Selective formation of chloroethane by the hydrochlorination of ethene using zinc catalysts

Article abstract of DOI:10.1016/j.jcat.2007.09.002

A detailed study of the hydrochlorination of ethene and higher alkenes using ZnCl₂/SiO₂ and ZnCl₂/Al₂O₃ catalysts is described and discussed. Based on earlier observations that supported gold can be an active catalyst for both ethyne hydrochlorination and oxidation reactions, we initially investigated using supported gold as catalysts for the oxychlorination of ethene. However, we found that oxychlorination did not occur in the presence of oxygen and, furthermore, that the gold acted as a poison/inhibitor during the initial reaction period, with the underlying reaction being ethene hydrochlorination. Supported Zn²⁺ was found to be a very effective catalyst for this reaction. The hydrochlorination of higher alkenes occurred, with high selectivity to a range of relatively complex chlorinated hydrocarbon products at rates of ca. 10-13 mol/(kg_cat h).

Full text of DOI:10.1016/j.jcat.2007.09.002

Journal of Catalysis 252 (2007) 23–29
www.elsevier.com/locate/jcat
Selective formation of chloroethane by the hydrochlorination of ethene
using zinc catalysts
a
a
a
b
Marco Conte , Thomas Davies , Albert F. Carley , Andrew A. Herzing ,
b
a,∗
Christopher J. Kiely , Graham J. Hutchings
a
School of Chemistry, Cardiff University, P.O. Box 912, Cardiff, CF10 3AT, UK
b
Department of Materials Science and Engineering, Lehigh University, 5 East Packer Avenue, Bethlehem, PA 18015-3195, USA
Received 16 May 2007; revised 5 September 2007; accepted 9 September 2007
Available online 3 October 2007
Abstract
A detailed study of the hydrochlorination of ethene and higher alkenes using ZnCl /SiO and ZnCl /Al O catalysts is described and discussed.
2
2
2
2 3
Based on earlier observations that supported gold can be an active catalyst for both ethyne hydrochlorination and oxidation reactions, we initially
investigated using supported gold as catalysts for the oxychlorination of ethene. However, we found that oxychlorination did not occur in the
presence of oxygen and, furthermore, that the gold acted as a poison/inhibitor during the initial reaction period, with the underlying reaction being
2+
ethene hydrochlorination. Supported Zn was found to be a very effective catalyst for this reaction. The hydrochlorination of higher alkenes
occurred, with high selectivity to a range of relatively complex chlorinated hydrocarbon products at rates of ca. 10–13 mol/(kgcat h).
©
2007 Elsevier Inc. All rights reserved.
Keywords: Ethene hydrochlorination; Zinc catalysts
1
. Introduction
chloride, today it is manufactured mainly by the hydrochlori-
nation of ethane, performed in a slurry phase batch reactor at
130–250 C and 1 bar, using AlCl3 as a catalyst [3]. Indeed,
◦
Chlorinated hydrocarbons remain important solvents and in-
termediates even with the current trend toward removing sol-
vents from chemical syntheses as one of the foundations of
green chemistry. Among chlorinated hydrocarbons, chloroeth-
ane in particular remains one of the most important. Of course,
a major application of chloroethane has been the manufac-
ture of tetraethyl lead, an antiknock additive for gasoline with
an average production of approximately 10 millions tons per
year [1]. Currently, chloroethane in particular is used in clean-
ing and degreasing applications; however, its most important
use is as a treatment agent for cellulose to manufacture ethyl-
cellulose, with an annual production of some thousands of
tons [1]. Chloroethane has been manufactured by the pyroly-
sis of ethanol and the chlorination of ethene, but both of these
routes have now been superseded [2]. Although chloroethane
can be synthesized as a byproduct from the production of vinyl
AlCl3 is the only catalyst that has been reported for this reaction
to date. Industrially, this process is highly efficient; however,
safety-related problems can be encountered due to the proper-
ties of AlCl . In fact, if traces of water are present, AlCl can be
readily hydrolyzed, which can be potentially explosive [4]. This
potential problem has stimulated an interest in finding alterna-
tive catalysts that are able to carry out the reaction under safer
conditions, while using smaller amounts of the active species.
