1). Relatively smaller
changes were observed in the longer wavelength region
with the most significant change being the attenuation
of the 650–700 nm bands (see Fig. 1). In DMF after
30–60 min, the Soret band due to FeII(TPP) could be
observed under anaerobic conditions. The reaction was
much faster when DMSO was the solvent, with all the
ferric porphyrin being converted to ferrous porphyrin
in about 5 min. The visible spectra of the ferric–OEP
complexes with thiosulfate in DMF and ethylene chlo-
ride (starting with Fe(OEP)(ClO4)) are also summarized
in Table 1. As with Fe(TPP)(Cl), the Soret bands for
3. Results and discussion
3.1. Visible and NMR spectroscopy of ferric porphyrin
complexes
The visible spectra of Fe(P)(SO4)− and [Fe(P)]2(SO4)
in various solvents are shown in Table 1. The sulfate
dimer was stable in non-coordinating solvents such as
methylene chloride. Some dissociation of the dimer
probably occurred in DMSO, especially at low concen-
trations of the ferric porphyrin. In DMF, the sulfate
dimer was probably present, as indicated by the shift in
the Soret band to 398 nm [17], as compared to the Soret
band for Fe(TPP)(NO3) in DMF which appears at 380
and 416 nm and for Fe(TPP)(SO4)− at 426 nm.
Table 1
Visible spectra of ferric porphyrins with anions
Compound
Fe(TPP)(Cl)
Solvent
Visible bands, nm (m M−1 cm−1
)
Reference
DMF
CH2Cl2
DMF
380 (50), 416 (95), 508 (12), 572 (4), 690
380s(59), 417 (110), 511 (13.4), 577s(3.3), 658s, 690 (3.2)
402 (95), 508 (15.6), 568, 606s
[37]
[38]
this work
[17]
Fe(TPP)(S2O3)−
Fe(TPP)(SO4)−
DMF
425 (187), 567 (15.8), 609 (7.8), 660s
DMSO
DMF
426 (272), 550s, 566 (15.3), 619 (80)
400 (93), 501 (8.2), 530 (10.6), 693 (4.3)
416, 490, 526, 571s, 617, 678
347 (43), 372 (51), 408 (113), 508 (11.0), 576 (3.5), 680 (2.9)
408 (99.5), 513 (9.5), 570 (2.4), 608 (1.4), 650 (1.2), 675 (1.3)
390 (116.7), 454 (7.3), 496 (8.8), 530s, 572 (4.4), 616 (5.9)
380 (88.4), 504 (9.1), 532 (9.1), 630 (5.2)
381 (85.1), 461 (5.75), 508 (7.4), 536 (7.76), 584 (2.14), 640 (3.80)
380 (113), 500 (9.2), 633 (3.2)
374, 502, 540s, 589s, 629
374, 503, 532s, 585s, 624s
377 (117), 504 (9.2), 533 (9.4), 583s, 634 (4.6)
387 (65), 500 (6.2), 530 (5.9), 627 (2.6)
391 (79), 473 (11), 582 (4.1)
[17]
[17]
this work
[18]
[39]
this work
this work
[41]
[Fe(TPP)]2SO4
DMSO
CH2Cl2
toluene
DMSO
DMF
CH2Cl2
CH2Cl2
DMSO
DMF
ethylene chloride
CHCl3
CHCl3
CHCl3
Fe(TPP)(ClO4)
Fe(OEP)(Cl)
Fe(OEP)(ClO4)
[40]
Fe(OEP)(S2O3)−
this work
this work
this work
this work
this work
[17]
[Fe(OEP)]2(SO4)
Fe(OEP)(SO4)−
Fe(OEP)(HSO4)
382 (77.9), 504 (7.8), 530s, 636 (1.0)

16
P.W. Crawford, M.D. Ryan / Inorganica Chimica Acta 328 (2002) 13–22
While the complex was high-spin at room tempera-
ture, the EPR evidence indicates that the complex shifts
to a low-spin complex in the frozen state at 77 K. In
DMF, the complex shifts completely to the low-spin
complex with g-values of 2.41, 2.27 and 1.93, which is
similar to the g-values of other ferric porphyrins with
thiolate ligands (Fe(OEP)(SPh)(DMF): 2.42, 2.27 and
1.92 [43]). The intensity of the high-spin EPR signal
decreased linearly with the concentration of thiosulfate,
reaching zero intensity with a 1:1 ferric porphyrin–thio-
sulfate ratio. As the high-spin signal decreased, the
low-spin signal appeared, with no evidence of an EPR-
silent thiosulfate bridged dimer.
