Deactivation factor of CuCl2-KCl/Al2O3 catalyst for ethylene oxychlorination in a commercial-scale plant

Article abstract of DOI:10.1016/j.apcata.2019.117205

Ethylene oxychlorination over a CuCl₂-KCl/Al₂O₃ catalyst was examined for 2 consecutive years in a commercial-scale plant. The oxychlorination performance of the CuCl₂-KCl/Al₂O₃ catalyst deteriorated gradually, which could seriously affect stable operation. To clarify such deactivation factors, the relationship between catalyst performance and physicochemical property change was investigated using X-ray fluorescence, electron probe microanalysis, and Brunauer–Emmett–Teller measurements. Sublimation of the CuCl component and subsequent increment of the ratio of K to Cu components were observed at the inlet of the catalyst bed, and were determined to be contributing factors in the deactivation of the CuCl₂-KCl/Al₂O₃ catalyst.

Full text of DOI:10.1016/j.apcata.2019.117205

AppliedCatalysisA,General589(2020)117205
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Deactivation factor of CuCl2-KCl/Al2O3 catalyst for ethylene oxychlorination
in a commercial-scale plant
Tomokazu Ohashia,*, Sae Someyaa, Yoshihiko Moria, Tetsuo Asakawaa, Makoto Hanayaa,
Motohiro Oguria, Ryo Watanabeb, Choji Fukuharab
a Functional Polymers Research Laboratory, Tosoh Corporation, 1-8 Kasumi, Yokkaichi, Mie, 510-8540, Japan
b Graduate School of Applied Chemistry and Biochemical Engineering, Shizuoka University, Shizuoka, 432-8561, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
Ethylene oxychlorination over a CuCl2-KCl/Al2O3 catalyst was examined for 2 consecutive years in a commer-
cial-scale plant. The oxychlorination performance of the CuCl2-KCl/Al2O3 catalyst deteriorated gradually, which
could seriously affect stable operation. To clarify such deactivation factors, the relationship between catalyst
performance and physicochemical property change was investigated using X-ray fluorescence, electron probe
microanalysis, and Brunauer–Emmett–Teller measurements. Sublimation of the CuCl component and subsequent
increment of the ratio of K to Cu components were observed at the inlet of the catalyst bed, and were determined
to be contributing factors in the deactivation of the CuCl2-KCl/Al2O3 catalyst.
Oxychlorination
Deactivation factor
Vinyl chloride monomer
Copper chloride
1. Introduction
undergoes a redox cycle during this process, which might influence its
catalytic performance. Kominami et al. investigated the rates of the
The balanced oxychlorination process is the predominant means of
manufacturing vinyl chloride monomer (VCM) [1]. The overall process
consists of three reactions, as shown in Scheme 1: (1) production of 1,2-
dichloroethane (EDC) by direct chlorination of ethylene, (2) production
of VCM by thermal decomposition of EDC, and (3) production of EDC
by oxychlorination of ethylene using HCl from (2). A significant benefit
of oxychlorination is that it recycles hydrogen chloride, which is a by-
product in various other reactions, such as the production of diphe-
nylmethanediisocyanate (MDI).
Traditionally, the performance of metal chlorides such as V, Cr, Mn,
Fe, Co, Ni, Cu as catalysts for oxychlorination has been studied [2–7].
Thus, in the industrial oxychlorination process, CuCl2 catalysts have
traditionally been used owing to their high activity and high selectivity
for EDC compared with other transition metal catalysts. For example, it
was reported that the reaction result of 99.0% conversion of HCl and
99.6% EDC selectivity was gaved in the catalysts supported 7.98 Cu,
chlorination of ethylene and the reoxidation of copper chloride by
feeding ethylene gas or a mixed gas of hydrogen chloride and oxygen as
pulses to various metal catalysts [6]. They revealed a correlation be-
tween the activity of the oxychlorination catalyst and the rate of EDC
formation, obtained from the rates of ethylene chlorination and copper
chloride reoxidation. Therefore, the oxidation state of Cu is a particu-
larly important parameter.
The oxychlorination of ethylene is an exothermic reaction with high
thermal energy (ΔrH゜=-239 kJ/mol) [13]. Various heat removal
techniques can be applied to reduce the heat of industrial processes
involving exothermic reactions; for example, a heat-exchange reactor
can be adopted [14]. Nevertheless, if heat removal is insufficient,
partial heat generation occurs in the catalyst layer, which deactivates
the Cu-based chloride catalyst. In a patent [15], it has been mentioned
that hotspots can reduce the selectivity of the reaction and lead to
carbonization of the organic reactants inside the catalyst pellets. Con-
sequently, the pellets can break down into fine powder, and this in-
creases resistance to gas flow along the tube and reduces the life of the
catalyst. Therefore, various methods of improving the catalyst dur-
ability have been reported, such as the addition of alkali metal chlor-
ides and/or alkaline earth metal chlorides [11,16–26]. However, the
durability of Cu-based oxychlorination catalysts is still unsatisfactory in
commercial plants. Therefore, it is imperative to investigate the catalyst
0.78 K and 0.05 Mg as chlorides respectively on alumina (HCl : C2H4
:
O2 = 70.3 : 234.3 : 21.1 N L/h, the reaction temperature was 493 to
560 K, the amount of catalyst was 303.9 g) [8]. The proposed oxy-
chlorination reaction mechanism with a CuCl2 catalyst is given in
Scheme 2. The reaction mechanism has been proposed to progress via
ethylene chlorination with CuCl2, followed by reoxidation of the re-
duced CuCl with O2 and HCl [3,6,3–7,9–12]. However, copper chloride
⁎ Corresponding author.
