Catalytic oxidation of thiol compounds by novel fuel cell-inspired co-porphyrin and co-imidazole catalysts

Article abstract of DOI:10.1560/8J05-UHG8-K6DN-VVB2

The catalytic performance of pyrolyzed carbon-supported cobalt-nitrogen donor (CoN₄) catalysts for the oxidation of thiol compounds by dioxygen in aqueous solution was studied. This paper continues our previous line of research, which was inspired by the electrocatalytic reduction of oxygen on pyrolyzed carbon-supported cobalt-porphyrins and related tetra-coordinated nitrogen donor-transition metal complexes (MeN₄, where Me stands for a transition metal atom). Both pyrolyzed carbon-supported Co-imidazole and Co-porphyrin exhibited fast catalytic oxidation of the different thiols. The rate of oxidation of different thiols on the pyrolyzed CoN₄ catalysts was compared to the homogeneous rate of oxidation using 5,10,15,20-tetrakis(4- sulfonatophenyl)porphyrin Co(II) tetrasodium salt as catalyst. Based on the cobalt content, the heterogeneous catalysts always outperformed the homogeneous one, and at times even exhibited 4,100-fold better catalysis. The dependence of the catalytic rate of oxidation on the preparation temperature was investigated, showing an optimal catalysis at ～650 °C for the cobalt-imidazole catalyst. The decrease in catalytic performance after heat treatment at elevated temperature was attributed to the formation of cobalt metal acting as a generator of carbon nanotubes.

Full text of DOI:10.1560/8J05-UHG8-K6DN-VVB2

Catalytic oxidation of thiol Compounds by novel fuel Cell-inspired Co-
porphyrin and Co-imidazole Catalysts
a,
AnAtoly GoifmAn,a Jenny Gun,a,c fAinA GelmAn,a irinA ekeltchik,a eckhArd Worch,b And ovAdiA lev *
aLaboratory of Environmental Chemistry, the Casali institute of Applied Chemistry, the Hebrew University of Jerusalem,
Jerusalem 91904, israel
binstitute of Water Chemistry, Dresden University of technology, D-01062 Dresden, Germany
cthe Hadassah College of Jerusalem, Jerusalem 91917, israel
(Received 15 September 2005)
Abstract. the catalytic performance of pyrolyzed carbon-supported cobalt–ni-
trogen donor (Con4) catalysts for the oxidation of thiol compounds by dioxygen in
aqueous solution was studied. this paper continues our previous line of research,
which was inspired by the electrocatalytic reduction of oxygen on pyrolyzed carbon-
supported cobalt-porphyrins and related tetra-coordinated nitrogen donor–transition
metal complexes (Men4, where Me stands for a transition metal atom). Both py-
rolyzed carbon-supported Co-imidazole and Co-porphyrin exhibited fast catalytic
oxidation of the different thiols. the rate of oxidation of different thiols on the
pyrolyzed Con4 catalysts was compared to the homogeneous rate of oxidation using
5,10,15,20-tetrakis(4-sulfonatophenyl)porphyrin Co(ii) tetrasodium salt as catalyst.
Based on the cobalt content, the heterogeneous catalysts always outperformed the
homogeneous one, and at times even exhibited 4,100-fold better catalysis. the
dependence of the catalytic rate of oxidation on the preparation temperature was
investigated, showing an optimal catalysis at ~650 °C for the cobalt-imidazole
catalyst. the decrease in catalytic performance after heat treatment at elevated
temperature was attributed to the formation of cobalt metal acting as a generator of
carbon nanotubes.
IntroductIon
oxygen electroreduction. Carbon-adsorbed metal-por-
phyrins reduced the oxygen reduction overpotential by
several hundred millivolts compared to blank carbon
electrodes. However, the vulnerability of tetrapyrroles
to free radical attack—at least relative to the demand-
ing requirements of the energy industry—hindered dis-
semination of the technology. the situation started to
change with the discovery that pyrolyzed carbon pow-
der-supported cobalt-porphyrin maintains the excellent
electrocatalytic reduction of dioxygen.5,6 this came as a
surprise to many in the industry since there was never
any doubt that the tetrapyrrole structure was destroyed
during the pyrolysis step. Still, electrocatalysis was
preserved even after over 900 °C heat treatment, while
the catalyst’s susceptibility to free radical and hydrogen
peroxide attacks was practically eliminated. Jasinsky’s
electrocatalyst soon found practical applications for
Metal-porphyrin catalysis is one of the outstanding
success stories of biomimetic chemistry. p450 mono-
oxygenases, heme protein–oxygen-binding mimics, and
peroxidase and catalase emulations are only a few of
the possibilities that were opened with the evolution of
porphyrin and phthalocyanine chemistry.1–3 one of the
earliest and most active scientific activities involved the
ability of metaloporphyrins to bind dioxygen and by
way of superoxo-metal-porphyrin complexes (por-Me-
oo.) to activate the otherwise inactive triplet oxygen.
this process can end by formation of either potent
short-lived hydroxyl and other oxo radicals, or by less
aggressive oxo and peroxo moieties that can function as
monooxygenases for epoxidation, hydroxylation, and
other fine chemistry catalyses.
