University of Southern California Dissertations and Theses

Methanol synthesis in a membrane reactor

(USC Thesis Other)

Methanol synthesis in a membrane reactor

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Methanol Synthesis in a Membrane Reactor

by

Sahar Soltani

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(CHEMICAL ENGINEERING)

July 2014

ABSTRACT
Methanol synthesis has undergone continuous improvements for nearly a century, as it
represents the starting raw material for the production of a variety of other chemicals and
solvents, including formaldehyde, methyl tertiary butyl ether, and acetic acid and fuel additives.
Methanol has a number of advantages as a fuel and a source of chemical products, such as being
more easily transportable than methane and other gaseous fuels, having a high energy density,
needing no desulphurization, and its reactions (e.g., steam reforming) proceeding at moderate
temperatures. Recent global energy shortages and more strict emission regulations have
motivated research and development of new fuel cells, among which a direct methanol fuel cell
is a prime candidate.
In the present work the CO/CO
2
conversion into methanol in both a traditional reactor
(TR) and a membrane reactor (MR) has been studied. The purpose of this study is to investigate
the possibility of using a MR to increase the total carbon conversion into methanol relative to
what a TR can convert.
MR with solvent purge, which is proposed in this work, incorporates the advantages of
the existing reactive separation systems that result from the in-situ product removal. In our
proposed process, we use a solvent as a sweep fluid. The membrane serves as an interface
contactor for the selective permeation of methanol. The solvent is chosen in a way that the main
product, methanol, has the highest solubility in it. Using TetraEthylene Glycol Dimethyl Ether as
an inert, high boiling point agent, the alcohol will be selectively removed in-situ from the reactor
using the membrane as an interface contactor between the methanol and the solvent. Due to the
very low solubility of H
2
, CO, and CO
2
, they will remain in the reactor, since the solvent blocks
the membrane’s pores.
ii | P a g e

Prior to the start of the membrane reactor experiments, the surface of a membrane was
successfully modified in order to increase membrane hydrophobicity and, therefore, to achieve
higher mass transfer rate that, in return, results in higher methanol production. The modified
membrane can be operated as a membrane contactor under the proposed operating conditions
without the loss of the inert solvent.
The kinetics of the methanol synthesis from CO, CO
2
and H
2
on a commercial
Cu/Al
2
O
3
/ZnO catalyst was investigated in an autoclave reactor. The effect of various
CO/(CO
2
+CO) ratios, stoichiometric number, pressure, temperature and flow rate was
investigated in the kinetic experiments.
Based on the experimental kinetics and the membrane surface modification results, we
developed a model of the MR to explore the effect of several factors on the total conversion of
carbon. The rate equation derived from the kinetic experiments was used in the model.
Membranes with various thicknesses were simulated, and their performance was compared. Our
results indicate great potential for application of membrane reactor in methanol synthesis using
CO, CO
2
and H
2.

iii | P a g e

ACKNOWLEDGMENTS
I would like to express my appreciation to all those who have supported me, in particular,
I would like to thank Professor Theodore T. Tsotsis, for his academic guidance, his expertise and
patience.
I am also indebted to Professor Muhammad Sahimi, who has been a source of knowledge
and experience and I am grateful for his support. I would also like to thank Professor G. K Surya
Prakash, for his kind guidance on my thesis.
I sincerely appreciate the help of all the Mork Family Department of Chemical
Engineering and Material Sciences staff, as well as all my friends and colleagues during the
years that I spent at USC.
Last but not least, I would like to thank my uncle, Mr. Ahmad Soltani, for his
encouragement and the most importantly, I am indebted to my parents, my sister, Gohar and my
brother, Ali. Their love was endless and their support was priceless.
This dissertation is dedicated to all the teachers in my life who taught me to fight for
success, and never give up.

Table of Contents
Chapter 1. Introduction – Motivation and a Brief History of Methanol ..................................... 1
1.1. Methanol Production ............................................................................................................ 3
1.1.1. The Current Process: Methanol Production from Methane ........................................... 3
1.1.2. A Potential Future Process: Methanol Production from Black Liquor Gasification
(BLG)....................................................................................................................................... 4
1.2. Conventional Methanol Process Design .............................................................................. 5
1.3. Industrial Methanol Synthesis Catalysts .............................................................................. 9
1.4. Novel Systems .................................................................................................................... 11
1.5. Scope of this Work ............................................................................................................. 13
References ................................................................................................................................. 15
Chapter 2. Methanol Kinetics Experiments ................................................................................ 17
2.1. Kinetics: Introduction and Review ..................................................................................... 17
2.2. Experiments ........................................................................................................................ 30
2.2.1. Kinetics Experiments: Experimental Set-up and Procedure ........................................ 30
2.2.2. Catalyst Preparation ..................................................................................................... 31
2.2.3. Kinetic Experiments: Parameter Estimation ................................................................ 33
2.3. Conclusions ........................................................................................................................ 37
References ................................................................................................................................. 38
Chapter 3. Membrane Reactor Experiments .............................................................................. 40
3.1. Introduction – Methanol Synthesis in a Membrane Reactor .............................................. 40
3.1.1. The Membrane Reactor Concept ................................................................................. 40
3.1.2. Different Types of Membrane Reactor ........................................................................ 44
3.1.2.1. Extractor-Type: Fischer Tropsch and Alcohol Synthesis ......................................... 44
3.1.2.3. Contactor-Type ......................................................................................................... 47
3.1.2.3. Distributor-Type .................................................................................................... 50
3.2. Membrane Reactor Experiments ........................................................................................ 53
3.2.1. Methodology ................................................................................................................ 53
3.2.2. Experimental Set up..................................................................................................... 54
3.2.3. Membrane Modification .............................................................................................. 56
3.2.3.1. Modification Method ................................................................................................ 59
v | P a g e

3.2.3.2. Modified Membrane Characterization...................................................................... 60
3.2.3.2.1. Break-Through Pressure ........................................................................................ 60
3.2.3.2.2. Contact Angle Measurements ................................................................................ 63
3.2.3.2.3. Membrane Morphology ......................................................................................... 65
3.2.3.2.4. FTIR – DRIFTS Test ............................................................................................. 66
3.2.3.2.5. Thermogravimetric Analysis (TGA) ..................................................................... 67
3.3. Results and Discussion ....................................................................................................... 68
3.3.1. Effect of Feed Flow Rate ............................................................................................. 68
3.3.2. Effect of Reaction Side Pressure ................................................................................. 70
3.3.3. Effect of Temperature .................................................................................................. 71
3.4. Conclusions ........................................................................................................................ 72
References ................................................................................................................................. 73
Chapter 4. Membrane Reactor Mathematical Modeling ............................................................ 92
4.1. Scope of Simulation ........................................................................................................... 92
4.2. Equations and Boundary Conditions .................................................................................. 93
4.2.1. Equations and Boundary Conditions ........................................................................... 93
4.2.2. Results and discussion: ................................................................................................ 99
4.2.2.1. Effect of temperature and pressure ........................................................................... 99
4.2.2.2. Effect of liquid thickness in the membrane ............................................................ 101
4.2.2.3. Effect of Flow Rate................................................................................................. 102
4.3. Conclusions: ..................................................................................................................... 104
References: .............................................................................................................................. 105
Chapter 5: Suggestions for Future Work ................................................................................. 106

1 | P a g e

Chapter 1. Introduction – Motivation and a Brief History of Methanol
Methanol (CH
3
OH or MeOH), which is also known as methyl alcohol or wood alcohol, is
the lowest molecular weight (MW) alcohol, and is a volatile, colorless, flammable, poisonous
liquid at room temperature. Methanol is produced naturally by many varieties of bacteria, as part
of their anaerobic metabolism, and it gets released into the atmosphere, where however after
several days it gets oxidized (in the presence of sunlight) by the oxygen into water and carbon
dioxide. Methanol is reported to have, most probably, first been discovered by Boyle in 1661
during the distillation of raw vinegar (Boyle named the new compound adiaphorus spiritus
lignorum). There are no written reports, however, about the use of this compound before the 19th
century.
The chemical formula and structure of methanol were discovered, independently, by
Dumas and Von Liebig in 1834. The term “methyl” was used for the first time in chemistry in
late 1835 owing to their research. During that time, several efforts were made by a number of
researchers to synthesize methanol. The first effort was by Berthelot in 1857 using dry-wood
distillation. As a result of his research methanol was commonly called then “wood alcohol.”
Methanol is mainly produced today as chemical grade (or AA grade), with specifications as
shown in Table 1.1.
Fuel grade methanol (e.g., for use as fuel in boilers or in turbines) is widely discussed
today, but apart from a few demonstration applications, is not widely utilized commercially. Air
Products and Chemicals, Inc. has, for example, an extensive active fuel-grade methanol test
program, using raw methanol produced through its liquid-phase MeOH technology [1] or
LPMeOH. Current interest in methanol for such applications comes from most recent attention
on its potential use in polymer electrolyte membrane (PEM) fuel cells as a result of recent global
2 | P a g e

energy shortages and progressively stricter emission regulations. The preferred fuel for these fuel
cells is hydrogen (H
2
) which, however, is not readily available today, and therefore a fuel
processor for the in situ production of H
2
is critical technology for the further development and
use of these systems for either mobile or stationary applications [2]. Methanol and gasoline are
considered the two most promising hydrogen energy carriers for such fuel cells [3]. Methanol is
attracting, in addition, particular attention here because of the potential for its direct use as fuel,
without having to be converted into H
2
first, in advanced types of fuel cells (known as direct
methanol fuel cells or DMFC).
Table 1.1. Chemical grade methanol specifications

Methanol is used also as a starting raw feed-stock for the production of a variety of other
chemicals and solvents, including formaldehyde, acetic acid, methyl tertiary butyl ether (MTBE)
and various fuel additives. Methanol finds, itself, use as a solvent, for example, in the
formulation of various antifreeze solutions, as well as to extract, wash, dry and crystallize
pharmaceutical and agricultural chemicals. It is used in the production of methylamines which
3 | P a g e

are used as intermediates in the production of a broad range of specialty chemicals with
applications in water treatment, shampoos, liquid detergents and animal feeds. It is also utilized
to produce methyl methacrylate (MMA), which is employed in the production of acrylic
polymers, and dimethyl terephthalate (DMT) used to make polyesters. It finds use in the
production of chlorine dioxide (a bleaching agent for the pulp and paper industries), of glycol
ethers (used in acrylic coatings, and in the new high-solid and waterborne coatings), and of
methyl mercaptan, which is used as an intermediate in the production of DL-methionine, an
amino-acid supplement in animal feeds [4]. Methanol has a number of advantages both as a fuel
but also as the raw material in the production of other chemicals, such as being more easily
transportable than methane (natural gas) and other gaseous fuels, having a high energy density,
needing no desulphurization, and its reaction to produce hydrogen (e.g., steam reforming)
proceeding at relatively moderate temperatures (200 – 400
◦
C) [5].
1.1. Methanol Production
1.1.1. The Current Process: Methanol Production from Methane
Today, methanol is principally manufactured by a two-step method, which involves first
the production of synthesis gas (syngas) from methane, via a number of routes but most
commonly the steam reforming (R1) or partial oxidation reactions (R2):
CH
4
+ H
2
O → CO + 3H
2
ΔH
o
= 206 (kJ/mol) (R1)
CH
4
+ ½O
2
→ CO + 2H
2
ΔH
o
= -36 (kJ/mol) (R2)
Methanol is then produced from the syngas according to the MeOH synthesis reactions below:
CO + 2H
2
→ CH
3
OH ΔH
o
= -90.6 (kJ/mol) (R3)
CO
2
+ 3H
2
→ CH
3
OH + H
2
O ΔH
o
= -49.5 (kJ/mol) (R4)
CO + H
2
O → CO
2
+ H
2
ΔH
o
= -41.2 (kJ/mol) (R5)
4 | P a g e

R1, the most common route to produce syngas, is a highly endothermic reaction and suffers from
high energy costs and thermal inefficiencies. Methanol synthesis (only two out of the three, R3-
R5, are independent reactions) is an exothermic process but needs a catalyst (typically copper-
based) and high temperature and high pressure conditions to achieve high yields of methanol.
The above commercial two-step methanol synthesis process from methane is energy
intensive, and the use of high-temperature and high-pressure conditions causes catalyst sintering
and deactivation. A more desirable alternative would be a direct, one-step partial oxidation of
methane into methanol process, since such one-step process would potentially have reduced
capital and operational costs. Such process is not currently commercially available, however, but
its development is attracting substantial current research interest. Among the methods and
techniques investigated, a non-thermal plasma chemical process is considered one of the most
promising technologies in synthesizing methanol [6]. In addition to mobile fuel cell uses, there is
today an increasing number of other applications which require that methanol be produced on
demand and in much smaller quantities [1,7], which motivate additional research into the
development of such smaller scale and flexible methanol synthesis processes. A comprehensive
review of current and potential future uses of methanol can be found in [8].
1.1.2. A Potential Future Process: Methanol Production from Black Liquor
Gasification (BLG)
Black liquor gasification (BLG) in modern pulp mills as a first step towards bio-fuel or
electricity production is an area of ongoing development, most of the current efforts driven by
climate change and fuel security concerns. Naqvi et al. [9] evaluated two different BLG systems,
regarded as the most promising future technology candidates for methanol production: (i) an
oxygen-blown, pressurized thermal BLG system; and (ii) a dry BLG system with direct
causticization. A key objective of their study was to assess integration possibilities of these BLG
5 | P a g e

technologies with a base-case Kraft pulp mill producing 1000 air-dried tons (ADt) of pulp/day in
terms of replacing the conventional recovery cycle. The study compared the system’s
performance in terms of methanol production, energy efficiency, and potential CO
2
reductions.
Their study indicated a larger potential of black liquor conversion into methanol from the
oxygen-blown, pressurized BLG system (~77 million tons/year of methanol) than the dry BLG
system (~30 million tons/year of methanol) utilizing identical amount of black liquor available
worldwide (220 million tons of dried solid (tDS)/year). The potential CO
2
emissions reduction
from the transport sector is substantially higher in the pressurized BLG system (117 million
tons/year CO
2
reductions) as compared to the dry BLG system (45 million tons/year CO
2

reductions). However, the dry BLG system with direct causticization showed better results when
considering the consequences of additional biomass import. The study, in addition, compared
methanol production via BLG with the production of other bio-refinery products, e.g., hydrogen,
dimethyl ether (DME) and bio-methane. The study has shown that there is a potential to replace
the conventional black liquor recovery cycle with the BLG systems in the pulp mills, thus co-
producing pulp and valuable energy products [9].
1.2. Conventional Methanol Process Design
Similar to Fischer Tropsch (FT) synthesis, one of the most difficult problems in designing
a reactor for methanol synthesis is removing the heat of reaction while maintaining precise
temperature control. This is important because catalyst life is seriously reduced by excessive
temperatures. Tijm et al. [1] reviewed the technical development of the methanol process during
the last few decades. The reactor technologies that were used for methanol synthesis in
commercial plants fall into two categories, namely, the gas-phase technologies and liquid-phase
technologies [1].
6 | P a g e

The gas-phase technologies operate using fixed-bed catalytic reactors, and many different
types of designs have been developed. A Lurgi methanol reactor is shown in Figure 1.1(a). The
catalyst is contained in fixed tubes, and the reactor temperature is controlled by the steam
pressure. An ICI (Imperial Chemical Industries) methanol reactor, as shown in Figure 1.1(b); it is
an adiabatic reactor with multiple catalyst zones and cold unreacted gas injected in between each
zone in order to increase the once-through conversion and also to lower the reaction temperature
to extend the catalysts life.

Figure 1.1. Fixed-bed reactors for methanol synthesis. (a) Multi-tubular fixed-bed. (b) Quench-
cooled fixed-bed.
This quench-cooled design was also used by the Haldor-Topsøe methanol synthesis
reactor and the TEC (Toyo Engineering Corporation) MRF-Z reactor. Conversion of syngas to
methanol in fixed-bed reactors is limited by the reaction equilibrium and the high-temperature
sensitivity of the catalyst. Temperature control is achieved by recycling large amounts of
hydrogen-rich gas, utilizing its heat capacity to moderate the gas temperatures, and the resulting
higher gas velocities to enhance the heat transfer.
7 | P a g e

To improve the heat transfer characteristics of the MeOH synthesis reactor, Sherwin and
Frank [10] developed a liquid-phase process for methanol synthesis in which a mineral oil acted
as a temperature moderator, which facilitated heat removal by transferring the heat of reaction
from the catalyst surface to boiling water in an internal tubular heat exchanger. This liquid-phase
process has a number of advantages: (1) The ability to operate with syngas rich in CO, as
obtained from modern coal gasifiers, and to produce a product that does not require further
purification before being used as a fuel. It was reported [10], for example, that CO
concentrations in the syngas in excess of 50% were routinely processed without any adverse
effects on catalyst activity. In contrast, fixed-bed, gas-phase methanol synthesis processes must
use a syngas rich in hydrogen. Such processes, as a result, require that a CO-rich syngas
feedstock undergoes stoichiometry adjustment via the water-gas shift reaction in order to
increase its hydrogen content. Typically, fixed-bed processes are limited to a feed syngas content
of ~16% CO to limit the once-through conversion in order to avoid excess heating; (2) enhanced
heat transfer of the highly exothermic heat of reaction; (3) a high once-through conversion of
CO.
A three-phase slurry reactor is typically being used for the liquid-phase methanol
synthesis process. The catalyst is used as a fine solids suspension in a high MW liquid. Slurry
reactors for the methanol synthesis reaction have been studied by many groups, e.g., Ozturk et al.
[11], Lishun et al. et al. [12], Van der Laan et al. [13], and Setinc et al. [14], among others.
Various types of slurry reactors are shown in Figure 1.2.
8 | P a g e

Figure 1.2. Types of slurry reactors utilized in methanol synthesis: (a) Bubble column. (b)
Internal-loop airlift reactor. (c) External-loop airlift reactor. (d) Spherical reactor.
During the liquid-phase methanol synthesis process a heat exchanger is typically
incorporated in the reactor. The heat transfer coefficient on the slurry side of the heat exchanger
is relatively large, and thus, a relatively small heat transfer area is required, and as a result the
heat exchanger occupies only a small fraction of the reactor volume. Further, the heat transfer
between the fine solid catalyst particles and the liquid phase in which they are suspended is
highly efficient, thereby keeping the temperature low and thus allowing high once-through
conversions without any loss of catalyst activity.
The aforementioned liquid-phase methanol (LPMEOH) technology, developed by Air
Products was tested for several years at a 10 (ton/day) pilot plant at La Porte, Texas. A
commercial-scale demonstration (260 ton/day) of the LPMEOH process was started in 1997, at
Eastman’s chemicals-from-coal complex in Kinsport, Tennessee. The results of the
demonstration process were satisfactory and verified the advantages of the use of the slurry
reactor [15] for the methanol synthesis process.

9 | P a g e

1.3. Industrial Methanol Synthesis Catalysts
Most of the catalysts used for modern-day, so-called low-pressure methanol synthesis
processes are copper-based. Among the most highly active and highly selective catalysts for
methanol production are Cu/ZnO/Al
2
O
3
catalysts [6]. These catalysts are quite active even for
temperatures as low as 200
◦
C. Typical compositions of some of the catalysts industrially utilized
for methanol synthesis are reported in Table 1.2.

