The role of weak Lewis acid sites for methanol thiolation

Article abstract of DOI:10.1039/c8cy02250j

Weak Lewis acid sites combined with strong base sites of Cs⁺ supported on WS₂ and γ-Al₂O₃ act as active sites in the thiolation of methanol. The acid-base pairs dissociate methanol upon adsorption. The formed surface alcoholate and the corresponding sulfuryl groups enable the substitution of oxygen for sulfur in a Langmuir-Hinshelwood mechanism. Stronger Lewis acid sites catalyze dimethyl ether formation via the Eley-Rideal mechanism in which methoxy groups react with gas phase methanol. The results demonstrate the importance of adjusting the acid-base strength in oxides to selectively catalyze substitution reactions.

Full text of DOI:10.1039/c8cy02250j

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
The role of weak Lewis acid sites for methanol
thiolation†
Cite this: DOI: 10.1039/c8cy02250j
a
Manuel Weber-Stockbauer,a Oliver Y. Gutiérrez,
‡
a
ab
*
*
Ricardo Bermejo-Deval
and Johannes A. Lercher
Weak Lewis acid sites combined with strong base sites of Cs+ supported on WS2 and γ-Al2O3 act as active
sites in the thiolation of methanol. The acid–base pairs dissociate methanol upon adsorption. The formed
surface alcoholate and the corresponding sulfuryl groups enable the substitution of oxygen for sulfur in a
Langmuir–Hinshelwood mechanism. Stronger Lewis acid sites catalyze dimethyl ether formation via the
Eley–Rideal mechanism in which methoxy groups react with gas phase methanol. The results demonstrate
the importance of adjusting the acid–base strength in oxides to selectively catalyze substitution reactions.
Received 30th October 2018,
Accepted 6th December 2018
DOI: 10.1039/c8cy02250j
rsc.li/catalysis
ments in order to derive a mechanism and the associated ki-
netic parameters. This insight should guide the next genera-
Introduction
The synthesis of methanethiol is a key intermediate step in
the industrial production of the amino acid methionine,1–4
petrochemicals, pesticides and pharmaceuticals.5,6 An effec-
tive route is the direct use of syngas and H2S, with the forma-
tion of chemical intermediate CS2 and other by-products.6–10
Another route is the thiolation of methanol, which involves
the simultaneous addition of the SH− nucleophile and the
elimination of the hydroxyl species in a concerted step (SN2
nucleophilic substitution). However, it is challenging to re-
place the strongly basic hydroxide ion (HO−) with a weaker
base, such as SH−.
A broad variety of mixed metal oxides have been proposed
as catalysts.11–17 High basic strength induced by alkali cat-
ions, such as Cs+, is required to achieve high selectivity.18,19
As the nature of the support for Cs+ appears not to alter its
catalytic activity in methanol thiolation and even sulfide ma-
terials have not induced marked changes, it is hypothesized
that the catalyzed steps take place on cesium chalcogenide
domains supported on oxides or sulfides.20 Thus, we decided
to explore the mechanism and kinetics of methanol thio-
lation on supported Cs+ on WS2 and γ-Al2O3. Detailed physi-
cochemical characterization of the surface properties by
probe molecules is combined with detailed kinetic measure-
tion of catalysts for the synthesis of thiols via nucleophilic
substitution reactions.
Experimental
Catalyst preparation
Catalysts were prepared by incipient wetness impregnation of
γ-Al2O3 (grain size of 0.15–0.25 mm) with aqueous solutions,
added dropwise to the agitated solid. For the synthesis of
CsW/Al2O3, the wt% composition was within the optimum ac-
tivity and selectivity to methanethiol.5 CsW/Al2O3 was synthe-
sized using 5.0 g of γ-Al2O3 with a target tungsten loading of
20.5 wt%, using a solution of 1.94 g of ammonium meta-
tungstate hydrate (Sigma Aldrich, 99.99%) in 4 mL of H2O.
The sample was dried at room temperature overnight. After
drying, the sample was calcined at 455 °C for 4 h, with an in-
crement of 5 °C min−1. Subsequently, 2 g of this solid sample
were impregnated with 640 mg of cesium acetate (Sigma Al-
drich, ≥99.99%) dissolved in 1.6 mL of H2O. The catalyst was
then dried and calcined as before. Cs/Al2O3 was prepared fol-
lowing the same procedure without the use of ammonium
metatungstate. All samples were activated by treatment in
H2S with a flow rate of 20 mL min−1 at 360 °C for 2 h.
Cs2WS4/Al2O3 was synthesized as follows: Cs2WS4 crystals
were formed by precipitation, mixing a solution of 350 mg of
(NH4)2WS4 in 20 mL of H2O and 325 mg of Cs2CO3 (Sigma Al-
drich, 99%) in 20 mL of H2O. A yellow precipitate was
formed. These solids were filtered and washed with ice cold
water and 1-propanol. Due to the low solubility of Cs2WS4,
450 mg of these were dissolved in 150 mL of water. Then 2 g
of Cs/Al2O3 were added to the solution. The water was elimi-
nated by evaporation in continuous rotation, precipitating
the Cs2WS4 crystals on the solid sample. The sample was
a Technische Universität München, Department of Chemistry and Catalysis
Research Center, Lichtenbergstrasse 4, D-84747 Garching, Germany.