E-mail address: tomokazu-oohashi-wv@tosoh.co.jp (T. Ohashi).
https://doi.org/10.1016/j.apcata.2019.117205
Received 6 February 2019; Received in revised form 8 August 2019; Accepted 9 August 2019
Availableonline15October2019
0926-860X/©2019ElsevierB.V.Allrightsreserved.

T. Ohashi, et al.
AppliedCatalysisA,General589(2020)117205
Scheme 1. Balanced oxychlorination process.
0.30 mL g−1
2.2. Commercial plant test
deterioration state by operation of the industrial plant from the view-
point of improving industrial catalysts and optimizing the method of
their application. In particular, it is important to understand the pro-
gression of catalyst degradation at each position of the reactor in order
to improve the dilution pattern of the catalyst and extend its life.
Nevertheless, there are only a few examples of detailed analysis of the
deactivation of the catalyst in industrial operation. Vetrivel et al. ana-
lyzed the Cu species of the deactivated catalyst [27,28], by in-
vestigating the change in catalyst composition by X-ray photoelectron
spectrometry. However, there is no information about the number of
runs a spent catalyst can be subjected to. Further, neither the dete-
rioration of the catalyst over time in a plant operation, nor the corre-
lation between the activity and composition at each reactor position
was considered.
.
The prepared catalyst was subjected to an endurance test in a
commercial plant reactor set up to renew the catalyst in a two-years.
The test catalyst was loaded into the central tube of the reactor con-
sisting of several thousand multi-tubes. The other tubes were filled with
a commercial catalyst having a life of 2 years. It was subjected to
commercial operation under a flow of reactant gas, and samples were
extracted from the reaction tube after 1 and 2 years. The samples were
divided into four or five sections, labeled from the inlet side of the
reaction tube as 1st, 2nd, 3rd, 4th, or 5th position (Fig. 1).
In this study, using CuCl2-KCl/Al2O3 catalyst samples taken from
actual plant equipment, deterioration factors were estimated from the
correlation between the physicochemical analysis (component, surface
area, element distribution, etc.) and the activity after long-term op-
eration.
2.3. Activity test
The activities of the fresh and used catalysts were evaluated using a
fixed-bed reactor (inner diameter = 29 mm). The catalyst was mixed
with an equal volume of diluent (14 mL, graphite pellets) and trans-
ferred to a Pyrex reaction tube. The remaining space in the reaction
tube was filled with glass beads. The reaction conditions were con-
trolled at a reactant ratio of HCl:C2H4:O2:N2 = 2.0:1.0:0.4:2.8 and a gas
hourly space velocity (GHSV) of 400 h−1. The reaction temperature
was controlled so that the inlet temperature of the catalyst layer was
493 K.
The catalytic activity was derived by dividing the amount of ethy-
lene conversion by the mass of the catalyst and reaction time for the
temperature conditions described above. The activity of the fresh cat-
alyst was considered the standard, and the activity of the used catalysts
was presented as the relative activity, based on an initial catalyst per-
formance of 100%.
2. Experimental method
2.1. Catalyst preparation
The CuCl2-KCl/Al2O3 catalyst was prepared by impregnating γ-
alumina (extruded cylinder, outer diameter: 5.0 mm, inner diameter:
2.0 mm, length: 5.0 mm, surface area 184 m2 g−1, pore volume 0.62 mL
g
−1) with aqueous solutions of copper chloride (224 g/L; Nihon Kagaku
Sangyo Co., Ltd) and potassium chloride (136 g/L; Otsuka Chemical
Co., Ltd). After impregnation, the catalyst was dried at 423 K for 2 h,
followed by calcination at 523 K (heating rate: 283 K/min) for 5 h in air
using a muffle furnace. Quantitative analysis of copper chloride and
potassium chloride was carried out using a scanning X-ray fluorescence
analyzer (Rigaku Corporation, (trade name) ZSX PrimusII). About 3 g of
the catalyst was pulverized, and then, a sample plate was prepared
using a pressure press; subsequently, this plate was subjected to mea-
surements in a Rh tube at a tube voltage of 50 kV and a tube current of
60 mA. The surface area and pore volume were measured using the N2
adsorption method (Microtrac bell BELSORP mini) at liquid-nitrogen
temperature and a nitrogen relative pressure of 0.001 to 0.995. Prior to
the measurements, the samples (0.1–0.2 g) were weighed and vacuum-
dried at 473 K for 4 h. The resultant 13.1 wt% CuCl2–4.9 wt% KCl/
Al2O3 catalyst had a surface area of 112 m2 g−1 and a pore volume of
2.4. Characterization of the catalyst
The shape retention rate of the used catalyst was determined by the
weight ratio of particles remaining on the top of a 2.8-mm mesh sieve.