Table 1.2. Composition of methanol synthesis catalysts
Producer Cu[%] Zn[%] Al[%]
BASF 65-75 20-30 5-10
Süd Chemie 65-75 18-23 5-10
ICI 61 30 9
Du Pont 50 19 31
Haldor-Topsøe 50-60 21-25 15-28

There are still arguments regarding the role of the catalyst active site during methanol
synthesis. However, synergy between the main catalyst the promoter and/or the support is now a
rather well-established phenomenon. This synergy is due to the transfer of species among the
catalyst components, which need not even be in direct physical or chemical contact. Varying the
catalyst composition has been shown [1] to results in differences in the morphology and
electronic effects among the catalyst investigated. Other important considerations include
catalyst particle size and surface area. Small particle size reduces the diffusion resistance of
reactant molecules entering into the catalyst pore allowing for full access and utilization of the
active sites. This means that small particle size can promote higher activity. Larger surface area
can also improve the dispersion of the active component, and provide added resistance to
sintering.
10 | P a g e

In the last decade, Synetix developed a new catalyst (Synetix 51-7). The catalyst contains
MgO which was reported to limit catalytic sintering [16]. Synetix reports that the catalyst has a
30% higher Cu surface area than other competitive methanol synthesis catalysts. Its research also
indicated that using the Synetix 51-7 (as well as a different MeOH synthesis catalyst, Synetix 51-
3) prevents a high concentration of CO
2
in the syngas from having a permanent effect on catalyst
performance. Research findings with other catalysts [17] have suggested that CO
2
-rich
conditions may cause irreversible damage.
Süd Chemie (in the USA, Union Catalyst, Inc.) has reported that the damaging effect of
high CO
2
levels on methanol catalysts is, in fact, a result of the water formed by the reverse
water gas shift (RWGS) reaction [18]. The company has also argued that the increasing diversity
of converter types, feed-stock and operating conditions found in methanol production creates the
need for tailor-made catalysts. With this in mind, Süd Chemie developed two new methanol
catalysts to compliment its existing C79-4 GL catalyst [19]. Süd Chemie reports that the two new
catalysts have a higher tolerance towards carbon oxides and are designed for an optimum balance
between activity, selectivity and lifetime, and to operate under a broad range of different
industrial conditions. C79-4 GL shows the best selectivity for isothermal reactors using syngas
obtained from the partial oxidation of oil fractions and/or coal. The new catalysts, C79-5 GL and
C79-6 GL, are reported [20] to have a different matrix structure, providing a more stable
crystalline Cu distribution. C79-5 GL is reported to have a long lifetime, claimed to be up to 4
years, and is particularly suited for isothermal and adiabatic converters that use syngas obtained
by methane steam reforming. C79-6 GL is designed specifically for use with syngas containing
high levels of olefin impurities (e.g., acetylene off-gases).
11 | P a g e

Haldor-Topsøe produces a multi-purpose methanol synthesis catalyst, MK-101 [21], that
has found extensive industrial use with most types of converters using syngas obtained from
variety of feed-stocks and reforming technologies. For example, Haldor-Topsøe reported that the
catalyst was recently used in an ammonia–methanol co-production plant. In addition, Haldor-
Topsøe has also developed a different methanol synthesis catalyst MK-121 with an optimized
copper dispersion, which according to them ensures improved activity (10-15% higher initial
activity) and stability (even under more severe operating conditions) over the MK-101, while at
the same time attaining a good selectivity. The MK-121 was reported capable to be used with a
wide range of gas compositions, to have a longer useful catalyst lifetime, higher conversion and
carbon efficiency, and lower by-product level. The stability of MK-121 is higher, even at severe
operating conditions. Since the higher activity of MK-121 allows operation at lower
temperatures, where conditions for by-product formation are less favorable, the total methanol
synthesis catalyst level of by-products is typically reduced relative to that of the predecessor,
MK-101. Consequently, even higher methanol production is achievable, due to both the higher
catalyst activity and to the reduction of product losses in the distillation section.
1.4. Novel Systems
Further significant improvement in methanol production must either come from catalyst
improvements, or via novel ways to carry out the reaction [22–25]. Prior works in the latter area
are reviewed in this section. Two such approaches involve:
a) recycling of the unconverted synthesis gas after products separation;
b) in situ product removal of the products of the reaction.
One of the earliest studies utilizing in situ product removal was by Westerterp et al. [23], who
proposed the selective adsorption of water and methanol on a solid in a trickle-bed reactor.
12 | P a g e

Removal of these reaction products is expected to benefit the reactor conversion by shifting to
the right the equilibrium of the two synthesis steps (R1 and R2 above). In addition the water
present is known to accelerate the crystallization of CuO and ZnO contained in the catalyst, thus
resulting in catalyst deactivation [9], and its removal is also likely to benefit maintaining catalyst
activity.
Continuous product removal from the reaction zone to improve both the reactants
conversion and methanol yield can also be accomplished via the use of membrane reactors [9]. A
membrane reactor (Figure 1.3) is a device that combines the separation via a membrane with
catalytic reaction in a single unit. The membrane selectively removes one or more product
species from the reaction system, typically resulting in higher conversion (this technology is
further described in Chapter 3).

Figure 1.3. Schematic of a membrane reactor
As it will be discussed further in Chapter 3, some of the studies, so far, have reported
higher reactant conversion and methanol yield via the use of selective membranes which remove
in situ the reaction products (methanol and/or water). Other membranes are used to deliver in situ
one of the reactants, e.g., hydrogen. Since synthesis gas from coal and biomass gasification is
13 | P a g e

low in H
2
, the addition of H
2
enhances methanol production [25]. Table 1.3 summarizes some of
the important MR studies, the improvements that can be achieved with respect to a traditional
reactor by using a membrane reactor (further discussion to be provided in Chapter 3). In
particular, the most promising results are obtained when methanol and/or water are selectively
removed from the reaction zone.

Table 1.3. Improvements in performance obtained using MRs for methanol production
Membrane type Improvement over the traditional reactor Ref.
Lithiated Nafion 40% methanol yield improvement 22
Pd-Ag 9% with respect to the thermodynamic equilibrium 25
Lithiated Nafion 40% methanol yield improvement 26
Silicon rubber/ceramic 22% conversion improvement 27
Zeolite 60% conversion improvement 28
Zeolite 132% conversion improvement 28

1.5. Scope of this Work
In our work we study a MR with solvent purge whose aim is the in-situ product
(methanol) removal in order to increase the product yield. In our approach, the membrane serves
as an interface contactor for the selective permeation and removal of methanol. The solvent
chosen is an inert high boiling point (B.P.) liquid in which the main reaction product, methanol,
has the highest solubility. The solvent, as a result, selectively removes in-situ from the reactor
via the membrane interface contactor the methanol, while the reactants (CO, CO
2
, H
2
) remain in
the reactor due to their very low solubility in the solvent which infiltrates the membrane and
blocks the passage of gases. In the MR experiments we utilize a commercial alumina ceramic
tube membrane with asymmetric structure. Prior to the start of the membrane reactor
experiments, the membrane surface is modified in order to increase membrane hydrophobicity,
in order to achieve desirable operating conditions, i.e., membrane blockage to prevent leak-
14 | P a g e

through of the gaseous reactants but not excessive membrane infiltration that would adversely
impact the mass transfer characteristics through the membrane. In the experiments reported in
this Thesis, we were able to operate the MR using this modified membrane under the proposed
operating conditions without the significant loss of the inert solvent.
In our research we also investigate the kinetics of the methanol synthesis reaction from
CO/CO
2
/H
2
mixtures on a commercial Cu/Al
2
O
3
/ZnO catalyst, specifically the effect on the
resulting rates of varying the CO/(CO
2
+CO) ratio, the stoichiometric number, pressure,
temperature and feed flow rate. The experimental rate data are utilized to derive appropriate
global rate expressions for the catalyst utilized.
In our study, we also develop a model of the MR in order to explore the effect of various
operating parameters such as temperatures, pressures, and feed flow rate on reactor performance,
including the total conversion of carbon. The reaction rate equations derived from the kinetic
experiments are used in the model. Membranes with different properties are simulated, and their
performances are compared.
In summary, in this Thesis an efficient syngas-to-methanol production process is
investigated. The remainder of this Thesis is organized as follows: Chapter 2 describes the
kinetic experiments and the derivation of the rate equations. Chapter 3 describes the MR
experiments, and the evaluation of the membranes as interphase contactors. Chapter 4 describes
the MR model and the results of the modeling investigations. Finally, Chapter 5 proposes
additional research that needs to be carried out to further validate the proposed technology.

15 | P a g e

References
1. P.J.A. Tijm, F.J. Waller, D.M. Brown, Applied Catalysis A: General, 221, 275–282 (2001)
2. Y.H. Chin, R. Dagle, J. Hu, A.C. Dohnalkova, Catalysis Today, 77, 79–88 (2002).
3. J. Han, I. Kim, K.S. Choi, Journal of Power Sources, 86, 223–227 (2000).
4. J. Jordan and associates, Annual Methanol Forum in Houston (2010).
5. K.Sekizawa, S.Yano, K. Educhi, H. Arai, Applied Catalysis A: General, 169: 291–297 (1998).
6. A. Indarto, IEEE, Transactions on Dielectrics and Electrical Insulation Vol. 15 (2008).
7. E.H. Stitt, Proc. NATO ASI, “Sustainable Strategies for the Upgrading of Natural Gas:
Fundamentals,Challenges and Opportunities”,(Derouane, E.G.;Parmon,V,;Lemos,F.;Ribeiro,
F.R.;eds), Springer, Dortrecht (2005) 185 (Sections 2.6.2,3.3.5) Strub,R.A.;Imarisio,G.”
Hydrogen as an Energy Vector’,Reidel,Dordrecht (1980).
8. G. A. Olah, Angewandte Chemie International Edition, 44, 2636-2639 (2009).
9. M. Naqvi, J. Yan, E. Dahlquist, Applied Energy, 90, 24–31(2012).
10. M.B. Sherwin, M.C. Frank, Hydrogen Processing, 55(11), 122-124 (1976).
11. S. Ozturk, Y. Shah, Chemical Engineering Journal, 37, 177-192 (1988).
12. H. Lishun, Chemical Engineering and Processing,46, 905-909 (2007).
13. G.Van der laan, Catalysis Today, 48, 93-100 (1999).
14. M. Setinc, Chemical Engineering Science, 54, 3577-3586 (1999).
15. T. Wang, Industrial & Engineering Chemistry Research, 46, 18 (2007).
16. T. Fitzpatrick, Erdol,Ergas/Kohle,117(9),408-412 (2001).
17. J. Wu, M. Saito, M.T. Takeuchi, Applied Catalysis A: General, 218, 235–240 (2001).
18. J. Ladebeck, Hydrocarbon Processing, 72(3), 89-91(1993).
19. S. Sa, J.M. Sousa, A. Mendez, Chemical Engineering Science. 66(20), 4913-4921 (2011).
16 | P a g e

20. J. Richardson, Nitrogen Methanol.
21. G. He, Cailiao Gongcheng/Journal of Materials Engineering,2, 21-24(2010).
22. R.P.W.J. Struis, S. Stucki, M. Wiedorn, Journal of Membrane Science, 113, 93–100 (1996).
23. K.R. Westerterp, M. Kczynski, T.N. Bodewes, M.S.A. Vrijland, Chemical Engineering &
Technology., 61, 193–199 (1989).
24. J. Wu, M. Saito, M.T. Takeuchi, Applied Catalysis A: General, 218, 235–240 (2001).
25. M.R. Rahimpour, S. Ghader, Chemical Engineering and Processing, 43 (9), 1181–1188
(2004).
26. G. Chen, Q. Yuan, Separation and Purification Technology, 34, 227–37 (2004).
27. F. Gallucci, L. Paturzo, A. Basile, Chemical Engineering and Processing, 43, 1029–33(2004).
28. F. Gallucci, L. Paturzo, A. Basile, International Journal of Hydrogen Energy, 32, 5050–
58(2007).

17 | P a g e

Chapter 2. Methanol Kinetics Experiments
2.1. Kinetics: Introduction and Review
The methanol synthesis reaction is described typically by reactions R3-R5, as discussed
in Chapter 1. As noted there, for reactor design purposes only two out of the three reactions are
needed. Early kinetic models were derived for the ZnO/Cr
2
O
3
catalyst which was utilized in the
high-pressure process, which has now almost been abandoned completely in favor of the low-
pressure technology: A classic example from this early work is the rate equation proposed by
Natta et al. [1]:
23
3
23
2*
2
3
/
(2 1)
()
CO H CH OH
CH OH
CO H CH OH
f f f K
r
A Bf Cf Df


  

In the above rate equation, f
i
denotes the fugacity of component i and A, B,C, and D are rate
constants to be estimated by experimental data. Natta et al. [1] assumed that methanol is
produced via the hydrogenation of CO. Their proposed Langmuir-Hinshelwood-Hougen-Watson
(L-H-H-W) type rate equation was derived under the assumption that the rate determining step
involves the reaction between one adsorbed CO molecule and two adsorbed H
2
molecules. Their
kinetic investigations were carried out in the absence of CO
2
, and at high temperature, as a result
of which catalyst deactivation was quite significant [1].
Leonov et al. [2] were the first to model the methanol synthesis kinetics over a low-
temperature Cu/ZnO/Al
2
O
3
catalyst. In their model they assumed that CO was the source of
carbon in the methanol produced, and they did not account for the influence of CO
2
in the feed
[2]:
3 2
3
2
3
0.34
0.5
0.66 0.5 *
2
( ) (2 2)
CH OH
CH OH
CO H
CH OH
CO H
P
PP
rk
P P P K
  
18 | P a g e

Klier et al. [3] considered a kinetic model that quantitatively describes the dependence of
methanol synthesis on the concentration of CO
2
. Though in their analysis CO is not the only
source of carbon in MeOH, it is still considered to be the most important source of carbon. Klier
et al. [3] demonstrated that there is a CO
2
/CO ratio in the syngas that maximizes the methanol
synthesis rate. An additional finding of significance was that selectivity changes with the syngas
composition. At concentrations of CO
2
between 0 and 10%, methanol is the sole product while at
higher concentrations of CO
2
methane is also formed as a by-product. They suggested the
following rate equation based on their experimental observations [3]:

(

)

(

) (2-3)
The equilibrium constant

and

are defined as follows:.

(2-4)

(2-5)
For the experimental reaction conditions employed in their work, the reverse reaction rate
terms are, however, negligible. Their proposed mechanism was able to explain the maximum in
the methanol synthesis rate observed when varying the CO
2
/CO ratio based on the presence of a
formate intermediate, which is produced from the reaction between CO
2
and H
2
or from CO and
H
2
O. They theorized that in the CO
2
-free synthesis gas, the surface formate would not form since
CO
2
and H
2
O are absent. Consequently, methanol would be formed at a low rate. In the CO
2
-rich
synthesis gas, the strong adsorption of CO
2
would block the sites active in formate generation,
and the methanol synthesis would then be retarded. At an intermediate CO
2
concentration, the
synthesis rates would be high because sufficient amounts of carbon dioxide and/or water would
19 | P a g e

be available to produce the surface formate. This concentration would not be enough to block the
active sites by adsorption.
Villa et al. [4] performed a complete kinetic investigation of methanol synthesis from
syngas over a commercial Cu/ZnO/Al
2
O
3
catalyst, and in their work both the methanol synthesis
reaction from CO and hydrogen and the reverse shift reaction of CO
2
and hydrogen were
considered. They performed forty kinetic experiments for pressures ranging from 30 to 95 atm
and temperatures in the range 215-254
o
C. They assumed that the hydrogenation of CO is the
only source of carbon in methanol. Their rate equations are shown below, with the fugacity
coefficients for the various species calculated by using the Redlich-Kwong-Soave equation of
state:

where (A,B,C,G, and M) are experimentally derived rate constants. The above reaction rate
expressions were derived based on the assumption that the generation of methanol and the water
gas shift reaction occur on two different types of sites.
Graaf et al. [5] performed kinetic experiments with a commercial catalyst (Cu-Zn-Al),
and with a CO, CO
2
and H
2
mixture as the feed. Their experiments showed that methanol can be
formed simultaneously from both CO and CO
2
. They described their results by a dual-site
Langmuir-Hinshelwood mechanism, based on dissociative hydrogen adsorption and three
independent reactions: methanol formation from CO, methanol formation from CO
2
, and the
water gas shift reaction (however, since only two of the three reactions are truly independent the
validity of their assumption and the further kinetic modeling are questionable). Statistical
20 | P a g e

discrimination among various mechanistic schemes (48 possible models) allowed them to select
the following final rate equations:

(

)

(

)

(

)

The kinetic parameters were determined from experimental data in the temperature range from
210 to 245
◦
C. They report [5] that for the commercial catalyst utilized in their study, their
proposed kinetic model explained the experimental results with a significantly improved
accuracy as compared with the kinetic model proposed by Villa et al. [5] (please see, however,
discussion above indicating that there is no real need to consider three rate expressions) .
Low-pressure methanol synthesis on a commercial Cu-ZnO-Al
2
O
3
catalyst was also
studied by Skrzypek et al. [6]. Two different kinds of feed were considered: a feed containing
only CO
2
/H
2
, and another containing CO
2
/CO/H
2
. They report the following two Langmuir-
Hinshelwood type reaction rate expressions:

⁄

(

)

⁄

(

)

(

)
21 | P a g e

(

)
In the above equations K
i
are the equilibrium constants.
Van den Bussche et al. [7] carried out experiments with an industrial Cu/ZnO/Al
2
O
3

catalyst, at pressures between 15 and 51 bar, temperatures varying between 180 and 280
◦
C, and

/

ratios in the feed ranging from 0 to 4.1. They reported the following two rate
expressions to describe their experimental results (The pressures are expressed in bar and the
reaction rates in mol/kg
cat
/s. The use of fugacities, rather than partial pressures, was also
considered by Van den Bussche et al. [7] by applying the Soave-Redlich-Kwong equation of
state. However, the compressibility factors were never outside the 0.99 to 1.01 range, and were
found to give negligible changes in the results):
2 2 2 2 2
3
2 2 2 2 2 2
22
2 2 2 2
2 2 2 2 2 2
22
' ' * 3
5 2 3 4 1
3
34
'
13
34
[1 (1 / )( / )]
(2 15)
(1 ( / )( / ) )
[1 ( )( / )]
(2 16)
(1 ( / )( / ) )
CH OH
H O H O
H O H O
a CO H H O H CO
MeOH
H H H H H O H O
CO CO H O H CO
RWGS
H H H H H O H O
k K K K P P K P P P P
r
K K K K P P K P K P
k P K P P P P
r
K K K K P P K P K P


  


  

The equilibrium constants K
1
*
and K
3
*
in the above expressions were taken from Graaf et al [5]:
*
10 1
*
10 3
3066
log 10.592 (2 17)
2073
log 1 / 2.029 (2 18)
K
T
K
T
  

  
while the rest of the rate constants were derived by nonlinear parameter fitting of their
experimental data using the following formulas:

(

(

))
in which T
av
equals 501.57 K, and where in the original context of Arrhenius or Van’t Hoff
expressions,
22 | P a g e

B(i) represents either E or (-∆H) or a combination of both, see Table 2.1 for the values of these
parameters. Van den Bussche et al. [7] in their mechanistic interpretation of their data adopted
the idea that CO
2
is the main reactant during methanol synthesis, while CO converts into CO
2

(and subsequently into methanol) via the WGS reaction:
23
. WGS hydrog
CO CO CH OH        

Vanden Bussche et al. [7] assumed that both reactions proceed on the copper surface,
with the role of ZnO limited to structural promotion. A reaction mechanism, occurring
exclusively on copper, was proposed and is shown in Fig. 2.1 [7].
Table 2.1. Parameter values for the steady-state kinetic model, parameter values for the
Arrhenius expression (k(i)= A(i) exp.(B(i)/RT);used in the rates by Vanden Bussche et al. [7]
√

A
B
0.499
17,197

A
B
6.62×10
-11

124,119

A
B
3,453.38
—

A
B
1.07
36,696

A
B
1.22×10
10

-94,765
As Fig. 2.1 indicates, both H
2
and CO
2
adsorb dissociatively on the copper surface [5]. The
oxidizing adsorption of CO
2
on metallic copper is promoted by traces of surface oxygen or
alkaline species. On the oxidized copper surface, carbonate structures are formed by further
adsorption of CO
2
[9-11].
23 | P a g e

Figure 2.1. Mechanism proposed by Vanden Bussche et al. [7]

These carbonates are quickly hydrogenated, first to bicarbonate structures and
subsequently to Cu formate, formaldehyde, methoxy species, and finally methanol. In this
sequence, shown in the left-hand side of Figure 2.1, the rate determining step is the
hydrogenation of the formate species, which is generally accepted to be the longest living
intermediate in methanol synthesis on copper [12-14]. During the two stages in the
hydrogenation of CO
2
into methanol, surface oxygen is released from the molecule. This species
is also hydrogenated by the available hydrogen atoms, yielding hydroxyl groups and
subsequently water, which is known to desorb relatively slowly. The right-hand side of Figure
2.1 describes the reverse water gas shift reaction, proceeding according to a redox mechanism. In
this sequence of reactions, the dissociative adsorption of CO
2
is rate determining, as was shown
by Nakamura et al. [15], Fujita et al. [16], and Ernst et al. [17]. Figure 2.2 shows the different
elementary reaction steps considered by Van den Bussche et al. [7]. Derivation of the rate
equations above was performed under the hypothesis of pseudo-steady-state of the concentration
of the different surface intermediates [18], and by assuming that the concentrations of the
24 | P a g e

bicarbonate, formaldehyde, methoxy, methanol, and hydroxyl species adsorbed under the
reaction conditions are negligible [5,9,11,14,17].

Figure 2.2. Reaction scheme for the synthesis of methanol and the reverse water gas shift
reaction proposed by Vanden Bussche et al. [7] (rds signifies the rate determining step)

Kubota et al. [19] developed a high-performance Cu/ZnO-based catalyst that was highly
active and selective in methanol synthesis. In agreement with Van den Bussche et al. [7], they
report that methanol is produced via CO
2
hydrogenation (and that direct methanol synthesis from
CO can be ignored), in combination with the reverse water gas shift (RWGS) reaction which
takes place simultaneously. Their rate equations for methanol synthesis (r
M
) and the RWGS
reaction (r
R
) below were derived on the basis of the assumption that methanol is produced via
formate and methoxy intermediate species, and that the surface reaction between the formate
species and adsorbed hydrogen atoms is the rate–determining step. The RWGS reaction is
reported to take place either via a formate intermediate or via the direct decomposition of CO
2

into CO on the copper surface.
r
M
= k
M
[P
CO2
P
H2
-P
CH3OH
P
H2O
/[K
M
P
2
H2
]}/A
2
(2-20)
25 | P a g e

r
R
= k
R
{P
CO2
-P
CO
P
H2O
/[K
R
P
H2
]}/A (2-21)
A= 1+K
H2
P
1/2
H2
+K
CO2
P
H2O
+K
H2CO2
P
1/2
H2
P
CO2
+K
H2OH
P
H2O
/P
H2
+K
H2O
P
H2O
(2-22)
Fundamental studies of methanol synthesis and decomposition, mainly over Cu-based
catalysts, were carried out by Rozovskii et al. [20] who performed kinetic experiments at small
contact times as well as tracer investigations. They report that for temperatures in the range (180
– 260
o
C) and pressures in the range (0.1 – 20 MPa) the reaction over Cu-based catalysts (but on
Zn-Cr-oxide catalysts as well) proceeds via an alternative pathway: CO
2
→CO→CH
3
OH. They
reported that the combination of these two reactions is a molecular chain reaction where excess
oxygen atoms migrate between CO
2
and H
2
O, supporting the methanol synthesis from CO/H
2

mixtures. Rozovskii et al. [20] provide the following surface reaction mechanism in which Z is a
Cu-containing active site (Figure 2.3):

Figure 2.3. Mechanism of Rozovskii et al. [20]
The following reaction rate is reported for the methanol synthesis step:

⁄

26 | P a g e

in which k
i
and K
i
are rate and surface equilibrium constants, K
p(m)
is

the methanol synthesis
equilibrium constant, and P
i
the partial pressures of the various components.
Shahrokhi et al. [21] studied the dynamic behavior and control of low-pressure methanol
synthesis fixed-bed reactors. For modeling the reactor, they utilized a transient heterogeneous
model, accounting also for mass and energy dispersion in the axial and radial directions, and the
Ergun equation to estimate the pressure drop. The pseudo-steady state reaction and diffusion
equation was utilized for the solid phase to calculate the effectiveness factor based on the rate
equations of Van den Bussche et al. [7] for the two independent reactions (hydrogenation of CO
2

and the RWGS reaction). Their steady state simulation results were in good agreement with
industrial data.
Lim et. al. [10] also studied methanol synthesis on Cu/ZnO/Al
2
O
3
/ZrO
2
catalyst, and
investigated the influence of carbon dioxide during hydrogenation. In addition to the three main
reactions for methanol synthesis, they considered one extra side reaction for the synthesis of
dimethyl ether (DME) from methanol in order to account for their experimental observations.
They used the elementary reactions shown in Table 2.2 for the hydrogenation of CO and CO
2
as
well as the RWGS reaction, and considered two different catalyst sites for the adsorption of CO
and CO
2
. They report that hydrogen and water adsorb on the Zn site, while the synthesis of
methanol from carbon dioxide occurs via a formate species which is adsorbed on copper. They
performed a total of 28 experiments, and the experimental data were analyzed via a non-linear
parameter estimation procedure, with the reaction equilibrium constants taken from the study of
Graff et al [5] (Eqn. 2-11, 2-12, 2-13).