E-mail: ricardo.bermejo@tum.de, Johannes.Lercher@ch.tum.de;
Fax: +0049 89 28913544; Tel: +0049 89 289 13560, +0049 89 28913540
b Institute for Integrated Catalysis, Pacific Northwest National Laboratory, 902
Battelle Boulevard, Richland, WA 99352, USA
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c8cy02250j
‡ Present address: Institute for Integrated Catalysis, Pacific Northwest National
Laboratory, Richland, Washington 99354, USA.
This journal is © The Royal Society of Chemistry 2018
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
dried at room temperature overnight, calcined and activated
in H2S.
to obtain each spectrum) and presented after background
subtraction and normalization to the mass of the wafer. The
concentrations of coordinating pyridine were calculated
using a molar extinction coefficient of 0.96 cm μmol−1 deter-
mined for the characteristic band at 1450 cm−1.21
Chemical and physicochemical characterization
The BET surface area and the pore size distribution of the
catalysts were determined by N2 adsorption–desorption at
−196 °C using a PMI Automated BET Sorptomatic 1900 Series
instrument. Prior to the adsorption, the samples were evacu-
ated at 250 °C for 2 h. Elemental analysis was carried out in
the Microanalytical Laboratory at the TU München. The crys-
talline structure of the catalysts was determined by powder
X-ray diffraction. XRD patterns were collected with a Philips
X'Pert System (Cu Kα radiation, 0.1542 nm) operating at 45
kV/40 mA, using a nickel Kβ-filter and solid-state detector
(X'Celerator). The measurements were carried out with a step
size of 0.017° and scan time of 0.31 s per step. Raman spec-
tra of active catalysts were recorded with a Renishaw Raman
Spectrometer (Type 1000), equipped with a CCD detector, a
Leica microscope and a 514 nm Ar laser under ambient con-
ditions shortly after sulfidation conditions. The calibration
was done using a Si (111) crystal prior to the measurements.
The wavenumber accuracy was within 1 cm−1. Adsorption
followed by temperature programmed desorption of H2S was
performed with a pulse technique using a flow apparatus
equipped with a mass spectrometer (QME 200, Pfeiffer Vac-
uum). A sample of 100 mg of catalyst was loaded in a quartz
reactor and activated in situ under 4.2 vol% H2S/He with a
flow of 6 mL min−1 at 360 °C for 2 h. For H2S adsorption, the
temperature was set to 360 °C and the sample was flushed
with He for 1 h prior to adsorption. Pulses of 4.4 vol% of H2S
in He (VTotal = 3.1 mL, duration ∼1 s) were introduced every
30 min (5.0 μmol min−1 of H2S) during continuous He flow (6
mL min−1). The total concentration of gas adsorbed was cal-
culated as the sum of the uptakes per pulse. 24 pulses were
performed per experiment.
The adsorption of pyridine and CH3OH was monitored by
IR spectroscopy in the transmission absorption mode (sam-
ples were pressed into self-supporting wafers). All studied
materials were treated in situ with H2S prior to the measure-
ments, using a gas flow system. Once the wafer was in place,
the cell was purged with He (10 mL min−1) and the tempera-
ture was increased from 50 °C to 360 °C (increment of 10 °C
min−1). After reaching 360 °C, sulfidation took place, using
10 vol% H2S in N2 (10 mL min−1) for 1.5 h. After sulfidation,
the cell was flushed with He for 0.1 h, followed by evacuation
to 10−5 mbar. Methanol was adsorbed at 50 °C, stepwise in-
creasing the methanol partial pressure (0.1 mbar, 0.5 mbar
and 1 mbar) followed by an increase in the temperature to
300 °C. For pyridine adsorption the cell was cooled down to
50 °C and the sample was exposed to 1 mbar of pyridine,
followed by decreasing the pyridine partial pressure. Further
evacuation to 10−5 mbar resulted in no pyridine adsorbed on
Cs/Al2O3 and CsW/Al2O3. Thus, spectra from different cata-
lysts were compared at 0.1 mbar. The spectra were recorded
on a Nicolet 6700 FTIR spectrometer (64 scans were collected
Catalytic testing and kinetic experiments
Kinetic experiments at varying temperatures were performed
using 0.125 g catalyst (granule diameter 0.15–0.25 mm) di-
luted with 1 g of SiC (granule diameter 0.075–0.100 mm) at 9
bar, in the temperature range of 300–360 °C. Standard calcu-
lations of the Weisz–Prater modulus showed that it was <1
for all the catalysts under all conditions, and, therefore, it
can be concluded that the kinetics results were unaffected by
internal mass transfer effects.22 For on-line analysis
a
Shimadzu GC 2014 equipped with a HSQ 80 column and a
TCD detector was used. GC samples were taken after 4 h of
the reaction running at a defined temperature, when steady
state was reached. The reactant flows were: CH3OH 10 mL
min−1 in the gas phase, H2S 20 mL min−1, and N2 20 mL
min−1. Before testing, the catalysts were activated by flushing
with 20 mL min−1 of pure H2S at 360 °C and 9 bar for 2 h.
