Vapor phase ethanol carbonylation over Rh supported on zeolite 13X

Article abstract of DOI:10.1016/j.apcata.2016.04.006

While methanol carbonylation has been extensively studied, higher alcohol carbonylation has received relatively little attention, even though, for example, ethanol carbonylation could be a useful route for the production of propionates. Here we use Rh/Na13X to investigate the vapor phase carbonylation of ethanol with an ethyl iodide co-feed. In the base case, the catalyst is ～40% selective to propionates, with the remainder forming ethylene and diethyl ether. Deposition of additional alkali can increase selectivities to ～60%. Isotopic labeling of ethyl iodide demonstrates reversible formation of ethyl iodide from ethanol, and that preferential incorporation of ethyl iodide initiates the Rh-catalyzed carbonylation cycle. XPS and in situ X-ray absorption spectroscopy are consistent an active anionic Rh^I iodide species at a zeolite exchange site. This proposed structure and the attendant catalytic reaction network are directly analogous to those of classic solution-phase Rh catalysts and other supported Rh catalysts. However, important differences are noted in the reaction orders and apparent activation barrier, which suggest that the rate of ethyl iodide formation is overall rate limiting under these conditions and for this catalyst.

Full text of DOI:10.1016/j.apcata.2016.04.006

Applied Catalysis A: General 520 (2016) 122–131
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Feature article
Vapor phase ethanol carbonylation over Rh supported on zeolite 13X
a
b
b
a,∗
Sara Yacob , Beata A. Kilos , David G. Barton , Justin M. Notestein
a
Northwestern University, Department of Chemical and Biological Engineering, Evanston, IL 60208, USA
The Dow Chemical Company, Core R&D, Midland, MI 48674, USA
b
a r t i c l e i n f o
a b s t r a c t
Article history:
While methanol carbonylation has been extensively studied, higher alcohol carbonylation has received
relatively little attention, even though, for example, ethanol carbonylation could be a useful route for
the production of propionates. Here we use Rh/Na13X to investigate the vapor phase carbonylation of
ethanol with an ethyl iodide co-feed. In the base case, the catalyst is ∼40% selective to propionates,
with the remainder forming ethylene and diethyl ether. Deposition of additional alkali can increase
selectivities to ∼60%. Isotopic labeling of ethyl iodide demonstrates reversible formation of ethyl iodide
from ethanol, and that preferential incorporation of ethyl iodide initiates the Rh-catalyzed carbonylation
Received 28 December 2015
Received in revised form 7 April 2016
Accepted 9 April 2016
Available online 13 April 2016
Keywords:
Ethanol
Ethyl iodide
Carbonylation
Zeolite 13X
Rhodium
I
cycle. XPS and in situ X-ray absorption spectroscopy are consistent an active anionic Rh iodide species at
a zeolite exchange site. This proposed structure and the attendant catalytic reaction network are directly
analogous to those of classic solution-phase Rh catalysts and other supported Rh catalysts. However,
important differences are noted in the reaction orders and apparent activation barrier, which suggest
that the rate of ethyl iodide formation is overall rate limiting under these conditions and for this catalyst.
©
2016 Elsevier B.V. All rights reserved.
1
. Introduction
there have been numerous supports investigated specifically for
Rh-catalyzed carbonylation, including various zeolites [12–15], sil-
ica [10], alumina [16,17], heteropolyacids [18], carbon [19,20], and
polymers [21,22]. Aside from Rh, other precious metals and tran-
sition metals have been investigated in methanol carbonylation,
including Ir [23], Ni [24–29], W [15], Co [30], and Mo [15,31] along
with several bimetallic systems [32–34]. Commercial processes
have been developed using Co (BASF), Rh (Monsanto), and Ir (BP-
Cativa) metals [35]. For all of these various methanol carbonylation
solid catalysts, an iodide remains a required co-catalyst. In classical
methanol carbonylation mechanisms, the alkyl iodide undergoes
oxidative addition to the carbonylation catalyst, and following CO
insertion, an acyl iodide reductively eliminates as a precursor to
acetic acid.
There have been far fewer studies on ethanol carbonylation
to yield propionates. Propionic acid inhibits the growth of both
mold and bacteria, which makes it attractive as a food preservative
[36]. Additionally, propionates are widely employed as a platform
chemical for many industries, including plastics, pharmaceuticals,
fragrances, and pesticides [37]. For example, methyl propionate is
a chemical intermediate to methyl methacrylate, and subsequently
the thermoplastic poly(methyl methacrylate) [38].
Carbonylation has widely been used to synthesize oxygenates
one carbon larger than the feedstock. Methanol carbonylation is
an important commercialized process and is currently the lead-
ing route to large scale manufacturing of acetic acid [1]. Prior to
the 1960s, methanol carbonylation was performed by BASF using a
◦
cobalt iodide catalyst [2]. The severe operating conditions (230 C,
6
1
00 atm) limited the economic advantages of the process and in
966, Paulik at American Monsanto commercialized a methanol
carbonylation process that used a homogenous Rh catalyst [3]. The
soluble Rh catalyst operates at comparatively milder conditions
ranging from 150 to 200 C and 30–60 atm. Progress in methanol
◦
carbonylation since then has been described extensively [2–5].
The largest commercialized methanol carbonylation processes
to date are all via soluble complexes that require a Group VIII tran-
sition metal with an iodide co-catalyst, typically methyl iodide.
There has been substantial effort to heterogenize or immobilize
the metal and conduct carbonylation in the vapor phase. Prior
methanol carbonylation work has shown that Rh supported on a
type-X zeolite is an effective heterogeneous catalyst in the pres-
ence of a methyl iodide co-feed [6–11]. Beyond the Rh/X system,
Similar to methanol carbonylation, ethanol carbonylation to
propionates has been demonstrated to occur over Group VIII
transition metals with iodide co-catalysts. Low valent Group VIII
transition metals are characteristically used for the oxidative
∗ Corresponding author.
E-mail address: j-notestein@northwestern.edu (J.M. Notestein).
http://dx.doi.org/10.1016/j.apcata.2016.04.006
0
926-860X/© 2016 Elsevier B.V. All rights reserved.

S. Yacob et al. / Applied Catalysis A: General 520 (2016) 122–131
123
addition of organic halides to form organometallic halide com-
plexes [39]. There have been several ethanol carbonylation studies
carried out with soluble catalysts very similar to the ones used com-
mercially for methanol carbonylation [40–49]. It is assumed that
the carbonylation mechanisms for ethanol in solution are analo-
gous to the established mechanisms for methanol. Regarding solid
catalysts, Nefedov et al. showed that Rh supported on type-X zeo-
lite is able to perform heterogeneous vapor-phase atmospheric
pressure carbonylation of ethanol when co-fed with ethyl iodide
in a mixture of hydrofluoric acid, hydrochloric acid, nitric acid, and
water. The standard solutions were obtained from Sigma-Aldrich at
1000 ppm concentration. ICP-AES standards were diluted from the
1000 ppm solution to known concentrations for calibration using
the same stock solutions that samples were dissolved in.
N physisorption was carried out on a Micromeritics 2010 ASAP
2
to determine Langmuir surface area and DFT micropore volume.
◦
Prior to analysis, samples were degassed under vacuum at 140 C
for at least 5 h. Langmuir surface areas were determined in the
range of 10–100 kPa.