3
3
Indeed, the role of AlCl is to act as a Lewis acid, but the
3
amount required to carry out the reaction is close to stoichio-
metric [5]. To some extent, this is because the reaction pathway
−
+
involves AlCl3 abstracting Cl , and then H reacts with the
−
−
alkene double bond and AlCl4 donates Cl , so for every HCl
that reacts an AlCl3 is required, however, because the rate is
very slow, stoichiometric amounts of AlCl3 are used, and the
AlCl3 is not a source of Cl in the absence of HCl. Against
this background, we initiated a study to identify a heteroge-
neous catalyst that could be used as a replacement for AlCl3.
*
Corresponding author. Fax: +44 29 2087 4075.
E-mail address: hutch@cf.ac.uk (G.J. Hutchings).
0021-9517/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2007.09.002

24
M. Conte et al. / Journal of Catalysis 252 (2007) 23–29
This paper describes the experimental work that forms the ba-
sis of the identification of ZnO and supported ZnCl2 as very
selective catalysts for the direct hydrochlorination of ethene to
chloroethane.
which the mixture was stirred for another 1 h. After cooling,
the slurry was filtered and washed using approximately 4 L of
◦
distilled water. The solid was then dried at 120 C for 16 h and
◦
calcined in static air (400 C, 4 h) before use.
2
. Experimental
2
.1.4. ZnCl2/SiO2 and ZnCl2/Al2O3 via impregnation
ZnCl2/SiO2 with 15 wt% Zn was prepared by incipient
2
.1. Catalyst preparation
wetness impregnation (BDH, ZnCl2, 265 mg per gram of sil-
ica). The solution was added to the support (7 mL of solu-
tion per gram of silica), and the product thud obtained was
The carbon-supported gold catalyst (1 wt% Au/C) was pre-
pared using an incipient wetness impregnation technique and
aqua regia as a solvent, as described previously [6].
◦
dried (16 h, 80 C). An identical procedure was used to pre-
pare ZnCl2/Al2O3.
2
.1.1. Preparation of Au/MO via deposition precipitation
ZnO (Aldrich, 3.0 g) was stirred in distilled water (150 mL),
x
2
.1.5. Catalyst characterization using transmission electron
microscopy and X-ray photoelectron spectroscopy
and HAuCl4·xH2O (Strem, 62 mg, assay 50.14% in 10 mL wa-
Samples were prepared for transmission electron microscopy
ter) was added dropwise over 2 min, to produce a final Au load-
(
TEM) analysis by dispersing the catalyst powders in high-
◦
ing of 1 wt%. The slurry was then heated to 80 C, and the pH
purity ethanol and allowing a drop of the suspension to dry
on a lacey-carbon film supported on a 300-mesh Cu TEM grid.
Bright-field (BF) imaging and selected-area ring diffraction pat-
terns were acquired using a Jeol 2000FX TEM operating at
was adjusted by adding a saturated Na2CO3 solution, to reach
a final value 9. After stirring for another 1 h and cooling, the
slurry was filtered and washed in approximately 4 L of distilled
◦
water. The solid was then dried at 120 C for 16 h. The same
2
00 keV with a LaB6 filament.
X-ray photoelectron spectra were recorded on a Kratos Axis
procedure was used for the preparation of Au/MgO, Au/La2O3,
Au/Al2O3, and Au/CeO2. All of these catalysts were used as
dried materials; in addition, all of the materials were calcined in
static air (400 C, 4 h) but these calcined materials were found
to be inactive as catalysts.
Ultra DLD spectrometer using a monochromatized AlKα X-ray
source (100 W). Spectra were recorded at analyzer pass ener-
gies of either 160 eV (survey scans) or 40 eV (detailed scans).
Binding energies are referenced to the C(1s) binding energy
of adventitious carbon contamination, which is taken to be
◦
2
.1.2. Preparation of Au/ZnO and Au/MgO via
2
84.7 eV.