The surface area and pore volume were measured using the N2 ad-
sorption method (Microtrac bell BELSORP mini). Prior to the mea-
surement, samples (0.1–0.2 g) were weighed following vacuum drying
at 473 K for 4 h. The fracture strength in the lateral direction of the
cylindrical catalyst used was measured using a digital push–pull gauge
(Aikoh Engineering MODEL-RX).
The amounts of Cu and K supported in the catalyst were measured
2

T. Ohashi, et al.
AppliedCatalysisA,General589(2020)117205
Scheme 2. Oxychlorination with a Cu(II) chloride catalyst.
using X-ray fluorescence (XRF, Rigaku ZSX Primus II). Five standard
samples were prepared with different amounts of Cu and K using the
pore-filling method; a calibration curve was prepared based on the
measurement results of these standard samples. In pretreatment, the
sample was crushed into powder and dried. Approximately 3 g of the
powdered sample was added into a 30-mm-φ plate ring followed by
pressing at 11 MPa for 40 s and 5.0 MPa for 20 s. Before measurement,
the sample plate was vacuum-dried to remove moisture.
2.5. Hydrothermal test
Hydrothermal tests were carried out on the γ-alumina substrate and
on the fresh catalyst using a stainless steel pressure vessel. The γ-
Alumina or fresh catalyst (5 g) and 1 mol/L hydrochloric acid (5 mL)
were loaded into a Teflon container (internal volume =37 mL), sepa-
rated by a Teflon mesh. The Teflon container was placed into a stainless
steel container and treated at 473 K for
a predetermined time.
Elemental mapping of the cross-section of the catalyst was per-
formed using electron probe microanalysis (EPMA; Shimadzu EPMA-
1610; accelerating voltage: 15 kV, illumination current: 30 nA, beam
diameter: 20 μm, analysis area: 10.752 × 10.752 mm). Prior to the
measurement, the upper surface of the cylindrical catalyst was polished
and coated with carbon.
Thereafter, the samples were dried at 473 K for 1 h. The surface area
and the pore volume of the samples were measured.
3

T. Ohashi, et al.
AppliedCatalysisA,General589(2020)117205
Fig. 1. Divided samples from the commercial plant after (a) 1 year and (b) 2 years.
3. Results and discussion
influence at the 1st position. Therefore, we speculate that the ratio of
shape retention was 90% or less despite maintaining the high fracture
strength of the 1st position catalyst. In our test catalysts, it was shown
that cracks and dust due to the decrease in fracture strength were not
dominant deterioration factors in plant operation.
3.1. Appearance and shape of catalysts used in the commercial plant
Fig. 2 shows the appearance of the fresh and used catalysts. The
upper side of each column is the appearance of a cylindrical outer
surface, and the lower side is of a cylindrical cross section. In the fresh
catalyst, the appearance of the cylindrical outer surface and cross-sec-
tion were likewise pale yellow-green. Vetrivel et al. reported a spent
catalyst with a dark core as compared to a spherical surface [27,28].
Furthermore, they reported that CuAlO2 was formed in the core of the
spherical spent catalyst. In contrast, all of our cylindrical shape used
catalysts were darker in color than the respective fresh ones. This color
change was more significant on the cylindrical outer surface. In addi-
tion, there was a difference between the appearance of the cylindrical
outer surface and the cross section in the used catalysts. We suggest that
the changes in the catalyst occurred mainly on the cylindrical outer
surface rather than in the bulk. Fig. 3 shows the shape retention, which
indicates mechanical durability of the catalyst after 2 years of opera-
tion. It has been reported that the large pressure loss caused by pow-
dering of the catalysts often hinders long-term plant operation [15].
Therefore, changes in strength and shape retention are important fac-
tors for commercial plant operation. The shape retention rate of the
used catalyst was determined by the weight ratio of particles remaining
on the top of a 2.8-mm mesh sieve. In the 1st and 2nd positions, the
retention rate fell below 90%, with the 2nd position falling the lowest.
This suggests that the load on the front side of the reaction tube was
high for the 2 years of operation. On the other hand, in the 3rd to 5th
positions, the shape retention rate was 90% or more, and the overall
average retention rate was ∼90%. The tested catalysts showed high
durability for 2 years of operation in the commercial plant [29].