from its early inception, the fuel cell industry4 real-
ized that the high affinity of oxygen to metaloporphyrins
and other metalo-tetrapyrrols can be utilized to promote
*Author to whom correspondence should be addressed. E-mail:
ovadia@vms.huji.ac.il
Israel Journal of Chemistry
Vol. 46
2006
pp. 17–26

18
mobile vehicles and electrical appliances.7 However, the Co-imidazole catalyst occurs for Co:imidazole ratio
despite the favorable oxygen reduction characteristics = 1:2, corresponding to a 1:4 Co:n ratio, similar to the
of the pyrolyzed carbon-supported cobalt-porphyrin Co:n ratio of Co-porphyrin. However, there were also
electrodes, their penetration into the main stream of the some differences between the two types of catalysts,
fuel cell industry was slowed down by the prohibitively the most significant one being the optimal pyrolysis
high costs of porphyrins and phthalocyanines. Several temperature. the best performance was obtained after
approaches were devised to replace the metal porphy- 880 °C treatment for the cobalt-porphyrin, while the
rins by less expensive nitrogen donors. Holoporphy- cobalt-imidazole catalyst started to lose activity already
rins and holophthalocyanines and then nitrogen-donor after 760 °C treatment. the primary task of the current
polymers such as polypyrrole and polyacrylonitrile study was to expand the scope of catalytic activities of
were investigated as electrode donors in place of the two model Men4-based catalysts, cobalt-imidazole and
costly porphyrins.8–10 next, it was demonstrated that cobalt-porphyrin, and to examine their performance for
the performance of the pyrolyzed carbon-supported the oxidation of different mercaptans by dissolved oxy-
forms of common, low molecular and inexpensive ni- gen under aqueous, room-temperature conditions.
trogen donors were comparable to the performance of
the removal of mercaptans and other sulfur com-
the carbon-supported metal-porphyrins.11 interestingly, pounds is an important environmental reaction due to
even simple and relatively cheap nitrogen donors such the abundance of thiols that are inevitably byproducts
as pyrrole, imidazole, and their methylated forms were of virtually all anaerobic wastewater treatments, with
proven valuable for oxygen electroreduction in acidic their volatility and malodor.14,15 the desulfurization of
media. However, one has to admit that the success of fuels is also continuing to be of major concern in oil
these catalysts is still limited, probably since they tend refineries.16 Cobalt-phthalocyanines were introduced in
to electrocatalyze the first 2-electron reduction of oxy- the sixties for the oxidation of sulfur-rich fuels, and dif-
gen to hydrogen peroxide better than the following step ferent variants of the process were commercialized.17,18
involving the reduction of hydrogen peroxide to water. More recently, several processes for the oxidation of
thus, a buildup of hydrogen peroxide in the electrolyte mercaptan compounds on copper(i) complexes were
and less than 4-electron exploitation of the dioxygen are proposed by Bagiyan et al.17,19 the homogeneous oxi-
observed, which limit further progress in this field.
dation of thiols by molecular oxygen without added
Despite the proven success of the cobalt-porphyrins catalysts (also called self-oxidation) has been amply
in the activation of triplet oxygen, this ability was never studied, mostly in connection with protein chemistry
transferred to the chemical catalysis arena, although and biology.20,21
conceptually they share the same target—the utiliza-
tion of the high oxidation potential of the conversion of
triplet dioxygen to water. In our first publications on the
ExpErImEntAl
subject, we have demonstrated that pyrolyzed carbon-
supported cobalt-porphyrins and their mimics are not
Materials
Diphenyldisulfide (≥95%) was synthesized from benzene-
thiol according to the solvent-free permanganate oxidation
procedure. Sodium sulfide hydrate was purchased from Merck
(Darmstadt, Germany). Dibasic ammonium phosphate, N,N-
limited to electrocatalysis and can indeed exhibit non-
electrochemical (environmental) catalysis as well. We
have demonstrated that the catalysts catalyze the oxida-
tion of hydrogen sulfide by dissolved dioxygen in aque-
dimethyl-p-phenylenediamine hemioxalate and 3,5-dinitro-
benzoic acid (DnBA) were from Sigma (Milwaukee, Wi,
ous media. the catalyst outperformed other granular
and catalytic activated-carbon powders that are widely
USA). imidazole, heptanethiol, dodecanethiol, 2-aminothio-
phenol (cysteamine), l-cysteine, p-chlorobenzenethiol, 2,2¢-
used for air and water dehydrosulfurization, and also dithiobis(5-nitropyridine) (Dtnp), and tetrabutylammonium
hydrogen sulfate (tBAHS) were from Aldrich (Milwaukee,
Wi, USA). thiosalicylic acid was from fluka (taiwan) and
p-thiocresol was from fluka. naphthalene and iron (iii)
chloride anhydrous were from BDH (poole, England). Sul-
furic acid (98%), hydrochloric acid (37%), phosphoric acid
(85%), methanol, ethanol, acetonitrile, dichloromethane, and
sodium hydrogen phosphate were from J.t. Baker (Deventer,
Holland). Sodium dihydrogen phosphate and sodium acetate
anhydrous were from Mallinckrodt (phillipsburg nJ, USA).