co-precipitation
Zn(NO3)2·6H2O (Aldrich, 7.3 g) was dissolved in distilled
2
.2. Catalyst testing
water (150 mL), and HAuCl4·xH2O (Strem 41 mg, assay
5
0.14%, in 10 mL water) was added dropwise over 2 min,
Catalysts were tested for ethene hydrochlorination and oxy-
to reach a final Au loading of 1 wt%. The solution was then
heated to 80 C, and saturated aqueous Na2CO3 was added
◦
hydrochlorination in a fixed-bed glass microreactor operating
just above atmospheric pressure as described previously for
ethyne hydrochlorination. Ethene or higher alkenes (5 mL/min)
and hydrogen chloride (5 mL/min) were fed though a mix-
until a pH of 9 was achieved. The mixture was then stirred
for 3 h, and the resulting precipitate was filtered and dried
◦
for 16 h at 120 C. The same procedure was used to prepare
◦
ing vessel/preheater (70 C) through calibrated mass flow con-
co-precipitated Au/MgO and Au/CeO2 with Mg(NO3)2·H2O
trollers to a heated glass reactor containing catalyst (200 mg),
and Ce(NO3)3·6H2O. Both catalysts were calcined in static air
◦
−1
(
400 C, 4 h).
giving a total GHSV of 6400 h . The pressure of the reactants
both HCl and C2H4) was in the range of 1.1–1.2 bar. The prod-
(
2
.1.3. Au/ZnO/SiO2 via co-precipitation and deposition
ucts were analyzed in two ways. First, the exit gas mixture was
passed through a Dreschel bottle containing NaOH at known
concentration for a predetermined time to measure the conver-
sion of HCl. Alternatively, the gas stream was analyzed by gas
precipitation
ZnO/SiO2 was prepared using a deposition–precipitation
method. Zn(NO3)2·6H2O (Aldrich, 14.6 g) was dissolved in a
slurry containing 250 mL of distilled water and 4 g of SiO2
◦
chromatography (GC). A reaction temperature of 250 C was
(
Aldrich). The amount of precursor was calculated to have a
chosen; blank tests using an empty reactor filled with quartz
wool did not display any catalytic activity with the reactants
under these flow conditions.
final ZnO/SiO2 ratio of 1:1 (by weight). The slurry was then
heated to 80 C and maintained at this temperature for 5 h. The
solid thus obtained was dried for 16 h at 120 C and then cal-
cined in static air (500 C, 5 h) to give ZnO/SiO2.
◦
◦
Oxychlorination of ethene was carried out under the same
temperature and pressure conditions but with a reactant ratio
C2H4:HCl:O2 of 1:2:0.7. Air was used as the source of oxy-
gen and reactant flows of C2H4, HCl, and air of 5, 10, and
17.5 mL/min, respectively, were used, giving a total GHSV
◦
ZnO/SiO2 (2 g) was stirred in 150 mL of distilled water, and
1 mg of HAuCl4·xH2O (Strem, assay 50.14%, in 10 mL wa-
4
ter) was added dropwise over 2 min, for a final Au loading of
1
◦
−1
wt%. The slurry was then heated to 80 C, and saturated aque-
of 20,800 h . Higher alkene reaction products were collected
ous Na2CO3 solution was added until pH 9 was attained, after
in a chloroform trap at the outlet of the catalytic reactor and

M. Conte et al. / Journal of Catalysis 252 (2007) 23–29
25
1
subsequently analyzed by H-NMR using an Avance DPX-400
Spectrospin Bruker instrument operating at 400 MHz.
3
. Results and discussion
3
.1. Oxychlorination of ethene
Initial experiments were carried out using Au/C as the cat-
alyst, because this has been found to be very effective for the
◦
hydrochlorination of ethyne [7–9]. At 180 C or temperatures
◦
up to 250 C, under our standard reaction conditions, this cat-
alyst was inactive for ethane hydrochlorination; indeed, in a
mixture of ethene and ethyne, only the ethyne was reactive. This
preference for the reaction of carbon–carbon triple bonds over
the reaction of carbon–carbon double bonds has been noted
with supported Au catalysts for hydrogenation reactions [10];
this effect is thought to be due to the preferential adsorption
of alkyne at the edges of the Au nanocrystals. In view of this,
we anticipated that oxychlorination could be a viable approach,
because supported Au catalysts have been found to be effec-
tive oxidation catalysts for alkenes [11–13]. We investigated
Fig. 1. Oxychlorination of ethene over Au/MOx catalysts. (2) Au/ZnO DP,
(E) Au/ZnO CP, (ꢀ) Au/ZnO CP clc, (") Au/MgO DP, (a) Au/MgO CP,
(
F) Au/MgO CP clc, (Q) Au/La O DP, (ꢁ) Au/Al O DP clc, (!) Au/CeO
2 3 2 3 2
DP, and (1) Au/CeO CP clc. Where: DP = deposition precipitation; CP =
2
coprecipitation; clc = calcined; all catalysts have a gold loading of 1% except
Au/MgO (3%).