The formation of the thiosulfate complex is thermo-
dynamically favored in that only a small excess of
thiosulfate is needed to form the spectrum. The most
characteristic feature of the visible spectrum in DMF is
the shift in the Soret band to higher energy (416–400
nm). This is contrary to the shift observed for the
Fe(P)(SO4)− complexes (416–426 nm). On the other
hand, iron–sulfur coordination has been observed to
shift the Soret band to higher energies, compared to
halide coordination [44–46]. The shift in the Soret band
to shorter wavelengths in the presence of thiosulfate
was unlikely to be due to dimer formation because the
dimer would not form in the presence of excess thiosul-
fate, and the sulfate studies showed that the dimer did
not form if chloride was also present (starting with
Fe(TPP)(Cl)).
Fig. 1. Solid line, visible spectrum of 0.10 mM Fe(TPP)(Cl) with 0.71
mM (Ph4P)2(S2O3) in DMF. Dashed line, visible spectrum obtained
by the oxidation of 0.50 mM FeII(TPP) at −0.30 V vs. Ag/AgNO3 in
DMF in the presence of 3.8 mM (Ph4P)2(S2O3).
the ferric–OEP complex shifted to shorter wavelengths
when thiosulfate was added. The spectrum of the
Fe(OEP)–thiosulfate complex in ethylene chloride
(starting with Fe(OEP)(ClO4)) in the 500–700 nm re-
gion was quite similar to Fe(OEP)(Cl) (Table 1).
The proton NMR spectra of Fe(TPP)(Cl) and
Fe(OEP)(Cl) in the presence and absence of thiosulfate
are summarized in Table 2. In both DMF-d7 and
DMSO-d6–toluene-d8 (at 0 °C), the proton NMR spec-
tra in the presence and absence of thiosulfate are
consistent with a high-spin ferric complex [17,18],
though small changes in the chemical shifts were ob-
served when thiosulfate was present. For example, the
pyrrole resonance of Fe(TPP)(Cl) in DMF-d7 was
shifted from 82.5 to 77.0 ppm when thiosulfate was
added. Similar shifts were observed in DMSO-d6–tolu-
ene-d8 at 0 °C. In the presence of thiosulfate, the
pyrrole resonance (77.0 ppm) was between that of the
sulfate dimer and Fe(TPP)(Cl).
3.2. Voltammetry of ferric porphyrin–sulfate and
thiosulfate complexes
The cyclic voltammetry of the sulfate dimer and
monomer in coordinating and non-coordinating sol-
vents has already been examined [17]. If there is an
Table 2
Proton NMR spectrum of ferric porphyrins
Compound
Solvent
l (ppm)
Reference
Pyrrole
Phenyl
Methylene
Fe(TPP)(Cl)
DMF-d7
DMSO-d6
DMF-d7
DMSO-d6
DMF-d7
DMF-d7
CDCl3
82.5
86.3
77.0
76.6
80.9
67.0
69.6
12.2, 13.2
9.9, 11.7, 12.7
11.4, 12.8
This work
This work
This work
This work
This work
This work
[18]
a
a
Fe(TPP)(S2O3)−
11.4, 12.9
Fe(TPP)(SO4)−
[Fe(TPP)]2(SO4)
13.0, 11.9, 10.3
13.7, 10.2, 10.1, 8.0
12.3, 10.5, 6.5
Fe(OEP)(Cl)
DMF-d7
CDCl3
DMF-d7
CDCl3
CDCl3
CDCl3
40.9, 43.7
39.6, 43.1
33.3, 37.3
38.8, 49.8
37.9
This work
[18]
This work
[17]
[18,42]
[17]
Fe(OEP)(S2O3)−
Fe(OEP)(SO4)−
[Fe(OEP)]2SO4
Fe(OEP)(HSO4)
39.7, 49.8
a 33% DMSO-d6-toluene-d8, 0 °C.

P.W. Crawford, M.D. Ryan / Inorganica Chimica Acta 328 (2002) 13–22
17
of the voltammetry is the irreversible electron transfer
kinetics of Wave Ib, as is indicated by DEp values
between 700 and 800 mV. This wave corresponds to the
reduction–oxidation of the ferric–thiosulfate complex.
FeIII(P)(S2O3)− +e− X FeII(P)+S2O32−
Wave Ib
The origin of such slow electron transfer kinetics is
the loss–gain of the axial ligand (thiosulfate) during the
redox process. The E1/2 value for Wave Ib can be
estimated based on the average of the anodic and
cathodic peak potentials. From the shift in the E1/2
values, a formation constant (KTS,Cl) for the displace-
ment of chloride by thiosulfate from the ferric por-
phyrin complex is estimated to be 4×103.
Unfortunately, the narrow range of thiosulfate solubil-
ity (up to 3–4 mM) prevents an accurate determination
of this value.
Fig. 2. (A) Cyclic voltammetry of 0.50 mM [Fe(TPP)]2(SO4) in DMF
with 10 mM (PNP)2(SO4). Scan rate, 100 mV s−1; (B) cyclic voltam-
metry 0.50 mM Fe(TPP)(Cl) in the presence of 2 mM (Ph4P)2(S2O3)
in DMF. Scan rates, 50, 100, 200 mV s−1. Working and auxiliary
electrodes were platinum. Supporting electrolyte, 0.10 M TBAP.