27 | P a g e

Table 2.2. Elementary reactions for the Cu/ZnO/Al
2
O
3
/ZrO
2
catalyzed methanol synthesis [10]
Adsorption

Surface reaction Elementary steps
(A) CO hydrogenation reaction

(B) Water –gas shift reaction

(C) CO
2
hydrogenation reaction

Xin et al. [11] studied methanol synthesis over a highly active Cu/Zn/Al/Zr fibrous
catalyst and used the kinetic model of Graaf et al. [5] to analyze the experimental data (Eqns. 2-
11,2-12,2-13). They were able to predict the impact of gas phase feed composition on their
measured rates
Table (2.3) summarizes the kinetic models and rate expressions that have been reported
on the methanol synthesis reaction that have been discussed, so far. For the remainder of this
Chapter we will utilize the rate expressions of Van den Bussche et al. [7], which qualitatively
(but not quantitatively – see further discussion below) agree with our preliminary experimental
investigations of the commercial catalyst (MK-121) that we utilize in our study.

28 | P a g e

Table 2.3. Kinetic Models on Methanol Synthesis
Catalyst T ( ◦C) P (atm) Kinetic equation Reference
Cu/ZnO/Cr
2
O
3
300-330
200-
315
23
3
23
2*
2
3
/
()
CO H CH OH
CH OH
CO H CH OH
f f f K
r
A Bf Cf Df


  
Natta (1955)
Cu/ZnO/Al
2
O
3
300-330
200-
315
23
2
23
2
1/2 3
( / )
(1 )
()
H CO CH OH eq
H
CO H CH OH
f f f K
r Kf
A Bf Cf Df


  
Pasquon(1960)
Cu/ZnO/Al
2
O
3
220-260 40-55
3 2
2
3
1
0.5
0.5 *
2
()
CH OH
CH OH
b
CO H
b
CO H
P
PP
rk
P P P K



Leonov(1973)
Cu-Zn 225-250 75
2
23
3
22
2 2 2
3
3 3 2 *
2
3
2
*3
1
( / ) ( / )
[1 ( / )] ( )
'( (1/ )( / ))
H CH OH
CH OH
redox CO CO CO
CH OH n
redox CO CO CO CO
CO H O H
K P P P P P K
r const
K P P F K P
k P K P P P





Klier et al
(1982)
Cu-Zn-Al 215-245 30-95
23
3
23
2 2 2
2*
2
3
*
3
2
/
()
CO H CH OH
CH OH
CO H CH OH
CO H CO H O
RWGS
f f f K
r
A Bf Cf Gf
f f f f K
r
M


  



Villa et al.
(1985)
Cu-Zn 235-265 80-140
23
2 3 2 2
2
1
2
1 2 3 4 5 6
/
()
o
CO H CH OH P
CO H CH OH H CO CO
f f f K
r
A A f A f A f A f f A f


    

Seyfert - Luft
(1985)
Cu-Zn-Al
Cu-Zn-Cr
N/A N/A
2 2 3 2 2 3
2 2 2 2
1/2 2
1 2 3
1/2 1/2 1/2
2 3 4 3
[ / ( )]
(1 )[1 / ( ) ]
o
CO H CH OH H O H P
CO H H O H
A A A f f f f f K
r
A f A f A f A f


  

Dybkjaer
(1985)
Cu-Zn-Al

210-245 15-50
2 2 2 3 2 2
3
2 2 2 2 2 2
2 3 2
3
2 2 2 2 2 2
' 3/2 3/2 *
,1 2
,1 1/2 1/2
' 3/2 1/2 *
,2 2
,2 1/2 1/2
[ / ( )]
(1 )[ ( / ) ]
[ / ( )]
(1 )[ ( / ) ]
ps c CO CO H CH OH H O H
CH OH
CO CO CO CO H H O H H O
ps c CO CO H CH OH H
CH OH
CO CO CO CO H H O H H O
RWGS
k K f f f f f K
r
K f K f f K K f
k K f f f f K
r
K f K f f K K f
r


  


  

2 2 2 2
2 2 2 2 2 2
'*
,1 3
1/2 1/2
[]
(1 )[ ( / ) ]
ps b CO CO H CO H O
CO CO CO CO H H O H H O
k K f f f f K
K f K f f K K f

  

Graaf et al
(1988)
ZnO-Al
2
O
3
200-240 50-70
2 2 2 2
3
22
2 2 2 2
33
2 2 2 2
3
22
2 2 2 2 2 2
33
2
1
2
11 3
2
1
2 2
( ) (1/ )( / )
[]
(1 )
( ) (1/ )( / )
2 [ ]
(1 )
CH OH
CH OH CH OH
CH OH
CH OH CH OH
CO H p H O H
H CO
H CO H O H O CO CO
CO H p H O H
H CO
H H H O H O CO CO
P P K P P P
r k K K
K P K P K P K P
P P K P P P
r k K K
K P K P K P K P


   


   

Skrzypek
(1991)
29 | P a g e

Cu-ZnO-
Al
2
O
3

180-280 15-51
2 2 2 2 2
3
2 2 2 2 2 2
22
2 2 2 2
2 2 2 2 2 2
22
' ' * 3
5 2 3 4
3
34
'
13
34
[1 (1/ )( / )]
(1 ( / )( / ) )
[1 ( )( / )]
(1 ( / )( / ) )
CH OH
H O H O
H O H O
a CO H H O H CO
MeOH
H H H H H O H O
CO CO H O H CO
RWGS
H H H H H O H O
k K K K P P K P P P P
r
K K K K P P K P K P
k P K P P P P
r
K K K K P P K P K P


  


  

VandenBussche
(1996)
Cu/ZnO/ZrO
2
/
Al
2
O
3
/SiO
2

200-275
r
M
=k
M
[P(CO
2
)P(H
2
)-P(CH
3
OH)P(H
2
O)/[K
M
P
2
(H
2
)]}/A
2

r
R
= k
R
{P(CO
2
)-P(CO)P(H
2
O)/[K
R
P(H
2
)]}/A
A= 1+K H2P H21/2+K CO2P H2O+K H2CO2P H2(1/2)P CO2+K H2OHP H2O/P H2+K H2OP H2O
Kubota T.
(2001)
Cu-based
Zn-Croxide
240 52
2
2
22
2 2 2
3 3
()
2 2 1
(1 )
1 / ( )
m H O
H
p m H CO
H O H O CO
PP
kP
K P P
r
K P K P K P





Rozovskii
(2003)
Cu/ZnO/ZrO
2
/
Al
2
O
3

250 50 The set of reaction rates in table 3.2
Lim et al.
(2009)
Cu/Zn/Al/Zr
Fibrous
230-260 25-48
2 3 2
2 2 2 2 2 2
2 2 2 2
2 2 2 2 2 2
2 2 3 2 2
3/2 1/2
11
1 1/2 1/2
22
2 1/2 1/2
3/2 3/2
2
3
[ / ( )]
(1 )[ ( / ) ]
[ / )]
(1 )[ ( / ) ]
3 [ / (
CO CO H CH OH H f
CO CO CO CO H H O H H O
CO CO H H O CO f
CO CO CO CO H H O H H O
CO CO H CH OH H O H f
k K f f f f K
r
K f K f f K K f
k K f f f f K
r
K f K f f K K f
k K f f f f f K
r


  


  


2 2 2 2 2 2
3
1/2 1/2
)]
(1 )[ ( / ) ]
CO CO CO CO H H O H H O
K f K f f K K f   

Xin et al.
(2009)

30 | P a g e

2.2. Experiments
2.2.1. Kinetics Experiments: Experimental Set-up and Procedure
A high-pressure autoclave-type flow reactor is employed in the experiments. The reactor
is heated by a ceramic oven and its temperature is controlled by a temperature controller
(OMEGA CN9000A). The temperature of the catalyst is measured by a thermocouple inserted in
the catalyst bed. The flow rates of the reactants are controlled by mass flow controllers (Brooks
5851E). These MFC have been calibrated by means of a bubble flow meter (Restek 20136). The
pressure in the reactor is controlled by a back-pressure regulator installed in the main product
stream (Meriam Instrument 3900GI0000). All tubes are heated by a heating tape (Amptek AWH-
051) to approximately 170 (
o
C) in order to prevent condensation of products. A slurry of dry ice
in acetone is used as a cold-bath (−78 °C) in order to remove the MeOH and water in the product
stream. The removed liquid products are then analyzed by a Gas Chromatograph (GC) with a
flame ionization detector (FID) (DB1 Agilent). The exit gas stream from the reactor, consisting
of unreacted H
2
, CO, and CO
2
, is analyzed by a Gas Chromatograph with a thermal conductivity
detector (TCD) (Carboxe 1000). Figure 2.4 shows a schematic of the experimental set-up.
31 | P a g e

Figure 2.4. Schematic of the set-up used for the kinetics experiments
2.2.2. Catalyst Preparation
As it was mentioned in Chapter 1, the MK-121 catalyst from Haldor-Topsøe is a
Cu/Zn/Al catalyst used for the synthesis of methanol product from hydrogen and carbon oxides
whose properties are shown in Table 2.4.

Table 2.4. MK-121 catalyst properties
Size , mm Cylinders with domed ends 6 × 4
Chemical Composition
Cu %
Zn%
Al%
Graphite, Oxygen in metallic oxides,
Carbonates, moisture, %

>43
20±3
5±1

Balance
Axial crush strength, kg/cm2
Expected filling density, kg/l
>220
1
To activate the catalyst prior to the reaction, the catalyst pellets were crushed and the
(650 µm-850 µm) portion was separated. 15 gr of the crushed catalyst was the mixed with the
same size and same amount of crushed quartz, and after loading into the reactor, the mixture was
32 | P a g e

heated-up in flowing N
2
under 18 bar pressure with a heating rate of 50
◦
C/hr. After a
temperature of 180
◦
C was reached, H
2
was introduced into the flowing N
2
stream, initially at a
concentration of 12 vol.% (at the same total flow rate as before, meaning the total flow rate of H
2

and N
2
is equal to the N
2
flow rate before introducing H
2
), which was increased gradually to 18
vol.% over a period of 8-12 hr. During this time an increase in the bed temperature was
observed. This increase was first observed as a jump in temperature when H
2
was first introduced
to the system (nd 10
◦
C), and then the temperature was fallen down after 5-10 (min), but not to
the initial bed temperature, and this case was repeated after every increase in H
2
volume
percentage. After this period, if the bed temperature was not reached to 230 (
◦
C) (according to
the manufacturer recommendation) the bed temperature was increased toward 230 (
◦
C) with the
rate of 10 (
◦
C/hr). After the temperature 230 (
◦
C) was reached, the H
2
volume percentage was
increased gradually to 22 volume percentage during 12 hours (at the same total flow rate as
before), and remained at 22 volume percentage for 2 (hr). After 2 (hr) the activation process was
finished. It is worth of mentioning that due to our limitation in the membrane reactor experiment,
this activation procedure was used with a small alteration. The temperature was increased until
200 (
◦
C)-instead of 230 (
◦
C) - was reached, and the hydrogen volume percentage was increased
gradually to 22 volume percentage during 18 hours (at the same total flow rate as before), and
remained at 22 volume percentage for 3 (hr). In order to keep the catalyst’s activity, it should
never get exposed to hydrocarbon at temperature below 190 (
◦
C), and while no synthesis gas is
flowing through the reactor, catalyst is kept pressurized using ultra high pure Nitrogen. The
method of catalyst activation is slightly different than what was recommended by the
manufacturer. The initial H2 percentage recommended by the manufacturer is as low as 2volume
%, but due to our MFC lowest flow rate limitations, we started from 12 vol.%.
33 | P a g e

2.2.3. Kinetic Experiments: Parameter Estimation
The Response Surface Methodology (RSM) approach is used in this research in order to
select the right number and type of kinetics experiments to be performed. Five experimental
parameters are considered (Table 2.5), namely the Carbon Factor (CF- by which the effect of
amount of CO in the feed to the reactor was investigated), the Stoichiometric Number (SN- by
which the effect of amount of Hydrogen in the feed to the reactor was investigated), Pressure (P),
Catalyst weight/Inlet flow-rate (W/F- by which the effect of inlet flow rate to the reactor was
investigated), and Temperature (T) as independent variables. For four of these parameters (CF,
SN, W/F, P) we considered two different values (levels), while we studied three different values
of temperatures (220
o
C, 230
o
C, and 245
o
C). Forty eight experiments were carried out as shown
in Table 2.6, which includes the values of the independent variables utilized and the measured
experimental conversion. We have used the Minitab 16 Statistical Software to carry out the
design of experiments (DOE) and for analyzing the results for DOE studies. Factorial design
allows for independent evaluation of the impacts of varying the experimental parameters and to
figure out which factor and which parameter has the most effect on the measured dependent
variable, in our case carbon conversion to methanol.

Table 2.5. Parameters and their values considered in the design of the kinetic experiments
Parameter Level 1 Level 2 Level 3
Carbon Factor (CF) =CO/(CO+CO
2
) 0.625 0.4 -
Stoichiometric Number (SN)=(H
2
-CO
2
)/(CO+CO
2
) 2 3 -
Pressure (bar) 20 30 -
Catalyst weight/Inlet flow-rate (W/F) 10 20 -
Temperature (
◦
C) 220 230 245

34 | P a g e

Table 2.6. Experimental runs using factorial design and their results
Run # T P W/F SN CF Experimental Conversion
1 245 30 10 3 0.4 35.7±2.3
2 245 30 20 3 0.6 33.2±1.9
3 230 30 20 3 0.6 34.2±2.3
4 245 20 10 3 0.6 29.2±1.7
5 220 20 10 2 0.4 39.1±2.96
6 220 20 10 2 0.6 40.1±1.65
7 230 20 10 2 0.4 48.5±2.96
8 220 30 10 3 0.6 38.5±1.32
9 230 20 20 2 0.4 29.2±3
10 220 30 20 3 0.4 29.7±1.5
11 245 20 10 2 0.4 28.4±2.95
12 245 20 20 2 0.4 40.7±4
13 220 30 10 2 0.6 33.2±3
14 230 20 10 2 0.6 35.5±4
15 220 20 20 2 0.4 38.9±3.96
16 230 30 20 2 0.4 30.9±1.65
17 245 30 10 2 0.6 23.7±2.31
18 220 30 20 2 0.4 26.7±3.13
19 220 20 20 3 0.6 14.6±1.64
20 230 20 10 3 0.6 16.1±1.16
21 245 20 20 3 0.4 27.7±3
22 220 20 10 3 0.4 16.1±4.18
23 230 20 10 3 0.4 11.1±1.95
24 245 30 10 2 0.4 19.9±2.97
25 230 30 10 2 0.4 33.5±1.32
26 230 30 10 2 0.6 28.6±1.32
27 245 30 20 3 0.4 35.7±3.63
28 245 20 20 3 0.6 44.8±3.59
29 220 20 10 3 0.6 38.2±3.96
30 230 30 20 2 0.6 29.9±1.78
31 245 30 10 3 0.6 39.3±3.96
32 220 20 20 3 0.4 39.2±4.16
33 220 20 20 2 0.6 31.9±2.164
34 220 30 10 3 0.4 31±2.97
35 230 30 10 3 0.4 40.1±2.95
36 245 30 20 2 0.6 23.5±2.97
37 220 30 20 3 0.6 27.3±1.98
38 245 30 20 2 0.4 14.2±1.198
39 220 30 10 2 0.4 27.1±1.9
40 245 20 20 2 0.6 36.2±1.32
41 230 20 20 3 0.4 6.03±1.3
42 245 20 10 3 0.4 18.3±0.66
43 220 30 20 2 0.6 12.4±2.97
44 230 20 20 2 0.6 15.9±1.98
45 230 30 10 3 0.6 27.2±2
46 230 30 20 3 0.4 23.6±2.3
47 230 20 20 3 0.6 22.4±1.96
48 245 20 10 2 0.6 18.3±1.16
35 | P a g e

Using the Minitab 16 software has provided guidance about the order via which to carry-
out the experiments, as indicated in Table 2.6. Further analysis indicates that the selected
independent variables (T, P, CF, SN and W/F) have all a significant and direct impact on the
measured reactor conversion.
For the analysis of the experimental data we utilize the rate expressions of Vanden
Bussche et al. [7 ], Eqns. (2-15) – (2-18). In which:

2
22
2
''
1 5 2 3 4 3
2 3 4 4
'
51
/
HO
aH
H H O
k k K K K k K
k K K K K k K
kk




The k values are in the form of the Arrhenius equation,

(

) .
For fitting the experimental data, we assume the reactor to be plug-flow and methanol
synthesis is described by two independent reactions: the reaction between CO
2
and hydrogen to
produce methanol and the shift reaction. The reactor is then described by the following
Equations (2-25) to (2-29):

In the above equations F
i
is the molar flow rate (mol/hr) of each component, and W (kg) is the
weight of catalyst variable. The carbon conversion is defined by the following equation (which
assumes no methanol in the feed as is the case with the experiments reported here):
36 | P a g e

x
carbon
=

=

The experimental conversion data have been used to fit the values of the rate parameters
(k
1
, k
2
, k
3
, k
4
, and k
5
) in the reaction rate expressions (2-24) – (2-28) above. For the parameter
fitting we have utilized three in-built MATLAB function: Genetic Algorithm (GA), “nlinfit”, and
“nlparci”. The GA algorithm was used to give us initial guess for “nlinfit”. The standard
optimization algorithms generate a single point at each iteration. The sequence of points
approaches an optimal solution whereas GA generates a population of points at each iteration,
and that population approaches the optimum solution. At each step then GA selects individuals
from this current population randomly to approach the optimum in the next step. The results
from GA, then were used in “nlinfit” in order to estimate the reaction rate coefficients (k
1
, k
2
, k
3
,
k
4
, and k
5
). The confidence limit of the parameters then were estimated from “nlparci”. Table 2.7
shows the values of the estimated parameters together with their 95% confidence limits. The
uncertainties in the values of the parameters are generally satisfactory (typically less than ±10%)
other that the activation energy for the rate constant k
1
for which the uncertainty is ~ ±17%.
Table 2.7. Parameter values
Parameter A B
k
1
0.626 (±4.040e-2) 76742.05(±13241.43)
k
2
3095.76(±143.39) 0. 0005(±2.14e-9)
k
3
1. 285(±0.132) 2638.2(±179.3)
k
4
9.47e-9(±6.36e-10) 7931.6(±405.8)
k
5
8.38e8(±3.56e7) -8.43e4(±5.59e3)

Figure 2.5 shows the agreement between the calculated conversion values using the
parameter values from Table 2.7 and the corresponding experimental conversions, indicating
once more the satisfactory goodness of fit.
37 | P a g e

Figure 2.5. Calculated conversion from model and experimental conversions

2.3. Conclusions
The kinetics of the methanol synthesis reaction over a commercial catalyst have been
investigated. The rate expression of Van Den Bussche et al. [7] were used to fit the experimental
data. The values of the rate constants were determined and the rate equations of Van Den
Bussche et al. [7] were shown to provide an adequate fit. These rate reactions are used in the
modeling of the behavior of the membrane reactor in Chapter 4.