◦
the oxychlorination reaction at 180 and 250 C but observed
only negligible activity. We also investigated Au/TiO2, which is
known to be an effective catalyst in alkene epoxidation [11–13],
but again we observed no reactivity. In view of this finding, we
also investigated Au/MgO (3 wt%) and Au/ZnO (1 wt%), which
are known to be efficient oxidation catalysts [14,15], along with
a range of supported gold catalysts; the results are shown in
Fig. 1. In all these cases, under the relatively short time scale of
our experiments, the surface of the supports and the gold could
be chlorided, arguing by analogy with gold catalysts for ethyne
hydrochlorination where Au3 is the active component [8]. In
addition, the supports were found to be inactive for the oxy-
chlorination of ethene. For both Au/ZnO and Au/MgO, the cat-
alytic activity was observed to increase with reaction time for
the catalyst obtained through deposition precipitation but not
for the catalyst obtained through co-precipitation, which dis-
plays lower activity. In all cases, the only product observed was
chloroethane (see supplementary data Fig. S1 for the data for
the Au/ZnO/SiO2 catalyst). Furthermore, negligible amounts of
CO2 and H2O were observed in these experiments, and no CO
was observed (within the limits of detection). We investigated
the oxidation of ethene under the same reaction conditions and
found only a negligible reaction.
+
Fig. 2. Conversion trend for the (2) Au/ZnO/SiO and (") Au/ZnO DP catalyst
for the hydrochlorination reaction of ethene; conversion trend of (Q) Au/ZnO
DP catalyst in presence of oxygen.
2
•
•
•
•
experiment A:
C2H4/HCl (2 h) → He/HCl (2 h) → C2H4/HCl (2 h),
experiment B:
He/HCl (2 h) → C2H4/HCl (2 h) → He/HCl (2 h),
experiment C:
3
.2. Hydrochlorination of ethene
C2H4/HCl (2 h) → C2H4/He (2 h) → C2H4/HCl (2 h),
experiment D:
In view of the lack of oxidation activity observed with ethene
C2H4/He (2 h) → C2H4/HCl (2 h) → C2H4/He (2 h).
and the observation that the only product observed was from
hydrochlorination, not oxychlorination, we subsequently inves-
tigated the Au/ZnO catalyst obtained through deposition pre-
cipitation and Au/ZnO/SiO2 catalysts, in the absence of O2. The
removal of oxygen from the reaction mixture led to enhanced
activity (Fig. 2). Consequently, we investigated the ethene hy-
drochlorination reaction over Au/ZnO and, in particular, we
carried out a systematic set of experiments to isolate and iden-
tify the effect of each reactant on activity, as follows:
The results, shown in Fig. 3, exhibit the opposite trend to that
observed for the same series of experiments carried out over
Au/C for the hydrochlorination of ethyne [6]. The most signifi-
cant observation concerns experiment C, in which the formation
of chloroethane was improved after treatment with inert gas and
ethene. In contrast, for ethyne hydrochlorination with Au/C [6],
the increase in the yield of the hydrochlorination product was

26
M. Conte et al. / Journal of Catalysis 252 (2007) 23–29
Fig. 4. Hydrochlorination of ethene over (2) ZnO/SiO , (") ZnCl /SiO and
2
2
2
◦
(Q) ZnCl /Al O using the standard reaction conditions at 250 C.
2 2 3
Fig. 3. Hydrochlorination of ethene over Au/ZnO DP: (2) experiment A:
C H /HCl (2 h) → He/HCl (2 h) → C H /HCl (2 h); (") experiment B: He/
2
4
2 4
HCl (2 h) → C H /HCl (2 h) → He/HCl (2 h); (Q) experiment C: C H /
2
4
2 4
Table 1
HCl (2 h) → He/C H (2 h) → C H /HCl (2 h); (a) experiment D: He/C H
2
4
2
4
2 4
Surface molar composition of the supported ZnCl and ZnO catalysts estimated
2
(
2 h) → HCl/C H (2 h) → He/C H (2 h).