Fe(TPP)(Cl)+S2O32− X Fe(TPP)(S2O3)− +Cl−
excess of sulfate, the first wave in all solvents was the
reduction of the ferric sulfate monomer to the ferrous
porphyrin, while the second wave was the reduction of
the ferrous porphyrin. The first wave was quite quasi-
reversible with DEp values of about 400 mV (Fig. 2(A)),
even though the sulfate remains complexed with the
ferrous porphyrin at high concentrations of sulfate (see
spectroelectrochemical data below).
When a stoichiometric amount of thiosulfate was
added to a solution of Fe(TPP)(Cl) in DMF, Wave Ia
almost completely disappeared and a new wave (Wave
Ib) occurred with an Epc value of about −1.1 V (Fig.
2(B)), which was typical of the behavior of ferric por-
phyrins with sulfate. The appearance of the new wave
correlated with the changes in the visible and NMR
spectra, as reported above. Thiosulfate readily dis-
placed chloride in DMF, as indicated by the appear-
ance of Wave Ib when substoichiometric amounts of
thiosulfate were added (Table 3). The significant feature
KTS,Cl
Because Fe(TPP)(Cl) was rapidly reduced by thiosul-
fate in DMSO, the voltammetry was carried out using
FeII(TPP) which was obtained by the zinc amalgam
reduction of Fe(TPP)(Cl) in toluene. The results were
analogous to those observed in DMF, with the disap-
pearance of Wave Ia when stoichiometric amounts of
thiosulfate were present.
The fact that no wave was observed at Wave Ia when
the concentration of thiosulfate exceeded stoichiometric
amounts indicates that the thiosulfate coordination
with FeIII(P) is stronger than sulfate coordination in
DMF. The similarity between the cyclic voltammetry of
Fe(P)(SO4)− in methylene chloride and the ferric–thio-
sulfate complex in DMF or DMSO is consistent with
a 1:1 ferric porphyrin–thiosulfate complex, Fe(P)-
(S2O3)−.
Table 3
Cyclic voltammetry of Fe(TPP)(Cl) in the presence of thiosulfate ion in DMF a
[S2O23−] (mM)
Wave Ia
Wave Ib
Wave II
b
b
b
E1/2
ipc (mA)
Epc
ipc (mA)
E1/2
ipc (mA)
0
−0.60
−0.60
−0.61
−0.61
77
62
44
19
−1.48
−1.48
−1.48
−1.47
−1.48
−1.48
−1.49
−1.47
84
103
98
95
90
108
83
78
0.13
0.25
0.50
0.75
1.0
−1.01
−1.06
−1.08
−1.17
−1.08
−1.18
−1.21
0.5
16
46
54
72
1.5
2.5
56
59
a [Fe(TPP)(Cl)]=0.50 mM, [TBAP]=0.10 M, platinum flag working electrode, scan rate=100 mV s−1, S2O32− added as the Ph4P+ salt.
b V vs. 0.10 M Ag/AgNO3 in CH3CN.

18
P.W. Crawford, M.D. Ryan / Inorganica Chimica Acta 328 (2002) 13–22
absence of added sulfate, only the Soret band for
Fe(TPP) can be observed). The longer wavelength re-
gion is also consistent with sulfate coordination, though
the shifts in this region are less dramatic (564 and 606
nm–569 and 609 nm, respectively).
A similar set of experiments was carried out using
DMSO as the solvent. The most significant difference
was the change in the Soret region. Two Soret bands
were observed at 403 and 426 nm. The longer wave-
length region was quite similar to the DMF results,
though some peaks appeared as shoulders. The same
spectrum can be obtained at 0 °C, using a DMSO–
DMF mixed solvent.
3.4. Kinetic study of the reduction of FeIII(P)(Cl) by
thiosulfate
Fig. 3. Visible spectra obtained during the reduction of Fe(TPP)(Cl)
by thiosulfate in DMSO. [Fe(TPP)(Cl)]=0.0355 mM; [S2O32−]=
0.88 mM.
As was indicated above, ferric porphyrins are re-
duced by thiosulfate. The reaction rate is substantially
faster in DMSO than in DMF. The reaction can be
conveniently followed by visible spectroscopy. In
DMSO, the Soret band for FeII(TPP) was observed
soon after mixing the ferric porphyrin with thiosulfate
(Fig. 3). The predominant feature in the absorbance–
time curves was a short induction period, which oc-
curred immediately after mixing. The length of the
induction period depended upon temperature, and was
substantially longer at 0 °C (in a mixed DMF–DMSO
solution). The presence of an induction period was
consistent with a multi-step mechanism.