38 | P a g e

References
1. G. Natta, in “Catalysis” (P. H. Emmett, Ed.), Vol. 3, p. 349. Reinhold, New York, (1955).
2. V. E. Leonov, M. M. Karavaev, E. N. Tsybina, G. S. Petrischeva, Kinetika i Kataliz, 14, 848
(1973).
3. K. Klier, V. Chatikavanij, R. G. Herman, G. W. Simmons, Journal of Catalysis, 74, 343
(1982).
4. P. Villa, P. Forzatti, G. Buzzi-Ferraris, G. Garone, I. Pasquon, Industrial & Engineering
Chemistry Process Design and Development, 24, 12 (1985).
5. G. H. Graaf, E. J. Stamhuis, A.A. C. M. Beenackers, Chemical Engineering Science, 43(12),
3185 (1988).
6. J. Skrzypek, M. Lachowska, H. Moroz, Chemical Engineering Science, 46(11), 2809-
2813(1991).
7. K. M. Vanden Bussche, G. F. Froment, Journal of Catalysis, 161,1-10 (1996).
8. G. Soave, Chem. Eng. Sci. 27, 1197 (1972).
9. R. A. Hadden, H. D. Vandervell, K. C. Waugh, G. Webb, Catalysis Letters, 1, 27 (1988).
10. H-W. Lim, M-J. Park, S-H. Kang, H-J. Chae, J. W. Bae, K-W. Jun, Industrial and
Engineering Chemistry Research, 48, 10448–10455(2009)
11. A.N. Xin, Catalysis, Kinetics and Reactors, Chinese Journal of Chemical Engineering, 17(1)
88-94 (2009).
12. A. Rozovskii, R. Ya, Chem. Rev. 58, 41 (1989).
13. S. G. Neophytides, A. J. Marchi, G. F. Froment, Applied Catalysis, 86, 45 (1992).
14. S. G. Neophytides, G. F. Froment, Industrial and Engineering Chemistry Research, 31, 1583
(1992).
39 | P a g e

15. J. Nakamura, J. M. Canmpbell, C. T. Campbell, Journal of the Chemical Society, Faraday
Transactions, 86, 2725 (1990).
16. S. Fujita, M. Usui, N. Takezawa, J. Catal. 134, 220 (1992).
17. K. H. Ernst, C. T. Campbell, G. Moretti, , J. Catal. 134, 66 (1992).
18. G. F. Froment, K. B. Bischoff, “Chemical Reactor Analysis and Design,” 2nd ed.,Wiley,
New York, (1990).
19. T. Kubota, I. Hayakawa, H. Mabuse, K. Mori, K. Ushikoshi, T. Watanabe , M. Saito,
Applied Organometallic Chemistry,15,121-126 (2001)
20. A. Y. Rozovskii ,G. I. Lin, Topics in Catalysis, 22(3–4), 155-160 (2003)
21. M. Shahrokhi, G.R. Baghmisheh, Chemical Engineering Science, 60, 4275 – 4286 (2005)

40 | P a g e

Chapter 3. Membrane Reactor Experiments
1

3.1. Introduction – Methanol Synthesis in a Membrane Reactor
3.1.1. The Membrane Reactor Concept
Process integration resulting from combining reaction and membrane separation in a
single unit promises numerous benefits compared to the more conventional processes [1-3]. In
recent years a multitude of concepts have been proposed [4-6] of how membranes can be applied
in combination with a chemical [4,5] or a biochemical reaction [7-10] in order to intensify the
process as a whole. In this Chapter, due to space limitations, the emphasis will be primarily on
conventional reaction systems; those interested in the application of the membrane reactor
concept to biochemical reaction systems are directed to a number of comprehensive review
articles [11,12]. One common feature of all membrane reactors (that differentiates them from
other reactive separation processes like reactive distillation [13] or reactive adsorption [14], etc.)
is their use of membranes [15]. These come in a variety of shapes and forms (flat, cylindrical,
hollow-fiber, spiral-wound [16-19]) and they divide the reactor volume into two chambers,
which are often referred to as the feed (or reject side) and the permeate side. In conventional
separation applications, membranes are called upon to permeate selectively one or more of the
components of the fluid mixture in contact with them, and in some instances to be totally
impermeable to one or more of the mixture components (this separation mechanism is known as
molecular sieving [20-24]). Membranes in membrane reactors do often separate preferentially
products and/or reactants from the reacting mixture in situ [25-27]; there are instances, however,
where the membrane is totally unselective and only functions as a contactor interface between

1 The discussion presented in the Introduction of this chapter has already been published as Soltani S.,
Sahimi M., Tsotsis T.T., Catalytic Membrane Reactors: A Brief Overview. Encyclopedia of Membrane
Science and Technology, Wiley-VCH, 2013.

41 | P a g e

two reactive streams [28-30], see further discussion to follow (in this case the classification of
the two membrane reactor volumes as feed and reject sides no longer applies). A great many
different types of membranes have found use in membrane reactors, including porous or dense
(e.g., Pd [31-35] or perovskite [36-39]), and made from a variety of materials (organic [40],
polymeric [41], inorganic [42,43], and metal [44]). The need for high throughput dictates often
the use of an asymmetric structure [45-47] with a thin separation layer on the top of a
mechanically strong support. For the cases where the membrane acts as a contactor interface
[48,49] the use of homogenous symmetric membranes may be preferable [50-52]. Transport
through the membrane is generally via convection and diffusion, driven by either the presence of
an overall pressure or a chemical potential gradient (to enhance the chemical potential gradient in
membrane reactors, particularly of the extractor-type – see discussion to follow – a sweep
stream, commonly low-pressure steam, is utilized). Membranes (particularly the inorganic ones)
can carry a positive or negative charge (or can be bipolar), so transport through the membrane
can be affected by an electric field [53], and since catalytic reactions take place at higher
temperatures thermal gradients may play a role as well [54,55].
There are different ways to classify the various types of membrane reactors [56]. One is
based on whether the membrane is catalytic or non-catalytic, selective or non-selective, and
whether or not a packed-bed or a fluidized-bed of catalysts is being used. Six different membrane
reactor configurations, based on such a classification, have been identified and are also shown in
Table 3.1. In the so-called CMR configuration, the membrane is catalytic and provides the sole
catalytic function in the reactor; it may or may not be permselective to any of the reactants
and/or products/intermediates (when the membrane is not permselective, the reactors are also
known as contactor-type membrane reactors, see further discussion to follow), however. When
42 | P a g e

the membrane is non-permselective and has no catalytic function either, this is known as the
CNMR configuration. Here, the membrane’s function is to distribute one (or more) of the
reactant(s) in a controllable way, and the catalytic function is provided by either a packed-bed or
a fluidized-bed of catalysts (taken, in its more “liberal” definition, to also mean a homogeneous
catalyst that the membrane helps to confine in that side [56]) on the other side of the membrane
(these are also known as distributor-type reactors, see further discussion below). The PBMR
configuration is the most commonly used one; here, the reaction function is provided by a
catalyst packed-bed while the membrane, which is permselective but not catalytic, provides only
the separation function (e.g., removing one of the products in order to improve conversion of a
reversible reaction via the Le Chatellier principle). If a fluidized-bed of catalysts is used, instead,
this is known as the FBMR configuration. When the membrane reactor employs a packed-bed
and a permselective membrane which is also catalytically active [56] this is known as the
PBCMR configuration (FBCMR, if a fluidized-bed is used, instead, for better control of the
process temperature or to improve the external mass transport characteristics).

Table 3.1. Membrane reactor configurations [56]
Acronym Description
CMR Catalytic membrane reactor
CNMR Catalytic non-permselective membrane reactor
PBMR Packed-bed membrane reactor
PBCMR Packed-bed catalytic membrane reactor
FBMR Fluidized-bed membrane reactor
FMCMR Fluidized-bed catalytic membrane reactor
Membrane reactors can also be categorized into three different types, depending on the
role the membrane is called upon to play during the process [57]: They are the extractor-type
[58,59], distributor-type [56,60], and contactor-type membrane reactors [28,30,48,49,61]. 1n the
43 | P a g e

extractor-type [58,59] membrane reactors, the membrane is there to selectively remove products
and/or intermediates from the reaction mixture, and in so doing to positively impact the
selectivity and/or yield of the reactor. In distributor-type membrane reactors, the membrane
provides an effective means to deliver, in a controllable fashion, one of the reactants [56,60], the
net result again being an increase in selectivity and/or yield via the avoidance of undesirable side
reactions [11,25,62,63]. In contactor-type membrane reactors, the membrane is itself catalytic
[64-67] and provides the interface where two (or more) reactants come together to react [68].
And though the membrane transport characteristics are important in determining the position of
the reactive interface [69], the membrane is not required to be permselective towards the
reactants and/or products. In the other two types of membrane reactors the membrane may be
reactive (sometimes not intentionally so), but generally it is not, as the reaction typically takes
place either in a packed catalyst bed [64,70] or with the aid of a homogeneous catalyst (that the
membrane helps to localize) in the membrane reactor feed side [56]. The latter is a particularly
promising area for membrane reactors [56]. Homogeneous catalysts are typically expensive and
are used in low concentrations. In conventional reactors catalyst recovery for reuse presents
challenges [71]. Catalyst immobilization is an alternative (e.g., in enzyme catalysis [11]) but is
often difficult and usually reduces the catalytic activity. The use of nanofiltration (NF)
membranes in membrane reactors to retain the catalyst [56] can be a promising alternative, when
chemically stable and highly retentive membranes can be identified. The use of membrane
reactors shows particular promise in chiral catalysis, where expensive metal complexes are
typically used as homogeneous catalysts with high reactivity [63]. In what follows, we briefly
describe the key features of these three different types of reactors.
44 | P a g e

3.1.2. Different Types of Membrane Reactor
3.1.2.1. Extractor-Type: Fischer Tropsch and Alcohol Synthesis
Extractor-type membrane reactors find the most common use [56,67]. The function of the
membrane in these reactors is, typically, to remove one or more of the reaction products [58]
(e.g., hydrogen in dehydrogenation reactions [72], water in esterification reactions [73]), and for
reactions which are thermodynamically-limited via the Le Chatelier’s principle to increase the
reactor conversion [74]. For highly selective membranes (e.g., Pd used in hydrogen production
[75]) the membrane serves the additional important role of in situ purifying the product (e.g.,
producing ultra-pure hydrogen for fuel cell applications), thus eliminating the need for additional
downstream separations. Other applications have also been discussed, for example, removing the
intermediate desired product in a consecutive-type reaction [76], where by doing so one
improves the reaction selectivity; or removing an intermediate and or product that may be
poisoning the catalyst (e.g., water for the Fischer-Tropsch reaction [77-80] or acids and/or esters
in biochemical reactions [81]).
In the technical literature, so far, the most frequent applications of extractor membrane
reactors are catalytic dehydrogenations of light alkanes or reactions for hydrogen generation,
such as the steam reforming or the water gas shift reaction (WGS), making use of H
2
-
permselective membranes, most commonly Pd or Pd-alloy type membranes [32-35,82,83], but
also other molecular sieve type membranes (e.g., carbon membranes [20-24]). Other common
reactions include esterifications [73,84,85], alcohol [86,87] and Fischer Tropsch synthesis [77-
80] all making use of water-permeable membranes (e.g., zeolite [88-91] or polymeric
pervaporation-type membranes [84,92,93]). In the following we will discuss membrane reactor
applications in Fischer Tropsch and Alcohol Synthesis briefly.
45 | P a g e

The high oil prices in recent years has spurred interest in the so-called gas-to-liquid
(GTL) processes [194,195] that seek to convert syngas produced from natural gas (particularly,
the so-called stranded gas in remote areas) into either alcohols or into transportation fuels (e.g.,
gasoline, diesel, jet fuel, etc.) via the Fischer–Tropsch (FT) synthesis process. Various types of
conventional reactors (including fixed-bed, fluidized bed, and slurry bubble column) have been
utilized making use mostly of either Co or Fe-based catalysts [196].
Multifunctional reactors have attracted through the years significant attention in the
alcohol and Fischer-Tropsch (FT) synthesis area. For example, a reactor for MeOH synthesis
from syngas, based on in situ product separation via condensation, was studied by Halloin and
coworkers [197-199]. High conversions were attained, but the space velocities were orders of
magnitude lower than in conventional designs, and scaling-up such a reactor is likely to be
challenging. Westerterp and coworkers [200-202] investigated a gas solid-solid trickle-flow
adsorptive reactor, in which a fine-particle solid adsorbent is trickled down, countercurrently to
the syngas, through a fixed catalyst bed. The adsorbent selectively removed the MeOH as it
formed, and 100% conversion at high production rates was achieved. Circulation of solids
encounters many problems on an industrial scale, however, especially for high-pressure
operations, such as attrition, the need for using a cumbersome solids-handling system, and
difficulty in process scaling-up. Kruglov [203] proposed, in a modeling study, to use instead for
MeOH synthesis a simulated countercurrent moving-bed chromatographic reactor (SCCMBR),
where countercurrent moving-bed operation is mimicked by periodically changing feed and
product locations sequentially along a fixed bed. The analysis indicates very high CO
conversions (96 - 99%) attainable in a single-pass operation. No experimental data were
presented, however, and scale-up of SCCMBR for large-volume, high pressure and temperature
46 | P a g e

applications will be technically challenging. A MeOH synthesis process making use of a liquid
absorbent was also proposed by Westerterp et al. [204]. In their studies they utilized a high-
pressure laboratory unit consisting of two packed-bed reactors (PBR) in series with a high-
temperature interstage, countercurrently-operated packed-bed absorber with tetraethylene glycol
dimethyl ether (TGDE) as the solvent for product removal. Process simulations were reported
with a four-reactor/absorber/flash system which was claimed to be economically competitive
with the conventional reactor.
Membrane reactors have also been suggested previously in the literature for MeOH
synthesis and for the Fischer-Tropsch reaction for overcoming both low conversion and thermal
management hurdles [205]. All three different types of membrane reactors (extractor, distributor
and contactor) types have been utilized. Basile and coworkers [206] used, for example, a
commercial Cu-Zn catalyst and a mesoporous zeolite-A membrane to synthesize MeOH in a
PBMR. The PBMR showed higher selectivity than the conventional PBR for the same level of
conversion; this was attributed to the removal by the membrane of the products (H
2
O and
CH
3
OH), which shifts the equilibrium to the right. Earlier, Barbieri et al. [87] had performed
simulations to show that removing H
2
O and CH
3
OH selectively results in improved yield and
selectivity. A Li-doped Nafion® tubular membrane packed with a Cu-Zn catalyst was tested by
Struis et al. [207] for the same reaction. At low gas hourly space velocities (GHSV), the PBMR
gave higher conversions than a conventional reactor of the same dimensions (4.2% vs. 2.5%).
The experiments were later validated by model simulations [208] using independently measured
reaction kinetics. A similar study was published later by Chen and Yuan [209], who used instead
a silicon rubber composite membrane supported on a macroporous alumina tube. Again modest
improvements in conversion (~20%) were observed for the PBMR over the conventional reactor,
47 | P a g e

but the experimental conversions were rather low (<10%). The long-term stability of these
polymeric membranes at the typical alcohol synthesis conditions (P>50 bar) is also a concern.
Water removal through membranes during FT synthesis over Co-based catalysts was first
proposed by Espinoza et al. [205] (but was not experimentally verified), the motivation coming
from prior studies showing significant catalyst deactivation due to reoxidation at high H
2
O/H
2

ratios, and H
2
O pressures. Fe-based FT catalysts also deactivate due to oxidation by CO
2
/H
2
O,
but additionally, H
2
O vapor strongly inhibits their reaction rate. Thus, the selective removal of
H
2
O during the FT synthesis could potentially be beneficial for both catalyst systems [210,211].
Fe-based catalysts are active for the water gas shift reaction (WGS) and Rhode et al. [211] have
shown that the selective removal of H
2
O during the FT reaction displaces the CO/CO
2
WGS
equilibrium in favor of CO, and enhances the hydrocarbon yield and the conversion of CO
2
to
long-chain hydrocarbons. Rhode et al. [210,211] studied this concept experimentally using SiO
2

membranes and a Fe/Cu/K/Al
2
O
3
catalyst. Under optimized conditions somewhat improved
conversions over the conventional reactor were observed in the PBMR.
3.1.2.3. Contactor-Type
In this configuration, as noted previously, the membranes is not permselective but simply
acts as an interface and a means to bring reactants effectively into contact [48, 29, 212]. In most
cases the membrane itself is catalytic, but in other cases it is not (e.g., CO
2
scrubbing from flue-
gas or raw natural gas using amine solvents in hollow-fiber membrane contactors [213]). The
latter membrane reactor configuration has been studied extensively [214] as an alternative way
of contacting gas and liquids, its main advantages over conventional packed (random or
structured) columns being (i) that the membrane provides a well-defined gas–liquid interfacial
area segregating the two phases and leading into the prevention of flooding and foaming of the
liquid (solvent); (ii) that the flow rate of the two streams can be independently controlled without
48 | P a g e

interference from each other; (iii) when using hollow fiber membranes high interfacial areas can
be achieved. Several studies [215-217] report enhanced overall mass transfer characteristics,
leading to a significant reduction of required unit size in comparison with traditional packed
columns.
Contactor-type membrane reactors have also been used to carry-out multi-phase catalytic
reactions, including the selective oxidation of light alkanes under mild conditions by means of
super acidic catalytic membranes and liquid phase hydrogenations [30,218-220], where
providing effective contact between the gaseous and liquid reactant is important, as hydrogen is
poorly soluble in organic liquids. The same reactors have also been used to perform phase-
transfer catalytic oxidations. The selective oxidation of hydrocarbons by phase-transfer catalysis
is commonly performed in stirred tank reactors with subsequent product separation [221,222].
The oxidant is located in the aqueous phase and the hydrocarbons form the organic phase. For
optimal performance, it is important to provide a large interfacial area between the two phases
for intensive phase contact, but to do so while avoiding substantial emulsification of the one
phase into the other, as this complicates phase separation upon completion of the reaction. The
use of a porous membrane attains both objectives in providing high interfacial area while
avoiding emulsification problems, and thus offers a convenient means to separate the two phases
[223-226].
An interesting application of the contactor-type membrane reactor concept is for cases
where control of the reactive interface is of critical importance. Here, two gaseous reactants are
separated by a non-selective porous catalytic membrane, forming a reaction front inside the
membrane. The exact position of the reaction zone is determined by the balance of the transport
rates of the two reactants within the membrane pores and, of course, the reaction rate itself. As
49 | P a g e

the transport rates vary (e.g., due to the fluctuating reactant concentrations) the position of the
reaction front adjusts accordingly. The ability to do so provides substantial process robustness.
Examples of the application of this concept include catalytic combustion for energy/power
generation [227], where the position of the reaction front can be conveniently adjusted by
controlling the pressure of the oxidant stream (this then prevents thermal runaway). The same
reactor type has also been used for the complete conversion of pollutants, for example, the
selective catalytic reduction of nitric oxides (NO
x
) with ammonia to form nitrogen and water
[228]. Again the ability to localize the reaction zone within the contactor membrane prevents
“slippage” of both the NOx and the ammonia.
A contactor-type MR operating in a “flow-through” mode (this type of reactor is known
as FTCMR) was studied for the FT reaction by Bradford et al. [65] who utilized a commercial,
tubular, alumina membrane coated (through slip-casting) with a powdered P/Pt-Co/γ-Al
2
O
3

catalyst. They report that the MR shows a higher C
2
+ hydrocarbon yield (albeit with a slightly
smaller selectivity), and a lower olefin/paraffin ratio than the PBR. They attribute the differences
to the increased H
2
/CO ratio within the catalytic membrane. A similar concept was also tested by
Khassin and coworkers [229]. They prepared their tubular catalytic membranes by mixing
powders of Co-Al co-precipitated catalyst, a pore-producing agent, and a dendritic metallic
copper powder as a reinforcing heat-conductive agent. Their MR operates as a radial-flow
reactor, and shows good FT activity with high selectivity towards C
5
+ hydrocarbons. The space-
time-yield, based on the exterior volume of the membrane, achieved with the FTCMR (200 kg of
liquid hydrocarbons/m
3
h) was higher than in traditional reactor designs for FT. The observed
catalytic activity was also up to three times higher than those in slurry reactor.
50 | P a g e

3.1.2.3. Distributor-Type
Distributor-type of membrane reactors have also received quite a bit of attention,
especially in recent years. In these reactors, as it was noted in several previous examples, one or
more of the reactants are added (distributed) through the membrane in a controlled fashion
(Figure 3.1). Both non-permselective (e.g., mesoporous or macroporous inorganic) and highly
permselective (oxygen and hydrogen permselective perovskite-type) membranes have been used
in these reactors. When using oxygen-permselective membranes in distributor-type membrane
reactors they perform a dual role, as a means to deliver the oxygen in a controlled fashion but
also allowing one to use air rather than pure oxygen, thus avoiding the need for a separate oxygen
plant [67].
Gas-phase partial oxidation of hydrocarbons is the most common reaction system the
distributor-type reactors are applied to [230] to. There are two key drivers motivating the use of
membrane reactors for such an application: (1) typically, the desired product is an intermediate
(e.g., ethylene oxide during ethylene partial oxidation [231]) which can easily be reacted further
to the undesirable total oxidation products (CO
2
and H
2
O); “starving” the reaction environment of
oxygen, via its controlled addition through the membrane, helps to minimize the undesired
reactions [230]; (2) a key concerns of running these partial reactions is the risk of entering the
homogeneous explosion limits [232,233], which dictates carrying them out in conventional
reactors at high dilutions of the hydrocarbon [232] which is uneconomical; membrane reactors,
particularly those using solid-oxide membranes, which are only permeable to oxygen, are
inherently safer to operate. The two key reactions that have been studied here (see further
discussion below) are motivated by the strong economic interest in developing processes that
upgrade methane (particularly the so-called “stranded quantities”) to higher valued products, and
include the oxidative coupling of methane (OCM) to produce C
2
and higher hydrocarbons
51 | P a g e

[234,235], and the partial oxidation of methane into syngas [232,233,236-238], the latter reaction
being studied by a large consortium in the USA led by Air Products [239] and thought to be
reaching its pre-commercial field-testing. Other types of reactions investigated include oxidative
dehydrogenations [240-243] where either pure oxygen (or air) is distributed through the
membrane to the catalytic side where the hydrocarbon is fed and a dehydrogenation catalyst is
present. Adding the oxidant is thought to provide two advantages [180,244,245]: (1) providing
the heat for the endothermic reaction; and (2) minimizing catalyst deactivation through coking.
Another important class of reactions, to which distributed-type reactors are applied, are
hydrogenation reactions [246,247]. Using the membrane as a coupling means between a
hydrogenation and dehydrogenation [248,249] reaction has been proposed (here, the
dehydrogenation reaction provides the reactant hydrogen for the hydrogenation reaction, while
the hydrogenation reaction provides the heat for the dehydrogenation), as is the coupling between
the endothermic steam reforming of methane with an exothermic partial oxidation.