2
4
2 4
from the XPS data
Molar ratios
observed only when the intermediate step involved treatment
with inert gas and HCl (i.e., equivalent to experiment A in this
set of experiments). In the present case, experiment A resulted
in no significant changes in the yield of the hydrochlorina-
tion product. From experiment B (He/HCl → C2H4/HCl →
He/HCl), a lower conversion to the hydrochlorination product
was seen than when the HCl pretreatment step was excluded.
This test is useful, because the trend for this catalyst was that
conversion increased with time on stream (Fig. 2), and this ex-
periment indicates that treatment with HCl alone was not the
sole cause of this effect with this catalyst. The observation that
the trends observed in ethene hydrochlorination were opposite
to those observed with the Au/C catalyst for hydrochlorination
of ethyne, suggested to us that the Au component was not an
active species in this reaction. In fact, the bulk structure of all
the materials was modified after the reaction, leading to forma-
tion of ZnCl2 when Au/ZnO was used (see supplementary data
Fig. S2, for the Au/ZnO catalyst obtained through deposition
precipitation).
Cl/Zn
Cl/(Si or Al)
Zn/(Si or Al)
C/(Si or Al)
ZnCl /SiO fresh
1.38
1.24
1.55
1.66
0.19
2.28
0.25
0.23
0.39
0.42
0.05
3.78
0.18
0.18
0.25
0.26
0.26
1.66
0.54
0.40
0.40
0.44
0.51
4.70
2
2
ZnCl /SiO used
2
2
ZnCl /Al O fresh
2
2 3
ZnCl /Al O used
2
2 3
ZnO/SiO fresh
2
ZnO/SiO used
2
ferent zinc component loadings (32 and 9.1 Zn wt% for the
for ZnO/SiO2 and ZnCl2/SiO2, respectively) and surface areas
2
2
(
53 m /g and 104 m /g for ZnO/SiO2 and ZnCl2/SiO2, re-
spectively) gave similar catalytic performance; in addition, ZnO
supported on SiO2 is able to evolve to ZnCl2 (see supplemen-
tary data Fig. S3). To further confirm that ZnCl2 is an active
species, ZnCl2/Al2O3 (10.3 Zn wt%, 57 m /g) was prepared; it
demonstrated the same activity (Fig. 4).
2
XPS analyses of the fresh and used silica- and alumina-
supported ZnCl2 and silica-supported ZnO catalysts corrobo-
rate these conclusions. All of the samples showed the presence
of Zn, Cl, O, C, and Si (Al) with no evidence of any contami-
nants. Quantification of the XPS data leads to the results given
in Table 1. The supported ZnCl2 materials show similar sur-
face compositions, with the alumina-supported samples having
somewhat higher Zn and Cl contents. The ZnO/SiO2 sample
shows very clear evidence for Cl uptake after reaction, leading
to a Cl/Si atomic ratio 1.5–2 times greater than that for the sup-
ported ZnCl2 catalysts. This may be due to enhanced wetting of
the silica induced by chlorination of the ZnO. This sample also
exhibits a significant increase in the C/Si ratio, indicating that
the chlorination of the surface may involve more than a simple
conversion of ZnO to ZnCl2 as phase transitions of ZnCl2.
We consider our observations significant, because there are
To investigate whether ZnCl2 is the active species, samples
of ZnCl2 (nominal amount 15 wt% Zn) supported on SiO2
and ZnO/SiO2 (nominal ratio 1:1 wt%) were tested as catalysts
for the hydrochlorination of ethene under the standard reaction
conditions. The results, shown in Fig. 4, demonstrate that both
catalysts exhibited high activity from the start of the reaction.
This observation implies that Au acted as a poison in this re-
action, because the corresponding Au catalysts were inactive at
the start (Fig. 2). Furthermore, tests carried out using unsup-
ported ZnO, ZnCl2, and Zn(OH)2 showed no activity. In view
of this finding, we consider the active catalyst to be supported
◦
ZnCl2. Because the catalyst is active at 250 C, and the melt-
◦
ing point of bulk ZnCl2 is 290 C, the formation of a supported,
almost liquid-like layer may be responsible for this catalytic
activity, emulating the nature of the current AlCl3 slurry phase
industrial process. This may help explain why the two mate-
rials, prepared using different methods, with significantly dif-
no literature reports of the use of Zn2 as a heterogeneous
+
catalyst for hydrochlorination reactions. But these findings are
consistent with the Lewis acid nature of Zn2 , and hence we
+

M. Conte et al. / Journal of Catalysis 252 (2007) 23–29
27
suggest that the supported zinc catalysts function as Lewis acid
catalysts for the hydrochlorination reaction. Because the cata-
lysts are operated at temperatures well below the boiling point
◦
of bulk zinc chloride (732 C), we do not believe that the cata-
lysts readily lose zinc chloride during use. Hydrolysis reaction
is the main pathway by which the current AlCl3 catalyst is deac-
tivated; for this reason, AlCl3 is used in stoichiometric amounts.