The analysis of the absorbance changes as a function
of time was carried out utilizing two chemometric
techniques: evolving factor analysis and multivariate
curve resolution (MCR) [48]. The first approach, (win-
dow) evolving factor analysis, is a useful technique for
determining the number of absorbing components
present at any time during the experiment. A typical set
of eigenvalues obtained using a 10-s window are shown
in Fig. 4. These results were consistent with the pres-
ence of only two absorbing components during the
reaction time (the very small third factor at the very
beginning of the experiment is an artifact; good isos-
bestic points are observed in that region). The same
conclusion was obtained for different concentrations of
ferric porphyrin and thiosulfate. Because all of these
experiments were carried out with an excess of thiosul-
fate, the two spectral species were FeIII(TPP)(S2O3)−
and FeII(TPP) (the latter species could be clearly iden-
tified at the end of the reaction).
Fig. 4. Window evolving factor analysis of the visible spectra for the
reduction of Fe(TPP)(Cl) in DMSO by thiosulfate. [Fe(TPP)(Cl)]=
0.0355 mM. [S2O32−]=1.4 mM.
3.3. Spectroelectrochemistry of Fe(P)(Cl) in the
presence of thiosulfate
The spectroelectrochemical oxidation of FeII(TPP) in
the presence and absence of thiosulfate was examined
in DMF. In the absence of thiosulfate, the oxidation of
FeII(TPP) in DMF yielded the spectrum for
FeIII(TPP)+ with bands at 416, 499, 530, 650s and 690
nm [47]. The original spectrum was regenerated by
stepping the potential back to the original value,
−1.30 V. Repeating the oxidation in the presence of
3.8 mM thiosulfate resulted in a spectrum with bands at
401, 424, 514 and 570 s nm (Fig. 1). The original
FeII(TPP) spectrum was obtained when the potential
was returned to −1.3 V. There was no evidence for
coordination between thiosulfate and FeII(TPP) in
DMF. This contrasts strongly with the spectroelectro-
chemistry of [Fe(TPP)]2(SO4) in the presence of 4 mM
sulfate. In this case, the Soret band for a 5-coordinate
anion complex can be observed (umax=438 nm), along
with the neutral Fe(TPP) band at 429 nm (in the
The spectra of the individual species were determined
using multivariate curve resolution (MCR). This tech-
nique allows one to extract the concentration profiles
and component spectra with as few assumptions about
the data as possible. A typical set of results are shown
in Fig. 5. In obtaining these results, it was assumed that
the concentrations and spectra of all the components

P.W. Crawford, M.D. Ryan / Inorganica Chimica Acta 328 (2002) 13–22
19
were non-negative. In addition, the concentrations of
FeII(TPP) and FeIII(TPP)(S2O3)− at long times were
equal to the initial concentrations of the ferric por-
phyrin and zero, respectively. The spectrum of species
two was clearly equal to FeII(TPP), and the spectrum of
species one was consistent with the spectrum of
Fe(TPP)(S2O3)− obtained by spectroelectrochemistry.
The concentration profiles for the ferric porphyrin with
different concentrations of thiosulfate are shown in Fig.
6. The predominant feature in all these deconvolutions
is that the induction period could be observed for both
the ferric species and the ferrous species. In addition,
the loss of the ferric porphyrin was equal to the gain in
the ferrous species, indicating that the final step in the
reaction was the reduction of FeIII(TPP)(S2O3)− to
FeII(TPP). Clearly, the initial step in this mechanism
does not involve the reduction of a significant amount
of ferric porphyrin, though the ferric porphyrin must be
present in order to initiate the reaction. Finally, while
the reaction rate does increase with the concentration
Fig. 7. Variation in the concentration of the ferric porphyrin species
as a function of time. [S2O32−]=0.44 mM. [FeIII(TPP)+]: (ꢀ)
0.00995 mM; (ꢁ) 0.0207 mM; (ꢂ) 0.041 mM; (2) 0.071 mM; (“)
0.101 mM.
of thiosulfate, it becomes less dependent upon thiosul-
fate at high concentrations. The concentration profile
for different concentrations of ferric porphyrin (main-
taining the concentration of thiosulfate constant) are
shown in Fig. 7. The rate of the reaction was clearly
dependent upon the concentration of ferric porphyrin,
decreasing substantially at high concentrations of ferric
porphyrin. It is unlikely that a dimeric species is present
at significant concentrations because of the use of
strongly coordinating solvent (DMSO) and the pres-
ence of reasonably strong ligands such as chloride and
excess thiosulfate.
Attempts to fit the kinetic data to a classical two-step
mechanism below were unsuccessful:
k
k
1
A X B“2 C
(1)
k
−1
While reasonable fits to the experimental data could be
obtained over a limited range of concentrations, it was
difficult to obtain reproducible rate constants or induc-
tion periods that were consistent with the data. A more
complex mechanism would be required to explain the
data.