[
50]. Similarly, Scurrell et al. carried out studies on the selectivity
of supported Rh on zeolite catalysts for carbonylation of alcohols
6,51]. Following this work more than 30 years ago, no additional
X-ray photoelectron spectroscopy (XPS) was carried out on a
Thermo Scientific ESCALAB 250 Xi. Prior to XPS, all powder sam-
ples were held under vacuum overnight. During analysis scans, the
beam size was 650 ␮m, the pass energy was 200 eV, the dwell time
was 500 ms, and 15–20 scans were taken for each sample analyzed.
All scans were normalized to the C 1s peak at 284.9 eV.
[
studies have been presented in the open literature to understand
the Rh speciation under reaction conditions or to support the pro-
posed mechanisms by which heterogeneous ethanol carbonylation
occurs. Following our own recent work disclosing a new catalyst for
ethanol carbonylation, [52] a detailed revisiting of the benchmark
catalyst was deemed necessary.
This work uses Rh/Na13X for vapor phase carbonylation of
ethanol to yield propionates. Extended X-ray absorption fine struc-
ture, in situ X-ray absorption near edge spectroscopy, and X-ray
photoelectron spectroscopy address the nature of the active Rh
catalyst. Isotopic labeling studies and catalytic tests enable us to
determine overall reaction barriers and propose a catalytic cycle for
ethanol carbonylation by comparison to other Rh catalysts. Finally,
we show the influence of adding additional alkali in improving the
behavior of these catalysts.
Temperature programmed desorption of ammonia (NH -TPD)
3
was performed on an Altamira Instruments AMI-200. Samples were
◦
first heated to 200 C under He flow before being cooled and intro-
duced to a 10% NH stream. After NH saturation, the samples were
3
3
◦
heated to 100 C again under He flow. The final TPD step was a
10 C/min ramp to 700 C under He flow. Desorbed NH was mea-
◦
◦
3
sured using a thermal conductivity detector (TCD) and quantified
by comparison to injections of known volumes of NH3.
Rh K-edge extended X-ray absorption fine structure (EXAFS)
analysis and X-ray absorption near edge structure (XANES) spec-
troscopy were performed at the Advanced Photon Source, Argonne
National Laboratory. The DuPont–Northwestern–Dow Collabora-
tive Access Team (DND-CAT) bending magnet D at Sector 5 was
used. All spectra were collected using a Si(111) double crys-
tal monochromator operating in transmission mode. For the
Rh K-edge, all EXAFS spectra were scanned in the range of
2
. Experimental details
2.1. Catalyst preparation
2
3000 eV to 24200 eV, and XANES spectra were scanned in the
1
wt% Rh/Na13X was synthesized using an ion-exchange
III
rage of 23100 eV to 23450 eV. Reference samples—Rh trichloro-
hydrate, Rh oxide, tris(triphenylphosphine)Rh chloride, and Rh
III
method between Rh trichlorohydrate, Rh Cl (H O)X and Na13X
molecular sieves (both procured from Sigma-Aldrich). The linear
formula for Na13X is Na86[(AlO )86(SiO2)106](H O)X giving an ele-
mental Si:Al ratio of 1.23. Rh Cl (H O)X, 20.5 mg, was dissolved in
0 mL of water and heated to 80 C with stirring, then 1 g of Na13X
molecular sieves was added to the solution. The ion-exchange solu-
tion was left with stirring overnight at 80 C. After ion-exchange,
the aqueous solution was filtered with Whatman grade 42 filter
paper (nominal particle retention in liquid = 2.5 ␮m) which yielded
the solid particles. The solid particles were washed with 200 mL of
purified water (Barnstead Nanopure system, purified to 18 Mꢀ cm
resistivity), collected, and dried in a 150 C oven overnight. After
drying, the solid particles were calcined at 400 C in static air, ramp-
ing from ambient at 5 C/min and then holding for 2 h. Prior work
has established that performance of this catalyst is only weakly
dependent on the ramp rate and ultimate temperature of the pre-
treatment [52].
3
2
III
I
III
iodide—were brushed onto Kapton tape and recorded at room tem-
perature. 2–5 scans of each reference were averaged to optimize
signal-to-noise ratio.
The EXAFS spectra were normalized and background subtracted
using Athena, a program in the iXAFS suite [53]. The Fourier trans-
2
2
III
3
2
◦
9
◦
2
form (FT) of the k -weighted EXAFS functions, ␹(k), were fitted
in the r space using Artemis, another program in the iXAFS suite.
The FT k range and fitted r range were 2–11 Å and 1–4 Å, respec-
tively. The amplitude reduction factor was fixed at 0.8; the free
parameters were interatomic distance (r), coordination number
◦
2
(CN), Debye-Waller factor (␴ ), and overall energy shift (E ). In the
◦
0
fitting procedure for bulk standards, the coordination number for
standards was fixed according to the known values of the ligands
around the Rh atom.
XANES experiments were performed using an in situ reaction
cell. Freestanding pellets of undiluted catalyst were supported in
the sample holder as gas and vapor feeds passed across the sample.
As for the catalytic runs, the fresh material was initially calcined
◦
Additional alkali was also impregnated onto Rh/Na13X using
Na CO , K CO , and CsHCO , as received from Sigma-Aldrich. The
2
3
2
3
3
added alkali:Na ratio during synthesis was 1000:1; final weight
loadings of the alkali are given in Table S1. The carbonates or bicar-
bonates were dissolved in 0.3 mL of purified water and impregnated
into 500 mg Rh/Na13X with hand mixing until well incorporated.
The resulting mixture was dried in a 150 C oven overnight. Once
dried, the solid material was calcined at 350 C in static air, ramping
◦
at 400 C in static air. No effort was made to protect the sample
from ambient conditions during the transfer to the in situ cell. The
in situ cell was temperature controlled using heating tape and an
external temperature controller. Spectra were acquired starting at
room temperature and under inert gas conditions. During reaction,
◦
◦
◦
the temperature was 200 C while CO, ethanol, and ethyl iodide
◦
from ambient at 5 C/min and then holding for 2 h.
flowed through the cell, with gas flow rates controlled by a glass
rotameter. A XANES spectrum was acquired every 10 min during
the 12 h reaction. XANES spectra were normalized and background
subtracted using Athena, and principal component analyses were
carried out using GRG nonlinear least squares fitting implemented
in MS-Excel.
2
.2. Catalyst characterization
Inductively coupled plasma atomic emission spectroscopy (ICP-
AES) was used to determine Rh, Na, K, and Cs content on the
catalysts. For all ICP-AES analyses, catalyst samples were dissolved

1
24
S. Yacob et al. / Applied Catalysis A: General 520 (2016) 122–131
2.3. Carbonylation experiments
3. Results and discussion
Gas phase reactions were carried out in ¼” OD plug flow
3.1. Ethanol carbonylation
microreactors constructed of Hastelloy C-276. Packed beds were
supported by quartz wool and contained 20–200 mg of 1 wt%
Rh/Na13X, depending on the desired space velocity, without addi-
tional diluent. The catalyst powder was sieved to between 100
and 120 ␮m. Ethanol and ethyl iodide were delivered by liquid
syringe pump and vaporized into the CO feed upstream of the reac-
tor. Ethanol and ethyl iodide were acquired from Sigma-Aldrich at
There are four major products (ethylene, DEE, EP, and PA)
detected in the vapor phase carbonylation reaction of EtOH with
EtI over Rh/Na13X catalysts under near-steady state conditions
(Fig. 1), arranged in two primary reaction pathways: dehydration
and carbonylation. H O is necessarily formed from the dehydra-
2
tion products, but not quantified. In the presence of H O and under
2
≥
99.5% purity and used as received without further purification.
reaction conditions, water-gas shift can also occur yielding H2 and
Ethanol to ethyl iodide ratios varied from 1:1 to 50:1; at standard
conditions the ratio was 10:1. CO to ethanol ratios varied from 5:1
to 16:1; at standard conditions the ratio was 15:1. CO was procured
from Air Products at ≥99% purity and used after passing through
molecular sieve and activated carbon beds to trap moisture and
metal carbonyls. Weight hourly space velocities (WHSV) were cal-
culated using gas and vapor mass flow rates and total weight of
CO , but the extent of water–gas shift activity was not quantified
2
with this detection scheme. [54,55] Carbonylation of EtOH does not
occur over the Na13X support alone—only dehydration to DEE and
ethylene (Table 1). Propionate species are formed when 1 wt% Rh
is supported on Na13X, and there is no evidence of EtOH carbony-
lation mechanisms enabled solely by acid zeolite sites under these
conditions, such as has been observed for dimethyl ether carbony-
lation [56–58]. All of the data in the discussion that follows have
been reproduced on more than a dozen syntheses of Rh/13X over
the course of several years, with variability no greater than that
seen within a given run.