Fig. 4 shows the fracture strength of the used catalysts. The fracture
strength of the fresh catalyst was 27 N, which decreased to 13–20 N
after 1 year and to 13–18 N after 2 years, indicating that there was no
significant change in fracture strength. The fracture strength was lowest
at the 2nd position for both years. At the 2nd to 5th positions, after 2
years a correlation was found between the fracture strength and shape
retention rate. It is estimated that the decrease in shape retention ratio
at the 2nd to 5th positions was dominated by physical cracking and
powdering due to the decrease in fracture strength. On the other hand,
the strength at the 1st position was higher, but the shape retention was
similar to those at the other positions. As described in Section 3.4,
weight loss due to sublimation of CuCl is presumed to have a major
3.2. Physicochemical properties of catalysts
Table 1 shows the Brunauer–Emmett–Teller (BET) specific surface
area and pore volume of the catalysts. Both properties were lower in the
used catalysts than the fresh ones at all positions. This indicates that the
decrease in pore volume and in BET specific surface area occurred at
the same time. The decrease in BET specific surface area was more
pronounced; the used catalysts were reduced to 68∼85 m2/g and the
fresh catalyst to 112 m2/g. In particular, the BET specific surface area in
the 2nd position decreased to nearly 60% of the initial value. This de-
creasing trend was similar to the decrease in fracture strength and
shape retention. Furthermore, such a decrease in BET specific surface
area was observed in spent catalysts by Vetrivel et al. [28]. To clarify
the cause, we conducted hydrothermal tests on the γ-alumina substrate
and the fresh catalyst. Fig. 5 shows the change in BET specific surface
area and pore volume after the hydrothermal tests. Both properties
decreased with increasing hydrothermal treatment time, although the
decrease was smaller in the catalyst than in the alumina substrate. This
is probably because of the reduced contact between water and the
alumina surface by the supported metal. Fig. 6 shows the pore dis-
tribution and pore volume of the fresh catalyst and the catalyst used for
2 years, alumina substrate, and hydrothermally treated sample. In the
catalysts used for two years in commercial plants, a decrease in pore
size of 3–10 nm and an increase in pore size of 10–50 nm were observed
in comparison with fresh ones. The changes in the catalysts used in
commercial plants were reproduced in the sample wherein the alumina
substrate was hydrothermally treated for 5 h. From these results, we
speculated that grain growth of γ-alumina due to hydrolysis was the
cause of the decrease in BET specific surface area and pore volume.
Therefore, we infer that in the used catalysts, the decrease in BET
specific surface area and pore volume occurs for the whole of the
downstream side, where the concentration of the by-produced water
increases.
3.3. Performance of catalysts used in the commercial plant
Table 2 shows the activity of the oxychlorination catalysts. The
4

T. Ohashi, et al.
AppliedCatalysisA,General589(2020)117205
Fig. 2. Appearance of fresh and used catalysts in the commercial plant.
activity given here is the relative activity, based on an initial catalyst
performance of 100%. In the used catalysts, the activity decreased with
increasing operation time. The catalysts sampled from the inlet side (1st
and 2nd positions) were clearly deactivated, but those from the middle
and outlet side (3rd, 4th, and 5th positions) were not notably reduced. In
particular, the specific activities decreased to 56% and 69% for the 1st
and 2nd positions, respectively, in the catalyst used for 2 years. This
suggests that the load on the front side of the reaction tube was high for
the 2 years of operation.
Fig. 7 shows the average temperature of the reaction tube during the
commercial plant test. The operating temperatures of commercial
plants are confidential and not reported. A sharp peak appeared at the
2nd and 3rd positions, which shows that the reaction progressed mainly
on the front side of the temperature peak position. Therefore, we infer
that the activity decline of the 1st and 2nd positions were significant.
Table 3 shows the selectivity of the oxychlorination catalysts. The
main product of the reaction of any catalyst was EDC. In addition to
EDC, trace amounts of chlorinated by-products, as well as carbon
monoxide and carbon dioxide, were produced, as found in other lit-
erature [30]. The EDC selectivity of the used catalysts decreased
slightly as compared to that of the fresh catalyst. With the used cata-
lysts, no significant change was observed in the formation rates of
carbon monoxide and carbon dioxide, but there was an increase in the
ethyl chloride (EC) and VCM selectivity.
3.4. Chemical composition of catalysts
Table 4 shows the chemical compositions of the catalysts. The
5

T. Ohashi, et al.
AppliedCatalysisA,General589(2020)117205
Table 1
Physical Physical properties of the oxychlorination catalysts talysts.
Catalyst
BETa) surface (m2/g)
Pore volume (ml/g)
Fresh
112
80
68
78
83
85
71
73
76
76
0.30
0.30
0.25
0.27
0.29
0.31
0.27
0.27
0.29
0.28
1 year
2 year
1st
2nd
3rd
4th
1st
2nd
3rd
4th
5th
a)
Brunauer–Emmett–Teller.
Fig. 3. Shape retention of the used catalyst after 2 years of operation.
Furthermore, iron oxide is known to react with hydrogen chloride to
form iron chloride [39]. The iron component observed in the used
catalyst is likely to have produced iron chloride because of the pro-
longed exposure to hydrogen chloride gas at high temperatures in the
plant. From this finding, we speculate that iron chloride, acting as a
Lewis acid, caused an increase in the amount of ethyl chloride and VCM
in used catalysts at the 3rd to 5th positions.
amount of K in the used catalysts was almost the same as that in the
fresh catalyst, while the amount of Cu decreased with operation year. It
has been found that CuCl formed in the oxychlorination catalytic cycle
was sublimable. It appears that this sublimation of CuCl reduced the
amount of Cu in the used catalysts [31–33]. This was most prominent
on the inlet side (1st and 2nd positions) where the reaction primarily
occurs. At the 1st and 2nd positions, the alumina surface could be ex-
posed owing to the decrement of Cu. Finocchio et al. [30] performed
pulse reaction experiments on catalysts with 1%–9% supported Cu.