Fig. 5. Variation in the concentration of the ferrous and ferric
porphyrin as a function of time as calculated by the multivariant
curve resolution method. [Fe(TPP)+]=0.0355 mM; [S2O32−]=0.88
mM. Solvent=DMSO.
Results similar to ours were reported by
Kalinichenko et al. [49], who investigated the reduction
of aqueous solutions of Fe(III) by thiosulfate. The
mechanism that they proposed was the initial genera-
tion of a catalyst that would oxidize thiosulfate. This
catalyst was then regenerated by the ferric ion:
Fe3+ +S2O32− “C+products
Fe3+ +C− “Fe2+ +C
(2)
(3)
(4)
(5)
S2O32− +C“S2O3− +C−
S2O3− +FeS2O3+ “S4O62− +Fe2+
Fig. 6. Variation in the concentration of the ferric porphyrin species
as a function of time. [Fe(TPP)+]=0.0355 mM. [S2O32−]: (ꢀ) 0.11
mM; (ꢃ) 0.59 mM; (ꢄ) 0.88 mM; (ꢅ) 1.4 mM; (ꢂ) 1.8 mM. k3k4
The ferric species in reactions (2) and (3) could also
be coordinated with thiosulfate. In addition, S2O3−
could react with S2O32− to form S4O63−, a more potent
reducing agent than S2O3−, which could be the reduc-
[C]=0.88 M−1
s
−2, k3/k4=32; (“) 9.8 mM sulfate, 0.38 mM
thiosulfate, 0.048 mM Fe(TPP)+. Line is theoretical line for a
pseudo-first reaction with k=0.026 s−1
.

20
P.W. Crawford, M.D. Ryan / Inorganica Chimica Acta 328 (2002) 13–22
tant in reaction (5) [50]. If we assume that the rate
constant for reaction (5) (k5) is large, the following rate
law can be obtained:
While the aqueous reduction potential for S2O32−
/
S4O62− has been found to be +0.08 V [56], the redox
potential for the S2O3− –S2O32− couple is much less
favorable (−1.30 V) [57]. The reaction is driven by the
coupling of S2O3− with S2O32−. The high potential for
the initial one-electron step is compensated for by the
low kinetic barrier for the self-exchange reaction [57].
The fast electron exchange rate favors the involvement
of a catalyst, via the mechanism described in reactions
(2)–(4). While the catalytic mechanism is consistent
with the kinetic data, the identity of the catalyst is less
clear. Kalinichenko et al. [49] postulated that the SꢀS−
couple, formed by the hydrolysis of thiosulfate, was the
catalytic couple.
d[Fe3+
]
k3k4[C][S2O32−][Fe3+
k3[Fe3+]+k4[S2O32−
]
]
=
(6)
dt
where k3 and k4 are the rate constants for reactions (3)
and (4), respectively. It will be assumed that C is
present at steady-state after an initial rise, and it is only
formed upon the mixing of thiosulfate with iron(III)
(reaction (2)) with a rate constant, k2. The induction
period was primarily determined by the rate at which
the catalyst C was generated. With this assumption, the
values of k2, k3 and k4 can be determined by using a
range of concentrations of ferric porphyrin and thiosul-
fate. The best fit for all the experimental data is shown
in Figs. 6 and 7 using the following set of kinetic
parameters: k3k4[C]=0.88 M−1 s−2 and k3/k4=32.
Without a knowledge of the identity or concentration
of the catalyst, C, only relative values of the rate
constant can be determined. The relative values,
though, are useful in providing a quantitative way to
evaluate the mechanism.
The mechanism described in reactions (2)–(4) gives
rise to a complex relationship between the rate con-
stant(s) and the concentrations of thiosulfate and ferric
porphyrin. In particular, the observed reaction rates
appear to decrease as the ferric porphyrin concentra-
tion increases, and the increase in the reaction rate with
thiosulfate is less than first order. This is consistent
with Eq. (6) in that both the ferric concentration and
thiosulfate is in the denominator of the equation. At
high concentrations of either reactant, the reaction
becomes zero order. This is because of the catalytic
nature of the reaction mechanism, in that both reac-
tions (reactions (3) and (4)) must be fast for the overall
reaction to be fast.
The autoreduction of ferric porphyrins in DMSO has
been noted for a number of anions, including hydroxide
[51,52], methoxide [53], cyanide [11,12], and mercap-
tides [14]. While the kinetics of the autoreduction has
not been examined for all of these species, kinetic
analysis of the cyanide reduction of Fe(TPP)+ indi-
cated that the reduction was pseudo-first order in cya-
nide [11,12]. These reactions have not been observed in
solvents other than DMSO [51]. The mechanism of the
metal reductions by thiosulfate is often more complex
than pseudo-first order reduction. Early work by Page
[54] indicated that the reduction of aqueous ferric ion
by thiosulfate was half-order in thiosulfate and 3/2
order in ferric ion. The reduction of (aq.) Co(III)
oxalate by thiosulfate was sluggish without trace
amounts of Cu(II) [55], which lead to a catalytic mech-
anism quite similar to reactions (2)–(4) with the CꢀC−
couple being Cu(II)ꢀCu(I).