−
1
catalyst. Calculated WHSV ranged from 6 to 1750 h . Reactor tem-
peratures ranged from 160 to 240 C and reactors were maintained
◦
at nominal atmospheric pressure. Tabulated data are averages over
a minimum of 5 h time on stream during stable operation.
Product concentrations were measured using a Shimadzu 2014
Gas Chromatograph-Flame Ionization Detector (GC-FID) with He
carrier and a 60 m Restek Stabilwax column programmed with a
◦
Product distributions at 200 C are in good agreement with
prior studies of EtOH carbonylation over Rh zeolites [51]. Despite
observing total propionate selectivity that is lower than what we
have reported over Rh-impregnated heteropolyacid (HPA) catalysts
under similar conditions, the Rh/Na13X catalyst is substantially
more active because of the higher temperature of maximum
◦
multi-step temperature ramp from 80 to 200 C. Ethylene, diethyl
ether (DEE), ethyl iodide (EtI), ethanol (EtOH), ethyl propionate
(
EP), and propionic acid (PA) were exclusively detected and iden-
−
1
tified by comparison to retention times of standards. In this
document, EP and PA are often referred to collectively as ‘propi-
onates’. Signals were correlated to injections of known amounts
of these species for calibration. For standard reaction conditions,
and unless reported otherwise, these species accounted for all sig-
nificant features in the chromatogram and together account for a
selectivity, giving higher propionate yield (at WHSV ≈ 22 h
Rh/Na13X propionate yield = 13.3% and Rh/Cs PW propi-
O
12 40
3
onate yield = 7.6%) [52]. Ammonia TPD shows the Rh exchanged
Na13X catalyst retains 850 ␮mol NH /gcat acidity, which is only
3
+
∼0.05H /Al, but it is much higher than reported for the fully
Cs-exchanged HPA systems that had total acidity less than
1
00 ± 10% mol balance on ethyl fragments. Side reactions with CO,
35 ␮mol NH /gcat. Ammonia TPD and additional characterization
3
such as water–gas shift, were not quantified. Conversion, selectiv-
ity, and mole balances are all based on ethyl fragments, and are
defined as follows, where n is the moles detected in each sample
injection:
are available in Table S1. The higher amount of residual acidity
remaining on the Na13X catalyst leads to proportionally more com-
petition from dehydration chemistry occurring in parallel with
carbonylation. Later we discuss the effect of additional alkali as a
route to reduce competitive dehydration.
nEtOH,in − nEtOH,out
Conversion of EtOH =
n
3.2. Effect of weight hourly space velocity
EtOH,in
The WHSV was controlled between 6 and 73 h 1 by changing the
catalyst weight and the mass flow rate of reactants while main-
taining a feed molar ratio of CO:EtOH:EtI at 150:10:1 (Table 1).
Except at the lowest WHSV, EtOH conversion increases nearly lin-
early with space-time across the packed bed, suggesting external
diffusion limitations are minimized; significant internal diffusion
limitations are not expected given the catalyst particle sizes used
and the rates observed. (Fig. S1) At high conversion, low WHSV,
the propionate selectivity begins to shift from ester to acid as a
result of the higher water partial pressure in the reactor due to
parallel EtOH dehydration. For example, molar concentrations in
the feed are approximately 0.6 mol% EtI, 6.2 mol% EtOH, and only
−
2
nEP,out
Selectivity for EP = n
+ 2nDEE,out + 2nEP,out + nPA,out
Ethylene,out
nEtOH,out + nEtI,out + nEthylene,out + 2nDEE,out + 2nEP,out + nPA,out
nEtOH,in + nEtI,in
Mole Balance =
2.4. Isotopic labeling study
−
1
trace water. At a WHSV of 6 h , where EtOH conversion is 59%,
the exit molar composition is approximately 0.4 mol% EtI, 2.8 mol%
During the isotopic labeling study, EtOH carbonylation was car-
ried out with CH CH I (*EtI) co-fed at 210 C, 1 atm nominal
1
3
◦
3
2
EtOH, and 2.4 mol% H O. Total selectivity to propionates (EP + PA)
2
−1
−1
pressure, 44 h WHSV, and a CO:EtOH:*EtI feed ratio of 163:10:1.
EtI was acquired from Sigma-Aldrich and used as received with
99% of the EtI containing the isotopic label. After 2 h time on
stream, the products were analyzed using an Agilent 7890A/5975C
Gas Chromatograph Mass Selective Detector with an Agilent VF-
WAXms column. Ethylene and CO were unable to be separated in
this experiment, and thus the mass balance of labeled ethyl frag-
ments could not be closed.
increases at WHSV 6 h up to 62%.
*
>
Reactor performance was recorded during an extended time
on stream (TOS) experiment to observe any changes in catalyst
behavior at standard conditions, Fig. 2A. During the 40 h exper-
iment, the EtOH conversion remained fairly constant near 35%,
indicating no significant deactivation. Selectivity also remained rel-
atively constant between 1.5 and 40 h. Even when operating at
low WHSV (longer residence times across the packed bed) extreme

S. Yacob et al. / Applied Catalysis A: General 520 (2016) 122–131
125
Fig. 1. The overall reaction network for EtOH dehydration and carbonylation.
Table 1
Performance of 1 wt% Rh/Na13X catalyst at 200 C .
◦
a
WHSV (hr 1
−
)
EtOH Conv. (%)
Ethylene Sel. (%)
DEE Sel. (%)
EP Sel. (%)
PA Sel. (%)
Catalyst
Na13X
22
6
23
59
45
31
21
54
32
31
42
33
46
6
8
14
21
0
0
13
9
1
0
Rh/Na13X
Rh/Na13X
Rh/Na13X
Rh/Na13X
49
52
42
45
12
22
b
73
a
00 ◦C, CO:EtOH:EtI = 150:10:1, Ethyl mol balance is 100 ± 10%. EtI conversion <5%.
2
b
Standard conditions.
deactivation is not observed. For example, at WHSV 6 h 1 (Fig. 2B)
the EtOH conversion decreases from 66% to 58% as the system
slowly reaches steady state over the first 6–10 h of TOS. However,
the remaining TOS of the trial showed relatively little change in
conversion or selectivity.
−
ies do not appear to have been run near differential conversion,
nor tested for diffusional limitations, so upward revision of the
apparent activation barrier is reasonable. The observed barrier is
discussed further below.