Benzenethiol and Co(ii) acetate tetrahydrate were from AC-
RoS. Co(ii) mesotetra-4-methoxy-phenylporphyrin (CoMt-
performed better than dissolved and carbon-adsorbed
cobalt-porphyrin that did not undergo the pyrolysis
step. We then showed that pyrolysis of carbon pow-
der impregnated with imidazole and cobalt(ii) acetate
provides a catalyst with equal, and at times superior,
catalytic performance for hydrogen sulfide oxidation
compared to the pyrolyzed cobalt porphyrin.
it was found that the cobalt-porphyrin and cobalt-
imidazole catalysts share much in common. notably,
the active sites of the two types of pyrolyzed catalysts
are very similar, and optimal catalytic performance of MPP) was donated by Frontier Scientific Advanced Discovery
Israel Journal of Chemistry 46 2006

19
Chemicals. 5,10,15,20-tetrakis(4-sulfonatophenyl)porphyrin cated for 10 min for better homogenization. the suspension
Co(ii) tetrasodium salt (CotSpp) was purchased from por- was dried by rotary evaporator. the catalysts were heated in
phyrin Systems (Lübeck, Germany).
vacuumed quartz ampoules for 4 h at t1 = 360 °C and then for
Acetylene black (AB50%, also named Shawinigan Black) an additional 4 h at the specified temperature (T2 = 760 °C un-
was purchased from Chevron phillips Chemical Company. All less otherwise specified). The catalysts were immersed in the
chemicals were of analytical grade and were used as received appropriate buffer for at least 2 h before usage.
unless otherwise stated. We used deionized water (conductiv-
Preparation of cobalt-porphyrin modified catalysts
ity < 0.1 mS/cm) purified by a Seradest SD 2000 system.
the preparation conditions of the pyrolyzed cobalt-por-
Analytical Procedures
phyrin catalysts were identical to those used in our previous
Some analytical procedures were described in our previ- study.12 Briefly, 100 mg of CoMTMPP was dissolved in 50
ous articles.12,13 Hydrogen sulfide was analyzed according to mL of CH2Cl2, and then 1 g of carbon was added to obtain
a Standard Methods protocol using the 4500-S2– D procedure. a suspension. the suspension was sonicated for 10 min for
nonvolatile thiols (cysteamine, thiosalicylic acid, and l-cys- better homogenization and then dried by rotary evaporator.
teine) were determined by a method based on Vairavamurthy The catalysts were heated for 5 h at the specified temperature
and Mopper’s procedure of derivatization with Dtnp, fol- (880 °C unless otherwise specified) in a flow of 99.999% pure
lowed by reversed phase HpLC separation. We used a ther- nitrogen. the catalysts were immersed in the appropriate buf-
moquest HpLC equipped with uv600lp pDAdetector operated fer for at least 12 h before usage.
at 390 nm for quantification of the thiols and at 250 nm for
Oxidation of Thiols
the determination of the internal standard. the separation was
conducted on a C-18 SYnERGi Hydro-Rp 80A 250 mm ´ 4.6
mm, 4 mm column using as eluant a gradient of acetonitrile
and aqueous solution containing 7.5 mM tBAHS and 0.05 M
sodium acetate at pH 3.5. Volatile thiols (benzenethiol, p-chlo-
robenzenethiol, p-thiocresol, heptanethiol, dodecanethiol) and
diphenyldisulfide were determined by GC/MS using Agilent
GC 6890n with MS 5973n, as detailed below.
Pyrolyzed carbon-supported cobalt-imidazole and cobalt-
porphyrin catalysts
125 mL of 0.01 mM thiol solutions were prepared in dark
bottles from air-saturated distilled water (containing 8.0 mg/L
dissolved oxygen). the solutions were kept at pH 7.0 by 0.02
M phosphate buffer, unless otherwise specified. To these so-
lutions we added 12 mg of modified Co-imidazole catalyst or
57 mg of Co-porphyrin catalyst, suspended in buffer. 100 mL
of 6.25 mM naphthalene in methanol was added as internal
standard. the bottles were stoppered and left to react with
magnetic stirring at ambient temperature (usually 23 ± 2 ºC).
After a set reaction period the solutions were quickly filtered
through 0.2-mm glass filters of Whatman GF/B (Maidstone,
England) to remove the carbon catalyst, extracted with 5 + 3 +
3 mL of dichloromethane, and analyzed by GC/MS. in order
to avoid catalyst loss or hydrogen sulfide evaporation during
sampling, each test was conducted using a dedicated bottle,
which was used for a single analysis only.
GC/MS analysis
A 125-mL sample was spiked with 100 mL of 6.25 mM
naphthalene in methanol as internal standard and extracted
with 5 + 3 + 3 mL of dichloromethane. A 2-mL aliquot of the
extract was injected to the GC/MS using J&W-DB1, 60 m, i.d.
0.32 mm, 0.25 mm column. When needed, extracts were stored
in brown glass vials for up to a week in a 4 °C refrigerator to
prevent further oxidation of the residual thiols. typically, the
following temperature program was used: 10 min at 100 °C,
10 °C/min T gradient, and finally 17 min at 250 °C. Other GC
conditions: Inlet, T = 200 °C; solvent delay, 11 min; He flow,
1.5 mL/min.
Homogeneous Co-porphyrin catalysis
the oxidation of different thiols by CotSpp was conduct-
ed under similar conditions, in all cases using 73 mg of cobalt-
porphyrin dissolved in buffer.