Fe(S2O3)++H2O“Fe2+ +SO42− +2H++S−
Because only a small amount of S− would be required,
the ferrous product from this reaction would be too
small to be observed. While this mechanism is plausi-
ble, it is probably too early to speculate on the identity
of the catalytic species.
The intriguing aspect of the reduction of Fe(P)(Cl)
with thiosulfate is the effect of sulfate ion on the
reaction. At high concentrations of sulfate (above 3–4
mM), the reaction was pseudo-first order with good
isosbestic points during the entire reaction time and was
faster than in the absence of sulfate (Fig. 6). An exam-
ination of the visible spectra during the reaction indi-
cated than most of the ferric porphyrin was complexed
with sulfate, rather than thiosulfate. The results indi-
cate that the identity of the axial ligand is not impor-
tant, but that the rate-determining step is the reaction
between the thiosulfate in solution and the complexed
ferric porphyrin. When the sulfate concentration was
less than 3 mM, the two-step reaction, observed in the
absence of sulfate, was observed. Previous spectral
studies with sulfate and ferric porphyrins in DMSO
indicated that dissociation of the complex appears be-
low this concentration.
No effect on the kinetics of the ferric porphyrin
reduction was observed when chloride was replaced
with perchlorate. This is consistent with the complete
conversion of the complex to Fe(TPP)+ in DMSO,
whether one starts with the chloro or perchlorate com-
plex. In addition, the results observed with ferric–OEP
were quite similar to the ferric–TPP complex. The
kinetics was consistent with a two-step mechanism,
though the reduction rate was slightly smaller with the
OEP-macrocycle.
In contrast with the reduction of Fe(TPP)(Cl) in
DMSO, the reduction in DMF is much more sluggish.
The complete reduction of 0.10 mM Fe(TPP)(Cl) with
0.70 mM thiosulfate takes about 12 h. Isosbestic points
were observed during the reduction, and the only spe-
cies observed were FeIII(TPP)(S2O3)− and FeII(TPP).
These results are consistent with previous studies of the

P.W. Crawford, M.D. Ryan / Inorganica Chimica Acta 328 (2002) 13–22
21
autoreduction of ferric porphyrins, which are signifi-
cantly faster in DMSO than in other solvents.
with an excess of thiosulfate, while only a low-spin
complex was observed in DMF. The residual high-spin
ferric porphyrin at 77 K is probably due to an equi-
librium concentration of an FeꢀO complex. The forma-
tion constants for the displacement of chloride by
thiosulfate (KTS,Cl) were nearly the same in DMF and
DMSO, based on the voltammetric data. As a result,
the EPR differences cannot be due to the ability of
ferric porphyrins to form the thiosulfate complex in the
two solvents.
In addition, the reduction of ferric porphyrins with
thiosulfate is also qualitatively different from the reduc-
tion by nitrite. In the latter case, the Fe–nitrite complex
is the species that is reduced, and the rate of the
reaction does not depend upon the concentration of
free (uncoordinated) nitrite [59]. In the case of nitrite,
though, the reduction of the ion must be accompanied
by an acid–base or oxygen transfer reaction to form
NO. The reduction of ferric porphyrins by nitrite has
characteristics of an inner-sphere reaction, while the
ferric porphyrin–thiosulfate reaction appears to be
outer-sphere in that coordination by thiosulfate is not
the rate-limiting step in the oxidation process. This was
indicated by the fact that the oxidation of thiosulfate
was actually accelerated by the addition of sulfate, and
that the rate of oxidation was controlled by the concen-
tration of free, not-coordinated thiosulfate. It is not
clear at this time as to why the mechanism of the
auto-reduction of ferric porphyrins changes in the pres-
ence of sulfate, but these results confirm the complexity
of the reaction of thiosulfate with ferric ion, as has been
observed in the literature.
3.5. Thiosulfate oxidation product isolation
Equimolar amounts of Fe(P)(Cl), where P=OEP or
TPP, were reacted with thiosulfate in DMF and
DMSO. The product, isolated following addition of
toluene to precipitate ionic compounds, yielded an in-
frared spectrum that closely corresponded to authentic
(Ph4P)2S4O6. In particular, bands were observed (au-
thentic (Ph4P)2S4O6 in parenthesis) at 592 (594), 629
(629) and 1230 (1232) cm−1, which were not present in
Ph4PCl. The tetrathionate ion could also be detected,
along with unreacted thiosulfate ion, by thin-layer
chromatography [58]. The visible spectroscopy of te-
trathionate in the presence of Fe(TPP)(Cl) was exam-
ined and there was no evidence for the coordination of
this species with the ferric porphyrin at the concentra-
tions used in this experiment.