They found that catalysts with a small amount of supported Cu had low
EDC selectivity and increased formation rate of ethyl chloride and VCM.
These results were consistent with the performance of used catalysts in
the 1st and 2nd positions. In addition, Muddada et al. [22], Shalygin
et al. [34], and Carmello et al. [35] showed that alumina serves as the
active site of VCM production by dehydrochlorination of EDC in ex-
periments feeding EDC to alumina. Muddada et al. [36] also examined
the relationship between the concentration of Lewis acid sites on the
catalyst and the by-products of the oxychlorination reaction. It was
confirmed that the by-productivity of ethyl chloride and VCM was in-
creased in catalysts with high Lewis acid concentrations. From these
reports and results, it is assumed that the exposed alumina surface acted
as a Lewis acid, which caused the increase in ethyl chloride and VCM in
the used catalysts at the 1st and 2nd positions.
Fig. 8 shows the results of EPMA of Cu and K in the cross-section of
the cylindrical catalyst. Cu and K were uniformly distributed in the
fresh catalyst, whereas in the 1-year-used catalyst at the 3rd position
and 2-years-used catalyst at the 3rd to 4th positions, Cu content in-
creased on the cylindrical surface layer with respect to the bulk. This
trend was particularly notable in the 1-year-used catalyst at the 3rd
position. In contrast, in the 1-year-used catalyst at the 1st to 2nd posi-
tions and in the 2-years-used catalyst at the 1st to 2nd positions, Cu
content was decreased throughout the catalyst, especially in on cy-
lindrical surface layer. In general, since the reaction mainly proceeds on
the surface of the solid catalyst, it was presumed that the Cu in the
cylindrical surface layer was remarkably sublimated. Vetrivel et al. [28]
studied the X-ray line profile of fresh and deactivated catalysts, and
reported that Cu was abundant on the fresh catalyst surface but reduced
on the deactivated catalyst surface. These findings were consistent with
our EPMA results for the used catalyst at the 1st and 2nd position.
Furthermore, the decrease in the Cu content by CuCl sublimation might
be related to the heat of reaction and the rate of formation of CuCl by
the reduction of CuCl2. From these results, we infer that in the 1st and
2nd positions, where the reaction primarily progressed, Cu was reduced
from the cylindrical surface layer by sublimation of CuCl, while at the
3rd and 4th positions, further downstream, sublimed CuCl was deposited
on the cylindrical surface layer. CuCl2 reduction would not have pre-
dominantly occurred at the 3rd position despite the high average tem-
perature. In contrast, no difference in the distribution of K was observed
In the used catalysts, iron was detected in addition to the compo-
nents of the fresh catalyst. In our commercial plant test, the origin of
the iron was unknown. However, Aharoni and Inbar [37], Kurta et al.
[38], and Flid [18] stated that in a fluidized bed oxychlorination re-
actor, iron penetration is usually due to the erosion of industrial reactor
walls. Todo et al. [3] evaluated the oxychlorination performance of
various metal chloride catalysts, and stated that the main product of
oxychlorination with FeCl3-supported catalyst was ethyl chloride.
Fig. 4. Fracture strength of the used catalysts after (a) 1 year and (b) 2 years.
6

T. Ohashi, et al.
AppliedCatalysisA,General589(2020)117205
Fig. 5. Results of hydrothermal tests. (a) Brunauer–Emmett–Teller (BET) specific surface area, (b) pore volume.
between the fresh catalyst and the used catalysts.
correlate with the activity, showing it was not the main deactivation
factor in catalysts used for 2 years. The BET specific surface area on the
inlet side (1st position) was high in contrast to the activity, and de-
creased to ∼70% in the 3rd–5th positions, whereas the specific activity
was maintained at 85% or more. In contrast, the tendency of activity
corresponds considerably with the order of Cu content. These findings
indicate that the sublimation of CuCl, caused by the high exothermic
energy profile at the inlet side was one of the factors of deactivation for
the CuCl2-KCl/Al2O3 catalyst. In addition, peak activity appeared at the
3rd position during the 1st year, and at the 4th position in the 2nd year
owing to deposition of sublimated CuCl from the upper stream. Changes
in the catalyst components due to CuCl sublimation and deposition
were presumed to be closely related to the catalyst activity.
3.5. Influence of changing chemical composition on catalytic performance
Fig. 9 shows the oxychlorination activity of the catalysts sampled
from each position of the reactor, in comparison with the BET specific
surface area and Cu content. The activity reported here is the relative
activity, based on an initial catalyst performance of 100%. As shown in
Fig. 9, the used catalysts for both the 1st and the 2nd years had similar
trends, except that inactivation on the inlet side was greater after 2
years than after 1 year. In both the 1st and 2nd year, specific activity
peaked slightly upstream from the last position. In other words, the
decrease in BET specific surface area (and hence, pore volume) did not
Fig. 6. Pore distribution and pore volume. (a) Pore distribution of fresh catalyst and catalyst used for 2 years, (b) Pore volume of fresh catalyst and catalyst used for 2
years, (c) Pore distribution of alumina substrate and hydrothermal test sample, (d) Pore volume of alumina substrate and hydrothermal test sample.