Figure 3.1. Distributor-type membrane reactor concept [67]
Distributor-type membrane reactors for the FT and alcohol synthesis reaction have also
been studied [250-252]. In their FT synthesis studies of Vanhove and coworkers [250-252], for
example, one of the reactants is fed through the permeate side of a tubular membrane, while the
other is fed through the catalyst bed located inside the membrane. The membrane utilized was
52 | P a g e

either inert Al
2
O
3
or Al
2
O
3
coated with a ZSM-5 zeolite layer. The results obtained in the MR
were compared to the PBR experiments operated under similar conditions, and were in
accordance with reaction kinetics for the Co-based catalysts, and the H
2
/CO ratio effect on
product selectivity [211]. With H
2
as the feed gas and CO as the distributed reactant, the H
2
/CO-
ratio stays high along the reactor length, resulting in diminished inhibition of the FT reaction and
increased conversions, when compared to the PBR. However, the formation of long-chain
hydrocarbons is not favored, and as a result C
1
–C
4
hydrocarbons were mainly produced. Using
CO as the feed gas resulted in lower conversions, as the inhibition by CO is high. Due to the low
H
2
/CO-ratio in the catalyst bed, however, an increase in the C10+ hydrocarbon yield and lower
CH
4
selectivity was observed. The ZSM-5 membrane showed similar results, but similarly to the
observations of He et al. [253,254] it altered the product distribution by secondary reactions at
the acidic zeolite sites, resulting in higher yields of short-chain hydrocarbons. A distributor-type
MR was also proposed for both the MeOH synthesis [e.g., 80,255,256] and the FT reaction [e.g.,
257-259] in a series of modeling investigations by Rahimpour and coworkers. They studied a
dual reactor system, in which the second stage is a MR using a Pd-Ag membrane that allows H
2

to permeate between the reactor side and the fresh syngas flowing in the permeate side that also
acts as a coolant for the second reactor. Simulations indicate that the presence of the membrane
helps maintain the optimal H
2
concentration in the second reactor, thus resulting in somewhat
improved yield and longer catalyst life. Other dual reactor systems have also been modeled by
the same Group including (i) a system that combines a FT synthesis MR with a water permeable
membrane coupled to an NH
3
decomposition MR [260,261,262] or a cyclohexane
dehydrogenation MR [263,264,265] making use of Pd/Ag membranes; (ii) a MeOH synthesis
system where a distributor-type methanol synthesis MR making use of Pd/Ag membranes is
53 | P a g e

coupled to an extractor-type cyclohexane dehydrogenation reactor [266,267,268] making use of
similar membranes. Unfortunately, none of these interesting reactor systems have received
experimental validation as yet.
3.2. Membrane Reactor Experiments
3.2.1. Methodology
Though MR have been previously utilized for methanol synthesis (and for the FT
reaction), none of the proposed concepts is particularly advantageous for methanol synthesis or
appropriate for the experimental conditions [3]. What we propose for study here, instead, is a
multi-functional methanol synthesis reactor concept, shown in Figure 3.2., which integrates
methanol synthesis with in-situ methanol separation via highly-permeable and robust inorganic
membranes with the aid of solvent purge (sweep). The membrane serves as an interface
contactor for the syngas and the solvent (e.g., various ionic liquids or more conventional high
boiling point (B.P.) solvents can be chosen) so that methanol has the highest solubility in it when

Figure 3.2. Our proposed membrane reactor concept
compared to the other methanol synthesis species. The solvent penetrates the membrane structure
and thus prevents the less soluble gas species (e.g., H
2
, CO, and CO
2
) from “bubbling” through
the membrane. A proper solvent penetration depth is critical for optimal reactor performance: too
small of a depth and the gases will escape through, and too large of a depth will negatively
54 | P a g e

impact membrane throughput. In our studies, optimal solvent penetration is attained by
appropriately manipulating the hydrophobicity of the membrane’s surface. A key focus of our
studies has been creating such hydrophobic coatings which are robust in the methanol synthesis
environment and conditions.
3.2.2. Experimental Set up
The experimental set-up is shown in Figure 3.3. The same reactor as the one which was
used for reaction kinetics experiments was used, and a ceramic membrane (Model &type-Table
3.2) was installed inside the reactor while the mixture of catalyst and quartz (the same size and
amount as in the reaction kinetics experiments) particles were placed around the membrane, so
that the reaction would take place in the shell-side. During the experiments, in the interior of the
membrane (tube-side) an inert liquid was used as a sweep stream.
Table 3.2. Ceramic membrane properties used in the membrane reactor experiments
Layer Material Thickness (µ) Average pore size ( ◦A)
Support α- Alumina 1100 2000- 4000
First layer α- Alumina 10-20 500
Second layer γ- Alumina 2-3 100
Outer Diameter(mm) 5.69
Inner Diameter (mm) 3.69
The membrane was sealed using Kalrez O-ring (DuPont K#010). Membrane porosity was
measured using two methods Helium Pycnometry and the Archimedes method, which are both
discussed in Appendix A. Back-pressure regulators were used on both the gas product and
permeate lines in order to control their pressure. The exit (reject) line from the reactor side was
heated by a heating tape. A high pressure pump (HPLC, Chrometech series1) was used to pump
the sweep liquid through the membrane. During reactor operation gas species from the reacting
mixture – products and un-reacted gases- transfer to the liquid immobilized in the membrane
55 | P a g e

pores in contact with the gas phase. Membrane pore size (for the membrane used in these
experiments) does not influence the transport, but membrane surface characteristics (e.g.,
hydrophobicity) and of course the species concentration difference in two phases play important
role.

Figure 3.3. Membrane reactor experimental set-up

Tetraethylene Glycol Dimethyl Ether (TGDE – CAS#) was used as a sweep liquid stream
in order to remove mainly the methanol and water products from the reaction zone. Literature
studies have demonstrated the suitability of the solvent selected in this work. Westerterp and co-
workers [5], for example, used TGDE as an absorbent during methanol synthesis. The solvent
showed excellent solubility toward methanol, but minimum solubility towards hydrogen and CO.
In this research, thermodynamic calculations have been carried out in order to calculate the
solubility of the various species found in methanol synthesis (H
2
, CO
2
, and CO, methanol and
H
2
O) into TGDE under a broad range of temperatures and pressures. The results of these
56 | P a g e

simulations are shown in Table 3.3, where they are also compared with experimental data from
Khosla et. al. [8], who carried out vapor liquid equilibrium (VLE) experiments of the same
synthesis gas components, methanol and water with TGDE in the temperature range 473-513 (K)
and pressure range 0.1-10 (MPa). In our simulations we use ProII (SimSciProII 9.1). A cold
stream (T=298 K) with the same composition as that used in the experiments by Khosla et. al. [8]
(indicated in the inlet composition column in Table 3.4) is fed to a flash drum, and vapor-liquid
equilibrium is established at various flash temperatures and pressures. Table 15.3 shows the
simulated as well as the experimental liquid phase composition in the flash drum (indicated
Table 3.3 as the outlet composition). As shown in Table 3.3, simulation and experimental results
are in good agreement, lending credence to our use of the model in further simulations of the MR
system (see Chapter 4), and also confirming that methanol has the highest solubility in TGDE
among all the methanol synthesis species.
3.2.3. Membrane Modification
As it was already pointed out, for the membranes utilized here it is the pore surface
characteristics rather than the pore size that determine the transfer rates through the membranes,
as long as the solvent invades the membrane pore structure and completely blocks the gas
transport. Figure 3.4 shows a schematic of a partially impregnated membrane in contact with a
gas phase during MR operation. Since the methanol (and other components) concentration in the
gas phase are typically higher than the concentrations in the liquid phase, transfer of methanol
and other species via the gas phase occupied part of the membrane to the liquid phase will occur.

57 | P a g e

Table 3.3. Solubility comparison, PROII simulations vs. experimental results
Temp.
(K)
Pressure
(bar)
TGDE
(exp,
mod)
CO (exp,
mod)
CO
2

(exp,
mod)
H
2
(exp,
mod)
H
2
O
(exp,
mod)
MeOH
(exp,
mod
473 52.72
0.7276,
0.7676
0.0141,
0.0184
0.0086,
0.0132
0.0273,
0.0333
0.0342,
0.0125
0.1848,
0.1503
493 52.72
0.7737,
0.78
0.0152,
0.0193
0.0111,
0.0127
0.0319,
0.036
0.0286,
0.0123
0.1362,
0.1349
513 52.72
0.7836,
0.7916
0.017,
0.021
0.0094,
0.0118
0.0355,
0.0401
0.0245,
0.012
0.1261,
0.124
473 78.58
0.7046,
0.7137
0.0232,
0.0272
0.0132,
0.0181
0.0409,
0.0499
0.0349,
0.0171
0.179,
0.1671
493 78.58
0.7681,
0.722
0.0203,
0.0286
0.0132,
0.0175
0.047,
0.0539
0.0273,
0.0169
0.1199,
0.1541
513 78.58
0.7284,
0.7347
0.0303,
0.0203
0.017,
0.0122
0.0583,
0.0524
0.0169,
0.0306
0.142,
0.146

In the MR experiments reported in this Chapter the liquid pressure in the interior (tube-
side) of the membrane was always kept higher than the gas phase (shell-side) pressure as to
avoid the gas from bubbling through. The hydrophobicity of the membrane material is thus a key
aspect determining membrane performance (Figure 3.4).

Figure 3.4. Gas-liquid contact in a hydrophobic membrane

58 | P a g e

Ceramic membranes exhibit excellent chemical, structural and thermal stabilities which
makes them appropriate for use in high-temperature MR applications, like the one described
here. However, most ceramic membranes available today, including the alumina membranes
used in this work, are hydrophilic in nature because of the presence of hydroxyl (OH-) groups on
the surface. Such membranes are not appropriate for use in the proposed process because the
solvent will completely wet the membrane including its support structure which results in
substantially diminished transfer rates. Another downside of using such membranes is that the
solvent may also completely wet and “pool” on the reactor-side surface of the membrane, thus
resulting in significant solvent losses via either evaporation or dispersion and carry-over of
droplets in the gas stream.
In order to make ceramic membranes, in general, and the alumina membranes used in this
study, in particular, appropriate for use in the proposed application their surface characteristics
must be changed and their hydrophilic nature needs to be modified. In our own studies we have
changed the nature of the alumina membranes from hydrophilic into hydrophobic via surface
modification using a fluoroalkylsilane (FAS) compound as a surface modifier following a
method originally developed by Lu et al. [257]. FAS are fluorinated organosilanes which are
composed of hydrolysable groups and hydrophobic ends [258]. It is believed that FAS are
attached to the membrane surface through the reaction of their hydrolysable groups with the
surface hydroxyl groups of the metal oxide membrane surface [257,258] (Figure 3.5).
59 | P a g e

Figure 1.5. FAS attachment to the membrane surface

3.2.3.1. Modification Method
To study the surface modification technique, we used tubular ceramic membranes of two
different lengths, 9 cm and 4 cm long, respectively. The modified 9 cm long membrane pieces
were used for break-through type pressure testing, while the 4 cm long pieces were used for a
variety of other tests that included Contact Angle (CA) measurements, Fourier Transform
Infrared Spectroscopy (FTIR), Scanning Electron Microscopy (SEM) and Transmission Electron
Microscopy (TEM), and Thermogravimetric Analysis (TGA). Prior to the deposition of the
surface modifying agent, the membrane pieces were cleaned via ultra-sonication in ethanol and
de-ionized water (DIW) for 30 min each. Then, the membrane were soaked in a solution of
ethanol and DIW (2:1 volume ratio) for 24 hr and dried in air at 333 K for 24 hr. The surface
modification solution (0.1 M) was prepared by dissolving 1H, 1H, 2H, 2H-
perfluorooctyltriethoxysilane (DYNASYLAN F8261) into Hexane (CAS#110-54-3, Merk cat.#
104371) under vigorous stirring for 12 hr at room temperature. The dry membranes were
immersed into the FAS/hexane solution at room temperature, ultra-sonicated in the solution for
30 min, and then left in the solution for an additional 24 hr to allow the coupling reaction
(grafting step) to complete. Afterwards the membranes were thoroughly rinsed with hexane (at
60 | P a g e

least 5 times) to remove any un-reacted FAS from the membranes, and were then dried at 373 K
for 12 hr. The grafting, rinsing and heating were repeated several times as needed (for most of
the membranes reported here 4 times). Upon completion of the grafting step, the membranes
were stored at room temperature for further characterization tests and membrane reactor
experiments.
3.2.3.2. Modified Membrane Characterization
The morphology and the surface properties of the modified samples were examined by
field emission scanning electron microscopy (FE-SEM) and TEM. The contact angle of the
grafted samples was measured and compared with that of the non-modified membranes by
means of a rame-hart instrument (Model # 290-F1). Diffuse Reflectance Infrared Fourier
Transform Spectroscopy (DRIFTS-FTIR) was used to confirm the existence of the Silane-OH
bond. In order to measure the highest temperature that the fluoroalkylsilane – OH bond could
resist, the weight loss of the grafted samples was measured using TGA. Also, the break-through
pressure of the modified membranes was measured, as described below.
3.2.3.2.1. Break-Through Pressure
The break through pressure depends on the surface tension, and contact angle as well as
the membrane’s pore size, as denoted by the Laplace equation, below:
Δp =

(3-1)
with σ being the surface tension, θ the contact angle, Δp is pressure difference at which gas
breaks-through the membrane, and r the pore radius. The performance of the membrane during
this test greatly depends on the hydrophobicity/hydrophilicity of the membrane (Figure 3.6). The
lower the contact angle and the higher the surface tension are, and the smaller the pore size is,
the higher is the breakthrough pressure, and therefore the lower the risk of gas bubbling through
the membrane.
61 | P a g e

Figure 2.6. (a) Hydrophilic membrane (θ < 90
◦
), (b) Hydrophobic membrane (θ > 90
◦
)
However, if the membrane is hydrophilic and is exposed to such a liquid it is likely to
completely wet the membrane and emerge on the other side and spread on the membrane surface
in contact with the gas phase; as previously noted this would then result in the potential loss of
the sweep liquid, but even if this does not happen, substantial infiltration of the sweep liquid
(Figure 3.7) through the membrane means reduced permeation rates through the membrane,
especially if the liquid completely invades the thick macroporous support layer.

Figure 3.7. Gas-Liquid contact in hydrophilic membrane when ΔP
Gas-Liquid
>P
breakthrough

Our approach instead here instead here is to modify the membrane surface so that it
becomes more hydrophobic, thus preventing the sweep liquid for completely wetting the
membrane structure. During operation we maintain the liquid in the membrane interior (tube-
side) at a higher pressure than the gas-phase side, hoping for the liquid to only partially invade,
but to completely block the membrane structure and thus prevent the gas from bubbling through.
Break-through pressure measurements (in this case the Laplace Equation above applies, but the
invading fluid now is a liquid against a “defending” gas phase within the membrane structure) in
this case measuring the pressure drop needed for the liquid to completely wet the membrane and
62 | P a g e

emerge on the other side, are again an important test of the success of the modification procedure
and of the ability of the modified membranes to function stably during the MR experiments. For
these tests, the modified membranes are connected to a HPLC pump via stainless steel tubing. A
back-pressure regulator installed at the end of the tubing connected at the exit of the membrane
allows us to pressurize the membrane using different liquids. At ambient temperature and
pressures of up to 30 bar (for these experiments the membrane shell-side is maintained at
atmospheric conditions), no liquid was observed on the outer surface of the modified membrane
for several hours. For non-modified membranes, even with a liquid pressure of only 4 psi both
DIW and TGDE wet the membrane and leaks-though to the outer surface of the membrane.
These tests then confirmed that modification was successful in terms of preventing the TGDE for
completely wetting the membrane structure.
The stability of the modified membrane to function properly and as intended was also
tested prior to the initiation of the MR experiments after the membrane was installed inside the
reactor. For that we used ultra-high pure nitrogen as the gas phase, flowing at a constant 1 ml/s
flow rate. While increasing the gas phase pressure, the liquid line pressure was kept always at 2
bar higher than the gas phase. The liquid flowing inside the modified membrane was recycled
back into the origin liquid reservoir, and the level of the liquid inside the reservoir was
monitored for several hours. No solvent loss, and no gas bubble were observed. These
observations then confirm that the liquid has successfully blocked the membrane pores and
prevents the gas to escape through, while there is no liquid leaking from the membrane pores on
the gas phase either. The above experiments were performed at ambient temperature. In order to
confirm the stability of the modified membrane at higher temperature, was heated up (gradually
up to 213 C) and liquid loss was monitored. This experiment showed 0.18 ml/min liquid loss,
63 | P a g e

when the inlet TGDE rate was constant 2 ml/min. This liquid loss was due to the liquid
evaporation.
3.2.3.2.2. Contact Angle Measurements
Figure 3.8 shows the shape of DIW droplet on the outer surface of a modified membrane.
Their shape is quite spherical indicative of a high contact angle.

Figure 3.8. DI water drops on the outer surface of the modified membrane.

The membranes utilized are of an asymmetric nature, the outside being an α- Al
2
O
3
support layer and coated on the inside surface an inner γ-Al
2
O
3
layer. To investigate the contact
angle on the inside layer of the grafted membrane, the modified membranes were crushed into
smaller pieces (3-5 mm in length), and the contact angle of the inner layer of each of these small
pieces was tested. As noted above, the contact angle was measured using a rame-hart (Model
#290) automated goniometer (Figure 3.9 set up).
64 | P a g e

Figure 3.9. (a) is the data acquisition system,(b) is the syringe pump to create drops of the
liquid,(c) is a high-speed camera,(d) is a sliding table to set the samples on, and (e) is the light
source.

As previously noted, an ideal membrane for the proposed application should be
hydrophobic with high porosity. Figure 3.10 shows a droplet on the inner surface of the
membrane before and after the modification which confirms the success of the modification.
(most of the contact angles in the literature are reported on a flat surface, but since our
membranes were commercial membranes, and tubular in shape) it was necessary to crush the
membranes in very small pieces to have a surface as close as possible to a flat surface.
65 | P a g e

Figure 3.10. Contact angle measurement, (a) non-modified membrane, (b) modified membrane
The contact angle of DIW, TGDE and two different ionic liquids on the modified
membrane are shown in Table 3.4.

Table 3.4. Contact angle results for the membrane modified using the FAS/Hexane solution
Liquid Used to Measure Contact Angle
BMIM
2
TGDE DI Water 1-MIM
3

one time immersion 77 45 94 153
two times immersion 109 48 136 148
three times immersion 94 85 116 133
four times immersion 110 94 123 118

3.2.3.2.3. Membrane Morphology
In order to compare the surface morphology of the membrane, scanning electron
microscopy (SEM) was utilized, using a JEOL JSM-7001F SEM/EDX instrument which has the
ability to magnify the samples by up to ~ 50,000 times. Our attention was focused primarily on
the inner layer before and after the modification. Figure 3.11 compares the SEM images before
and after modification. It can be seen that there is no clear difference between the non-grafted
and grafted membrane. The grafted FAS molecules were most likely bound to the Al
2
O
3

2
1-Methylimidazolium Hydrogen Sulfate
3
1-Butyl-3-methylimidazolium Hexafluorophosphate
66 | P a g e

membrane surface but could not polymerize to form an observable layer. Therefore, it is difficult
to distinguish the difference in surface morphologies between these two membranes.
Figure 3.11.(a) SEM image of the cross-section of non-modified and (b) SEM image of the
modified surface membrane.
3.2.3.2.4. FTIR – DRIFTS Test
The modified membrane was crushed into fine powder (## siz) and was mixed with
Potassium Bromid (KBr) ( 1 % weight). The mixture was then placed in the sample cup of a
DRIFTS cell (Collector II), and was analyzed by FTIR (Nicolet 10). The spectra are shown in
Figure 3.12.
Although figure 3.12, shows the silane vibration, the vibration is not very clear as it was
reported by Picard et. al. [259]. The possible explanation for this matter could be the fact that the
amount of FAS seated on the last layer of the surface of the membrane is very small compared to
the amount of support.
67 | P a g e

Figure 3.12. FTIR results shows existence Silane vibration in the modified membrane

3.2.3.2.5. Thermogravimetric Analysis (TGA)
Since methanol synthesis takes place at a temperature of 200
o
C and above it is important
to confirm that the modified membrane keeps its hydrophobic character at these temperatures.
The stability of the surface layer at different temperatures was investigated via TGA analysis.
Figure 3.13 shows the weight loss in flowing He for the membrane sample modified by FAS.
The membrane weight stays constant up to 100
o
C (indicative of the very hydrophobic nature of
these membranes and no condensed water being present) but some decomposition of the FAS is
indicated after that, but with the weight stabilized after the temperature of 175
o
C and below 225
o
C. The initial weight gain, due to FAS grafting was 1.35%, thus the FAS weight loss due to
decomposition is ~ 25%. .