We consider zinc chloride to be more stable toward hydroly-
sis, and consequently the catalysts are more stable. Although
at present the longest run time that we have investigated is
only 4 h, we have found no changes in the bulk structure of
the catalysts (see supplementary data Figs. S4 and S5 for the
ZnCl2/SiO2 and ZnCl2/Al2O3 before and after reaction). This
increased stability ensures that the supported zinc chloride ma-
terials can be used in catalytic rather than stoichiometric results.
Having established the nature of the ZnCl2-like active
species for the ethene hydrochlorination reaction, we carried
out a set of experiments using the ZnCl2/SiO2 catalyst to iden-
tify the effect of each reactant on activity, as follows:
Fig. 5. Hydrochlorination of ethene using ZnCl2/SiO2 as catalyst: (2) exper-
iment E: C H /HCl (2 h) → He/HCl (2 h) → C H /HCl (2 h); (") experi-
2
4
2 4
ment F: He/HCl (2 h) → C H /HCl (2 h) → He/HCl (2 h); (Q) experiment
2
4
G: C H /HCl (2 h) → He/C H (2 h) → C H /HCl (2 h); (a) experiment H:
2
4
2
4
2 4
He/C H (2 h) → HCl/C H (2 h) → He/C H (2 h).
2
4
2
4
2 4
•
•
•
•
experiment E:
data Fig. S7). To test whether the isomerization was a catalyzed
or a thermal reaction, we further investigated the reaction over
SiC and found it to be ineffective, indicating that the isomeriza-
tion observed was due to a catalyzed process.
C2H4/HCl (2 h) → He/HCl (2 h) → C2H4/HCl (2 h),
experiment F:
He/HCl (2 h) → C2H4/HCl (2 h) → He/HCl (2 h),
experiment G:
Subsequently, we reacted 2-methylpropene over ZnCl /SiO
2
2
C2H4/HCl (2 h) → C2H4/He (2 h) → C2H4/HCl (2 h),
using the same reaction conditions. A 15% conversion level was
observed, with the major product being 2-chloro-2-methylpro-
pane, as determined by NMR analysis, along with traces of
experiment H:
C2H4/He (2 h) → C2H4/HCl (2 h) → C2H4/He (2 h).
1
-chloro-2-methylpropane (see supplementary data Fig. S8).
The results, shown in Fig. 5, are very different from those when
Au was present on the catalyst surface, confirming that Au had
a negative effect for this reaction. But there are similarities to
the trends observed for the Au/C catalyst for the hydrochlorina-
tion of ethyne [6]. Indeed, when Au/ZnO was used, the average
conversion was 3–4%, whereas for ZnCl2/SiO2, it was 7–8%;
however, these two catalysts exhibited opposing effects toward
the different reactant treatments. For ZnCl2, the active species
was already present, whereas for Au/ZnO, the active species
had to be created in situ by surface chloriding during the ini-
tial reaction period. This proposal is supported by the trend
in conversion, which increased to a steady state for Au/ZnO,
whereas for ZnCl2/SiO2, the activity started out high and re-
mained steady throughout.
When isoprene was used as the substrate, two well-defined
chlorinated products were obtained: 2-methyl-4-chloro-2-butene
and 2-methyl-2-chloro-3-butene (see supplementary data
Fig. S9); conversions were 7–8%. This is a significant result,
because there are no previous reports on the use of ZnCl to
2
synthesize these chlorinated products, and, furthermore, they
can be generated without forming other byproducts using the
ZnCl /SiO catalyst. Thus, we can conclude that hydrochlori-
2
2
2+
nation reactions of alkenes using supported Zn catalysts are
selective. This finding is clear when the reaction products from
higher alkenes are considered, but even for ethane, we do not
observe other potential products, such as ethene oligomers and
dichloroethane.