Experiments at varying residence times were performed at
360 °C with a catalyst loading of 0.05–0.2 g and a total pres-
sure of 9 bar, using the same concentration of reactants as
described before. Reaction orders were determined at 360 °C.
For reaction order in H2S, the partial pressure of methanol
was kept constant at 2.2 bar, while the H2S partial pressure
was varied between 1.1 and 5.6 bar. To measure the metha-
nol reaction orders, the H2S partial pressure was set to 4.5
bar and the methanol partial pressure varied from 0.6 to 2.2
bar gaseous methanol. The N2 gas flow was adjusted to com-
pensate the volume flow changes and keep the total volume
flow constant at 80 mL min−1. The amount of the catalyst
used in each experiment was adjusted accordingly to ensure
a methanol conversion lower than 10%. Reaction orders for
cesium-modified materials were measured with 0.01 g of cat-
alyst, while 0.001 g was sufficient for γ-Al2O3. To avoid
channeling effects, γ-Al2O3 was physically mixed with SiO2,
being known to be inactive in the studied reaction, in a ratio
of 1 : 9.
Results and discussion
Physicochemical properties
The corresponding X-ray diffractograms of the catalysts inves-
tigated are compiled in the ESI (Fig. S1†). All of the materials
exhibited reflections of the γ-Al2O3 support (PDF no. 96-101-
0462). The X-ray diffractogram of Cs2WS4/Al2O3 showed a pat-
tern consistent with the Cs2WS4 phase (PDF no. 96-221-4428),
while CsW/Al2O3 exhibited both the Cs2WS4 and the WS2
(PDF no. 96-591-0004) phases.
The presence of sharp reflections in the Cs2WS4/Al2O3
diffractogram indicates the presence of large crystalline
Cs2WS4 domains.
A sharp peak was observed in the
diffractogram of Cs2WS4/Al2O3 at 27.9 2Θ corresponding to
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Catalysis Science & Technology
Paper
the 004 basal plane, indicative of crystal growth on that
plane. The average particle diameter of the Cs2WS4 crystals
was approximately 100 nm (calculated from the XRD data by
the Scherrer equation). Due to the large particle size of
Cs2WS4, these crystals were concluded to be precipitated on
the surface of Cs/Al2O3. Such large crystals would not cover
the alumina surface. In contrast, the lower intensity of
Cs2WS4 and WS2 for CsW/Al2O3 suggests the presence of
much smaller crystals. Cs2CO3 bands were observed on all
catalysts prior to sulfidation (not shown), however, H2S treat-
ment removed these phases.
In the Raman spectrum of Cs2WS4/Al2O3 (Fig. 1), the
bands at 482 and 459 cm−1 have been attributed to νasIJWS)
and νsIJWS).23–25 This confirms the presence of the Cs2WS4
phase in addition to Cs+ supported on the Al2O3 support, as
observed in the X-ray diffractogram. In the case of CsW/
Al2O3, two additional bands appear at 418 and 352 cm−1,
which are characteristic of WS2 vibration modes.26,27 The first
band corresponds to the first-order A1g (T) optical mode,
while the latter corresponds to an overlap between the first-
order E12g (T) optical mode and the second order longitudinal
phonons (2LA (M)). Both the X-ray diffractograms and Raman
spectra of CsW/Al2O3 showed a much higher intensity of WS2
than of Cs2WS4. This suggests that during sulfidation Cs+ is
segregating, forming an apparent separate phase after
sulfidation.7,28 Surprisingly, this segregation into two phases
reversed after reaction (Fig. S4†). The concentration of
Cs2WS4 was significantly higher than that of the WS2 phase
as observed from the Raman spectra. It is unclear at present
whether this change is atomistically related to the methanol
thiolation. In addition to the Cs+ cation balancing the nega-
tive charge of WS4−2, other sulfide species could be formed
interacting with the WS2 phase.20
In the absence of W, other bands appear for Cs/Al2O3
(Fig. 1) in the region between 1100–200 cm−1 which are
assigned to different sulfur oxyanions such as sulfite (SO3−2),
−2
thiosulfate (S2O3−2), dithionate (S2O6−2), pyrosulfite (S2O5
)
and dithionite (S2O4−2).29–32 As it has been shown for differ-
ent alkali cations supported on γ–Al2O3,18 these sulfur
oxyanions can be formed in the presence of water as an oxi-
dizing agent following the elementary reactions proposed in
Fig. S2.† These sulfur oxyanions were formed on the γ-Al2O3
support only in the presence of H2S.32 Additional bands be-
tween 800 and 500 cm−1 of the CsW/Al2O3 catalyst correspond
to additional S–O bonds of the sulfur oxyanions in the pres-
ence of WS2 and Cs2WS4 phases.33 These bands were much
weaker in intensity than the W–S bands previously discussed.