68 | P a g e

Figure 3.13 TGA test of the modified ceramic membrane in flowing Helium.
To avoid substantial changes of the membrane surface during the experiments most of the MR
runs were done for temperatures of 210
o
C and below.
3.3. Results and Discussion
A series of experiments have been performed in order to explore the effect of different
experimental conditions on the total carbon conversion in the membrane reactor. The
experimental parameters studied include the reactor pressure, temperature, and W/F (gr/mol/hr)
the ratio of weight of catalyst weight to the total feed flow rate, at a constant stoichiometric
number (SN=2), constant carbon dioxide-carbon monoxide ratio in the feed (

=0.6). The
membrane which was used in this set of experiments was 9 cm length, modified by FAS as
previously described, with the same characteristics as indicated in Table 3.2.
3.3.1. Effect of Feed Flow Rate
The effect of total feed flow rate to the reaction zone was explored for three different
W/F (weight of catalyst to total feed flow rate ratio) equal to 10, 15, and 20 (gr/mol/hr) at
constant reactor pressure (30 bar), constant sweep liquid pressure (32 bar), constant temperature
69 | P a g e

(200
◦
C) and constant stoichiometric number (SN=2). The results are shown in Figure 3.14. The
total carbon conversion for the PBR is defined by Eqn. (2-30), while the conversion for the MR
is defined by the following Eqn. (which assumes that gas feed and the sweep liquid contain no
methanol):
Conversion
carbon,MR
=

=

Figure 3.14. Comparison of the feed flow rate effect on total carbon conversion in the PBR and
MR
The figure shows that MR attains conversions which are higher than those for the packed
bed reactor. As it was expected, by increasing the W/F (which in our case was accomplished by
decreasing the inlet gas flow rate with constant weight of the catalyst ~15 gr), conversion was
increased gradually in both the packed-bed reactor and the membrane reactor. In order to
perform the packed-bed reactor experiments, the sweep liquid flow was stopped and the
permeate line was closed. The relatively small conversion improvements is due to limitations
with the size of the commercial membrane that was installed inside the reactor. Also due to our
70 | P a g e

limitations caused by lower limits of the mass flow controllers (MFCs) that we used in the
laboratory, the maximum W/F ratios that we could attain was limited to lower values.
3.3.2. Effect of Reaction Side Pressure
As it was mentioned in the reaction kinetics studies with increase in pressure the
conversion of the syngas to methanol will increase. In the membrane reactor experiments the
pressure difference between gas phase and liquid phase has been kept constant, and equal to 2
bar, where sweep liquid has higher pressure. Due to the limitation that we had with respect with
temperature, in all these experiments temperature was set to be 200
◦
C. The W/F was set equal to
20 gr/mol/hr, the stoichiometric number SN was set equal to 2, while the carbon dioxide-carbon
monoxide ratio in the feed (

) was set equal to 0.6. The membrane which was used in
this series of experiments was 9 cm length, modified by FAS, and had the same characteristics as
indicated in Table 3.2. Figure 3.15 indicates that at higher pressures the carbon conversion is
higher, as expected, and the membrane reactor shows better performance in terms of carbon
conversion. By increasing the pressure methanol synthesis reaction conversion will progressively
shift to the right side. The more methanol is produced and removed from the reaction zone, the
higher the conversion in the MR would be.
71 | P a g e

Figure 3.15. Reaction side pressure effect on membrane reactor performance

3.3.3. Effect of Temperature
The Methanol synthesis reactions are exothermic, after reaching equilibrium, reducing
the temperature has positive effect on conversion. As it was mentioned before, the temperature
that we could operate the membrane reactor was limited due to concerns about FAS stability. In
addition, the Kalrez O-rings, used to seal the membrane also showed deformation at higher
temperature.

Figure 3.16. Temperature effect on membrane reactor performance
72 | P a g e

So, for these reasons, we decided to investigate the temperature effect limited to very narrow
changes. Experiments were performed at constant pressure equal to 30 bar, SN=2, W/F=20,

=0.6. Figure 3.16 clearly shows that at lower temperatures MR has a higher conversion.
By increasing the reactor temperature the MR shows performance closer to the PBR.
3.4. Conclusions
In this chapter the application of a novel MR for methanol synthesis has been studied.
The process studied uses high B.P. sweep liquid in order to remove methanol, the main product,
from the reaction zone. A commercial ceramic membrane was used in this study, whose surface
was modified via a fluorosurfactant (FAS) to become more hydrophobic. Different
characterization tests have been performed on the modified membrane to confirm the success of
the surface modification. MR experiments were performed using the modified membrane, and
the results were compared with the PBR performance at the same operating conditions. The MR
experiments showed an improvement in the total carbon conversion.

73 | P a g e

References
1. M. Mulder, Basic Principles of Membrane Technology, Kluwer Academic Publishers, 1996.
2. T. Melin, R. Rautenbach. Membranverfahren - Grundlagen der Modulund Anlagenauslegung.
VDI-Buch Chemische Technik / Verfahrenstechnik. Springer, 3 edition, 2007.
3. F.V. Lima, P. Daoutidis, Mi. Tsapatsis, J.J. Marano, Industrial and Engineering Chemistry
Research, 51, 5480-5489 (2012).
4. U.S. Pat. 09/905651 (2002), R. Mackay, and A.F. Sammells.
5. D. Mendes, A. Mendes, L.M. Madeira, A. Iulianelli, J.M. Sousa, A. Basile, Asia-Pacific
Journal of Chemical Engineering, 5, 1, 111-37 (2010).
6. K. Babita, S. Sridhar, K.V. Raghavan, International Journal of Hydrogen Energy, 36, 11,
6671-6688 (2011).
7. U.S. Patent: 4937196, W.J. Wrasidlo, F.K. Hofmann, , D.M. deWinter.
8. N.S. Abo-Ghander, F. Logist, J.R. Grace, J.F.M. Van Impe, S.S.E.H. Elnashaie, C.J. Lim,
Chemical Engineering Science, 65, 10, 3113-27 (2010).
9. H.H. Tsai, V. Ravindran, M. Pirbazari, Chemical Engineering Science, 60, 20, 5620-5636
(2005).
10. N.S. Abo-Ghander, J.R. Grace, S.S.E.H. Elnashaie, C.J. Lim, Chemical Engineering Science,
63, 7, 1817-1826 (2008).
11. G.M. Rios, M.P. Belleville, D. Paolucci, J. Sanchez, Journal of Membrane Science, 242, 1-2,
189-196 (2004).
12. S. Mozia, Separation and Purification Technology, 73, 2, 71-91 (2010).
13. S. Engell, Logistic Optimization of Chemical Production Processes, Wiley-VCH, 2008.
14. M. Sharma, R.K. Vyas, K. Singh, Adsorption. 19(1), 161-188 (2013).
74 | P a g e

15. G. Schembecker, S. Tlatlik, Chemical Engineering and Processing, 42, 3, 179-189 (2003).
16. J.-H. Roh, I. Lee, Computer Modeling in Engineering & Sciences, 26, 1, 13-30 (2008).
17. X. Feng, J. Ivory, Journal of Membrane Science, 176, 2, 197-207 (2000).
18. A. R. Ricardo, S. Velizarov, J. G. Crespo, M. A.M Reis, Process Biochemistry, 47, 12, 1832-
1838 (2012).
19. M. Younas, S. Druon-Bocquet, J. Sanchez, Journal of Membrane Science, 382, 1-2, 70-81
(2011).
20. S. Sa, H. Silva, J. M. Sousa, A. J. Mendes, Membrane Science, 339, 1-2, 160-70 (2009).
21. K. Briceño, A. Iulianelli, D. Montané, R. Garcia-Valls, A. Basile, International Journal of
Hydrogen Energy, 37, 18, 13536-13544 (2012).
22. M. Sheintuch, I. Efremenko. Chemical Engineering Science, 59, 22-23, 4739-4746 (2004).
23. G., Sznejer, M. Sheintuch, Chemical Engineering Science, 59, 10, 2013-2021 (2004).
24. K.Varoon, Zh. Xueyi, B. Elyassi, D.D. Brewer, M. Gettel, S. Kumar, J.A. Lee, S.
Maheshwari, A. Mittal, Chun-Yi Sung, M. Cococcioni, L.F. Francis, A.V. McCormick, K.A.
Mkhoyan, M. Tsapatsis, Science, 333, 6052, 72-5 (2011).
25. J. Wöltinger, K. Drauz, A.S. Bommarius, Applied Catalysis A: General, 221, 1-2, 171-185
(2001).
26. M. Ulbricht, Polymer, 47, 7, 2217-62 (2006).
27. M.O. Daramola, E.F. Aransiola, T.V. Ojumu, Materials, 5, 2101-2136 (2012).
28. B. Lauterboeck, M. Ortner, R. Haider, Water Research, 46, 15, 4861-4869 (2012).
29. J.M. Jani, M. Wessling, R.G.H. Lammertink, Microfluidics and Nanofluidics, 13, 3, 499-509
(2012).
75 | P a g e

30. W.K. Chan, J. Jouet, S. Heng, K.L. Yeung, J.C. Schrotter, Journal of Solid State Chemistry,
189, 96–100 (2012).
31. S.H. Chen, P.F. Xing, P.J. Zhao, W.M Chen, Separation and Science Technology, 37, 2701-
2716 (2002).
32. R. Koc, N.K. Kazantzis, Y. Hua Ma, International Journal of Hydrogen Energy, 36, 8, 4934-
4951 (2011).
33. M. Miyamoto, C. Hayakawa, K. Kamata, M. Arakawa, S. Uemiya, Journal of Hydrogen
Energy, 36, 13, 7771-5 (2011)
34. A. Brunetti, G. Barbieri, E. Drioli, Chemical Engineering Science, 64, 15, 3448-3454 (2009).
35. G. Barbieri, A. Brunetti, G. Tricoli, E. Drioli, Journal of Power Sources, 182, 1, 160-167
(2008).
36. H. Luo, Y. Wei, H. Jiang, W. Yuan, Y. Lv, J. Caro, H. Wang, Journal of Membrane Science,
350, 1-2, 154-160 (2010).
37. Z. Gong L. Hong, Journal of Membrane Science, 380, 1-2, 81-86 (2011).
38. E.G. Babakhani, J. Towfighi, Z. Taheri, A.N. Pour, M. Zekordi, A. Taheri, Journal of Natural
Gas Chemistry. 21, 5, 519-525 (2012).
39. Z. Yuwen, L. Jiao, L. Yong, D. Weizhong, L. Xionggang, Rare Metals, 9, 3, 231-237 (2010).
40. M. Pontié, D. Lemordant, Journal of Membrane Science, 141, 1, 13-17 (1998).
41. R. Wong, S. Atchariyawut, C. Feng, R. Jiraratananon, D.T. Liang, Journal of Membrane
Science, 285, 1-2, 272-81 (2006).
42. S. Heinrich, S.A.R.K. Deshmukh, L. Morl, Mv.S. Annaland, J.A.M. Kuipers, Chemical
Engineering Science, 62, 1-2, 416-36 (2007).
76 | P a g e

43. C.A. Scholes, K.H. Smith, S.E. Kentish, G.W. Stevens, International Journal of Greenhouse
Gas Control, 4, 5, 739-755 (2010).
44. D.S. Scholl, Y.H .Ma, MRS Bulletin, 31, 10, 770-3 (2006).
45. T. Xu, Y. Fu, X. Wang, Desalination, 171, 2, 155-165 (2005).
46. R. Chen, Y. Jiang, W. Xing, Industrial & Engineering Chemistry Research, 50, 8, 4405-4411
(2011).
47. Z. Liu, G. Zhang, X. Dong, W. Jiang, W. Jin, N. Xu, Journal of Membrane Science, 415-416,
313-319 (2012).
48. A. Gabelman, S.T. Hwang, W.B. Krantz, Journal of Membrane Science, 257, 1-2, 11-36, 15
(2005).
49. M.G. Buonomenna, E. Drioli, Organic Process Research & Development, 12, 5, 982-988
(2008).
50. Z. Jia, Q. Chang, J. Qin, X. Hong, Journal of Membrane Science, 342, 1-2, 1-5 (2009).
51. A. Criscuoli, E. Rossi, F. Cofone, E. Drioli, Clean Technologies and Environmental Policy,
12, 1, 53-61 (2010).
52. S. Cardea, A. Gugliuzza, M. Sessa, M.C. Aceto, E. Drioli, E. Reverchon, ACS Applied
Materials and Interfaces, 1, 171-180 (2009).
53. M. Kohout, P. Hasal, I. Schreiber, Chemical Engineering Science, 83, 110-118 (2012).
54. C. Wu, T. Xu, W. Yang, Journal of Membrane Science, 224, 1-2, 117-125 (2003).
55. C. Wu, T. Xu, M. Gong, W. Yang, Journal of Membrane Science, 247, 1-2, 111-118 (2005).
56. J.G. Sanchez Marcano, T.T. Tsotsis, Catalytic Membranes and Membrane Reactors, Wiley-
VCH, Weinheim, 2002.
57. A. Julbe, D. Farrusseng, C. Guizard, Catalysis Today, 104, 102–113 (2005).
77 | P a g e

58. A.G. Dixon, International Journal of Chemical Reactor Engineering, 1 (R6), 1-35 (2003).
59. R. Dittmeyer, V. Hollein, K. Daub, Journal of Molecular Catalysis. A: Chemical, 173 (1-2)
135-184 (2001).
60. O. Schramm, A. Seidel-Morgenstern, Chemical Engineering Science, 54(10), 1447-1453
(1999).
61. M. Pera-Titus, M. Fridmann, N. Guilhaume, Journal of Membrane Science, 401, 204-216
(2012).
62. L. Giorno, E. Drioli, Trends in Biotechnology, 18, 8, 339-349 (2000).
63. U. Kragl, T.D Wars, Trends in Biotechnology, 19, 11, 442-449 (2001).
64. G Barbieri, A Brunetti, T Granato,Paola Brnardo, Enrico Drioli, Industrial and Engineering
Chemistry Research, 44, 20, 7676-7683 (2005).
65. C.J.M. Bradford, A.P. Mure Te, Applied Catalysis A: General, 283, 1-2, 39-46 (2005).
66. M.P. Pina, G. Picasso, M. Gutierrez, J. Herguido, Chemical Engineering Journal, 126, 2-3,
119-30 (2007).
67. T. Westermann, T. Melin, Chemical Engineering and Processing, 48, 17-28 (2009).
68. D. deMontigny, P. Tontiwachwuthikul, A. Chakma, Industrial and Engineering Chemistry
Research, 44, 15, 5726-5732 (2005).
69. A. Mansourizadeh, A.F. Ismail, Journal of Hazardous Materials, 171, 1-3, 38-53 (2009).
70. G.J. Hwang, K. Onuki, Journal of Membrane Science, 194, 2, 207-215 (2001).
71. R.T. Jacobsen, Chemical Engineering Progress, 101, 2, 20-3 (2005).
72. Seyed A. S. Rezai Y. Traa, Journal of Membrane Science, 319, 1-2, 279-285 (2008).
73. T.A. Peters, N.E. Benes, J.T.F. Keurentjes, Industrial and Engineering Chemistry Research,
44, 25, 9490-9496 (2005).
78 | P a g e

74. J.N. Armor, Applied Catalysis A: General, 222, 91-99 (2001).
75. S. Tosti, L. Bettinali, V. Violante, International Journal of Hydrogen Energy, 25, 4, 319-325
(2000).
76. J. M. Sousa , A. Mendes, Industrial and Engineering Chemistry Research, 45, 6, 2094-2107
(2006).
77. M. Bayat, M.R. Rahimpour, B. Moghtaderi, Journal of Natural Gas Science and Engineering,
3, 4, 555-570 (2011).
78. M.P. Rohde, G. Schaub, S. Khajavi, J.C. Jansen, F. Kapteijn, Materials, 115, 1-2, 123-136
(2008).
79. F.A.N. Fernandes, Journal of Natural Gas Chemistry, 16, 2, 107-114 (2007).
80. M.R. Rahimpour, A. Mirvakili, K. Paymooni, Energy, 36, 2, 1223-1235 (2011).
81. T.A. Rogers, A.S. Bommarius, Chemical Engineering Science, 65, 6, 2118-24 (2010).
82. G. Barbieri, A. Brunetti, A. Caravella, E. Drioli, RSC Advances, 1,4, 651-661 (2011).
83. J.N. Armor, T.S. Farris, P.L. Silveston, J.W. Hightower, W.O. Haag, J.R.H. Ross, S.
Kaliaguine, A. Sarkany, J.W. Geus, Studies in Surface Science and Catalysis, 75(B), 1362-1375
(1993).
84. T.M. Aminabhavi, B.V.K. Naidu, S. Sridhar, R. Rangarajan, Journal of Applied Polymer
Science, 95, 5, 1143-1153 (2005).
85. S.Y. Lim, B. Park, F. Hung, M. Sahimi, T.T. Tsotsis, Chemical Engineering Science, 22-23,
4933-4946 (2002).
86. M. R. Rahimpour, S. Ghader, Chemical Engineering and Processing, 43, 1181-1188 (2004).
87. G. Barbieri, G. Marigliano, G. Golemme, E. Drioli, Chemical Engineering Journal, 85, 53-59
(2002).
79 | P a g e

88. A. Alshebani, M. Pera-Titus, E. Landrivon, T. Schiestel,S. Miachon, J.A. Dalmon,
Microporous and Mesoporous Materials, 115, 197–205 (2008).
89. M.O. Daramola, A.J. Burger, M. Pera-Titus,A. Giroir-Fendler, L. Lorenzen, J.A. Dalmon,
Separation Science and Technology, 45, 21–27 (2010).
90. Z. Deng, C.-H. Nicolas, M.O. Daramola,J. Sublet,Th. Schiestel, A.J. Burger, Y. Guo, A.
Giroir-Fendler, M. Pera-Titus, Journal of Membrane Science, 364, 1–8 (2010).
91. M.O. Daramola, A.J. Burger, M. Pera-Titus, A. Giroir-Fendler, L. Lorenzen, S. Miachon,
J.A. Dalmon, Journal of Membrane Science, 337, 106–112 (2009).
92. J.E. Castanheiro, A.M. Ramos, I.M. Fonseca, J. Vital, Applied Catalysis A: General, 311, 1-
2, 17-23 (2006).
93. K.J. Kim, S.H. Park, W.W. So, S.J. Moon, Journal of Applied Polymer Science, 79, 8, 1450-
1455 (2001).
94. "Market Study: Propylene, Ceresana Research, February 2011". ceresana.com. Retrieved
2011-02-13.
95. OECD SIDS Initial Assessment Profile — Ethylene" (PDF). inchem.org. Retrieved 2008-05-
21.
96. P. Hacarlioglu, Y. Gu, S.T. Oyama, Journal of Natural Gas Chemistry, 15, 2, 73-81 (2006).
97. M.H. Khademi, A. Jahanmiri, M.R. Rahimpour, International Journal of Hydrogen Energy,
34, 12, 5091-5107 (2009).
98. G.F. Froment, Catalysis Reviews - Science and Engineering, 50, 1, 1-18 (2008).
99. R.L. Verson, American Institute of Chemical Engineers, National Meeting, 7B 27 (1988).
100. S.S.E.H. Elnashaie, B.K. Abdalla, R. Hughes, Industrial and Engineering Chemistry
Research, 32, 11, 2537-2541 (1993).
80 | P a g e

101. L. Wang, K. Murata, M. Inaba, Catalysis Today, 82, 99-104 (2003).
102. Z. Wu, I.M.D. Hatim, B.F.K. Kingsbury, E. Gbenedio, K. Li, AIChE Journal, 55, 9, 2389-
2398 (2009).
103. W. Liang, R. Hughes, Catalysis Today, 104, 2-4, 238-243 (2005).
104. T.M. Raybold, M.C. Huff, Catalysis Today, 56, 35-44 (2000).
105. N. Itoh, E. Tamura, S. Hara,T. Takahashi, A. Shono, K. Sato, T. Namba, Catalysis Today,
82, 119-125 (2003).
106. S. Yolcular, O. Olgun, Energy Sources. Part A, 32, 15, 1437-45 (2010).
107. N. Itoh, W.C. Xu, S. Hara, K. Kakehida, Y. Kaneko, A. Igarashi, Industrial Engineering
Chemistry Research, 42, 25, 6576-6581 (2003).
108. A.W. Wang, B.A. Raich, B.K. Johnson, H.C. Foley, American Chemical Society, Division
of Petroleum Chemistry, Preprints. 37, 3, 743-751 (1992).
109. J. Zaman, A. Chakma, International Journal of Hydrogen Energy, 20, 1, 21-8 (1995).
110. C.P. Badra, International Journal of Hydrogen Energy, 20, 9, 717-21 (1995).
111. S. Kar, R.C. Bindal, S. Prabhakar, P.K. Tewari, Journal of Hydrogen Energy, 37, 4, 3612-
20 (2012).
112. M. Nomura, S. Kasahara,H. Okuda, N. Shin-ichi, International Journal of Hydrogen Energy,
30, 13-14, 1465-73 (2005).
113. X. Wu, K. Onuki, Tsinghua Science and Technology, 10, 2, 270-6 (2005).
114. F.J. Macdonald, E.A. Fletcher, Energy, 7, 12, 953-60 (1982).
115. B.-K. Sea, K. Kusakabe, S. Morooka, Industrial and Engineering Chemistry Research, 37,
6, 2509-2515 (1998).
81 | P a g e