The discovery of supported Zn2 as an effective alkene hy-
+
In a subsequent set of experiments, we investigated the hy-
drochlorination of higher alkenes over the ZnCl2/SiO2 cata-
lyst. We first investigated propene (propene:HCl = 1:1 using
drochlorination catalyst is not overly surprising, because the
present industrial production uses Al3 , and both of these
cations can act as Lewis acid catalysts. Consequently, we sug-
gest that the reaction mechanism may be similar in both cases,
probably involving the alkene adsorbing on the cation and HCl
forming a complex before being added to the carbon–carbon
double bond. However, it should be noted that other cations
that we have investigated also could act as Lewis acid catalysts
but were found to be inactive. Indeed, until our observation that
+
5
mL/min for each reactant, a relative inlet pressure of 1.1 bar,
◦
and a reaction temperature of 250 C). Products were collected
after the reactor in a chloroform trap and characterized by NMR
spectroscopy. The conversion was observed to be 20%. Al-
though the main product was the expected Markovnikov prod-
uct 2-chloropropane, a significant amount of anti-Markovnikov
product also was produced (see supplementary data Fig. S6),
with a main product: secondary product ratio of ca. 5:1. Sub-
sequent experiments involving feeding 1-chloropropane con-
firmed that isomerization occurred over the catalyst, which was
the origin of the two chlorinated products (see supplementary
2
+
Zn is an active catalyst, only AlCl had been used; moreover,
3
2+
Zn is considerably more active.
To date, we have not attempted to optimize the prepara-
tion of the supported ZnCl catalysts. Even so, rates of product
2
formation of ca. 10–13 mol/(kgcat h) were observed, although

28
M. Conte et al. / Journal of Catalysis 252 (2007) 23–29
Fig. 6. Bright field electron micrographs of the (a) Au/ZnO and (b) ZnO/SiO catalysts.
2
the conversions per pass that we achieved were relatively low
ca. 10–20%). We believe that this performance can be im-
Whereas discrete Au crystallites of ca. 20–100 nm diameter
are readily visible on the bulk ZnO particles (Fig. 6a), proper
characterization of the Au species in the supported ZnO/SiO2
catalyst from bright field imaging is not possible due to the
anticipated similarity in size between the nanoscale ZnO par-
ticles and those of the deposited Au. STEM analyses based
on high-angle annular dark-field imaging and XEDS spectrum
imaging to further investigate the Au dispersion on the complex
(
proved considerably. Our observation that supported ZnCl2
acts as a very selective heterogeneous catalyst for hydrochlo-
rination reactions is perhaps unique in the chemistry of zinc
catalysts, because there have been no previous disclosures con-
2+
cerning the use of Zn for this reaction. There are numerous
examples of Zn2 acting as a homogeneous catalyst for reac-
+
tions with carbon–carbon double bonds [16–18], but examples
ZnO/SiO support are ongoing; we will report the results in a
2
of supported Zn2 acting as a heterogeneous catalyst are less
common. There is a recent report that ZnCl2/Al2O3 can act as
a Lewis acid catalyst for the O-methylation of catechol [19];
and more notably there are a number of studies concerning the
use of zinc for the oxychlorination of ethene to form vinyl
chloride [20,21] and the oxychlorination of ethane to vinyl
chloride [22].
+
future paper. Thus, even though our catalytic results strongly
imply that the Au is acting to block the active sites on the
ZnCl /SiO catalysts, the precise microstructural details of how
2
2
this actually occurs have yet to be elucidated.
4. Conclusion
We have shown that supported Zn2 acts as a very selec-
tive heterogeneous catalyst for the hydrochlorination of ethene
and higher alkenes. In the case of ethene hydrochlorination,
chloroethane is the exclusive product. Higher alkenes demon-
strate similar reactivities to ethene and also give relatively few
by-products. With the current nonoptimized catalysts, synthesis
rates of ca. 10–13 mol/(kgcat h) have been achieved; we an-
ticipate that these rates can be improved through appropriate
catalyst design.
+
We further investigated two findings. First, we attempted to
determine why bulk ZnO was inactive as opposed to supported
ZnO, which was very active. This finding was somewhat sur-
prising, because it could be expected that the surface of bulk and
supported ZnO would both rapidly react with HCl, thereby cre-
ating the active catalyst. In particular, our results demonstrate
that supported ZnO and supported ZnCl are equally effec-
2
tive as catalysts and are effective without any induction period.