3.4. Effect of CO and EtI partial pressure and the addition of water
3.3. Effect of temperature
CO and EtI partial pressures were independently varied from
Fig. 3 shows the performance of a 1 wt% Rh/Na13X catalyst oper-
standard conditions and rates measured. For all tests in Table 2,
the total pressure was nominally 1 atm and the temperature was
fixed at 200 C. When the CO partial pressure was decreased below
◦
◦
ating within a temperature envelope of 160 C to 240 C and at 2–3%
conversion. To control conversion, the WHSV was varied while
maintaining atmospheric pressure and standard feed conditions
◦
standard conditions, He gas was used as a balance to maintain a
−
1
(
Table S2). Each temperature point is an average of 5 h TOS at near-
constant WHSV of 22 h for all tests. The final two entries of Table 2
show the effect of water addition to the feed.
steady state conditions. Under standard conditions, EtI and EtOH
are 0.6 mol% and 6.2 mol% of the inlet composition, respectively. At
all temperatures, net consumption of EtI was low (<10% conversion)
indicating no significant buildup of iodide in the reactor or loss as
With increasing CO partial pressure from 0.3–0.9 atm, car-
bonylation vs. dehydration selectivities are stable, while EtOH
conversion decreases monotonically. This is in contrast to Rh/HPA
[52], where propionate yields increased monotonically with
increased CO partial pressures. The former suggests reversible inhi-
bition by CO coordinated to the Rh center, while the latter is
consistent with CO addition occurring at the rate limiting step. In
the complete absence of EtI (CO/EtOH feed) the conversion of EtOH
was less than 1%, with the only detectable products being ethylene
and DEE. This was true at all temperatures and WHSV’s tested when
EtI was excluded, and is unsurprising given the well-documented
need for iodide co-catalysts in this system. Alternatively, when
EtOH was removed completely from the feed (CO/EtI feed) there
also were no carbonylation products detected and only ethylene
was found in the reactor effluent. Finally, the increase of water
in the system drove propionate selectivity towards PA, however,
total propionate selectivity decreased at the highest levels of water
addition and the total yields suffered.
The ethylene yield plateaus at moderate EtI partial pressures.
The relative influences of EtI and CO on ethylene yields indicates
that ethylene formation is not only acid-catalyzed, but also cat-
alyzed by Rh and competes with the carbonylation reaction. In
contrast, the yield of propionates is nearly linearly dependent on
EtI partial pressures until high EtI partial pressures (see Fig. S3). In
prior work with Rh/HPA [52], yields were largely independent of
EtI feed ratios, suggesting rapid EtI formation and addition to Rh.
other iodine-containing species relative to the total ethyl feed rate
◦
(
EtI + EtOH). At all temperatures above 180 C, ethylene is the prod-
uct detected in greatest abundance. Selectivity towards ethylene
increases with temperature at the expense of DEE, trending in the
same direction as the free energies of reaction that favor ethylene
◦
at 200 C (ethylene ꢁGr = −14.0 kJ/mol, DEE ꢁGr = − 4.7 kJ/mol; all
values per mol ethanol reacted [59]). Ethylene has been previously
observed to be the predominant side product [50].
◦
The selectivity for EP was highest at 200 C (42%), while the PA
was formed at 2% selectivity or lower at all temperatures, con-
sistent with the high ratio of EtOH to water at these low EtOH
conversions. This maximum in selectivity has not been previously
reported in related catalysts; earlier systems reported monotonic
decreases in selectivity to propionates with increasing temperature
[
7]. Because of this maximum in selectivity, a reactor tempera-
◦
ture of 200 C was used for the remainder of this study. In spite
of the maximum in selectivity, propionate yields increased mono-
tonically with temperature, giving a satisfactorily linear Arrhenius
plot for total propionate production (Fig. S2). The observed acti-
vation energy towards propionates was 138 kJ/mol. This barrier is
substantially higher than the barrier previously reported for vapor-
phase ethanol carbonylation of 94 kJ/mol [51] and also higher than
reported values for the homogenous system [40,42]. Prior stud-

1
26
S. Yacob et al. / Applied Catalysis A: General 520 (2016) 122–131
Table 2
a
Effect of changing reactant partial pressures over 1 wt% Rh/Na13X .
Feed Conditions
Product Distribution
CO:EtOH:EtI:H2O Molar ratio
PCO (atm)
0.903
0.602
0.301
0
PEtOH (atm)
0.088
0.088
0.088
0.088
PEtI (atm)
0.009
0.009
0.009
0.009
EtOH Conv. (%)
Ethylene Sel. (%)
42
43
49
DEE Sel. (%)
EP Sel. (%)
PA Sel. (%)
1
1
5
0
50: 10: 1: 0b
00: 10: 1: 0
0: 10: 1: 0
: 10: 1: 0
31
36
45
<1
14
14
9
42
41
37
0
1
1
5
0
100
0
1
1
1
1
50: 10: 0: 0
50: 10: 0.2: 0
50: 10: 1: 0b
50: 10: 5: 0
0.911
0.910
0.903
0.873
0.089
0.080
0.088
0.085
0
<1
12
31
62
5
95
19
14
3
0
0
0
1
27
0.002
0.009
0.042
34
42
36
47
42
34
1
1
50: 10: 10: 0
50: 0: 1: 0
0.837
0.990
0.081
0
0.081
0.010
66
<5
30
100
2
0
40
0
29
0
c
1
1
50: 10: 1: 20
50: 10: 1: 55
0.752
0.588
0.080
0.062
0.008
0.006
24
19
48
81
11
6
33
0
9
4
a
b
c
◦
2
00 C. WHSV = 22 h.
Standard conditions.
EtI conversion; no EtOH is present in feed.
◦
Fig. 3. Selectivity of products when using 1 wt% Rh/Na13X catalyst in a 160 C to
◦ −1
40 C temperature envelope. WHSV = 12–1750 h . CO:EtOH:EtI = 150:10:1. Error
2
bars represent standard deviation during 5 h TOS. ᮀ ethylene selectivity, ꢁ DEE
selectivity, × EP selectivity, ꢀ PA selectivity. EtOH conversion is 2–3% for all tem-
peratures.
step is EtI addition to the Rh complex [40,42]. EtI is formed through
the reversible condensation reaction between EtOH and HI or other
iodides present in the system. Thus, if the overall rate-limiting step
is the formation of EtI, one step earlier than for classic soluble phase
processes, the net reaction will still be first order with respect to
added iodide. Using this rate-limiting step also agrees better with
the high overall activation energy of 138 kJ/mol, which is similar
to reported values for ethanol dehydration over acid zeolites [60].
While we cannot rule out a higher barrier for EtI addition due to
steric or other constraints by the zeolite framework, higher barri-
ers and rate-limiting formation of the alkyl iodide have also been
observed in the synthesis of anhydrides under low-water condi-
tions [61], whereas in typical solution phase carbonylation, or over
the highly polarizable HPA [52], the reaction of EtOH and iodide
would be expected to be rapid.
◦
Fig. 2. Extended time on stream of 1% Rh/Na13X catalyst. 200
CO:EtOH:EtI = 150:10:1. GC sampling every 0.25–0.5 h.
ethylene selectivity, ꢁ DEE selectivity, × EP selectivity, ꢀ PA selectivity. (A) oper-
C and
♦
EtOH conversion,
ᮀ
−1
−1
ating at a WHSV of 22 h , (B) operating at a WHSV of 6 h .
