S.A. Schmidt et al. / Applied Catalysis A: General 490 (2015) 117–127
119
2.3. Methyl chloride synthesis
lower than two, indicating that chlorine was removed from the
catalyst during calcination and released as HCl. With an increas-
ing zinc loading the Cl/Zn ratio was lowered, from 0.78 at 2.4 wt%
Zn to 0.31 at a nominal zinc loading of 25 wt%. This indicates
that most of the zinc is not present as ZnCl2 on the surface but
reacts with the alumina surface or forms separate particles of an
oxygen-containing zinc species as ZnO or a zinc chloride hydrox-
The reaction was performed in a quartz-made tubular reactor
with an inner diameter of 1 cm. 0.05 g of catalyst were loaded into
the reactor as a fixed bed between quartz sand and quartz wool.
The reactor was equipped with a thermocouple positioned close
to the catalytic bed to determine the reaction temperature. The
reaction temperature was 210 ◦C for catalyst screening and 300 ◦C
for catalyst stability tests. The flow rates 10 mL/min He/HCl (AGA,
20.000%), 10 mL/min He and 0.0034 mL/min MeOH (J.T. Baker, HPLC
gradient grade) (volumetric flow rates at room temperature). These
flow rates gave the stoichiometric HCl:MeOH ratio 1:1 and a helium
content of 83% to prevent corrosion. The gas flow was controlled
by mass flow controllers (Bronkhorst EL-flow). The liquid methanol
was pumped using an HPLC pump (Shimadzu LC-20AD) and evap-
orated in the heated lines. All the lines and equipment from the
methanol inlet to the GC were isolated and heated by electrical
heating wires to 96 ◦C. To guarantee homogeneous methanol evap-
oration, a helium flow was used. After the reactor, a neutralization
bottle (106 ◦C) filled with calcium oxide (Fischer, general-purpose
grade) was installed to remove HCl and water from the product gas.
In this way, the corrosion problem was minimized, since water/HCl
injections into the GC were prevented. Samples were withdrawn
with a heated gas-tight syringe (100 ◦C) through a septum in the
sampling section. Prior to the reactor outlet, a washing bottle filled
with a volume based 2:6:2 mixture of water/methanol/ethanol
amine (Sigma–Aldrich, ≥98%) at room temperature was placed
to strip methyl chloride from the reaction product gas. Swagelok
stainless steel lines and valves were used in the equipment. All
conversions and selectivities reported were obtained from steady-
state experiments. Several samples were withdrawn over at least
200 min to observe catalyst stability and guarantee that the reac-
tor operates at a steady state. Selected experiments were repeated
to ensure the reproducibility of the results. The standard devia-
tion of the methanol conversion was lower than 6% and for the
selectivity lower than 1% in all experiments. The main error is
attributed to the GC analysis and integration of the GC spec-
tra.
zinc species on the surface, forming e.g. Al
by Tovar et al. [5]. The nature of zinc on the catalyst surface is further
discussed in the following sections.
The surface areas and pore volumes of neat and zinc-modified
the surface area and the pore volume are steadily lowered with
increasing zinc loading due to pore blocking by bulk zinc species
or eventually also due to alumina hydrolysis during impregnation.
Hydrolysis of alumina during preparations in aqueous solutions has
been reported in the literature [11,12].
The acidity of the neat and zinc-modified aluminas was mea-
sured by FTIR spectroscopy using pyridine as the probe molecule.
The strength of the acid sites is classified upon the temperature
that pyridine desorbs from the catalyst, i.e. desorption between
250 and 350 ◦C for weak acid sites, 350 and 450 ◦C for medium
acid sites. Strong acid sites retain pyridine at 450 ◦C. Both neat and
zinc-modified aluminas are Lewis acidic and no Brønsted acid sites
are present in the catalyst. The acidity profiles of the aluminas are
given in Table 3. Upon zinc modification with 2.4 wt% zinc, the share
of medium strength acid sites is slightly increased, while upon a
further increase of zinc loading, the percentage of weak acid sites
increases. The total amount of Lewis acid sites increases upon zinc
modification and goes through a maximum at a loading of 5.0 wt%.
The decrease in the number of acid sites at higher loadings can be
explained by the lowering of total surface area.
3.2. Characterization by FTIR spectroscopy, X-ray powder
diffraction (XRD) and TEM analysis
3.2.1. FTIR spectroscopy of OH-stretching region and pyridine
adsorbed on the catalyst
The gaseous reaction products obtained were manually injected
to the GC. The gas chromatographic analysis of the samples was
carried out using an Agilent 6890N GC equipped with an FI detec-
tor and an HP-Plot/U Column (30 m, I.D. 0.530 mm, film thickness:
20 m, T = 110 ◦C). The system was calibrated for methyl chloride,
dimethyl ether and methanol using mixtures of known concen-
trations. The peaks of all products were calibrated with respect
to the methanol peak. The concentration of the separate compo-
nents were thus calculated using the obtained response factors,
the methanol/component peak ratio and the initial methanol flow.
The carbon balances are consequently 100% closed.
To get a more detailed picture of the alumina surface modifica-
tion by zinc, the OH-stretching region of neat and zinc-modified
dine adsorbed on the catalyst. The surface hydroxyl groups of
transitional aluminas have been extensively studied in the liter-
ature using IR spectroscopy and quantum chemical calculations in
order to understand its catalytic properties [13–15]. Furthermore,
the adsorption of pyridine on the Lewis acid sites of alumina has
tions have been carried out in this field, the precise assignment
of the vibrational bands to a specific surface site is not unequiv-
ocally established. The field has recently been reviewed by Busca
[19] and the interpretation of the observed vibrational modes in
the following discussion is based on this review. As observed by
Liu and Truitt [18] and further described by Lundie et al. [16] Lewis
acid sites on alumina have neighboring OH groups that have spe-
cific vibrational modes. Lewis acid sites with a specific strength can
be correlated to a surface hydroxyl mode with a specific wavenum-
ber. Thus, observed OH vibrational modes can serve as an indirect
indicator for Lewis acidity of alumina. The association of surface
on the work of Lundie et al. [16].
2.4. Calculation of methanol conversion and methyl chloride
selectivity
The methanol conversion and the methyl chloride selectivity
were calculated as follows (no products were present in the feed):
XMeOH = (c0MeOH − cMeOH)/c0MeOH and SMeCl = cMeCl/(cMeCl + 2cDME).
3. Results and discussion
Zn/alumina catalysts were prepared with nominal Zn loadings
from 2.5 to 25 wt% using ZnCl2 as precursor. The content of Zn in
wt% as well as the Cl/Zn and Al/Zn ratios determined by EDX are
given in Table 2. For all catalysts, the Cl/Zn ratio is significantly
3.2.2. Pyridine vibrational modes
The IR-spectra of pyridine absorbed on neat and zinc-modified
alumina are shown in Fig. 1. It is clearly visible that neat alu-
mina has mainly two kinds of acid sites, characterized by two