7

T. Ohashi, et al.
AppliedCatalysisA,General589(2020)117205
Table 2
Table 4
Activities of the oxychlorination catalysts.
Chemical compositions of the catalysts.
Catalyst
Amount of
(mmol)
Amount of ethylene
Activity
(%)
Catalyst
Amount
(mmol/g・h)
Cu
K
Fe
others
(wt%)
K/Cu
(mol/mol)
(wt%)
(wt%)
(wt%)
Fresh
1
44
3.8
2.7
3.1
4.3
3.5
2.1
2.6
3.6
3.6
3.2
100
70
82
112
93
56
69
96
94
85
1st
29
45
64
46
38
47
61
45
41
Fresh
year
6.2
5.0
2.6
2.5
N.D.
N.D.
91.3
92.5
0.68
0.81
1
2
1st
2nd
3rd
4th
1st
2nd
3rd
4th
1st
5.0
6.0
5.7
4.6
2.5
2.5
2.5
2.4
N.D.
0.1
0.7
92.5
91.4
91.1
93.0
0.81
0.68
0.71
0.85
2
2nd
3rd
4th
5th
N.D.
years
2nd
3rd
4th
5th
4.8
6.1
6.2
5.8
2.4
2.4
2.4
2.5
0.1
0.3
0.3
0.4
92.7
91.2
91.1
91.3
0.81
0.64
0.63
0.70
Reaction condition: a reactant ratio of HCl:C2H4:O2:N2 = 2.0:1.0:0.4:2.8, a gas
hourly space velocity (GHSV) of 400 h−1, 493 K -2 cm from inlet of catalyst
layer in atmospheric pressure.
CuCl2/γ-Al2O3 catalyst. Arcoya et al. [16] and Dmitrieva et al. [40]
examined the effect of the addition of alkali chlorides on the CuCl2/α-
Al2O3 catalyst. They found that activity decreased linearly with in-
creasing KCl content in the catalysts with high KCl loadings. They
considered that the decrease in activity was due to KCl being inactive
and interfering with the contact between the reactants and CuCl2. In
addition, Prasad and Rao [41] measured the chemisorption rate of
ethylene to CuCl2/γ-Al2O3 or CuCl2-KCl/γ-Al2O3 at −78 °C. They
showed that the amount of ethylene chemisorption decreased linearly
as the K/Cu ratio decreased from 0.5 to 0.14 (Cu/K atomic ratio from 2
to 7 [41]), and concluded that the amount of ethylene chemisorption at
low-temperature may be related to the active sites on the oxychlor-
ination catalysts.
These reports show good agreement with our results. In the used
catalysts, the K/Cu ratio increased due to the decrease in Cu caused by
sublimation. Furthermore, the activity decreased concurrently with the
increasing K/Cu ratio. These results suggests that the main cause of
used catalyst deactivation was the increase in the K/Cu ratio due to Cu
sublimation. In the catalysts used for 1 year, the activity was equivalent
to the fresh ones with low K/Cu ratios, but was lower than the fresh
catalysts with high K/Cu ratios. However, in the 1-year-used catalyst at
the 3rd position, the activity was higher than the fresh catalyst.
Elemental mapping for the 1-year-used catalyst at the 3rd position
(Fig. 8) show that Cu content increased on the cylindrical surface layer
with respect to the bulk, implying that the K/Cu ratio was lower on the
surface of the 1-year-used catalyst at the 3rd position than on the surface
of the fresh catalyst. Therefore, in the 1-year-used catalyst at the 3rd
position, the activity was higher than that of the fresh catalyst. In the
catalysts used for 2 years, the activity of all layers declined further; in
particular, the activities of catalysts with high K/Cu ratios were greatly
reduced compared with the catalysts used for 1 year. Elemental map-
ping in the 2-years used catalyst at the 2nd position (Fig. 8) showed that
Cu was distributed more toward the surface than the center of the ring
in the fresh catalyst, but was reduced at the surface in the used catalyst.
This means that the K/Cu ratio was higher on the surface of the used
catalyst than on that of the fresh catalyst, especially at the 1st and 2nd
positions. Therefore, the activity of the 2-years used catalyst was in-
ferred to be lower than that of fresh catalyst. These findings suggest that
changes resulting from the use of catalysts in commercial plants occur
primarily on the surface of the catalyst. Therefore, it is important to
obtain the composition of the surface, which contains the active sites, to
investigate the performance of catalysts for use in commercial plants.
From these results, we concluded that the main cause of catalyst de-
activation during operation was the increase in the K/Cu ratio due to Cu
sublimation. A strong correlation between K/Cu and activity was con-
firmed, suggesting that the change in the K/Cu ratio may have had a
dominant influence on the active site. Lamberti et al. reported that the
Fig. 7. Average temperature of the reaction tube during the commercial plant
test.
Table 3
Selectivities of the oxychlorination catalysts.