IR spectroscopy confirmed the presence of the sulfur
oxyanions in all three catalysts (Cs2WS4/Al2O3, CsW/Al2O3 and
Cs/Al2O3, Fig. S3†). The surface sulfates on γ-Al2O3 and WS2
(in CsW/Al2O3) are balancing the Cs+, W4+ or Al3+ cations,
providing oxygen atoms to form Lewis acid–base pairs with
different charge distributions. Small cesium clusters were not
observed with the current physicochemical characterization
methods.
Using IR spectroscopy, previous studies have shown that
H2S adsorbs dissociatively on both Cs/Al2O3 and CsW/Al2O3.20
In addition, Cs atoms on γ-Al2O3, WS2 or Ru increased the
surface basicity, enhancing the adsorption of the H2S
molecule.18,20,34–36 The Cs/Al2O3 catalyst resulted in the
highest uptake of H2S per unit area (0.20 μmol m−2, Table 1).
The addition of the Cs2WS4 crystals to Cs/Al2O3 (Cs2WS4/
Al2O3) does not seem to enhance the H2S adsorption capacity
(0.13 μmol m−2), however these crystals decrease the concen-
tration of H2S on the surface. The uptake of H2S on CsW/
Al2O3 (0.16 μmol m−2) was lower than that on the Cs/Al2O3
catalyst. The involvement of some of the Cs+ in the formation
of Cs2WS4 is hypothesized to decrease the availability of Cs+
to adsorb H2S in the CsW/Al2O3 catalyst. We also hypothesize
that in Cs2WS4 Cs+ interacts only weakly with H2S, due to the
low reactivity of the Cs2WS4 phase in methanol thiolation
(see below).
Acid–base sites
The IR spectra of adsorbed pyridine were used to evaluate
qualitatively and quantitatively the acidity of the different
materials (Fig. 2). For qualitative evaluation, the assignment
by Morterra et al.37 was used. The bands between 1580 and
1620 cm−1 are assigned to the 8a ring vibration mode of
coordinatively bound pyridine to LAS in γ-Al2O3, i.e., strong
LAS associated with a vacancy in tetrahedral coordination
Fig. 1 Raman spectra of a) Cs2WS4/γ-Al2O3, b) CsW/γ-Al2O3 and c) Cs/
γ-Al2O3. The symbols (*) and (§) are assigned to the WS4 and WS2
phases. The following symbols are for the anions: sulfite (green
triangle), thiosulfate (blue square), dithionate (black diamond) and
pyrosulfite (orange dot).
2−
This journal is © The Royal Society of Chemistry 2018
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Table 1 Physical and adsorption properties of Cs2WS4/Al2O3, CsW/Al2O3, Cs/γ-Al2O3 and γ-Al2O3. From left to right: W and Cs content, pore volume,
BET surface area and H2S adsorbed
Catalyst
W (wt%)
Cs (wt%)
Pore vol. (cm3 g−1
)
BET SA (m2 g−1
)
Ads. H2S (μmol g−1
)
Ads. H2S (μmol m−2
)
Cs2WS4/Al2O3
CsW/Al2O3
Cs/Al2O3
5.1
16.4
—
20.6
15.2
15.6
—
0.20
0.11
0.24
0.34
141
68
117
222
18
11
22
—
0.13
0.16
0.20
—
γ-Al2O3
—
(>1600 cm−1); weak LAS associated with a vacancy in octahe-
dral coordination (<1600 cm−1) and H-bond pyridine,
adsorbed on weakly acidic surface hydroxyl groups. The band
at 1573 cm−1 is attributed to the 8b vibrational mode. After
the addition of Cs+ to the γ-Al2O3 support, a new band at a
lower wavenumber (1581 cm−1) was observed. It is attributed
to pyridine adsorbed on weak LAS, induced by the interaction
of the Cs+ cation with the Al2O3 support. This is also observed
on CsW/Al2O3, showing the dominance of the weak LAS of
the Cs+ cation with the support.
methanol thiolation selectivity, since these sites have differ-
ent chemical affinity to adsorb the reactants CH3OH and H2S
(Chart 1). The presence of Cs+ increases the basicity, due to
its lower Sanderson electronegativity.38,39 In addition, the
presence of large Cs+ cations blocks the access to stronger
Lewis acid sites of γ-Al2O3 and/or WS2.
Due to the changes in the acid–base properties of γ-Al2O3
and WS2 upon Cs addition, we would like to discern the
chemical interaction of methanol with these Lewis acid–base
pairs prior to the mechanistic studies of methanol
conversion.