116. K. Kusakabe, S. Morooka, Memoirs of the Kyushu University, Faculty of Engineering, 58,
3, 163-181 (1998).
117. A. Kogan, International Journal of Hydrogen Energy, 25, 11, 1043-50 (2000).
118. Y.F. Gu, S.T. Oyama, Advanced Materials, 19, 1636-1640 (2007).
119. P. Hacarlioglu, D. Lee, G.V. Gibbs, S.T. Oyama, Journal of Membrane Science, 313, 277-
283 (2008).
120. A.K. Prabhu, S.T. Oyama, Journal of Membrane Science, 176, 233-248 (2000).
121. W.J. Koros, G.K. Fleming, Journal of Membrane Science, 83, 1-80 (1993).
122. S. Tosti, L. Bettinali, S. Castelli, F. Sarto, S. Scaglione, V. Violante, Journal of Membrane
Science, 196-241 (2002).
123. A. Basile, F. Gallucci, S. Tosti, A. Iulianelli, E. Drioli, International Journal of Hydrogen
Energy, 32, 9, 1201-10 (2007).
124. J.N. Keuler, L. Lorenzen, Journal of Membrane Science, 195, 203-213 (2002).
125. P.D.Vaidya, A.E. Rodrigues, Chemical Engineering Journal, 117, 39-49 (2006).
126. Ch.Y. Tsai, S.Y. Tam, Y. Lu, C. J. Brinker, Journal of Membrane Science, 169, 2, 255-268
(2000).
127. Y. Shu, M. Ichikawa, Catalysis Today, 71, 1-2, 55-67 (2001).
128. B.S. Liu, P.Y. Lian, X.H. Zhao, Separation and Purification Technology, 32, 281-287
(2003).
129. W.H. Lin, H.F. Chang, Catalysis Today, 97,181-188 (2004).
130. J.N. Keuler, L. Lorenzen, Industrial and Engineering Chemistry Research, 41 1960-1966
(2002).
82 | P a g e

131. A. Trianto, W.L. Qiao, T. Kokugan, Journal of Chemical Engineering Japan, 34, 1065-1068
(2002).
132. J.N. Keuler, L. Lorenzen, Journal of Membrane Science, 202, 17-26 (2002).
133. T. Sato, H. Yokoyama, H. Miki, N. Itoh, Journal of Membrane Science, 289, 1-2, 97-105
(2007).
134. M.P. Gimeno, Z.T.Wu, J. Soler, J. Herguido, K. Li, M. Menéndez, Chemical Engineering
Journal, 155, 298-303 (2009).
135. A.M. Martini, L.M. Walter, T.C.W. Ku, J.M. Budai, J.C. McIntosh, M. Schoell, AAPG
Bulletin, 87, 8, 1355-1375 (2003).
136. P. Häussinger, R. Lohmüller, A.M. Watson, Hydrogen. Ullmann's Encyclopedia of
Industrial Chemistry, Weinheim: Wiley-VCH, 2005.
137. J. Tong, Y. Matsumura, H. Suda, K. Haraya, Industrial and Engineering Chemistry
Research, 44,5, 1454-1465 (2005).
138. J.R. Lattner, M.P. Harold, International Journal of Hydrogen Energy. 29, 4, 393-417 (2004).
139. Mv.S. Annaland, T.P. Tiemersma, C.S. Patil, J.A.M. Kuipers, Chemical Engineering
Science, 61, 5, 1602-16 (2006).
140. M. Abdollahi, J. Yu, P.K.T. Liu, R. Ciora, M. Sahimi, T.T. Tsotsis, Journal of Membrane
Science, 32, 390-391 (2012).
141. A. Basile, S. Campanari, G. Manzolini, A. Iulianelli, T. Longo, S. Liguori, M. De Falco, V.
Piemonte, International Journal of Hydrogen Energy, 36, 2, 1531-9 (2011).
142. N. Mori, T. Nakamura, K.I. Noda, O. Sakai, A. Takahashi, N. Ogawa, H. Sakai,Y.
Iwamoto, T. Hattori, Industrial and Engineering Chemistry Research, 46, 7, 1952-1958 (2007).
83 | P a g e

143. T. Tsuru, K. Yamaguchi, T. Yoshioka, M. Asaeda, AIChE Journal, 50, 11, 2794-2805
(2004).
144. T. Tsuru, Journal of the Japan Petroleum Institute, 54, 5, 277-86 (2011).
145. F. Gallucci, A. Basile, International Journal of Hydrogen Energy, 33, 6, 1671-87 (2008).
146. T. Kume, International Gas Union Research Conferences, Japan, 146-152, 2011.
147. R.W. van den Brink, F.A. de Bruijn, AIP Conference Proceedings, 1169, 119-26 (2009).
148. E. Engwall, J. Saukaitis, M. Joshi, A. Matzakos, P. Del, A. Paggio, AIChE Annual Meeting,
Conference Proceedings, 2007 AIChE Annual Meeting, 2007.
149. N. Pomerantz, Y.H. Ma, AIChE Journal, 56, 3062 – 3073 (2010).
150. C.H. Chen, Y. H.Ma, Journal of Membrane Science, 362(1-2), 535-544 (2010).
151. N. Pomerantz, Y.H. Ma, Industrial & Engineering Chemistry Research, 48(8) 4031-4039
(2009).
152. T. Tsuru, T. Morita, H. Shintani, T. Yoshioka, M. Asaeda, Journal of Membrane Science,
316, 53-62 (2008).
153. K.S. Patel, A.K. Sunol, International Journal of Hydrogen Energy, 32, 2344-2358 (2007).
154. T. Kim, S. Moon, S.I. Hong, Applied Catalysis A: General, 224, 1-2, 111-120 (2002).
155. J.M. Ginsburg, J. Piña, T. El Solh, H.I. De Lasa, Industrial and Engineering Chemistry
Research, 44, 14, 4846-4854 (2005).
156. J. Guo, H. Lou, H. Zhao, D. Chai, X. Zheng, Applied Catalysis A: General, 273, 1-2, 75-82
(2004).
157. A.A. Adesina, Current Opinion in Chemical Engineering, 1, 272-280 (2012).
158. S.T. Oyama, P. Hacarlioglu, Y. Gu, D. Lee, International Journal of Hydrogen Energy, 37,
10444-10450 (2012).
84 | P a g e

159. S.T. Oyama, H. Lim, Chemical Engineering Journal, 151, 1-3, 351-358 (2009).
160. J. Kniep, M. Anderson, Y.S. Lin, Industrial and Engineering Chemistry Research, 50, 22,
12426-12432 (2011).
161. A. Basile, L. Paturzo, Catalysis Today, 67,55-64 (2001).
162. D.R. Palo, R.A. Dagle, J.D. Holladay, Chemical Reviews, 107, 3992-4021 (2007).
163. G.Q. Lu, J.C.D. da Costa, M. Duke, S. Giessler, R. Socolow, R.H. Williams, T. Kreutz
Journal of Colloid and Interface Science, 314,589-603 (2007)
164. R.E. Uhrig, Proceedings of the International Hydrogen Energy Congress Exhibition, IHEC
2005, Istanbul (Turkey), 13–15 July 2005
165. H.F. Chang, W.J. Pai, Y.J. Chen, W.H. Lin, International Journal of Hydrogen Energy, 35,
23, 12986-92 (2010).
166. F. Gallucci, M. Van Sint Annaland, J.A.M. Kuipers, Topics in Catalysis, 51, 133-145
(2008).
167. F. Gallucci, A. Basile, S. Tosti, A. Iulianelli, E. Drioli, International Journal of Hydrogen
Energy, 32, 9, 1201-1210 (2007).
168. A. Lima Da Silva, I.L. Müller, International Journal of Hydrogen Energy, 36, 3, 2057-2075
(2011).
169. J. Agrell, H. Birgersson, M. Boutonnet, Journal of Power Sources, 106, 249-257 (2002).
170. J. Han, I.S. Kim, K.S. Choi, Journal of Power Sources, 86, 223-227 (2000).
171. T. Takahashi, M. Inoue, T. Kai, Applied Catalysis A- General, 218, 189-195 (2001).
172. S. Tanaka, K.S. Chang, K.B. Min, D. Satoh, K. Yoshida, M. Esashi, Chemical Engineering
Journal, 101, 1-3, 143-9 (2004).
173. G. Chen, Q. Yuan, H. Li, Sh. Li, Chemical Engineering Journal, 101, 1-3, 101-106 (2004).
85 | P a g e

174. R.E. Buxbaum, Separation Science and Technology, 34, 2113-2123 (1999).
175. S. Wieland, T. Melin, A. Lamm, Chemical Engineering Science, 57(9), 1571-1576 (2002).
176. Y.M Lin, M.H Rei, International Journal of Hydrogen Energy, 25, 211-9 (2000).
177. A. Basile, S. Tosti, G. Capannelli, G. Vitulli, A. Iulianelli, F. Gallucci, E. Drioli, Catalysis
Today, 14, 118-237 (2006).
178. A. Basile, F. Gallucci, L. Paturzo, Catalysis Today, 104, 244-250 (2005).
179. A. Basile, F. Gallucci, A. Iulianelli, F. Borgognoni, S. Tosti, Journal of Membrane Science,
311, 46-52 (2008).
180. N. Liu, Z. Yuan, S. Wang, C. Zhang, S. Wang, D. Li, International Journal of Hydrogen
Energy, 33, 1643-51 (2008).
181. R.P. Omorjan, R.N. Paunovic, M.N. Tekic, Journal of Membrane Science, 154, 273-280
(1999).
182. D.K. Liguras, D.I. Kondarides, X.E. Verykios, Applied Catalysis, 43, 345-354 (2003).
183. S. Freni, N. Mondello, S. Cavallaro, G. Cacciola, V.N. Pardon, V.A. Sobyanin, Reaction
Kinetics Catalysis Letters, 71, 143-152 (2000).
184. S. Consonni, F. Viganò, International Journal of Hydrogen Energy, 30, 7, 701-718 (2005).
185. A. Brunetti, A. Caravella, G. Barbieri, E. Drioli, Journal of Membrane Science, 306, 1-2,
329-340 (2007).
186. A. Harale, H. T. Hwang P.K.T. Liu, M. Sahimi, T.T. Tsotsis, Chemical Engineering
Science, 62, 15, 4126-37 (2007).
187. P.K.T. Liu, M. Sahimi, T.T. Tsotsis, Current Opinion in Chemical Engineering, 1, 342-351
(2012).
86 | P a g e

188. M. Abdollahi, J. Yu, P.K.T. Liu, R. Ciora, M. Sahimi, T.T. Tsotsis, Journal of Membrane
Science, 363, 160-169 (2010).
189. M. Abdollahi, J. Yu, P.K.T. Liu, R. Ciora, M. Sahimi, T.T. Tsotsis, Industrial and
Engineering Chemistry Research, 49, 10986-10993 (2010).
190. Z. Tang, S.J. Kim, G. K. Reddy, J. Dong, P. Smirniotis, Journal of Membrane Science, 354,
1-2, 114-122 (2010).
191. Y. H. Ma ,D. Driscoll, J. Hissam, Composite Pd and Alloy Porous Stainless Steel
Membranes for Hydrogen Production and Process Intensification , DOE Contract Number: DE-
FC26-07NT43058 (2011).
192. S. Tosti, A. Basile, G. Chiappetta, C. Rizzello, V. Violante, Chemical Engineering Journal,
93,23-30 (2003).
193. H. Yoshida, H. Takeshita, S. Konishi, H. Ohno, T. Kurasawa, H. Watanabe, Nuclear
Technology Fusion, 5, 178-188 (1984).
194. H. Schulz, Zh. Nie, F. Ousmanov, Catalysis Today, 71, 351–360 (2002).
195. K. R. Hall, Catalysis Today, 106, 243-246 (2005).
196. Y.N. Wang, W.P. Ma, Y.J. Lu, J. Yang, Y.Y. Xu, H.W. Xiang, Y.W. Li, Y.L. Zhao, B.J.
Zhang, Fuel, 82,195–213 (2003).
197. B. Haut, V. Halloin, H.B. Armor, Chemical Engineering and Processing, 43, 979-986
(2004).
198. H. Ben Amor, V.L. Halloin, Chemical Engineering Science, 54, 1419-1424 (1999).
199. V.L. Halloin, S.J. Wajc, Chemical Engineering Science, 49, 4691-4698 (1994).
200. M. Kuczynsky, R. Oyeavaar, R. Pieters, K. Westerterp, Chemical Engineering Science, 42,
1887-1898 (1987).
87 | P a g e

201. K.R. Westerterp, M. Kuczynski, Hydrocarbon Process, 65, 80-83 (1986).
202. K.R. Westerterp, M. Kuczynski, Chemical Engineering Science, 42, 1871-1885 (1987).
203. A. Kruglov, Chemical Engineering Science, 49, 4699-4716 (1994).
204. K. R. Westerterp, M. Kuczynski, C.H.M. Kamphuis, Industrial and Engineering Chemistry
Research, 28, 763-771 (1989).
205. R.L. Espinoza, E. du Toit, J. Santamaria, M. Menendez, J. Coronas, S. Irusta, Studies in
Surface Science and Catalysis, 30, 389-394 (2000).
206. F. Gallucci, L. Paturzo, A. Basile, Chemical Engineering and Processing, 43, 1029-1036
(2004).
207. R.P.W.J. Struis, S. Stucki, M. Wiedorn, Journal of Membrane Science, 113, 93-100 (1996).
208. R.P.W.J. Struis, S. Stucki, Applied Catalysis A: General, 216, 117-129 (2001).
209. G. Chen , Q. Yuan, Separation and Purification Technology, 34, 1-3, 227-237 (2004).
210. M. P. Rohde, D. Unruh, G. Schaub, Industrial and Engineering Chemistry Research, 44,
9653-9658 (2005).
211. M. P. Rohde, D. Unruh, G. Schaub, Catalysis Today, 106, 143-148 (2005).

212. J.M. Zheng, Y.Y. Xu, Z.K. Xu, Separation Science and Technology, 38, 6, 1247-1267
(2003).
213. R. Faiz, M. Al-Marzouqi, Journal of Membrane Science, 342, 1-2, 269-78 (2009).
214. A. Gabelman, S-T. Hwang, Journal of Membrane Science, 159, 1-2, 61-106 (1999).
215. D. W. Johnson, M. J. Semmens, J. S. Gulliver, Journal of Membrane Science, 128, 1, 67-81
(1997).
216. Z. Ren, W. Zhang, H. Li, Wei Lin, Chemical Engineering Journal, 146, 2, 220-6 (2009).
88 | P a g e

217. P. Wang, M. M. Teoh, T.S. Chung, Water Research, 45, 17, 5489-5500 (2011).
218. A.P. Korikov, K.K. Sirkar, Journal of Membrane Science, 246, 27-37 (2005).
219. R. Klaassen, P.H.M. Feron, A.E. Jansen, Chemical Engineering Research and Design, 83
234 – 246 (2005).
220. V.V. Usachov, V.V. Teplyakov, A.Y. Okunev, N.I. Laguntsov, Separation and Purification
Technology, 57, 502-506 (2007).
221. R. Molinari, P. Argurio, S. M. Carnevale, T. Poerio, Journal of Membrane Science, 366, 1-
2, 139-147 (2011).
222. S. Miachon, V. Perez, G. Crehan, E. Torp, H. Rder, R. Bredesen, J.A. Dalmon, Catalysis
Today, 82, 1-4, 75-81 (2003).
223. I.H. Kumakiri, P. Jorunn, A. Thijs, A.G. Melbye, H. Rder, Journal of Petroleum Science
and Engineering, 79, 1-2, 37-44 (2011).
224. M. Gutiérrez, P. Pina M.Torres, M.A.Cauqui, J. Herguido, Catalysis Today, 149, 3-4, 326-
333 (2010).
225. A.D. R. Pashkova, N. Kaltenborn, H. Richter, Chemical Engineering Journal, 165, 3, 924-
933 (2010).
226. A. Julbe, D. Farrusseng, C. Guizard, Industrial Ceramics, 20, 1, 18-21 (2000).
227. S.K. Ryi, S.H. Kim, J.S. Park, S.H. Choi, S.H. Cho, Chemical Engineering Journal, 113, 1,
47-53 (2005).
228. V.T. Zaspalis, W. van Praag, K. Keizer, J.G. van Ommen, J.R.H. Ross, A.J. Burggraaf,
Applied Catalysis, 74, 2, 249-260 (1991).
229. A.A. Khassin, A.G. Sipatrov, T.M. Yurieva, G.K. Chermashentseva, N.A. Rudina, V.N.
Parmon, Catalysis Today, 105, 362-366 (2005).
89 | P a g e

230. A. Seidel-Morgenstern, Membrane Reactors: Distributing Reactants to Improve Selectivity
and Yield, Wiley-VCH, Germany,2010.
231. V.I. Bukhtiyarov, A.I. Boronin, V.I. Savchenko, Journal of Catalysis, 150, 262-267 (1994).
232. W. Jin, X. Gu, S. Li, P. Huang, N. Xu, J. Shi, Chemical Engineering Science, 55, 2617-
2625 (2000).
233. C. Delbos, G. Lebain, N. Richet, C. Bertail, Catalysis Today, 156, 3-4, 146-152 (2010).
234. X. Tan, Z. Pang, Z. Gu, S. Liu, Journal of Membrane Science, 302, 109–114 (2007).
235. J.A. Hugill, F.W.A. Tillemans, J.W. Dijkstra, S. Spoelstra, Applied Thermal Engineering,
25, 1259 – 1271 (2005).
236. A. Kleinert, A. Feldhoff, T. Schiestel, J. Caro, Catalysis Today, 118, 44 – 51 (2006).
237. W. Zhang, J. Smit, M. van Sint Annaland, J.A.M. Kuipers, Journal of Membrane Science,
291, 19 – 32 (2007).
238. X. Yin, L. Hong, Z. Liu, Journal of Membrane Science, 311, 89 – 97 (2008).
239. R. J. Copeland, Y. Gershanovich, B. Windecker, Final Report: for US Department of
Energy, 2005.
240. D. Milne, T. Seodigeng, D. Glasser, D. Hildebrandt, B. Hausberger, Catalysis Today,156, 3-
4, 237-245 (2010).
241. H. Wang, Y. Cong, X. Zhu, W. Yang, Reaction Kinetics and Catalysis Letters, 79, 2, 351-
356 (2003).
242. C. Téllez, M. Menéndez, J. Santamaría, Chemical Engineering Science, 54, 13-14, 2917-
2925 (1999).
243. S. Assabumrungrat, T. Rienchalanusarn, P. Praserthdam, S. Goto, Chemical Engineering
Journal, 85, 1, 69-79 (2002).
90 | P a g e

244. H. Wang, Y. Cong, W. Yang, Catalysis Today, 104, 160-167 (2005).
245. Y.T. Chua, A.R. Mohamed, S. Bhatia, Applied Catalysis: A General, 343, 142-148 (2008).
246. S. Ji, T. Xiao, S. Li, C. Xu, R. Hou, K.S. Coleman, M.L.H. Green, Applied Catalysis: A
General, 225, 271-284 (2002).
247. L. Greiner, S. Laue, J. Woltinger, A. Lies, Topics in Current Chemistry, 276, 111-124
(2007).
248. J. Peureuxa, M. Torresb, H. Mozzanegaa, A. Giroir-Fendlera, J.A. Dalmon, Catalysis
Today, 25, 3-4,409-415 (1995).
249. S.S.E.H. Elnashaie, T. Moustafa, T. Alsoudani, S.S. Elshishini, Computers & Chemical
Engineering, 24, 2-7, 1293-300 (2000)
250. H. Jiang, Z. Cao, S. Schirrmeister, T. Schiestel, J. Caro, Angewandte Chemie- International
Edition, 49, 33, 5656-5660 (2010).
251. S.T. Leonard, S. Miachon, D. Vanhove, Recents Progres en Genie des Procedes, 89, 226-
231 (2003).
252. D. Vanhove, S. Leonard, Recents Progres en Genie des Procedes, 89, 194-198 (2003).
253. L. Guillou, S. Leonard, V. Le Courtois, E. Payen, D. Vanhove, Proceedings of the ICCMR-
6, Lahnstein, 6–9 July, 2004.
254. M.H. Khademi, M.R. Rahimpour, A. Jahanmiri, International Journal of Hydrogen Energy,
35, 5, 1936-1950 (2010).
255. M.H. Khademi, P. Setoodeh, M.R. Rahimpour, A. Jahanmiri, International Journal of
Hydrogen Energy, 34, 16, 6930-6944 (2009).
256. M.H. Khademi, M.R. Rahimpour, A. Jahanmiri, Chemical Engineering Communications,
199, 7, 889-911 (2012).
91 | P a g e

257. J. Lu, Y. Yu, J. Zhou, L. Song, X. Hu, A. Larbot, Applied Surface Science, 255,9092-9099
(2009).
258. Ch. Ch. Wei , K. Li, Industrial & Engineering Chemistry Research, 48, 3446–3452 (2009)
259. C. Picard, A. Larbot, F. Guida-Pietrasanta, B. Boutevin, A. Ratsimihey, Separation and
Purification Technology 25, 65–69 (2001).
92 | P a g e

Chapter 4. Membrane Reactor Mathematical Modeling
4.1. Scope of Simulation
As it was mentioned in chapter 3, in our proposed membrane reactor the tubular
membrane reactor concept is counter-current flow mode. This membrane-base configuration can
have several advantages; for example having much higher surface area per unit contactor
volume, and having an independent flow for both gas and liquid phase. Also membrane reactor
configuration can increase conversion for the reactions which are equilibrium-limited similar to
methanol synthesis. Some applications of hollow fiber membrane contactors have been reviewed
by others. Membrane distillation, a thermally driven process for transferring a solvent across a
non-wetted membrane, was reviewed by Sirkar [1]. Membrane perstraction also, a process where
one or more components of a liquid feed solution diffuse across a non-porous membrane into a
stripping or sweep liquid, was discussed by Sirkar [2].
The main challenge usually in designing and operating these devices is to maximize the
mass transfer rate by producing as much interfacial area as possible. Due to membrane material
and properties, theoretically membrane can be filled with gas, which results in non-wetting
model, or can be filled with liquid, which results in wetted (or overall wetted) model.
Ideally in the gas absorption process pores of the membrane should be completely gas-
filled, to minimize any mass transfer resistance due to the presence of the membrane. Therefore
the membrane itself usually does not have any selectivity for the gases to be separated. One of
the most important factors for absorbent selection is the surface tension of the liquid absorbent.
Although the membrane used for gas absorption is generally hydrophobic microporous,
absorbent solutions with low surface tension can penetrate inside the membrane pores and cause
the membrane wetting gradually with time.
93 | P a g e

4.2. Equations and Boundary Conditions
4.2.1. Equations and Boundary Conditions
The reactor feed side (indicated by F superscript) and the permeate side (indicated by P
superscript) are described by the following mass balance equations

∑

(4-1)

(4-2)
or in expanded form:

(4-3)

(4-4)

(4-5)

(4-6)

(4-7)

(4-8)

(4-9)

(4-10)
94 | P a g e

(4-11)

(4-12)

(4-13)

(4-14)
together with the following boundary conditions at z=0:
 
22
22
22
33
0, ,
0, ,
0, ,
0, ,
0, ,
0, ,
(4 15)
(4 16)
(4 17)
0 (4 18)
0
0 (4
1
)
4
0
9
2
F feed F
CO CO
F feed F
CO CO
F feed F
HH
F feed F
H O H O
F feed F
CH OH CH OH
F feed F
TGDE TGDE
FF
FF
FF
FF
FF
FF










2
2
2
3
0,
0,
0,
0,
0,
0,
0 (4 21)
0 (4 22)
0 (4 23)
0 (4 24)
0 (4 25)
(4 26)
P
CO
P
CO
P
H
P
HO
P
CH OH
P feed
TGDE TGDE
F
F
F
F
F
FF







where

(m
2
/m
3
,

) is the membrane geometric area per unit volume on
the feed-side,

(m
2
/m
3
,

) is the membrane geometric area per unit volume on the
permeate-side.