Catalyst
Selectivity(%)
EDCa)
VCMb)
ECc)
TCEd)
CO
CO2
Fresh
99.5
99.5
99.6
99.4
99.5
99.4
99.4
99.3
99.4
99.3
0.0
0.0
0.0
0.0
0.0
0.0
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.2
0.2
0.2
0.2
0.2
0.3
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.2
0.1
0.1
0.1
0.1
0.1
0.1
0.2
0.2
0.1
0.2
0.1
0.2
0.1
0.2
0.1
0.1
one
two
1st
2nd
3rd
4th
1st
2nd
3rd
4th
5th
Reaction condition: a reactant ratio of HCl:C2H4:O2:N2 = 2.0:1.0:0.4:2.8, a gas
hourly space velocity (GHSV) of 400 h−1, 493 K -2 cm from inlet of catalyst
layer in atmospheric pressure.
a)
EDC; ethylene dichloride.
VCM; vinyl chloride monomer.
EC; ethyl chloride, d) TCE; trichloroethane.
b)
c)
Fig. 10 shows the relationship between the K-to-Cu ratio (K/Cu) and
the activity of the fresh and used catalysts. The K-to-Cu ratio was de-
termined from the amount of each element measured by X-ray fluor-
escence. The fresh catalysts shown here were prepared to have different
K/Cu ratios, with approximately the same amount of Cu (5.9–6.2 wt%),
separately from the catalyst loaded in a commercial plant. A strong
linear correlation was found between the specific activity and the K/Cu
ratio in the fresh catalysts. Rouco [21] compared the oxychlorination
catalytic activities of CuCl2/γ-Al2O3 and CuCl2-KCl/γ-Al2O3 and re-
ported that the catalyst with added KCl had lower activity than the
8

T. Ohashi, et al.
AppliedCatalysisA,General589(2020)117205
Fig. 8. Cu and K distribution in the fresh and used catalysts. Data from electron probe microanalysis (EPMA). Samples were cut along the length of the cylinder.
active phase of CuCl2-KCl/Al2O3 is a mixed chloride (KxCuCl2+x) phase
that reduces the ability of the active surface to adsorb ethylene and/or
transfer two Cl atoms to each ethylene molecule [17,20]. Therefore, it is
presumed that the amount of the (KxCuCl2+x) phase increased by an
increase in the K/Cu ratio in our used catalysts as well. In the future, we
will carry out a detailed evaluation of the change of catalyst by
9

T. Ohashi, et al.
AppliedCatalysisA,General589(2020)117205
Fig. 9. Oxychlorination performance change of used catalysts after (a) 1year and (b) 2 years. The reaction conditions were controlled at a reactant ratio of
HCl:C2H4:O2:N2 = 2.0:1.0:0.4:2.8, a gas hourly space velocity (GHSV) of 400 h-1, the reaction temperature of 493 K.
of the active site. We conclude that this state change is the dominant
factor in the depreciation of plant operation life. Therefore, it is im-
portant to control the K/Cu ratio and adjust the dilution rate of the
catalyst at each position in the plant for preventing loss of catalytic
activity due to the sublimation of Cu.
References
[1] K. Bungaku, Hitotsubashi J. Soc. Sci. 16 (1991) 39–71.
[2] P.G. Hall, P. Heaton, D.R. Rosseinsky, J. Chem. Soc., Faraday Trans. 1 (80) (1984)
3059–3070.
[3] N. Todo, M. Kurita, H. Hagiwara, J. Chem. Soc. Jap. 69 (1966) 1463–1466.
[4] J.A. Allen, J. Chem. Technol. Biotechnol. 12 (1962) 406–412.
[5] J.A. Allen, A.J. Clark, Rev. Pure Appl. Chem. 21 (1971) 145–166.
[6] N. Kominami, K. Kawarazaki, Y. Yamazaki, T. Sakurai, Catalyst 7 (1965) 359–362.
[7] N. Kominami, K. Kawarazaki, Y. Yamazaki, T. Sakurai, Catalyst 7 (1965) 363–366.
[8] O. Carlo, C. Francesco, C. Marco, EP2198958A1.
[9] K. Miyauchi, Y. Sato, K. Higuchi, K. Fujimoto, Kogyo Kagaku Zasshi 71 (1968)
Fig. 10. Activity vs. K/Cu ratio. The reaction conditions were controlled at a
reactant ratio of HCl:C2H4:O2:N2 = 2.0:1.0:0.4:2.8, a gas hourly space velocity
(GHSV) of 400 h-1, the reaction temperature of 493 K.
695–699.
[10] Y. Xie, H. Zhang, R. Wang, SCIG 23 (1980) 979–991.
[11] V.M. Zhernosek, I.B. Vasileva, A.K. Avetisov, A.I. Gel’bshtein, Kinet. Katal. 12
(1971) 353–358.
[12] M.P. Dmitrieva, Y.M. Bakshi, A.I. Gel’bshtein, Kinet. Katal. 26 (1985) 1359–1364.
[13] Z. Czarny, M. Repelewicz, Chema Stosowana 3–4 (1990) 227–231.
implementing a kinetic study and electronic state analysis of Cu in the
used catalysts.
[14] R.W. McPherson, C.W. Starks, G.J. Fryar, J. Hydrocarbon Process. 58 (1979) 75–88.
[15] D. Lorette, L. Kevin, M. Andrea, V. Sandro, WO2012052199A1.