The bands at 1450 cm−1 were used for quantitative and
qualitative analysis of the concentration of LAS on the cata-
lysts.20 The bands at 1448 and 1444 cm−1 correspond to
strong LAS of γ–Al2O3, while the presence of Cs+ results in a
band of lower wavenumber (1440 cm−1). The Lewis acid site
concentration decreased from 680 μmol g−1 for γ-Al2O3 to 490
μmol g−1 for Cs/Al2O3. In the case of CsW/Al2O3, a lower con-
centration of acid sites (38 μmol g−1) was observed.20
Adsorption of CH3OH
The IR bands of methanol adsorbed on CsW/Al2O3, Cs/Al2O3
and γ-Al2O3 between 2950 and 2800 cm−1 are assigned to the
asymmetric and symmetric CH3 stretching vibrations (νas and
νs, respectively), split up by Fermi resonance with overtones of
the methyl bending vibrations (Fig. 3).40 The spectra suggest
The high LAS heterogeneity observed in γ-Al2O3 was also
observed previously on the surface of WS2/Al2O3.20 This
heterogeneity in Lewis acid strength will be reflected in the
Chart 1 Schematic representation of the adsorption of the substrates
methanol and H2S on the different catalysts A) Cs/Al2O3 and B)
γ-Al2O3. The H, C, O, S, Al and Cs atoms are presented in white, grey,
red, yellow, blue and purple, respectively.
Fig. 2 Spectra of pyridine adsorbed on a) γ-Al2O3, b) Cs/Al2O3 and c)
CsW/Al2O3 (50 °C and evacuated). An inset shows an expansion of the
spectrum of CsW/Al2O3.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Catalysis Science & Technology
Paper
Fig. 4 Methanol consumption rate for Cs2WS4/γ-Al2O3, CsW/γ-Al2O3,
Cs/Al2O3 and γ-Al2O3 between 300 and 360 °C.
Kinetics of methanethiol formation
The rates of methanol consumption and methanethiol forma-
tion between 300 and 360 °C for Cs2WS4/Al2O3, CsW/Al2O3,
Cs/Al2O3 and γ-Al2O3 are shown in Fig. 4 and 5, respectively.
The highest rate of methanol consumption was observed for
γ-Al2O3 (2.3–5.0 × 10−5 molCH OH s−1 gcat−1), i.e. 2–3 times
3
Fig. 3 Spectra of methanol adsorbed on a) CsW/Al2O3, b) Cs/Al2O3
and c) γ-Al2O3 (50 °C and 0.1 mbar methanol partial pressure). An inset
at the top left shows the adsorption of methanol on Lewis acid–base
pairs.41
higher than that for the other catalysts. The rate of formation
of all products increased with temperature, except for di-
methyl ether (DME) that passed through a maximum be-
tween 320 and 340 °C (Fig. S8†). This rate was at least 2 or-
ders of magnitude higher with γ-Al2O3 (1.8–2.4
mol(CH ) O s−1 gcat−1) than with CsW/Al2O3 (0.0–1.3 × 10−7
mol(CH ) O s−1 gcat−1). DME was not observed with Cs/Al2O3
×
10−5
3
2
that methanol adsorbs on both strong Lewis acid and strong
basic sites in CsW/Al2O3, Cs/Al2O3 and γ-Al2O3 (Fig. S5–S7).3
The molecular adsorption of methanol is concluded to take
place on strong Lewis acid sites at 50 °C.42 Upon heating, disso-
ciation is favored, leading to the formation of bridging
methoxides in the strong Lewis acid sites (Species I, 2940 cm−1
and 2840 cm−1), mainly observed in γ-Al2O3 (Fig. S7†).
At 50 °C, the acid–base pairs with a weaker Lewis acid cat-
ion and stronger base anion result in the dissociation of the
methanol O–H group and the alcoholate formation (Species
II, 2800–2820 cm−1 and 2930–2945 cm−1).43,44
3
2
and Cs2WS4/Al2O3 catalysts. γ-Al2O3, CsW/Al2O3 and Cs/γ-Al2O3
followed similar trends for the rate of methanethiol (CH3SH)
formation with temperature. The rates of formation of di-
methyl sulfide (DMS, Fig. S9†), dimethyl disulfide (DMDS,
Fig. S10†) and methane (Fig. S11†) were orders of magnitude
lower than that of methanethiol (Fig. 5).
Fig. 6 shows the product yield variations as a function of
methanol conversion at 360 °C on γ-Al2O3. The linear increase
Upon heating, the intensity of the bands characteristic for
methoxy decreased without other major changes in the IR spec-
tra. Upon adsorption at 50 °C and at 10−1 mbar, the alcoholate
formation on the stronger base anion dominated on all cata-
lysts (Fig. 4), in contrast to parent γ-Al2O3. The variation in
wavenumbers between Species II on CsW/Al2O3 and Cs/Al2O3 is
attributed to the differences between the electronegativity of W
and Al. We attribute this to the fact that the base strength of
the former catalysts is significantly higher than that of γ-Al2O3,
allowing for a more facile methanolate formation.
Indeed, the formation of bridging methoxides on strong
Lewis acids and alcoholate on strong bases had equal contri-
butions on WS2/Al2O3 upon evacuation,20 similar to γ-Al2O3.
We hypothesize at this point that the relative concentrations
of methoxide and alcoholate on the catalyst surface influence
the selectivity to methanethiol and dimethyl ether.