4. Conclusions
The reduction of ferric porphyrins with thiosulfate is
complex because of the high barrier for the formation
of S2O3−. Coordination of the ferric porphyrin with
sulfate does not impede the reaction, indicating that
prior coordination is not necessary for the reaction,
though six-coordinate intermediates cannot be ex-
cluded. This is consistent with reaction (3) in that the
rate limiting step is the reaction of the ferric porphyrin
(sulfate or thiosulfate coordinated) with the catalyst.
The reaction of S2O3− (S4O63−) with the ferric por-
phyrin is not rate limiting. The change in the reaction
to pseudo-first order in the presence of sulfate may not
necessarily indicate a change in mechanism, but may
mean changes in the relative rate constants or concen-
tration of the catalyst.
The difference in the reactivity of ferric porphyrins
with thiosulfate between DMF and DMSO is quite
intriguing. The difference between these two solvents is
generally not large, though DMSO does coordinate
more strongly. The spectroscopic and voltammetric evi-
dence is quite strong that thiosulfate coordinates well
with ferric porphyrins in either solution. The cyclic
voltammetry, in particular, shows that there is little
dissociation of the complex, even with a small excess of
thiosulfate. The similarity of the visible spectra between
Fe(TPP)(SO4)− and Fe(TPP)(S2O3)− in DMSO could
indicate that the coordination changed from a FeꢀS to
an FeꢀO bonded complex, perhaps driven by axial
coordination by DMSO. The EPR results were also
indicative of a difference in iron coordination in these
two solvents. In DMSO, the high spin complex does
not completely disappear in the EPR spectrum even
Acknowledgements
The authors would like to thank William Antholine
of the Medical College of Wisconsin for obtaining the
EPR spectra, and Scott Reid for helpful discussions on
chemical kinetics.
References
[1] H.D. Peck Jr., J. LeGall (Eds.), Inorganic Microbial Sulfur
Metabolism, Academic Press, New York, 1994.
[2] R. Meulenberg, J.T. Pronk, W. Hazeu, J.P. Vandijken, J. Frank,
P. Bos, J.G. Kuenen, J. Gen. Microbiol. 139 (1993) 2033.
[3] W.-P. Lu, D.P. Kelly, J. Gen. Microbiol. 130 (1984) 1683.
[4] J.B. Fernandes, D. Feng, A. Chang, A. Keyser, M.D. Ryan,
Inorg. Chem. 25 (1986) 2606.
[5] D. Feng, M.D. Ryan, Inorg. Chem. 26 (1987) 2480.
[6] I.-K. Choi, Y. Liu, D. Feng, K.J. Paeng, M.D. Ryan, Inorg.
Chem. 30 (1991) 1832.
[7] Y.M. Liu, M.D. Ryan, J. Electroanal. Chem. 368 (1994) 209.
[8] L.M. Epstein, D.K. Straub, C. Maricondi, Inorg. Chem. 6 (1967)
1720.
[9] G.N. LaMar, J. Del Gaudio, Adv. Chem. Ser. 162 (1977) 207.

22
P.W. Crawford, M.D. Ryan / Inorganica Chimica Acta 328 (2002) 13–22
[37] D. Lexa, M. Momenteau, J. Mispelter, Biochim. Biophys. Acta
338 (1974) 151.
[10] J. Del Gaudio, G.N. LaMar, J. Am. Chem. Soc. 100 (1978)
1112.
[38] E.B. Fleischer, J.M. Palmer, T.S. Srivastava, A. Chatterjee, J.
Am. Chem. Soc. 93 (1971) 3162.
[39] C.A. Reed, T. Mashiko, S.P. Bentley, M.E. Kastner, W.R.
Scheidt, K. Spartalian, G. Lang, J. Am. Chem. Soc. 101 (1979)
2948.
[40] D.H. Dolphin, J.R. Sams, T.B. Tsin, Inorg. Chem. 16 (1977)
711.
[41] H. Sugimoto, N. Ueda, M. Mori, Bull. Chem. Soc. Jpn. 55
(1982) 3468.
[42] M.A. Phillippi, H.M. Goff, J. Chem. Soc., Chem. Commun. 1
(1980) 455.
[43] S.C. Tang, S. Koch, G.C. Papaefthymiou, S. Foner, R.B.
Frankel, J.A. Ibers, R.H. Holm, J. Am. Chem. Soc. 98 (1976)
2414.
[44] H. Oshio, T. Ama, T. Watanabe, K. Nakamoto, Inorg. Chim.
Acta 96 (1985) 61.
[45] H. Aissaoui, S. Ghirlanda, C. Gmur, W.-D. Woggon, J. Mol.
Catal. A 113 (1996) 393.
[46] D.R. English, D.N. Hendrickson, K.S. Suslick, C.W. Eigenbrot
Jr., W.R. Scheidt, J. Am. Chem. Soc. 106 (1984) 7258.