[16] A. Arcoya, A. Cortes, X.L. Seoane, Can. J. Chem. Eng. 60 (1982) 55–60.
[17] C. Lamberti, C. Prestipino, F. Bonino, L. Capello, S. Bordiga, G. Spoto, A. Zecchina,
S.D. Moreno, B. Cremaschi, M. Garilli, A. Marsella, D. Carmello, S. Vidotto,
G. Leofanti, Angew. Chem. Int. Ed. 41 (2002) 2341–2344.
4. Summary and conclusions
[18] M.R. Flid, Catal. Use Indium 8 (2016) 23–31.
An oxychlorination catalyst was investigated after operation for 2
years in a commercial plant, to determine the cause of deterioration
during long-term operation. The oxychlorination performance of the
CuCl2-KCl/Al2O3 catalyst gradually declined, which would seriously
affect the stable operation of a commercial plant. In samples of used
catalysts, cracking and powdering were confirmed as the main reasons
for declining catalyst activity due to operation of the plant. However,
our test catalyst had sufficient mechanical durability after 2 years of
operation. Furthermore, by investigating the relationships between
catalyst performance and changes in physicochemical characteristics, a
clear correlation was found between the sublimation of Cu and activity,
whereas a decrease in BET specific surface area and pore volume did
not contribute to the decrease in activity. We speculate that the increase
in the K/Cu ratio due to sublimation of Cu caused a change in the state
[19] N.B. Muddada, U. Olsbye, G. Leofanti, D. Gianolio, F. Bonino, S. Bordiga,
T. Fuglerud, S. Vidotto, A. Marsella, C. Lamberti, Dalton Trans. 39 (2010)
8437–8449.
[20] N.B. Muddada, U. Olsbye, L. Caccialupi, F. Cavani, G. Leofanti, D. Gianolio,
S. Bordiga, C. Lamberti, Phys. Chem. Chem. Phys. 12 (2010) 5605–5618.
[21] A.J. Rouco, J. Catal. 157 (1995) 380–387.
[22] N.B. Muddada, T. Fuglerud, C. Lamberti, U. Olsbye, Top. Catal. 57 (2014) 741–756.
[23] K. Zurowski, Pol. J. Chem. 69 (1995) 1718–1728.
[24] Wm.C. Conner Jr, W.J.M. Pieters, A.J. Signorelli, Appl. Catal. 11 (1984) 59–71.
[25] D. Gianolio, N.B. Muddada, U. Olsbye, C. Lamberti, Nucl. Instrum. Methods Phys.
Res. B 284 (2012) 53–57.
[26] Y.M. Sorokin, Y.M. Bakshi, A.I. Gel’bshtein, Kinet. Katal. 17 (1976) 888–893.
[27] R. Vetrivel, K. Seshan, K.R. Krishnamurthy, T.S.R. Prasada Rao, Bull. Mater. Sci. 9
(1987) 75–80.
[28] R. Vetrivel, K.V. Rao, K. Seshan, K.R. Krishnamurthy, T.S.R. Prasada Rao, 9th
International Congress on Catalysis, (1987), pp. 1766–1773.
[29] C.H. Bartholomew, Chem. Eng. 91 (1984) 96–112.
10

T. Ohashi, et al.
AppliedCatalysisA,General589(2020)117205
[30] E. Finocchio, N. Rossi, G. Busca, M. Padovan, G. Leofanti, B. Cremaschi, A. Marsella,
D. Carmello, J. Catal. 179 (1998) 606–618.
G. Busca, J. Catal. 191 (2000) 354–363.
[36] N.B. Muddada, U. Olsbye, T. Fuglerud, S. Vidotto, A. Marsella, S. Bordiga,
[31] T. Kekeshi, K. Miura, M. Isshiki, Bull. Inst. Adv. Mater. Process. Tohoku Univ. 50
(1994) 16–26.
[32] P.S.S. Prasad, K.B.S. Prasad, M.S. Ananth, Ind. Eng. Chem. Res. 40 (2001)
5487–5495.
[33] M.M. Mallikarjunan, S. Zahed Hussain, J. Sci. Ind. Res. 43 (1984) 94–97.
[34] A.S. Shalygin, L.V. Malysheva, E.A. Paukshtis, Kinet. Katal. 52 (2011) 305–315.
[35] D. Carmello, E. Finocchio, A. Marsella, B. Cremaschi, G. Leofanti, M. Padovan,
D. Gianolio, G. Leofanti, C. Lamberti, J. Catal. 284 (2011) 236–246.
[37] C. Aharoni, J. Inbar, J. Appl. Chem. Biotechnol. 23 (1973) 333–338.
[38] S.A. Kurta, I.M. Mykytyn, M.V. Khaber, J. Appl. Chem. 78 (2005) 1088–1092.
[39] K. Sato, M. Tomita, J. Mining Metall. Inst. Jpn. 88 (1013) (1972) 435–440.
[40] M.P. Dmitrieva, Y.M. Bakshi, A.I. Gel’bshtein, Kinet. Katal. 37 (1996) 79–83.
[41] P.S.S. Prasad, P.K. Rao, J. Chem. Soc. Chem. Commun. 12 (1987) 951–953.
11