Fig. 5 Methanethiol formation rate for Cs2WS4/Al2O3, CsW/Al2O3, Cs/
γ-Al2O3 and γ-Al2O3 between 300 and 360 °C.
This journal is © The Royal Society of Chemistry 2018
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Fig. 7 Product yields during reaction of methanol and H2S over Cs/
Al2O3 (360 °C, 9 bar).
Fig.
6 Product yields during reaction of methanol and H2S over
γ-Al2O3 (360 °C, 9 bar).
of CH3SH and DME at low conversion shows that both are
primary products with γ-Al2O3, as well as with WS2/γ-Al2O3.20
For γ-Al2O3, CH3SH formation from DME was observed at
methanol conversions above 50% (Reaction 6, Scheme 1).
DMS formation is suggested to form via H2S elimination
from CH3SH (Reaction 3, Scheme 1). Similarly, DMDS and
methane (not shown in Fig. 6 and 7) were secondary products
(Reactions 4 and 5, Scheme 1). For Cs/Al2O3, methanethiol
was the only primary product (Fig. 7), since its yield increases
linearly with methanol conversion. At high conversions, the
yields of DMS and DMDS were higher with Cs/Al2O3 than
with γ-Al2O3. The higher selectivity to methanethiol with Cs/
Al2O3 is hypothesized to lead to higher concentrations of
DMS and DMDS. The presence of the W sulfide phases in
CsW/Al2O3 (WS2 and WS42−) and Cs2WS4/Al2O3 (WS42−) in-
creased the hydrogenolysis of methanethiol13,45,46 (Reaction
5, Scheme 1) or methanol to methane.13 However, the rates
were three orders of magnitude lower than those to methane-
thiol. The similarity in catalytic properties of all Cs+
containing catalysts led us to conclude that the strong basic
sites induced by Cs+ were the key factor to achieve high cata-
lytic activity. Surprisingly, the apparent activation energies of
both CsW/Al2O3 and Cs/Al2O3 were lower (67 and 69 kJ mol−1
,
respectively), than that observed with γ-Al2O3 (113 kJ mol−1)
(Table 2).
The observed reaction order of 1.4 in CH3OH (Fig. S12†)
for the formation of dimethyl ether with γ-Al2O3 suggests a
partial coverage of methanol in the Eley–Rideal type bimolec-
ular reaction (ESI† Section 3). The reaction is taking place be-
tween a methoxy group bound to a strong Lewis acid site and
weakly adsorbed methanol.43,47 Upon formation (Step 4,
Scheme 2), DME desorbs from the surface, leaving behind a
proton that recombines with an OH group to form water
(Step 8). The rate determining step (r.d.s) involves the nucleo-
+
philic attack of the weakly sorbed methanol with the CH3
goup.48 Therefore, the rate expression for the formation of
DME is (complete derivation in Section 3 of the ESI†):
1.5
k K CH OH
1 

4
3
(1)
rCH OCH

3
3
a
with a = (1 + K10.5ijCH3OH]0.5 + [H2O]0.5/K80.5).
The observed reaction order was 0.4 for both CH3OH and
H2S in the formation of methanethiol with γ-Al2O3. These
values are close to those expected for a reaction involving dis-
sociated methanol and dissociated hydrogen sulfide (0.5 for
both reactants, ESI† Section 3) in a Langmuir–Hinshelwood
type mechanism. In the formation of CH3SH the r.d.s. in-
volves the nucleophilic substitution of the oxygen atom in
methanolate with SH−(Step 5, Scheme 2).20
Table 2 Apparent activation energies and rates of methanethiol forma-
tion normalized per gram and surface area at 360 °C and 9 bar
Rate CH3SH
Rate CH3SH
(mol s−1 m2)
Ea
Scheme 1 Reaction network for the reaction of methanol with H2S
over Cs-free (the dominant reactions are (1) and (2)) and Cs-modified
catalysts (the dominant reaction is (1)). The numbers in parentheses
correspond to the reaction numbers as described in the text. This fig-
ure is based on Scheme 1 of our previous work.20
(mol s−1 g−1
)
(kJ mol−1
)
CsW/Al2O3
Cs/Al2O3
γ-Al2O3
4.3 × 10−5
4.5 × 10−5
3.1 × 10−4
6.4 × 10−7
3.9 × 10−7
1.4 × 10−6
67
69
113
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Catalysis Science & Technology
Paper
wood reaction mechanism (Scheme 3). Upon addition of Cs+
on the surface of γ-Al2O3 we are able to alter its Lewis acid–
base properties, forming a stronger base (anion) next to the
cation.38,39
To test the hypothesis that only the Cs+ induced basic
sites are causing the catalytic activity, Cs/SiO2 was also ex-
plored. Pure silica did not show activity in methanol thio-
lation. The presence of Cs+ cations on silica, however, in-
duced activity for methanol thiolation (Fig. S14†). We
conclude, therefore, that basic sites associated with alkali cat-
ions, and in particular here with Cs+ suffice to generate
strong basic sites, independent of the specific nature of the
support. Therefore, we are currently working on dispersing
Cs+ on supports without strong LAS and a strong interaction
with the Cs+ cation.