The requirement for EtI is demonstrated by exposing Rh/Na13X
◦
−1
to CO and EtI (150:10 molar ratio) at 200 C and a WHSV of 22 h
for 4 h, giving ∼15% EtI conversion and 100% selectivity to ethylene
(Fig. 4). By mole balance, the catalyst bed has been loaded with
230 mol of HI/mol of Rh. The feed was then switched to CO/EtOH
at the same conditions. Carbonylation products were detected for
Here, the apparent near first order dependence of carbonylation at
low EtI partial pressure at first glance is consistent with the classic
solution phase Rh-catalyzed carbonylation, where the rate-limiting

S. Yacob et al. / Applied Catalysis A: General 520 (2016) 122–131
127
RhIII oxide standards. Principal component analysis revealed that
III
the fresh catalyst was >99% Rh in an oxygen coordination environ-
−
ment, indicating that the Cl ions from the precursor were cleanly
removed by the calcination treatment (Fig. S4). Since there is no
III
Rh trichlorohydrate contribution in the fresh catalyst, consistent
−
with XPS studies showing no remaining Cl ions on the catalyst
III
after calcination (Fig. S5), the Rh trichlorohydrate standard was
removed from inclusion in further principal component analyses
or EXAFS fitting.
Structural parameters obtained from EXAFS analysis are shown
2
in Table S3. Fig. 6shows the k -weighted FT of the EXAFS spec-
tra (uncorrected) for standards Rh foil, Rh O , and RhI at the
2
3
3
K-edge. Fits capture expected scattering distances and number
of neighbors: Rh-O at 2.04 Å and Rh–Rh at 2.72 Å for the Rh O3
2
bulk material [62], Rh-I at 2.65 Å for RhI3 [63,64], and 2.69 Å
for Rh–Rh in bulk Rh foil [64,65]. The molecular Rh(I) complex
I
tris(triphenylphosphine)Rh Cl was also considered (Fig. S6), but
the Rh-Cl and Rh-P interactions are not representative of any to
2
be found in the experimental materials. Fig. 6 also shows the k -
weighted FT of the EXAFS spectra for the fresh Rh/Na13X catalyst.
The first peak is attributed to the Rh-O scattering and has a bond-
distance of 2.02 Å, within error of that of Rh O . No additional
Fig. 4. Performance of 1 wt% Rh/Na13X catalyst with feed switched from EtI and CO
◦
−1
to EtOH and CO after 4 h. 200 C. WHSV = 20 h . CO: EtI = 15:1 or CO:EtOH = 15:1. ♦
EtOH conversion, − EtI conversion, ᮀ ethylene selectivity, ꢁ DEE selectivity, × EP
selectivity. ꢀ PA selectivity is <2%.
2
3
features are seen, giving no evidence that the fresh catalyst con-
tains Rh–Rh interactions, consistent with isolated Rh cations in
an oxygen coordination environment, likely as a mixture of aquo,
hydroxo, and framework oxygens; no attempt was made to fit to
higher shells that could further refine this fit. This conclusion differs
substantially from an older EXAFS study of supported Rh catalysts
prepared by a similar method, where large Rh O crystallites or Rh
3
h but in decreasing selectivity as iodide is slowly flushed from the
reactor. Despite the over-saturation of HI, the catalyst is unable to
sustainably retain the halide and turn over the carbonylation cycle
with only an EtOH and CO feed, demonstrating the requirement for
EtI to be present in the feed.
2
3
nanoparticles were seen after calcination, depending on the pre-
cursor [64]. Unfortunately, the origin of the improved dispersion in
this manuscript is not obvious.
3.5. Isotopic labeling study
The near edge region of the Rh catalyst was studied during
two pretreatment stages and an in situ reaction. To assign pseudo-
character to the catalyst during the pretreatment and reaction, a
principal component analysis was performed on each XANES spec-
trum taken every 10 min (Fig. 7, representative individual fit Fig.
S7). From these principal component analyses, changes in aver-
age oxidation state with time were extracted. Fitting references
were the Rh(III) oxide, Rh(III) iodide, and Rh(0) metal. It is under-
stood that the conclusions from this analysis may depend on the
standards chosen, and future work may include other molecu-
lar Rh0 and Rh standards as references. This study did consider
tris(triphenylphosphine) Rh(I) chloride as a fitting reference, but
its K-edge energy of 23227.8 eV was unusually high for its formal
oxidation state of +1 and it was not used as a reference in the figures
that follow. However, comparably good overall fits were obtained
when it was included (Fig. S8) and conclusions were unchanged
regarding the evolution of the average oxidation state with time or
fraction iodide character.
Isotopically labeled *EtI was used to probe how the ethyl
species is inserted into the carbonylation catalytic cycle and
−
1
the origin of side products. After 2 h TOS at a WHSV of 44 h
CO:EtOH:*EtI = 163:10:1), the reactor products were analyzed
(
using GC–MS. The experiment was duplicated using unlabeled EtI
for comparison. The normalized mass fragmentation patterns for
EtI, EtOH, DEE, and EP are shown in Fig. 5A–D. Under these con-
ditions, the EtOH conversion is 26% and the net EtI conversion is
less than 5%. Even at near zero net consumption of EtI, the efflu-
1
2
13
I
ent EtI was 81% C and 19% C EtI, indicating significant reaction
of unlabeled ethanol with liberated iodide. Likewise, the EtOH car-
1
3
ried a 5% C label (m/z 46 against 47). Given the 10:1 feed ratio of
unlabeled and labeled ethyl fragments, the EtOH and EtI are formed
reversibly, but they have not reached equilibrium over the length
of the reactor at this space velocity.
EP showed 71% C and 29% 13C incorporation into the propi-
onate fragment. The greater label incorporation in EP compared to
the level found in the effluent EtI is consistent with EtI initiating
the carbonylation catalytic cycle faster than the timescale required
for equilibration with EtOH, leading to preferential incorporation
of the label. Label incorporation into the ethoxy fragment of EP was
12
To begin, the fresh catalyst was heated from ambient to reac-
◦
tion temperature, 200 C, under inert He gas flow. During this initial
heating period, principal component analyses of the XANES region
show that the average oxidation state of the Rh species remained
near +3. Once reaction temperature was reached, the gas flow was
switched to CO, then vaporized EtOH and EtI were added to the
9
%, similar to the total fraction of labeled ethyl fragments present in
the exiting EtI and EtOH, calculated to be ∼7%. This indicates that
ester formation is near equilibrium, reacting equally with EtI or
EtOH. DEE also showed ∼30% single incorporation of the label and
a negligible 2% double incorporation, indicating substantial forma-
tion through cross-reaction of EtI and EtOH. Like carbonylation, the
label is not fully equilibrated.
−
1
CO flow at a WHSV of 6 h (time = 0 h). At time = 0 h, when EtOH
III
and EtI are introduced, the catalyst gains some Rh iodide char-
acter; this remains fairly constant at 30% over the course of the
reaction, in agreement with the unchanging partial pressure of the
III
added iodide. Rh oxide character drops over the course of the reac-
tion but never disappears, while the contribution from reduced Rh
increases over the first 6 h, then levels out. All together, the aver-
age oxidation state of the catalyst drops from +2.6 to +1.7 over the
first 6 h, then stabilizes. This 6 h induction period is very similar
3
.6. EXAFS and in situ XANES
X-ray absorption spectra of Rh/Na13X after initial calcination
0
III
(
before reaction) were compared to Rh , Rh trichlorohydrate, and

1
28
S. Yacob et al. / Applied Catalysis A: General 520 (2016) 122–131
Fig. 5. Key fragments of GC–MS spectra, normalized to each of their most intense peaks for (A) EtI, (B) EtOH, (C) DEE, and (D) EP, detected when operating at WHSV = 44 h−1,
◦
2
10 C, and CO:EtOH:EtI = 163:10:1 using either labeled *EtI or unlabeled EtI.
to the induction period in product distributions observed with the
catalyst when operating at these relatively low WHSV conditions.