Electrochemistry and microscopy
Cyclic voltammetry was carried out using a pARC 263 po-
tentiostat/galvanostat (EG&G, princeton, nJ, USA). We used
a conventional three-electrode cell with a 3-mm i.d. carbon
paste working electrode, pt counter, and KCl satd Ag/AgCl
reference. All tests were carried out in 0.5 M sulfuric acid so-
lution. the carbon paste electrodes were prepared by mixing
100 mg of catalyst-loaded carbon powder with 30 mg mineral
oil (Aldrich) and loading the paste into a 3-mm-diameter glass
tube. All potentials are reported versus KCl satd Ag/AgCl ref-
erence. Scanning electron microscope imaging was performed
by JEoL JXA-8600 Superprobe.
rEsults
The first round of tests included kinetic oxidation tests
that were conducted for the set of thiol compounds
listed in table 1, in order to determine the dominant
substrate-dependent rate law and the relevant rate
constants. this initial study was not meant to give a
comprehensive account of the reaction rate at a large
parametric range; it was merely used to provide a basis
for comparison and ranking of the individual rate con-
stants and to allow a comparison to the rate constants of
homogeneous cobalt-porphyrin catalysis. All the tests
were conducted under identical starting conditions: 0.02
Preparation of cobalt-imidazole modified catalysts
the preparation conditions of the pyrolyzed cobalt-por-
phyrin catalysts were identical to those used in our previous
study.13 Briefly, set quantities of cobalt acetate and imidazole
were dissolved in 50 mL of ethanol, and then 1 g of carbon
was added to obtain a suspension. the suspension was soni- M phosphate buffer, pH 7, 0.01 mM thiol, and 8 mg/L
Goifman et al. / Oxidation of Thiols by Co-Porphyrin and Co-Imidazole Catalysts

20
table 1
Kinetics of the catalytic conversion of thiol compounds by dioxygen on pyrolyzed carbon-supported cobalt-imidazole, pyrolyzed
carbon-supported cobalt-porphyrin (CoMtMpp), and homogeneous catalysis by the water-soluble Co-porphyrin (CotSpp). All
rate constants are based on the cobalt content (i.e., normalized to 1 mg Co/L) of the catalysts and measured in min–1 (mg Co/L)–1
kCo-imidazole
kCoMTMP
0.6
kCoMTMP
kCoTSPP
—
compound name
molecular formula
kCo-imidazole
kCoMTMP
kCoTSPP
pKaSH*
hydrogen sulfide
benzenethiol
H2S
C6H5SH
0.0016
0.027
0.026
0.027
0.017
0.013
0.020
0.014
0.015
0.0025
0.049
0.034
0.033
0.014
0.0050
0.019
0.031
0.0064
—
7.0
6.6
0.0023
0.0062
0.0045
0.0062
0.0014
0.6
22
p-chlorobenzenethiol
p-thiocresol
p-ClC6H4SH
p-CH3C6H4SH
C7H15SH
6.1
0.8
6
6.8
0.8
7
heptanethiol
10.8
10.6
1.2
2
dodecanethiol
2-aminoethanthiol
thiosalicylic acid
l-cysteine
C12H25SH
2.5
3
HSCH2CH2nH2
HSC6H4CooH
HSCH2CHnH2CooH
0.00048 10.5
0.00030 8.9
0.0000016 11.5
1.0
40
0.5
103
4100
2.4
*pKaSH values correspond to the proton dissociation constants of the thiol groups.
concentrations a Dean–Hinshelwood or more complex
kinetics may prevail, but this was not searched for in
the current phase of this study. in all cases, the oxida-
tion rate constants of the different thiols by dioxygen, in
the absence of the catalysts, were negligible compared
to the much higher rate of reaction in the presence of
either cobalt–porphyrin or pyrolyzed cobalt-imidazole
catalysts.
the fact that the kinetics of disappearance of the
different thiols followed a first-order rate law allowed
direct comparison between the rate constants for the
oxidation of the different thiols on the different cata-
lysts. figure 2 and table 1 detail the observed reaction
kinetics (based on pH 7 and 8 mg/L oxygen) for the ho-
mogeneous cobalt-porphyrin and the two heterogeneous
Men4 catalysts.
table 1 (right column) demonstrates that in all cases
fig. 1. (a) time trace of the oxidation of different thiol com- the heterogeneous catalysts outperformed the homoge-
pounds by dissolved oxygen on pyrolyzed carbon-supported
Co-imidazole catalyst. (b) Logarithmic form of the time
traces. (0.02 M phosphate buffer, pH 7; initial concentration of
dissolved oxygen, 8 mg/L; initial concentration of the differ-
ent thiols = 0.01 mM; 96 mg/L catalyst.) Legend: 1—benzene-
thiol, 2—2-aminoethanethiol (cysteamine), 3—thiosalicylic
acid, 4—diphenyldisulfide.
neous catalysis of thiols. the difference in performance
is quite substantial and at times, e.g., for thiosalicylic
acid, for the heterogeneous catalysts we observed as
much as 100-fold faster heterogeneous catalysis com-
pared to the homogeneous catalysts, and for l-cysteine
a 4,100-fold increase was observed. the two types of
pyrolyzed Men4 behaved somewhat differently and ex-
hibited a different degree of catalysis for different thiols.