Conclusion
Scheme 2 Elementary reaction steps for the formation of dimethyl
ether (DME) and CH3SH, with (*) being the anion next to a strong
Lewis acid and (c) the cation next to a strong basic site. Both of these
are Lewis acid–base pairs.
Strong basic sites are concluded to be the active sites for
methanol thiolation. The addition of Cs+ enhances the base
strength, independent of the support and the presence of
Al–O and/or W–S bonds. In the absence of Cs+ cations,
γ-Al2O3 also catalyzes the formation of DME. The dehydration
of methanol into DME with γ-Al2O3 follows an Eley–Rideal
mechanism between a methoxy group and a weakly sorbed
methanol on strong Lewis acid sites. Pyridine and methanol
adsorption, as well as the reaction order close to 0.5 in meth-
anol thiolation with γ-Al2O3 and Cs/Al2O3, clearly indicate that
the dissociative adsorption of methanol and hydrogen sulfide
on strong basic sites is necessary for the SN2 nucleophilic
substitution to methanethiol. However, the similar apparent
activation energy and pyridine bands in the IR spectrum with
both Cs/Al2O3 and CsW/Al2O3 (69 and 67 kJ mol−1, respec-
tively), and the higher apparent activation energy for γ-Al2O3,
113 kJ mol−1, show the intrinsic properties of the strong basic
sites to determine the energy barriers in the methanol thio-
lation. Thus, future design of catalysts for methanol thio-
lation should entail only the presence of strong basic sites
and strong interaction of the Cs+ cation with the support.
Scheme 3 A) WS2 phase with Cs+ cations, B) γ-Al2O3 with Cs+ cations,
and C) the proposed mechanism for the thiolation of methanol on
Lewis acid–base pairs formed, regardless of the anion (A) interacting
with the Cs+ cation.
Conflicts of interest
Therefore, the rate expression for the methanethiol forma-
tion is
There are no conflicts to declare.
0.5
0.5
Acknowledgements
k5K2K3 CH OH
H S
2




3
(2)
rCH SH

b2
3
RBD would like to thank the Alexander von Humboldt foun-
dation for financial support.
with b = (1 + K20.5ijCH3OH]0.5 + K30.5ijH2S]0.5 + [CH3SH]0.5/K6
+ [H2O]0.5/K70.5).
0.5
References
The observed reaction order was 0.7 and 0.6 for CH3OH
and H2S for the formation of CH3SH with Cs/Al2O3 (Fig.
S13†), indicating a similar mechanism to the one proposed
with γ-Al2O3. This was also observed previously with WS2/
Al2O3 and CsW/Al2O3,20 with reaction orders of 0.5 in both
CH3OH and H2S (Fig. S12†). Therefore, the methanolate and
the SH species follow the bimolecular Langmuir–Hinshel-
1 Y. Zhang, S. Chen, M. Wu, W. Fang and Y. Yang, Catal.
Commun., 2012, 22, 48–51.
2 D. Kumar and J. Gomes, Biotechnol. Adv., 2005, 23, 41–61.
3 W. Leuchtenberger, K. Huthmacher and K. Drauz, Appl.
Microbiol. Biotechnol., 2005, 69, 1–8.
4 T. Willke, Appl. Microbiol. Biotechnol., 2014, 98, 9893–9914.
This journal is © The Royal Society of Chemistry 2018
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
5 J. Sauer, W. Boeck, L. Von Hippel, W. Burkhardt, S.
Rautenberg, D. Arntz and W. Hofen, US Pat., US5852219,
1998.
6 W. Taifan and J. Baltrusaitis, Catal. Sci. Technol., 2017, 7,
2919–2929.
López-Urías, J.-C. Charlier, H. Terrones and M. Terrones, Sci.
Rep., 2013, 3, 1–8.
27 M. Thripuranthaka, R. V. Kashid, C. Sekhar Rout and D. J.
Late, Appl. Phys. Lett., 2014, 104, 81911.
28 R. R. Chianelli, M. H. Siadati, M. P. De la Rosa, G. Berhault,
J. P. Wilcoxon, R. Bearden and B. L. Abrams, Catal. Rev.: Sci.
Eng., 2006, 48, 1–41.
7 O. Y. Gutiérrez, C. Kaufmann, A. Hrabar, Y. Zhu and J. A.
Lercher, J. Catal., 2011, 280, 264–273.
8 O. Y. Gutiérrez, C. Kaufmann and J. A. Lercher,
ChemCatChem, 2011, 3, 1480–1490.
29 H. Takahashi, N. Kaneko and K. Miwa, Spectrochim. Acta,
Part A, 1982, 38, 1147–1153.
9 C. Kaufmann, O. Y. Gutiérrez, Y. Zhu and J. A. Lercher, Res.
Chem. Intermed., 2010, 36, 211–225.
10 O. Y. Gutiérrez, C. Kaufmann and J. A. Lercher, ACS Catal.,
2011, 1, 1595–1603.