(
Fig. 2B). The formation of Rh-I species with time is consistent with
2
the EXAFS analysis of the used catalyst. The k -weighted FT of the
EXAFS spectra for the used Rh/Na13X catalyst, Fig. 6, shows sig-
nificant gains in the second scattering peak, which we assign to
Rh-I interactions, giving a Rh-I distance of 2.65 Å and a coordination
number of 1.1, in addition to a Rh-O distance of 2.07 Å and a coor-
dination number of 1.1. (Table S3) The interaction at 2.07 Å is again
in contrast with older work [64], where only the larger distance
peak was observed, more consistent with Rh metal nanoparticles.
Elsewhere, Rh supported on zeolite Y has been observed reversibly
form metal–metal bonds when under cycles of reducing and oxi-
dizing conditions [66]. Given these reports, the second scatterer
could also be assigned to a Rh–Rh interaction, but the overall fit
was worse (Table S3) and TEM of the used catalyst was featureless,
giving no evidence of Rh nanoparticles.
To reconcile the observed oxidation state and ligand environ-
ment, the total Rh on the catalyst is hypothesized to be present
as a mixture of two Rh species under reaction conditions (Fig. 8).
I
Specifically, ∼80% of the Rh is hypothesized to be an anionic Rh I
and the active species for carbonylation. The second species is an
III
oxidized, anionic Rh I representing ∼20% of the Rh on the catalyst.
2
This latter species would be expected to be catalytically inert and to
form more readily in the presence of excess iodide. With these two
hypothesized Rh complexes, the average Rh-I and Rh-O coordina-
tion numbers are each 1.2, consistent with the used catalyst EXAFS
results (Table S3). The average oxidation state of 1.4 is in reason-
able agreement with the observed average oxidation state of 1.7
2
Fig. 6. k -weighted Fourier transformed Rh K-edge EXAFS spectra (fitted in the r
◦
space) of Rh foil, Rh2O3, RhI3, and fresh and used Rh/Na13X all acquired at 25 C.
reduced rhodium. The +1 oxidation state reference material,
tris(triphenylphosphine)Rh Cl, could not be used in fitting the
experimental XPS spectra because it has a non-zero vapor pres-
sure. Other Rh complexes are typically <0.4 eV shifted from the
III
I
from in situ XANES. The Rh species requires two exchange sites in
close proximity to each other within a supercage, compared to the
I
I
Rh I species that only requires a single exchange site. The Na13X
0
support contains a Si:Al ratio of 1.23, and proximate exchange sites
are common. Zeolite supports with higher Si:Al ratios may lead to
a greater fraction of the catalyst present in an active form. The net
negative charge on these surface complexes would be balanced by
the Na+ or other alkali present in these materials.
Rh features [67], they are hard to distinguish from the metal in
our experimental samples, and we consider only one set of features
for ‘reduced’ Rh sites in the 0/ + 1 oxidation state. The XPS curves
for the fresh and used catalyst are shown in Fig. 9. The fresh cata-
III
III
lyst only contains contributions from Rh species, presumably Rh
in oxygen coordination in agreement with the EXAFS and XANES
principal component analysis of the fresh catalyst. The used catalyst
contains a very broad Rh feature that can be fit with contributions
from multiple Rh species. To reduce the number of free parame-
ters, the oxidized and reduced 3d3/2 components were constrained
to represent 60% of the 3d5/2 peak in the used catalyst, as was the
3
.7. XPS results
XPS was performed on the fresh and used catalyst after a
−
1
catalytic test at a WHSV of 6 h . The XPS spectra were fit
to Gaussians representing contributions from oxidized (+3) and

S. Yacob et al. / Applied Catalysis A: General 520 (2016) 122–131
129
Fig. 7. Principal component analyses of in situ XANES and average oxidation state over time for a catalyst under WHSV = 6 h−1, 200 ◦C, CO:EtOH:EtI = 150:10:1. ♦ Rh(0), ꢁ
Rh(III) in an oxygen coordination environment, X Rh(III) in an iodide coordination environment, ᮀ average oxidation state (right axis).
case in the fresh catalyst. The used catalyst XPS spectra were fit
to be approximately 58% reduced and 42% oxidized character, or
an average oxidation state of 1.3–1.8 (depending on whether the
reduced species is counted as Rh0 or Rh ), not dissimilar to the 1.7
average oxidation state estimated from in situ XANES. The average
oxidation state as determined from XPS is also in agreement with
the hypothesized mixture of two Rh species on the catalyst in Fig. 8
that would give an average oxidation state of 1.4.
I
3.8. Hypothesized mechanism
Fig. 8. Hypothesized Rh species present during reaction conditions. Not shown are
Na counter-cations.
Reports on soluble Rh catalyzed methanol carbonylation have
used FTIR and other techniques to show that the Rh transforms
of EtI and EtOH. As described earlier, the overall rate-limiting step
was assigned to be the formation of EtI, shown in step 1 as the reac-
tion of EtOH and HI, but numerous other salts could play the role
of iodide shuttle. Finally, the propionyl iodide ejected from the Rh
cycle is converted to EP or PA in step 2, depending on the partial
pressure of EtOH or H2O.
III
I
from a Rh pre-catalyst to its active form, an anionic Rh species,
−
Rh(CO) I ] , as in established mechanisms [68,69]. Here, EXAFS
2 2
[
analysis and XPS of the catalyst after reaction conditions were con-
I
sistent with an active complex that is a (zeolite)Rh I. Under reaction
conditions, the Rh species likely also has 2 or more coordinated CO,
I
−
giving a [(zeolite)Rh (CO) I] located at an exchange site on the
2
1
3X.
The mechanism proposed for EtOH carbonylation (Fig. 10) is
3.9. Alkali Addition
analogous to that of Rh-catalyzed methanol carbonylation. The Rh
cycle is entered through oxidative addition of EtI to the active
The alkali present on the support is hypothesized to play a role
in stabilizing the active anionic Rh species, titrating away acid
sites, and in shuttling the iodide ions in the rate limiting step.
As such, three materials were synthesized with additional alkali
impregnated into the Rh/Na13X catalyst. After addition of Na2CO3,
K2CO3, or CsHCO3, only moderate amounts of additional alkali were
retained, and the acidity decreased slightly. (Table S1) All overex-
changed materials showed substantial decreases in dehydration
I
Rh species, A, which is also the primary species proposed from
III
spectroscopy. This addition yields a Rh -ethyl complex, B, from
III
which CO rapidly inserts to the Rh -ethyl bond, forming the
III
(
zeolite)Rh (CO) I -propionate species, C. After CO adds, D, propi-
2
2
I
onyl iodide reductively eliminates, regenerating the active Rh form
of the catalyst, A. We have proposed here beta hydride elimination
of ethylene from B to form species F to account for the influence of
Rh and CO on rates of ethanol dehydration. Species F is also simi-
lar to the minority, inactive species proposed from spectroscopy.
Finally, we also propose for formation of species E in the pres-
+
+
+
activity due to the reduction in acidity, with Na > K > Cs . Over-
all, the catalyst selectivity towards total propionates reached as
high as 62% (Table 3) with only a slight decrease in EtOH conver-
+
sion. K showed improvements in carbonylation yields, suggesting
I
ence of excess CO, a catalytically inactive Rh complex, to explain
a direct role in increasing the rate of iodide addition. The lackluster
+
the inverse dependence of yields on CO partial pressure. Given the
steady conversion with time on stream (after any observed induc-
tion periods) the formation of E appears to be reversible.
performance of Cs may be associated with the decrease in surface
and pore volume relative to the parent Rh/Na13X (Fig. S10) that
accompanied addition of this large cation.