[47] S.E. Jones, G.S. Srivatsa, D.T. Sawyer, T.G. Traylor, T.C.
Mincey, Inorg. Chem. 22 (1983) 3903.
[48] R. Tauler, A.K. Smilde, J.M. Henshaw, L.W. Burgess, B.R.
Kowalski, Anal. Chem. 66 (1994) 3337.
[49] I.E. Kalinichenko, E.P. Rodina, V.V. Trachevskii, Zh. Obshch.
Khim. 67 (1997) 47.
[50] R. Sarala, S.B. Rabin, D.M. Stanbury, Inorg. Chem. 30 (1991)
3999.
[51] K. Shin, S.K. Kramer, H.M. Goff, Inorg. Chem. 26 (1987) 4103.
[52] G.S. Srivatsa, D.T. Sawyer, Inorg. Chem. 24 (1985) 1732.
[53] T. Otsuka, T. Ohya, M. Sato, Inorg. Chem. 24 (1985) 776.
[54] F.M. Page, Trans. Faraday Soc. 56 (1960) 398.
[55] M. Kimura, M. Ishibashi, Inorg. Chim. Acta 129 (1987) 69.
[56] W.H. Latimer, The Oxidation States of the Elements and Their
Potentials in Aqueous Solutions, Prentice Hall, Englewood
Cliffs, NJ, 1952.
[57] R. Sarala, D.M. Stanbury, Inorg. Chem. 31 (1992) 2771.
[58] A. Westley, J. Westley, Anal. Biochem. 142 (1984) 163.
[59] J.B. Fernandes, Spectroscopic and electrochemical study of iron
porphyrins and chlorins in the presence of nitrite in non-aqueous
solvents, Ph.D. Thesis, Marquette University, 1988.
[11] J. Del Gaudio, G.N. LaMar, J. Am. Chem. Soc. 98 (1976) 3014.
[12] S. Modi, V.P. Shedbalkar, D.V. Behere, Inorg. Chim. Acta 173
(1990) 9.
[13] C.J. Swan, D.L. Trimm, Adv. Chem. Ser. 76 (1968) 182.
[14] C.K. Chang, D. Dolphin, Proc. Natl. Acad. Sci. USA 73 (1976)
3338.
[15] K. Nakamoto, Infrared and Raman Spectra of Inorganic and
Coordination Compounds, Wiley, New York, 1978.
[16] Y.O. Kim, B.H. Song, H.M. Goff, Inorg. Chem. 32 (1993) 1304.
[17] P.W. Crawford, M.D. Ryan, Inorg. Chim. Acta 179 (1991) 25.
[18] M.A. Phillippi, N. Baenziger, H.M. Goff, Inorg. Chem. 20
(1981) 3904.
[19] W.R. Scheidt, Y.J. Lee, T. Bartzcak, K. Hatano, Inorg. Chem.
23 (1984) 2552.
[20] M.S. Reynolds, R.H. Holm, Inorg. Chim. Acta 155 (1989) 113.
[21] R.G. Wilkins, The Study of Kinetics and Mechanism of Reac-
tions of Transition Metal Complexes, Allyn and Bacon, Boston,
1974.
[22] A. McAuley, Coord. Chem. Rev. 5 (1970) 245.
[23] C.E. Ophardt, J. Chem. Ed. 64 (1987) 716.
[24] W.K. Wilmarth, D.M. Stanbury, J.E. Byrd, H.N. Po, C.-P.
Chua, Coord. Chem. Rev. 51 (1983) 155.
[25] S.M. Chen, J. Mol. Catal. A 138 (1999) 1.
[26] S.M. Chen, J. Electroanal. Chem. 432 (1997) 101.
[27] S.M. Chen, Inorg. Chim. Acta 244 (1996) 155.
[28] A.D. Adler, F.R. Longo, V. Varadi, Inorg. Syn. 16 (1976) 213.
[29] R. Larsson, B. Falkesson, R. Lykvist, Chem. Scr. 13 (1979) 178.
[30] A.J. Banister, L.F. Moore, J.S. Padley, in: G. Nickless (Ed.),
Inorganic Sulphur Chemistry, Elsevier, Amsterdam, 1968, pp.
137–198.
[31] K. Buijs, J. Inorg. Nucl. Chem. 24 (1962) 229.
[32] A. Martinsen, J. Songstad, Acta Chem. Scand. Ser. A A31
(1977) 645.
[33] C.A. Reed, J.K. Kouba, C.J. Grimes, S.K. Cheung, Inorg.
Chem. 17 (1978) 2666.
[34] J.T. Landrum, K. Hatano, W.R. Scheidt, C.A. Reed, J. Am.
Chem. Soc. 102 (1980) 6729.
[35] X.Q. Lin, K.M. Kadish, Anal. Chem. 57 (1985) 1498.
[36] R.A. Hasty, S.L. Sutter, Can. J. Chem. 47 (1969) 4537.