However, linear correlation of the two catalysts exhib-
ited a significant linear dependence (correlation coeffi-
cient, R2 = 0.81) between the k-values for the oxidation
of the different thiols on the two catalysts.
oxygen. figure 1 demonstrates that, at least for the low
concentration range addressed in the current studies, the
time trace of the conversion of the different thiols fol-
lowed a first-order rate law with respect to the different
substrates. it is indeed plausible that at higher reactant
Israel Journal of Chemistry
46
2006

21
fig. 2. (a) Comparison of the catalytic performance of homogeneous cobalt-porphyrin, CotSpp (blank bars), and pyrolyzed
carbon-supported cobalt-porphyrin, CoMtMpp (striped bars) and cobalt-imidazole catalysts (dotted bars). for comparison, the
rate constants were normalized by the cobalt content in the different catalysts. (b) A correlation between the relative efficiencies
of the Co-imidazole and CoMtMpp catalysts.
figure 2b shows that when the absolute catalysis provement in catalysis by the pyrolysis step, which is
was large, then the CoMtMpp exhibited better cataly- at least partly explained by the very high pKa’s of these
sis, and for the more refractive thiols the Co-imidazole compounds, which implies that only a negligible frac-
catalyst exhibited better performance. this, in fact, is tion of these compounds is present in the active deprot-
an encouraging observation, showing that the much onated form at pH 7.
cheaper Co-imidazole catalyst exhibits competitive per-
formance, and more so when the catalysis is slow. for pH Effects
simplicity, the abscissa of fig. 2b delineates the cata-
In order to clarify the influence of the pH on the
lytic rate for CoMtMpp, but a very similar curve was catalytic conversion rate, we investigated the pH depen-
obtained when the x-coordinate delineated the average dence of the rate constant of several thio-compounds.
of CoMtMpp and Co-imidazole rates.
the rate constants of benzenethiol, cysteamine, and
the most pronounced catalysis was observed for the thiosalicylic acid are depicted in fig. 3 as a function of
aryl compounds (benzenethiol, chlorobenzenethiol, and the pH of the solution. The figure demonstrates three
thiocresol), though the internal pecking order amongst different behaviors. Benzenethiol’s rate of conversion
the aryls followed neither the electrophilicity of the is almost constant up to pH ~ 7–8, and then the rate is
thiols nor their dissociated fraction. the good chelat- gradually increased. the threshold of the kinetic in-
ing agents, aminothiols and thiosalicylic acid, also ex- crease coincides with the pKa of the benzenethiol (pH
hibited rather fast reaction rates, which are ascribed to 6.6, see Table 1, column 6). The kinetic coefficient of
the bidentate coordination of the active cobalt center. the thiosalicylic acid was indifferent to the pH level,
Bagiyan et al.19 observed a similarly high reaction rate regardless of the pH “distance” to the pKa values of the
of aminothiols on copper catalysts and attributed the thiosalicylic acid. note that pKa,1, corresponding to the
improved catalysis to the bidentate chelating ability of deprotonation of the carboxyl group = 3.5, and pKa,2,
these ligands. the alkylthiols exhibited the poorest im- corresponding to the deprotonation of the thiol = 8.9.
Goifman et al. / Oxidation of Thiols by Co-Porphyrin and Co-Imidazole Catalysts

22
Similar indifference of oxidation kinetics of strong che- that at high pH (>8.5) there is a significant deviation
lating agents to pH change was also noted by Bagiyan from mass balance, and diphenyldisulfide is not the
et al.19 the kinetics of the conversion of cysteamine sole product. However, we were unable to identify the
(fig. 3, curve 2) followed bell-shape dependence, with missing oxidation product. Comparison of the HpLC
an optimal rate around pH 8. A similar dependence was chromatogram of the products with the retention time
noted by Bagiyan et al. for copper complexes and other of benzenesulfonic acid reference material precluded
catalysts.19,20 the peak performance coincides with the the possibility that this is the missing reaction product.
dissociation constant (pKa 7.9) of the amino group. negative mode ESi/MS also precluded the formation of
Presumably the lower affinity of the ammonium moiety benzenesulfinic acid.
compared to the deprotonated amino coordinating group
dependence of the Catalytic Performance on the
is responsible for lower binding of the thiol to the Con4
Pyrolysis Temperature of the Co-Imidazole Catalyst
centers, which reduces the rate constants as the pH is
in our previous publications on the pyrolyzed Men4
lowered below the pKa of the amino group.
catalysts12,13 we have noted that the catalytic perfor-
for the weak chelating agents (aromatic and alkyl
mance is optimal after 880 °C treatment for Co-porphy-
thiols), the change in catalytic activity is associated
rin and after 760 °C for the cobalt-imidazole catalysts.
with the deprotonation of the thiol and formation of the
it was interesting to compare the catalytic performance
thiolate anion. formation of the negatively charged thi-
with regard to the thiol oxidation with the catalytic
olates accelerates the rate of the reaction, probably due
rates for conversion of hydrogen sulfide on the same
to increased association with the Con4 center. Similar
pyrolyzed carbon-supported Men4 catalyst. figure 4
observations were indeed reported for the homogeneous
provides such a comparison. The figure shows that there
oxidation of alkylthiols on Co-porphyrin25 as well as for
is a good qualitative correlation between the observed
the homogeneous and heterogeneous catalytic oxida-
rate constants for hydrogen sulfide conversion (curve
tion of alkylthiols.17,20 the bell-shape dependence of the
3 in the upper frame) and for the conversion of ben-
oxidation rate of the aminothiol compounds on the pH
zenethiol (curve 2, right axis). Consistent with our prior
was previously reported by Bagiyan et al.,20 and it was
knowledge about the oxidation of hydrogen sulfide, the
attributed to the chelation of transition metal impurities
catalyst exhibited an optimal catalytic performance after
by the different aminothiols. the authors noted that the
heat treatment at intermediate temperatures (~650 °C)
maximal activity cannot be explained by the pKa’s of
for benzenethiol (curve 2) and ~760 °C for hydrogen
the aminothiols. presumably the chelation ability of the
sulfide (curve 1). Interestingly, the temperature depen-
thiosalycilic acid is so strong that it does not constitute
dence is more pronounced for the benzenethiol conver-
a rate-determining step, and therefore a pH change does
sion compared to the hydrogen sulfide. To complete the
not influence the observed kinetics.