There are a number of acid-catalyzed reactions that occur in
parallel. The dehydration of EtOH and the dehalogenation of EtI are
both acid-catalyzed routes to ethylene, and both would be essen-
tially irreversible at these conditions. The isotopic labeling study
suggests that DEE is formed through the bimolecular cross-reaction
4. Conclusions
EtOH carbonylation to propionates has been demonstrated over
Rh/Na13X with co-catalytic EtI. Rh was necessary for carbonylation

1
30
S. Yacob et al. / Applied Catalysis A: General 520 (2016) 122–131
A
B
3
20
315
310
305
300
320
315
310
305
300
Binding Energy (eV)
Binding Energy (eV)
−
1
Fig. 9. XPS of the fresh and used Rh/Na13X catalyst following reaction at WHSV = 6 h . (A) Fresh Rh/Na13X, (B) Used Rh/Na13X. The XPS spectra were fit to Gaussians
representing 3d5/2 and 3d3/2 contributions from oxidized (+3) and reduced (0/ + 1) Rh species.
Table 3
◦
a
Performance of support and catalysts at 200 C .
Catalyst
EtOH Conv. (% )E thylene Sel. (% )D EE Sel. (% )E P Sel. (% )P A Sel. (%)
Na13Xb
23
31
54
42
36
34
42
46
14
4
4
10
0
0
1
7
10
4
Rh/Na13Xb
42
53
52
44
Rh-Na2CO3/Na13X20
Rh-K2CO3/Na13X 27
Rh-CsHCO3/Na13X27
a
00 ◦C, WHSV = 22 h, CO:EtOH:EtI = 150:10:1. Ethyl mol balance 100 ± 5%. EtI
2
conversion <5%.
b
Repeated from Table 1.
reactions to proceed; only products from dehydration of EtOH were
formed otherwise. Carbonylation selectivity was monitored while
varying several operating parameters, including temperature, CO
partial pressure, EtI partial pressure, and water addition. CO was
found to have an inhibitory impact on EtOH conversion, presum-
ably caused by reversible coordination and formation of an inactive
Rh complex. Propionate yields increase with EtI partial pressure,
but the effect diminishes above the standard 1:10 EtI:EtOH feed
ratio. Therefore, the proposed rate-limiting step under base case
conditions is EtI synthesis, as also consistent with the high appar-
ent activation energy of 138 kJ/mol. We note that the barrier, and
indeed the rate-limiting step, may be very sensitive to the level of
acidity, the amount and nature of the added alkali, and the rela-
tive amounts of iodide and water present at reaction conditions.
Continuous feeds of EtI are necessary, as has been seen in earlier
systems, and isotopic labeling indicates that EtI was incorporated
preferentially into the propionates and diethyl ether.
Fig. 10. Hypothesized rhodium-catalyzed EtOH carbonylation cycle. Not drawn are
the Na or other alkali counter-cations to the anionic Rh complexes. X represents a
zeolite exchange site.
+
DMR-1121262), the Keck Foundation, the State of Illinois, and
Northwestern University. The CleanCat Core facility acknowledges
funding from the Department of Energy (DE-FG02-03ER15457)
used for the purchase of the Altamira AMI-200. This EXAFS and
in situ XANES work was performed at the DuPont-Northwestern-
Dow Collaborative Access Team (DND-CAT) located at Sector 5
of the Advanced Photon Source (APS). DND-CAT is supported
by Northwestern University, E.I. DuPont de Nemours & Co., and
The Dow Chemical Company. This research used resources of the
Advanced Photon Source, a U.S. Department of Energy (DOE) Office
of Science User Facility operated for the DOE Office of Science
by Argonne National Laboratory under Contract No. DE-AC02-
Through XPS, EXAFS, and XANES, the Rh oxidation state and
coordination environment were determined. The catalyst appears
III
to transform from an isolated Rh to an average oxidation state of
1
.4 under reaction conditions and afterwards. The working catalyst
I
is proposed to be a mixture of an active anionic Rh I supported at
III
an exchange site on the zeolite (∼80%) and an inactive Rh species
(
∼20%). Overall, the catalyst is proposed to carry out carbonylation
by a simple modification of classical mechanisms, where the limit-
ing step in the vapor phase under differential conversion (very low
water) is the formation of the alkyl iodide rather than its oxidative
addition. In the base case, carbonylation selectivity is maximized
at ∼40%, but this was increased to ∼60% with aggressive over-
exchange with additional alkali. Balancing acidity and activation
of the Rh within different zeolite frameworks and levels of alkali
exchange present opportunities for future catalyst optimization.
0
6CH11357.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.apcata.2016.04.
0
06.
Acknowledgements
References
The Dow Chemical Company supported this work. This XPS
work was performed in the Keck-II facility of NUANCE Center
at Northwestern University. The NUANCE Center is supported
by the International Institute for Nanotechnology, MRSEC (NSF
[
[
1] G.J. Sunley, D.J. Watson, Catal. Today 58 (2000) 293–307.
2] P.M. Maitlis, A. Haynes, G.J. Sunley, M.J. Howard, J. Chem. Soc. -Dalton Trans.
(1996) 2187–2196.
[3] F.E. Paulik, J.F. Roth, Chem. Commun. (1968) 1578.

S. Yacob et al. / Applied Catalysis A: General 520 (2016) 122–131
[36] U.S. Environmental Protection Agency, R.E.D. Facts: Propionic Acid, in
131
[
[
[
4] D. Forster, J. Am. Chem. Soc. 98 (1976) 846–848.
5] J.H. Jones, Platin. Met. Rev. 44 (2000) 94–105.
Pesticides and Toxic Substances, pamphlet 7508W (1991).
[37] Technical Data Sheet: Propionic Acid, The Dow Chemical Company, 2012.
[38] J.J. Spivey, M.R. Gogate, J.R. Zoeller, R.D. Colberg, Ind. Eng. Chem. Res. 36
(1997) 4600–4608.
[39] S. Murai, Activation of Unreactive Bonds and Organic Synthesis, 1999.
[40] J. Hjortkjaer, J.C.A. Jorgensen, J. Mol. Catal. 4 (1978) 199–203.
[41] C. Lai, Q. Chen, Q. Wu, X. Liu, Tianranqi Huagong 33 (2008) 60–64.
[42] S.B. Dake, D.S. Kolhe, R.V. Chaudhari, J. Mol. Catal. 24 (1984).
[43] A.L. Lapidus, E.Z. Gildenberg, A.R. Sharipova, N.F. Kononov, Y.T. Eidus, Kinet.
Catal. 17 (1976) 1274–1277.
[44] A.L. Lapidus, O.L. Eliseev, T.N. Bondarenko, N.N. Stepin, Bull. Acad. Sci. USSR,
Div. Chem. Sci. 31 (1983) 1852–1858.
[45] R.P. Patil, A.A. Kelkar, R.V. Chaudhari, J. Mol. Catal. 47 (1988).
[46] C. Lai, J. Zeng, Y. Yong, J. Liao, Q. Chen, Tianranqi Huagong 33 (2008) 40–43.
[47] M. Khan, S. Halligudi, S. Shukla, J. Mol. Catal. 48 (1998) L69–L71.
[48] T.W. Dekleva, D. Forster, J. Am. Chem. Soc. 107 (1985) 3565–3567.
[49] G. Jenner, G. Bitsi, J. Mol. Catal. 40 (1987) 71–82.