Second, we wished to elucidate the way in which Au was act-
ing as an inhibitor or poison during the initial phase of the
reaction. To resolve these issues, we used transmission elec-
tron microscopy (TEM) to characterize some of the catalysts
explored in this study. Electron microscopy of unsupported,
calcined ZnO (Fig. 6a) showed that it was composed of rel-
atively large crystals of zincite (ca. 0.1–3.0 µm) whereas the
supported catalysts (Fig. 6b) contained much smaller (2–5 nm)
ZnO crystallites supported on ca. 50-nm amorphous silica parti-
cles. Thus, the increased activity of the supported ZnO relative
to its bulk counterpart can be correlated with a significant in-
crease in the specific surface area of the ZnO species. The
decreased particle size also means that the fraction of the to-
tal surface area represented by nonbasal plane ZnO surfaces in
the crystallites is much higher, and these may be either catalyt-
ically more reactive or more easily converted to ZnCl2.
Acknowledgments
This work was supported by the U.K. Engineering and
Physical Sciences Research Council and Johnson Matthey plc
(project ATHENA) and the European Union (project AURI-
CAT; Contract HPRN-CT-2002-00174). The authors thank the
World Gold Council for sponsoring the studentship of M.C. un-
der the GROW scheme.
Supporting material
The online version of this article contains additional support-
ing material.
Please visit DOI: 10.1016/j.jcat.2007.09.002.

M. Conte et al. / Journal of Catalysis 252 (2007) 23–29
29
References
[9] B. Nkosi, N.J. Coville, G.J. Hutchings, M.D. Adams, J. Friedl, F.E. Wag-
ner, J. Catal. 128 (1991) 378.
[
[
[
[
[
[
10] Y. Segura, N. López, J. Pérez-Ramírez, J. Catal. 247 (2007) 383.
11] M. Haruta, Gold Bull. 37 (2004) 27.
12] G.J. Hutchings, Gold Bull. 29 (1996) 123.
13] A.S.K. Hashmi, G.J. Hutchings, Angew. Chem. Int. Ed. 45 (2006) 7896.
14] D. Syomin, B.E. Koel, Surf. Sci. 53 (2002) 498.
[
[
1] Chloroethane, National Safety Council, Itasca, IL, 2005.
2] J.G. Speight, Chemical and Process Design Handbook, McGraw–Hill,
New York, 2002.
[
[
3] J. March, Advanced Organic Chemistry, Wiley, New York, 1992.
4] F.G. Mann, B.C. Saunders, Practical Organic Chemistry, Longman Inc.,
New York, 1978.
15] Z. Yan, S. Chinta, A.A. Mohamed, J.P. Fackler, D.W. Goodman, J. Am.
Chem. Soc. 127 (2005) 1604.
[
[
[
[
5] C. Chambers, A.K. Holliday, Modern Inorganic Chemistry, Butterworths,
[16] C.-C. Wang, P.-S. Lin, C.-H. Cheng, J. Am. Chem. Soc. 124 (2002) 9696.
[17] J.-X. Wang, X. Jia, T. Meng, L. Xin, Synthesis (2005) 2669.
[18] P. Shukla, Y.-C. Hsu, C.-H. Cheng, J. Org. Chem. 71 (2006) 655.
[19] Z. Fu, Y. Yu, D. Yin, Y. Xu, H. Liu, H. Liao, Q. Xu, F. Tan, J. Wang,
J. Mol. Catal. A 232 (2005) 69.
London, 1975.
6] M. Conte, A.F. Carley, C. Heirene, D.J. Willock, P. Johnston, A.A. Herz-
ing, C.J. Kiely, G.J. Hutchings, J. Catal. 250 (2007) 231.
7] B. Nkosi, N.J. Coville, G.J. Hutchings, J. Chem. Soc. Chem. Commun.
(1988) 71.
[20] M.F. Lemanski, F.C. Leitert, C.G. Vinson, US Pat. 4115323 (1978).
8] B. Nkosi, N.J. Coville, G.J. Hutchings, M.D. Adams, J. Friedl, F.E. Wag-
ner, J. Catal. 128 (1991) 366.
[21] C.N. Kenney, Proc. Int. Congr. Catal. 2 (1971) 501.
[22] E.G. Prichert, US Pat. 3720723 (1973).