comparison we have included in fig. 4 a curve delineat-
Identification of the Oxidation Products
ing the electrocatalytic performance, i.e., dependence
GC/MS and ESi/MS analyses of the oxidation prod- of the half wave potential for oxygen electroreduction
ucts of the different thiols on pyrolyzed carbon-sup- on pyrolyzed carbon-supported cobalt-imidazole carbon
ported Co-imidazole catalysts at pH 7 revealed that in paste electrode. in this curve, lower oxidation potential
all cases the primary oxidation product was the disul- is a manifestation of better catalysis. the bold curve de-
fide form, RSSR. Only for the benzenethiol we noticed lineating E1/2 dependence on the heat treatment tempera-
Fig. 3. First-order kinetic coefficients for the oxidation of 1—benzenethiol, 2—2-aminoethanthiol (cysteamine), and 3—thio-
salicylic acid by air oxygen with pyrolyzed carbon-supported cobalt-imidazole catalyst as a function of pH.
Israel Journal of Chemistry
46
2006

23
0.16
0.12
0.08
0.04
0
0.0025
0.0020
0.0015
3
2
a
-1
-1
1
(µgCo/L)
(µgCo/L)
-1
-1
0.0010
, min
1,3
,2min
k
k
0.0005
-0.04
0
350
500
650
800
950
-800
-600
2
b
1
-400
, mV
E
-200
0
350
500
650
800
950
Temperature, °C
fig. 4. (a) Catalytic performance of heat-treated cobalt-imidazole (curves 1, 2) and cobalt-porphyrin catalysts (curve 3). the
curves depict the observed kinetic coefficient in [min–1(mg Co/L)–1] calculated based on 8 mg/L oxygen as a function of the heat
treatment temperature for hydrogen sulfide (curves 1 and 3) and benzenethiol oxidation (curve 2). (b) Electrocatalytic perfor-
mance of heat-treated cobalt-imidazole (curve 1) and cobalt-porphyrin (curve 2) catalysts. the curves depict the dependence of
the half wave potential, (E1⁄2 vs. Ag/AgCl, satd reference) for O2 electroreduction on the heat treatment temperature.
ture goes through a shallow maximum at around 600 °C, of the two different Men4 catalysts we examined the
similar to the catalytic performance curves, though the surface morphology of the catalysts by electron micros-
electrocatalysis curve is even less steep than for the hy- copy. figure 5a–d depicts the SEM micrographs of Co-
drogen sulfide catalysis. The comparison between the imidazole catalysts that were prepared at different tem-
curves indicates that the electrocatalytic and catalytic peratures. While the morphology of the 360 and 465 °C
mechanisms are somewhat different, which came as no catalysts is very similar, the morphology of the catalysts
surprise. But the overall tendency is maintained. the changes dramatically above 760 °C. Starting at 760 °C
performance of both electrocatalysts is improved for we witness the evolution of carbon nanotubes, and at
higher pyrolysis temperature.
even higher temperature the carbon nanotubes become
more abundant and practically cover the whole catalyst.
Close examination of fig. 5c,d shows that at the base of
SEM Analysis
the temperature dependence of the rate constants for each carbon nanotube there is a bright spot. We attribute
the hydrogen sulfide oxidation as well as for benzene- the bright spots to melted cobalt metal dots, which serve
thiol electrooxidation (curves 1 and 2 in fig. 4) differed as catalyst for the pyrolysis of the organic substrate
from the temperature dependence of the cobalt porphy- (imidazole and its degradation products) and nanotube
rin catalysts (curve 3). the optimal temperature for the growth from this center. in contrast, fig. 5e, which de-
pyrolysis of the cobalt-porphyrin was much higher com- picts the morphology of the cobalt-porphyrin catalyst,
pared to the cobalt-imidazole catalysts. A similar ten- does not contain any nanotube, although this catalyst
dency was observed for the electrocatalytic behavior.12,13 was prepared at 880 °C. Apparently, the Co-imidazole
in order to examine the reason for the different behavior catalyst loses its active Co sites at high temperature,
Goifman et al. / Oxidation of Thiols by Co-Porphyrin and Co-Imidazole Catalysts

24
of the test, 0.43 mg/L of pyrolyzed carbon-supported
CoMtMpp catalyst was added. two hours after the
introduction of the thiosalicylic acid its concentration
was measured, and the oxygen was bubbled for 15 min.
then, another round of thiol oxidation starting with 0.01
mM thiosalicylic acid dosing and its catalytic removal
was initiated. We did not notice any degradation of the
catalyst performance after 15 repeated tests, which cor-
respond to ~660 thiol–disulfide turnovers (based on Co
loading). the repeatability tests are shown in fig. 6.
a
b
0.5
m
1.0 m
c
d
Some Mechanistic Aspects
0.5
m
0.5 m
the catalysis of thiol oxidation by transition metals
or by metal-porphyrins often proceeds through a com-
plex mechanism involving the generation of active oxo
radicals or hydrogen peroxide. Hydrogen peroxide is an
(undesirable) byproduct of oxygen electroreduction on
pyrolyzed, as well as non-pyrolyzed metal-porphyrin-
modified electrodes. In order to explore the possibility
that the catalytic oxidation of thiols on pyrolyzed car-
bon-supported cobalt-imidazole involves a free radical
mechanism, we compared the benezenthiol catalytic
oxidation rate with and without the additions of 0.1 M
Ki or 0.1 M nH3, keeping the constant pH 7 conditions.
iodide was reported to inhibit radical chain reactions,
while ammonia accelerates free radical reactions.19
However, the observed rate of reaction remained prac-
tically unaltered, within our experimental error, by the
presence of these catalytic effectors. Similarly, our at-
tempt to detect traces of hydrogen peroxide during the
catalytic oxidation of benzenethiol by the Enhanced
ChemoLuminescence (ECL) using the luminol and
HRp procedure, detailed in ref 26 was unsuccessful.
e
0.5
m
fig. 5. SEM micrographs of pyrolyzed carbon-supported co-
balt-imidazole catalysts prepared at (a) 360 °C, (b) 465 ºC, (c)
760 ºC, (d) 1020 ºC, and (e) cobalt-porphyrin catalyst prepared
at 880 °C. Accelerating voltage = 5.0 keV (b, e) and 15 keV
(a, c, d).
some of the cobalt is melted, and it serves as a catalytic
site for the conversion of the imidazole to carbon. the
cobalt centers in the heat-treated Co-porphyrin are bet-
ter ligated, which elevates the melting temperature. Ad-
ditionally, the lower volatility of the tetrapyrrole does
not support evaporation and nanotube formation.
Repeatability Studies
We have carried out stability tests in order to eluci- other observations that were already detailed above
date the stability of the catalysts under prolonged usage. also support the conclusion that the reaction does not
the repeatability studies were carried out by repeated involve free radical formation. first, radical chain reac-
introduction of 0.01 mM thiosalicylic acid to a phos- tions invariably result in nonlinear concentration depen-
phate-buffered solution held at pH 7.0. At the beginning dence, whereas fig. 1 shows simple linear concentration
Turnovers
0
40 80 130 170 210 250 290 340 380 420 460 510 550 590 660
10
8
M
6
µ
,
C4
2
0
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15
Repetitions
fig. 6. Repeatability test: time trace of thiosalicylic acid oxidation on Co-imidazole catalyst at pH 7. Catalytic turnovers, cal-
culated based on the number of thiol molecules that were oxidized on each Co atom, are depicted in the upper section of the
curve.
Israel Journal of Chemistry
46
2006

25
Scheme 1. Dioxygen activation, intraparticle charge transfer, and oxidation of four thiols on supported Con4 sites.
dependence. free radical chain reactions should have lates). Otherwise it would be difficult to conceive how
yielded a variety of oxygenated sulfur byproducts, while three species (oxygen and two thiolates) adsorb on the
we have observed predominantly the disulfide product. same or very close by sites and how oxygen transfer is
Admittedly, none of these observations alone is conclu- so scarcely observed.
sive, yet their cumulative weight points against a free
radical chain mechanism.
Scheme 1 depicts a postulated mechanism. for sim-
plicity, each of the six arrows denotes several elementa-
Scheme 1 provides a qualitative postulate for the ry steps rather than a single one. Arrow 1 marks the first
reaction mechanism. it is based on the following guid- step in a catalytic cycle, with the adsorption of dioxygen
ing observations: (1) the reaction does not involve free on one of the Co (or cobalt-n4) sites. then, by intra-
radical formation. (2) Adsorption of the thiols is essen- particle two-electron transfer steps, two remote cobalt
tial for their catalytic removal, which is manifested in sites are oxidized, while the adsorbed oxygen is con-
the observed pH dependence and the higher oxidation verted to hydrogen peroxide (or its deprotonated form
rate of better chelating agents. (3) Substrate oxygen- or to a bound superoxo species). However, the bound
ation is rather scarce. (4) the conductivity of the carbon hydrogen peroxide is not released to the solution and
substrate is essential for efficient catalysis. The second does not start chain radical reactions or give oxygenated
and third observations lead us to the postulate that the products. Rather, it is dehydrated, leaving a bound oxo
oxygen activation site may be remote from the thiol species. Arrow 3 signals the subsequent (and probably
adsorption site, and the reaction may proceed through consecutive) adsorption of two thiolates on the Co(iii)
intraparticle electron transfer between these sites. A sites (though it is also possible that the adsorption of
similar observation was used to explain photocatalytic the thiols precedes and stimulates the intramolecular
oxidations by dioxygen on titania photocatalysts. in a charge transfer step). Subsequently, the two adsorbed
similar manner we postulate that the carbon particle be- thiolates can undergo charge exchange with the Co(iii)
haves as a miniature electrochemical cell where oxygen sites, stimulating the formation of a disulfide bond by
bonding and its reduction to H2o occurs on one site and the reaction of the two adsorbed thiol radicals (marked
the thiolate adsorption takes place elsewhere on two by arrows 4 and 5). The disulfide is then released from
nearby sites (to allow dimerization of the bound thio- the active sites (arrow 6).