11 M. Ziolek, J. Kujawa, J. Czyzniewska, I. Nowak, A. Aboulayt,
O. Saur and J. C. Lavalley, Appl. Catal., A, 1998, 171,
109–115.
30 P. D. Clark, N. I. Dowling, M. Huang, O. Okemona, G. D.
Butlin, R. Hou and W. S. Kijlstra, Appl. Catal., A, 2002, 235,
61–69.
31 D. Santamara-Pérez, A. Vegas, C. Muehle and M. Jansen,
J. Chem. Phys., 2011, 135, 1–8.
32 J. S. Church and D. J. Evans, Spectrochim. Acta, Part A,
2008, 69, 256–262.
12 M. Ziolek, J. Kujawa, O. Saur and J. C. Lavalley, J. Phys.
Chem., 1993, 97, 9761–9766.
13 T. J. Paskach, G. L. Schrader and R. E. Mccarley, J. Catal.,
2002, 211, 285–295.
33 A. J. Van der Vlies, G. Kishan, J. W. Niemantsverdriet, R.
Prins and T. Weber, J. Phys. Chem. B, 2002, 106, 3449–3457.
34 E. Schachtl, E. Kondratieva, O. Y. Gutiérrez and J. A. Lercher,
J. Phys. Chem. Lett., 2015, 6, 2929–2932.
14 A. V. Mashkina, Russ. Chem. Rev., 1995, 64, 1131–1147.
15 H. Redlingshöfer, C. Weckbecker, L. Dörflein and B.
Rückriegel, US Pat., 7691776B2, 2010.
16 W. M. Wang, J. J. Li, Q. J. He, S. Peng, Y. Q. Yang and M. Li,
ChemistrySelect, 2018, 3, 9663–9671.
17 A. Eftekhari, J. Mater. Chem. A, 2017, 5, 18299–18325.
18 R. Bermejo-Deval, R. M. Walter, J. A. Gutierrez and J. A.
Lercher, Catal. Sci. Technol., 2017, 7, 4437–4443.
19 D. Kiani, G. Belletti, P. Quaino, F. Tielens and J. Baltrusaitis,
J. Phys. Chem. C, 2018, 122, 24190–24201.
35 C. Petit, F. Mauge and J. Lavalley, Stud. Surf. Sci. Catal.,
1997, 106, 157–166.
36 A. Ishihara, J. Lee, F. Dumeignil, R. Higashi, A. Wang, E. W.
Qian and T. Kabe, J. Catal., 2003, 217, 59–68.
37 C. Morterra and G. Magnacca, Catal. Today, 1996, 27,
497–532.
38 R. T. Sanderson, J. Am. Chem. Soc., 1983, 105, 2259–2261.
39 R. T. Sanderson, J. Chem. Educ., 1988, 65, 112–118.
40 J. Derouault, J. Le Calve and M. T. Forel, Spectrochim. Acta,
Part A, 1972, 28, 359–371.
20 A. V. Pashigreva, E. Kondratieva, R. Bermejo-Deval, O. Y.
Gutiérrez and J. A. Lercher, J. Catal., 2017, 345, 308–318.
21 S. M. Maier, A. Jentys and J. A. Lercher, J. Phys. Chem. C,
2011, 115, 8005–8013.
22 K. Bischoff and G. Froment, Chemical Reactor analysis and
design, New York, 1990.
41 L. J. Burcham, L. E. Briand and I. E. Wachs, Langmuir,
2001, 17, 6164–6174.
42 G. Busca, P. F. Rossi, V. Lorenzelli, M. Benaissa, J. Travert
and J. Lavalley, J. Phys. Chem., 1985, 89, 5433–5439.
43 J. H. de Boer, R. B. Fahim, B. G. Linsen, W. J. Visseren and
W. F. N. M. de Vleesschauwer, J. Catal., 1967, 7, 163–172.
44 G. Busca, P. Forzatti, E. Tronconi and J. C. Lavalley, Stud.
Surf. Sci. Catal., 1985, 20, 15–24.
23 B. R. Srinivasan, C. Näther and W. Bensch, Acta Crystallogr.,
Sect. E: Struct. Rep. Online, 2007, 63, 167.
24 M. J. F. Leroy and G. Kaufmann, Bull. Soc. Chim. Fr.,
1966, 3090.
45 T. Todorova, R. Prins and T. Weber, J. Catal., 2005, 236,
190–204.
25 I. R. Beattie, P. J. Jones, A. D. Willson and N. A. Young,
J. Chem. Soc., Dalton Trans., 1988, 2047–2049.
26 A. Berkdemir, H. R. Gutiérrez, A. R. Botello-Méndez, N.
Perea-López, A. L. Elías, C.-I. Chia, B. Wang, V. H. Crespi, F.
46 R. L. Wilson and C. Kemball, J. Catal., 1964, 3, 426–437.
47 S. Buchang and B. H. Davis, J. Catal., 1995, 157, 359–367.
48 Z. Zuo, W. Huang, P. Han, Z. Gao and Z. Li, Appl. Catal., A,
2011, 408, 130–136.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2018