[50] B.K. Nefedov, R.V. Dzhaparidze, O.G. Mamaev, Bull. Aca. Sci. USSR, Div. Chem.
Sci. 27 (1978) 1448–1450.
[51] M.S. Scurrell, T. Hauberg, Appl. Catal. 2 (1982) 225–238.
[52] S. Yacob, S. Park, B.A. Kilos, D.G. Barton, J.M. Notestein, J. Catal. 325 (2015) 1–8.
[53] B. Ravel, M. Newville, J. Synchrotron Rad. 12 (2005) 537–541.
[54] Y. Lei, N.W. Cant, D.L. Trimm, J. Catal. 239 (2006) 227–236.
[55] E. Trabuco, P.C. Ford, J. Mol. Catal. A 148 (1999) 1–7.
[56] A. Bhan, E. Iglesia, Acc. Chem. Res. 41 (2008) 559–567.
[57] P. Cheung, A. Bhan, G.J. Sunley, E. Iglesia, Angew. Chem. Int. Ed. 45 (2006)
1617–1620.
6] B. Christensen, M.S. Scurrell, J. Chem. Soc. -Faraday Trans. I 73 (1977)
2
036–2039.
7] B. Christensen, M.S. Scurrell, J. Chem. Soc. -Faraday Trans. I 74 (1978)
313–2319.
8] B.K. Nefedov, R.V. Dzhaparidze, Y.T. Eidus, Bull. Acad. Sci. USSR, Div. Chem. Sci.
6 (1977) 1310–1312.
9] B.K. Nefedov, R.V. Dzhaparidze, O.G. Mamaev, N.S. Sergeeva, Bull. Acad. Sci.
USSR, Div. Chem. Sci. 28 (1979) 348–353.
10] S.L.T. Andersson, M.S. Scurrell, J. Catal. 71 (1981) 233–243.
11] S.L.T. Andersson, M.S. Scurrell, J. Catal. 59 (1979) 340–356.
12] J. Bodis, J. Zsako, C. Nemeth, J. Mink, Vib. Spectrosc. 9 (1995) 197–202.
13] M.V. Luzgin, A.G. Stepanov, A. Sassi, J. Sommer, Chem. Eur. J. 6 (2000)
368–2376.
14] S. Bischoff, A. Weigt, M. Kant, U. Schulke, B. Lucke, Catal. Today 36 (1997)
73–284.
15] G.C. Tustin, J.R. Zoeller, H.L. Browning, A.H. Singleton, Carboxylic acids and
esters by carbonylation of methanol in a heterogeneous vapour phase,
provides improved reaction rates at lower pressure and lower cost, to
Eastman Chem Co, (1999) U.S. Patent US5900505-A.
16] A. Krzywicki, G. Pannetier, Bull. Soc. Chim. Fr. (1975) 1093–1098.
17] A. Krzywicki, G. Pannetier, Bull. Soc. Chim. Fr. (1977) 64–68.
18] L.D. Dingwall, A.F. Lee, J.M. Lynam, K. Wilson, L. Olivi, J.M.S. Deeley, S.
Gaemers, G.J. Sunley, ACS Catal. 2 (2012) 1368–1376.
19] F. Li, Q. Qian, F. Yan, G. Yuan, Mater. Chem. Phys. 107 (2008) 310–316.
20] G. Zehl, S. Bischoff, B. Lucke, Catal. Lett. 19 (1993) 247–255.
21] N. De Blasio, E. Tempesti, A. Kaddouri, C. Mazzocchia, D.J. Cole-Hamilton, J.
Catal. 176 (1998) 253–259.
[
[
[
2
2
[
[
[
[
2
[
[
2
[
[
[
[
[
[
[58] P. Cheung, A. Bhan, G.J. Sunley, D.J. Law, E. Iglesia, J. Catal. 245 (2007) 110–123.
[59] CRC Handbook of Chemistry and Physics, 95th edition ed., Taylor and Francis
Group, LLC, 2014–2015.
[
[
22] E. Tempesti, A. Kaddouri, N. De Blasio, J. Therm. Anal. Calorim. 53 (1998)
1
77–187.
23] C.S. Garland, M.F., Giles, J.G., Sunley, K.E., Clode, S.G., Carl, F.G., Martin, G.S.
John, Carbonylation of an alcohol by reaction of methanol with carbon
monoxide using a ruthenium or osmium promoted iridium catalyst, to BP
Chem. Ltd., (1997) U.S. Patent 5672743.
[60] H. Xin, X. Li, Y. Fang, X. Yi, W. Hu, Y. Chu, F. Zhang, A. Zheng, H. Zhang, X. Li, J.
Catal. 312 (2014) 204–215.
[61] J.R. Zoeller, J.D. Cloyd, N.L. Lafferty, V.A. Nicely, S.W. Polichnowski, S.L. Cook,
Adv. Chem. (1992) 377–394.
[
[
[
[
24] S. Bischoff, H.E. Maneck, H. Preiss, K. Fujimoto, Appl. Catal. 75 (1991) 45–56.
25] S. Yao, C. Yang, Y. Tan, Y. Han, Catal. Commun. 9 (2008) 2107–2111.
26] N. Zhang, D. Zhou, Y. Zhu, Chin. J. Catal. 27 (2006) 591–595.
[62] R.S. Mulukutla, T. Shido, K. Asakura, T. Kogure, Y. Iwasawa, Appl. Catal. A 228
(2002) 305–314.
[63] B.E. Cowie, D.J.H. Emslie, H.A. Jenkins, J.F. Britten, Inorg. Chem. 49 (2010)
4060–4072.
27] K. Fujimoto, T. Shikada, K. Omata, H. Tominaga, Ind. Eng. Chem. Prod. Res. Dev.
2
1 (1982) 429–432.
[64] D.R. Denley, R.H. Raymond, S.C. Tang, J. Catal. 87 (1984) 414–423.
[65] A. Gayen, K.R. Priolkar, R. Sarode, V. Jayaram, M.S. Hegde, G.N. Subbanna, S.
Emura, Chem. Mater. 16 (2004) 2317–2328.
[66] P. Serna, B.C. Gates, Acc. Chem. Res. 47 (2014) 2612–2620.
[67] J. Kugai, V. Subramani, C.S. Song, M.H. Engelhard, Y.H. Chin, J. Catal. 238
(2006) 430–440.
[
[
[
28] K. Fujimoto, S. Bischoff, K. Omata, H. Yagita, J. Catal. 133 (1992) 370–382.
29] T.C. Liu, S.J. Chiu, Ind. Eng. Chem. Prod. Res. Dev. 33 (1994) 488–492.
30] W.J. Smith, Process for producing carbonylation products, to BP Chem. Ltd.
(
2009) U.S. Patent 7491842.
[
[
[
31] T. Shikada, H. Yagita, K. Fujimoto, H. Tominaga, Chem. Lett. (1985) 547–550.
32] F. Peng, F. Xiao-Bao, Catal. Today 93–95 (2004) 451–455.
33] H.X. Wang, W.W. Guo, L.J. in: T. Sun (Ed.) Zhu, Energy Eng. Env. Eng. (2013)
[68] R. Golhosseini, A. Naderifar, A. Mohammadrezaei, M.J. Nasr, Int. J. Chem.
React. Eng. 10 (2012).
9
83–986.
[69] N. Yoneda, S. Kusano, M. Yasui, P. Pujado, S. Wilcher, Appl. Catal. A 221 (2001)
253–265.
[
[
34] T.C. Liu, S.J. Chiu, Appl. Catal. A 117 (1994) 17–27.
35] A. Haynes, Adv. Catal. 53 (2010) 1–45.