Goifman et al. / Oxidation of Thiols by Co-Porphyrin and Co-Imidazole Catalysts

26
electric-fuel.com/
(8) franke, R.; ohms, D.; Wiesener, K.J. Electroanal.
Chem. 1989, 260, 63–73.
(9) Gupta, S.; tryk, D.; Bae, i.; Adred, W.; Yeager, E. J.
Appl. Electrochem. 1989, 19, 19–27.
(10) okada, t.; Gotou, S.; Yoshida, M.n.; Yuasa, M.; Hi-
rose, t.; Sekine, i. J. Inorg. Organometal. Polym. 1999,
9, 199–204.
(11) Bouwkamp-Wijnoltz, AL.; Visscher, W.; van Veen,
J.A.R. Electrochim. Acta 1999, 45, 379–386.
(12) Goifman, A.; Gun, J.; Gitis, V.; Kamyshny, A.; Lev,
o.; Donner, J.; Bornick, H.; Worch, E. Appl. Catal. B:
Environ. 2004, 54, 225–235.
(13) Goifman, A.; Gun, J.; Gelman, f.; Ekeltchik, i.; Lev,
o.; Donner, J.; Börnick, H.; Worch, E. Appl. Catal. B:
Environ. 2006, 63, 296–304.
(14) MacRae, A.W.; farrimond, M.; Hodsoll, J.; falahee,
M.; Simms, J.; Hopper, R. Water Sci. Technol. 1999, 40,
149–152.
Concluding Remarks
pyrolyzed carbon-supported nitrogen-donor cobalt
catalysts are a mature, well-researched class of electro-
catalysts and an emerging class of chemical catalysts.
the article shows that in addition to the already proven
efficient hydrogen sulfide catalysis and the proven elec-
trocatalytic activity of these materials, they are efficient
oxidation catalysts for thiol–disulfide conversions. Ap-
parently, the catalysis does not go through free radical
formation, which, on the one hand, precludes the ap-
plication of this catalysis for organic degradation and
decontamination of complex mixtures of contaminants,
but, on the other hand, may open the door for fine chem-
istry-targeted catalytic reactions.
it is known that cobalt is the least active transition
metal catalyst for thiol oxidation, and the fast, efficient
catalysis demonstrated in this study is indeed most en-
couraging because it may be extended to other metals
and other nitrogen donors as well.
(15) Ginzburg, B.; Dor, i.; Lev, o. Water Sci. Technol. 1999,
40, 65–71.
(16) Steudel, R. Ind. Eng. Chem. Res. 1996, 35, 1417–
1423.
(17) Bagiyan, G.A.; Koroleva, I.K.; Soroka, N.V.; Ufimtsev,
A.V. Russ. J. Appl. Chem. 2003, 76, 90–96.
(18) Sharipov, A.K.; Kovutenko, S.V. Neftekhimiya 1997,
37, 563–567.
Acknowledgments. JG, EW, and AG thankfully acknowledge
the financial help of the BMBF Germany–MOS Israel water
technology program. fG and iE thank the Valazzi-pikovsky
Fellowship Fund. OL and AG gratefully acknowledge the fi-
nancial support of the israel Science foundation.
(19) Bagiyan, G.A.; Koroleva, I.K.; Soroka, N.V.; Ufimtsev,
A.V. Kinet. Catal. 2004, 45, 372–380.
rEfErEncEs And notEs
(1) Sheldon, R.A., Ed. Metalloporphyrins in Catalytic Oxi-
dations; Marcel Dekker: new York, 1994.
(2) Mansuy, D.; Battioni, p. in Porphyrin Handbook;
Kadish, K.M.; Smith, K.M.; Guilard, R., Eds.; Aca-
demic press: San Diego, 2000; Vol. 4, pp 1–15.
(3) Boulatov, R. Pure Appl. Chem. 2004, 76, 303–319.
(4) Jasinsky, R. Nature 1964, 201, 1212.
(5) Jahnke, H.; Schoenborn, M.; Zimmermann, G. Top.
Curr. Chem. 1976, 61,133–181.
(6) Bagotzky, V.S.; tarasevich, M.R.; Radyushkina, K.A.;
Levina, o.A.; Andrusova, S.i. J. Power Sources 1977,
2, 233–240.
(20) Bagiyan, G.A.; Koroleva, I.K.; Soroka, N.V.; Ufimtsev,
A.V. Russ. Chem. Bull., Int. Ed. 2003, 52, 1135–1141.
(21) Chaudhuri, A. Curr. Sci. 1995, 68, 692–698.
(22) Shaabani, A.; Lee, D.G. Tetrahedron Lett. 2001, 42,
5833–5836.
(23) APHA, AWWA, WEF Standard Methods for the Analy-
sis of Water and Wastewater; American public Health
Association: Washington, DC, 1998.
(24) Vairavamurthy, A.; Mopper, K. Anal. Chim. Acta 1990,
236, 363–370.
(25) Hu, X.; Huang, Z.; Gu, G.; Wang, L.; Chen, B. J. Mol.
Catal. A: Chem. 1989, 132, 171–179.
(7) for example, the israeli company, Arotech http://www.
(26) Birman, S. Biochem. J. 1985, 225, 825.
Israel Journal of Chemistry
46
2006