(mol/s), (i= CO,CO
2
, H
2
,H
2
O, CH
3
OH, TGDE) is the molar flow rate of each
component in the feed-side and

(mol/s), (i= CO,CO
2
, H
2
,H
2
O, CH
3
OH, TGDE) is the molar
flow rate of each component in the permeate-side, r (m) is the membrane thickness.

is the flux (mol/m
2
s) of species i at the outer membrane side (r
membrane out
)
and

membrane is the flux (mol/m
2
s) of species i at the inner (r
membrane
) membrane side.
To describe the transport of each component through the part of the membrane that is
occupied by the gas phase, the Dusty Gas Model (DGM) is used. The diffusion process is due to
95 | P a g e

two regimes: molecule–wall interactions (Kundson diffusion), which is more prevalent in small
pores and molecule-molecule interactions (molecular diffusion) with D
ij
and

as diffusion
coefficients that can be calculated. The DGM, is described by the following equation (4-27):
∑

(

)

in which;

√

(m
2
s
-1
) (4-28)

(4-29)
and

=

(4-30)
x is the mole fraction of the components, is the total pressure gradient,

is the component
partial pressure gradient, R is the gas constant (J/mole.K), d
p
the average pore size of the porous
layer, ε is the porosity, τ is the tortuosity,

is atomic diffusion volume, M
i
is molecular weight
(g) , T and P are the operating temperature (K) and pressure (bar) relatively, B
0
is permeability
coefficient (m
2
), and µ is dynamic viscosity (µP) which has been calculated using Corresponding
States Methods [3]. In case that the pressure difference over the part of the membrane that is
occupied by gas is absent (which is the assumption we make in this study), equation (4-27) will
change into equation (4-31) as follows:
∑

Concentration profile for different components in the gas filled membrane section then would be
as follows:
96 | P a g e

∑

and to calculate fluxes, we need to calculate the concentration profiles using steady-state form of
equation of continuity:

Having six unknown component partial pressures/concentrations and six unknown fluxes (

,
i=CO, CO
2
, H
2
O, H
2
, CH
3
OH,TGDE) as well, twelve differential equations are needed to be
solved. The system of differential equations in the gas filled section of the membrane (equations
32-43) would be:

For each component bulk term in gas filled section of the membrane (equations 34-a- to 39-a)
would be:
97 | P a g e

1 3 1 5 5 1 1 6 2 5 1 2 1 4 1 1 4
3 2 2 1 5 1 2 2 1 4 2 2 4
31 2
1,2 1,3 1,4 1,5 2,5
2,1 2,3 2,4
(1) (34 )
(2)
g mg g mg g mg g mg g mg g mg g mg g mg g mg g mg
e e e e e
g mg g mg g g mg g mg g mg g mg
e e e
y N y N y N y N y N y N y N y N y N y N
Bulkterm a
D D D D D
y N y N y y N y N y N y N
Bulkterm
D D D
   
     
 
   
2 2 5 2 2 6
1 3 3 1 2 1 3 2 4 1 3 4 5 1 3 5 6 3
1 4 4 1
6
2,5 2,5
36
3,1 3,2 3,4 3,5 3,5
4,1
(35 )
(3) (36 )
(4)
mg g mg g mg g mg
ee
g mg g mg g mg g mg g mg g mg g mg g mg g mg g mg
e e e e e
g mg g mg
e
N y N y N y N
a
DD
y N y N y N y N y N y N y N y N y N y N
Bulkterm a
D D D D D
y N y N y
Bulkterm
D


    
     


3 4 4 1 5 4 4 5 4 4 6 2 4 4 2
1 5 5 1 2 5 5 2 3 5 5 3 4 5 5 6 5 5
6
4,2 4,3 4,5 4,6
46
5,1 5,2 5,3 5,4
(37 )
(5)
g mg g mg g mg g mg g mg g mg g mg g mg
e e e e
g mg g mg g mg g mg g mg g mg g mg g mg g mg g mg
e e e e
y N y N y N y N y N y N N y N
a
D D D D
y N y N y N y N y N y N y N y N y N y N
Bulkterm
D D D D D
 
   
    
    
2 6 6 2 3 3 4 6 6 11
5,6
6 6 6 4 5 6 5 66
6,1 6,2 6,3 6,4 6,5
(38 )
(6) (39 )
e
g mg g mg g mg g mg g mg g mg g mg g mg g mg g mg
e e e e e
a
y N y N y N y N y N y N y N y N y N y N
Bulkterm a
D D D D D

 
     

The above equations are coupled with the following boundary conditions.
at r = r membrane-out

2 2 2 3
, , , , i CO CO H O H CH OH 
at r=r
liquid filled membrane section /gas filled membrane section interphase

2 2 2 3
, , , , i CO CO H O H CH OH 
in which

is the equilibrium concentration between gas phase and liquid phase:
2 2 2 3
*
( , , , , ) (4 48)
**
mg
equilibrium ii
i
i
p
C i CO CO H O H CH OH
RT 

  

both Φ (fugacity coefficient) and γ (activity coefficient) were calculated from SRK (Soave Redlich
Kwong) equation of state.
Using mass conservation equation, in cylindrical coordinates for tubular membranes (i.e.

(

) ), the equations describing transport through the liquid part of the membrane
are as follow:
98 | P a g e

(4-49)

(4-50)

(4-51)

(4-52)

(4-53)

(

)
(

)
(4-54)
where

is the effective diffusion coefficient (cm
2
/s) calculated using the Wilke-Chang [4]
equation (4-55) multiply by (

) , the liquid penetration through the membrane thickness (cm),
and

the concentration difference (mol/cm
3
) across the liquid filled
membrane section, and concentration for different components in the bulk liquid (permeate).

√

(4-55)
in which;

is the molar volume (cm
3
/gr.mol), is the viscosity in centipoises, is an associate
parameter for solvents (2.6 for water, 1.9 for methanol, and 1 for other components), and T is
temperature (K).
These equations (49-54) can then be solved knowing this fact that:
at r
membrane
= r
liquid filled membrane section /gas filled membrane section interphase,

99 | P a g e

(for

) (4-56)

is species concentrations (mol/m
3
) in the permeate side (equations. It was also assumed that
the molar density on the permeate side to be ρ
P
(mol/m
3
) and to be equal to the molar density of
the solvent (ρ
s
, mol/cm
3
). Then:

=

∑

(4-57)

Where
ρ
mix
= Gas/liquid mixture density at flowing pressure and temperature, kg/m3
p = Operating pressure, kPa (abs)
d
L
= Liquid relative density (water = 1, use average value for hydrocarbon-water mixtures) at
standard conditions, dimensionless;
R = Gas/liquid ratio at normal conditions;
T = Operating temperature, K
d
g
= Gas relative density (air = 1), dimensionless;
Z = Gas compressibility factor, dimensionless.
4.2.2. Results and discussion:
4.2.2.1. Effect of temperature and pressure
Figure 4.1 and 4.2 show the effect of temperature and pressure on total carbon conversion
both in membrane reactor. The membrane used in this simulation has the same characteristics
mentioned in Table 3.2 and Appendix A, and also the total carbon conversion was calculated
using equation 3-2.
100 | P a g e

Figure 4.1. W/F vs. conversion at T=220
o
C, and sweep flowrate= 1ml/min.
Methanol synthesis reactions are exothermic so that reducing the temperature has positive
effect on conversion. The total number of moles in the methanol synthesis reduces as the
reactions proceed. Hence increasing the pressure of reaction raises carbon conversion.

Figure 4.2. W/F vs. conversion at 30 bar and sweep flowrate= 1ml/min
101 | P a g e

4.2.2.2. Effect of liquid thickness in the membrane
As it was discussed before in the chapter 3, the membrane-based process can be carried
out in non-wetting, and wetting mode. In the non-wetting mode, with hydrophobic membrane,
and the absorbing liquid flowing through inside the membrane, the gas pressure has to be lower
than the liquid pressure. This fact is necessary to prevent dispersion of gas as bubbles in liquid.
In this mode liquid does not enter the pores, and therefore liquid does not wet the membrane and
membrane pores are gas filled. In this mode the gas components first diffuse through themselves
(gas phase) and then they dissolve in the liquid. The gas phase mass-transfer coefficient is much
higher than the liquid-phase mass transfer coefficient. The other mode of operation is where the
membrane pores are filled with liquid. This mode is called wetted mode. In this mode, with
hydrophilic membrane, the membrane pores are filled with liquid, and the gas pressure is higher
than liquid pressure to prevent the liquid from dispersing as drops in the gas. In this mode gas
components must diffuse in the liquid phase, and much lower mass-transfer through liquid is
expected.

Figure 4.3. The effect of liquid thickness penetrated in the pores on conversion at P=30 bar,
T=220
o
C, and sweep flowrate= 1 ml/min.
102 | P a g e

Figure 4.3 shows the effect of liquid thickness in the membrane pores. The membrane
used in this simulation has the same characteristics mentioned in Table 3.2. It is obvious from
Figure 4.3 that with non-wetted mode the conversion is much higher than the wetted-mode, in
which the liquid fills the pores and the liquid thickness is the same of the membrane diameter.
4.2.2.3. Effect of Flow Rate
The calculated total carbon conversion for feeds with different compositions at 220
o
C
and 30 bar are depicted in figure 4.4. In this figure CF=F
CO
/(F
CO
+F
CO2
) and SN=(F
H2
-
F
CO2
)/(F
CO
+F
CO2
), where, F
CO
, F
CO2
, and F
H2
are respectively the molar flow rate of carbon
monoxide, carbon dioxide, and hydrogen. The equilibrium total carbon conversions for the same
operational pressure (30 bar), temperature (220
o
C), and feed composition are also plotted. It
would be noticed in figure 4.4 that the feed with CN=0.714 and SN=3 has the highest conversion
for the entire W/F range. This is due to the higher equilibrium conversion and also the reaction
rate corresponding the aforementioned feed composition.

Fig. 4.4. Conversion vs. W/F for different feed compositions at P=30 bar, T=220oC, and sweep
flowrate= 1ml/min.
103 | P a g e

4.2.2.4. Membrane Reactor Experiment and Mathematical Model Results Comparison
To validate the model, we compared our simulation results with the experimental data
obtained and discussed in chapter 3. Figures 4.5 and 4.6 discuss the comparison between the
theoretical calculation and experimental data. The membrane used in this study is intensively
hydrophilic, and has been modified to hydrophobic membrane using the method mention in
chapter 3, section 3.2.3.1. As it can be seen in the figures, the theoretical calculation and
experimental data are in good agreement.

Figure 4.4. Comparison between the modeling and experiment: the effect of feed flow rate
20
25
30
35
40
45
7.5 12.5 17.5
Total Varbon Conversion
W/F (gr/mol/hr)
Conversion Comparison: Effect of Feed Flow Rate
PFR Modeling MR
104 | P a g e

Figure 4.5. Comparison between the modeling and experiment: the effect of feed flow rate
4.3. Conclusions:
A mathematical model was developed in order to describe and predict the application of
our proposed membrane reactor. The results from mathematical modeling and experiments were
compared to each other and comparison showed that they are in good agreement. Although the
membrane used in this study is hydrophobic, and it is expected to resist the wetting of a wetting
sweep liquid. The aqueous solution of organic compound (TGDE), can penetrate partially in to
the pores, and membrane will be gradually wetted over prolonged periods of operational time
[5]. The difference between the mathematical modeling and experimental data can be explained.

20
25
30
35
40
45
50
20 25 30 35
Total Carbon Conversion
Pressure (bar)
Conversion Comparison: Effect of Pressure
PFR Modeling MR2
105 | P a g e

References:
1. K.K. Sirkar, Other new membrane processes, in: W.S.W. Ho, K.K. Sirkar (Eds.), Membrane
Handbook, Chapman & Hall, New York, 1992, pp. 899±904.
2. K.K. Sirkar, Other new membrane processes, in: W.S.W. Ho, K.K. Sirkar (Eds.), Membrane
Handbook, Chapman & Hall, New York, 1992, pp. 904±909.
3. B.E. Poling, J.M. Prausnitz, J.P. O’Connell, “The properties of gases and liquids” McGraw-
Hill Professional, USA (2000), ISBN: 0-07011-682-2
4. C. R. Wilke, P. Chang, AIChE Journal, 1, 264–270 (1955).
5. J.-G. Lu, Y.-F. Zheng, M.-D. Cheng, Journal of Membrane Science, 308, 180-190 (2008).

106 | P a g e

Chapter 5: Suggestions for Future Work
In our work so far we studied a MR with solvent purge whose aim is the in-situ product
(methanol) removal in order to increase the product yield. In our studies we used alumina
membranes, supplied by our industrial collaborator (Media and Process Technology, Inc.) with
the nominal pore size of 100 Å used for these experiments. Since the organic solvent that was
used in this study (TGDE) was extremely wetting, the ceramic membrane was modified in order
to increase membrane hydrophobicity. This membrane based gas-liquid process, offered
independent control of gas and liquid flow rates in a compact devise, which resulted in
improvements in total carbon conversion compared to the conventional PFR reactor. The
experiments were performed at different temperature, pressures, and inlet gas flow rates.
Poirier to the membrane reactor experiments, the kinetics of the methanol synthesis reaction
from CO/CO
2
/H
2
mixtures on a commercial Cu/Al
2
O
3
/ZnO catalyst was also investigated.
As it was discussed before one important criteria that was considered to choose this
solvent was not only high boiling point as well as having high solubility for methanol compared
to other components. One important aspect that we noticed, was TGDE evaporation that was
verified using condensed liquid phase analysis which was done by GC-FID (Model **).
Both our simulations and the experimental data confirmed the vaporization of the TGDE at
various temperatures. Figure 6.1 shows the amount of TGDE in the gas phase (which eventually
would be condensed at the same time as methanol, and water in the bubblers) from the
mathematical simulations and TGDE vapor pressure curve (Figure 6.2).
107 | P a g e

Figure 5.1. TGDE evaporation rate, calculated using mathematical calculations

Our experimental data also confirmed 8%-12% of TGDE evaporation in the MR
experiments discussed in chapter 3.

Figure 5.2. TGDE Vapor pressure, calculated by PROII simulator

Solvent loss could be an important problem due to economical constrains that might
cause in two ways: first, TGDE make up stream should be added to the system that increases the
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
460 480 500 520 540
TGDE evaporation (mol/min *10^-7)
Temperature (K)
TGDE Evaporation
0
0.2
0.4
0.6
0.8
473.15 483.15 493.15 503.15 513.15 523.15 Vapor Pressure (bar)
Temperature (K)
TGDE vapor pressure
108 | P a g e

costs of operation, and second, instead of one steps of purification, two steps would be needed:
one for the methanol from the TGDE reservoir, and second from the possible condensed product.
Using a solvent with low vapor pressure, for example ionic liquids that were mentioned
in Table 3.4, could be a possible solution the TGDE vaporization problem. Ionic liquid as it is
discussed in Table 3.4, can be could candidate to be used as sweep liquid, due to the non-wetting
nature, very low vapor pressure, and having high solubility for Methanol.

Abstract (if available)

Abstract Methanol synthesis has undergone continuous improvements for nearly a century, as it represents the starting raw material for the production of a variety of other chemicals and solvents, including formaldehyde, methyl tertiary butyl ether, and acetic acid and fuel additives. Methanol has a number of advantages as a fuel and a source of chemical products, such as being more easily transportable than methane and other gaseous fuels, having a high energy density, needing no desulphurization, and its reactions (e.g., steam reforming) proceeding at moderate temperatures. Recent global energy shortages and more strict emission regulations have motivated research and development of new fuel cells, among which a direct methanol fuel cell is a prime candidate. ❧ In the present work the CO/CO₂ conversion into methanol in both a traditional reactor (TR) and a membrane reactor (MR) has been studied. The purpose of this study is to investigate the possibility of using a MR to increase the total carbon conversion into methanol relative to what a TR can convert. ❧ MR with solvent purge, which is proposed in this work, incorporates the advantages of the existing reactive separation systems that result from the in‐situ product removal. In our proposed process, we use a solvent as a sweep fluid. The membrane serves as an interface contactor for the selective permeation of methanol. The solvent is chosen in a way that the main product, methanol, has the highest solubility in it. Using TetraEthylene Glycol Dimethyl Ether as an inert, high boiling point agent, the alcohol will be selectively removed in‐situ from the reactor using the membrane as an interface contactor between the methanol and the solvent. Due to the very low solubility of H₂, CO, and CO₂, they will remain in the reactor, since the solvent blocks the membrane’s pores. ❧ Prior to the start of the membrane reactor experiments, the surface of a membrane was successfully modified in order to increase membrane hydrophobicity and, therefore, to achieve higher mass transfer rate that, in return, results in higher methanol production. The modified membrane can be operated as a membrane contactor under the proposed operating conditions without the loss of the inert solvent. ❧ The kinetics of the methanol synthesis from CO, CO₂ and H₂ on a commercial Cu/Al₂O₃/ZnO catalyst was investigated in an autoclave reactor. The effect of various CO/(CO₂+CO) ratios, stoichiometric number, pressure, temperature and flow rate was investigated in the kinetic experiments. ❧ Based on the experimental kinetics and the membrane surface modification results, we developed a model of the MR to explore the effect of several factors on the total conversion of carbon. The rate equation derived from the kinetic experiments was used in the model. Membranes with various thicknesses were simulated, and their performance was compared. Our results indicate great potential for application of membrane reactor in methanol synthesis using CO, CO₂ and H₂.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Methanol synthesis in the membrane reactor

PDF

Methanol synthesis in a contactor-type membrane reactor

PDF

The use of carbon molecule sieve and Pd membranes for conventional and reactive applications

PDF

A flow-through membrane reactor for destruction of a chemical warfare simulant

PDF

A hybrid adsorbent-membrane reactor (HAMR) system for hydrogen production

PDF

On the use of membrane reactors in biomass utilization

PDF

A high efficiency, ultra-compact process for pre-combustion CO₂ capture

PDF

Process intensification in hydrogen production via membrane-based reactive separations

PDF

Studies of transport phenomena in hydrotalcite membranes, and their use in direct methanol fuel cells

PDF

An integrated 'one-box' process for hydrogen production

PDF

Silicon carbide ceramic membranes

PDF

Development of polystyrene sulfonic acid-polyvinylidene fluoride (PSSA-PVDF) blends for direct methanol fuel cells (DMFCS)

PDF

A study of the application of membrane-based reactive separation to the carbon dioxide methanation

PDF

A process-based molecular model of nano-porous silicon carbide membranes

PDF

Novel methods for landfill gas and biogas clean-up

PDF

Preparation of polyetherimide nanoparticles by electrospray drying, and their use in the preparation of mixed-matrix carbon molecular-sieve (CMS) membranes

PDF

The purification of contaminated air streams via recative separation techniques

PDF

Chemical and mechanical deformation of porous media and materials during adsorption and fluid flow

PDF

Fabrication of nanoporous silicon carbide membranes for gas separation applications

PDF

In situ studies of the thermal evolution of the structure and sorption properties of Mg-Al-CO3 layered double hydroxide

Asset Metadata

Creator Soltani, Sahar (author)

Core Title Methanol synthesis in a membrane reactor

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Chemical Engineering

Publication Date 07/16/2014

Defense Date 04/18/2014

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag membrane reactor,methanol,OAI-PMH Harvest,synthesis

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Tsotsis, Theodore T. (committee chair), Prakash, G. K. Surya (committee member), Sahimi, Muhammad (committee member)

Creator Email sahars81@gmail.com,ssoltani@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-444631

Unique identifier UC11287775

Identifier etd-SoltaniSah-2709.pdf (filename),usctheses-c3-444631 (legacy record id)

Legacy Identifier etd-SoltaniSah-2709.pdf

Dmrecord 444631

Document Type Dissertation

Format application/pdf (imt)

Rights Soltani, Sahar

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA