Вы находитесь на странице: 1из 19

DOI: 10.1002/cssc.

201100447

A Comparative Review of Petroleum-Based and Bio-Based


Acrolein Production
Lu Liu,*[a] X. Philip Ye,*[a] and Joseph J. Bozell[b]

1162

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemSusChem 2012, 5, 1162 1180

Acrolein is an important chemical intermediate for many


common industrial chemicals, leading to an array of useful end
products. This paper reviews all the synthetic methods, including the former (aldol condensation) and contemporary (partial
oxidation of propylene) manufacturing methods, the partial oxidation of propane, and most importantly, the bio-based glycerol-dehydration route. Emphasis is placed on the petroleum-

1. Introduction
1.1. Acroleinan important and versatile intermediate
Acrolein (IUPAC name: prop-2-enal), the simplest unsaturated
aldehyde, is a colorless liquid (b.p. 53 8C) at room temperature
with a pungent and piercing odor. The conjugation of the carbonyl group with a vinyl group provides acrolein with a high
degree of reactivity. Acrolein is a critical intermediate for the
chemical industry and is converted to acrylic acid and its
esters, glutaraldehyde, methionine, polyurethanes and polyester resins (Figure 1).

Figure 1. Chemicals derived from acrolein.[14]

Acrolein derivatives offer high industrial values. Acrylic acid


is a valuable commodity chemical with a lucrative market.
Polyacrylic acid can be formulated into super-absorbent materials for diapers, acrylate plastics, and paints and coatings. As
one of the fastest growing commodity chemicals, acrylic acid
is in increasingly higher demand with an annual global growth
rate of 4 %. A shortage of acrylic acid supply was reported in
2010 due to the closure of some plants despite the growing
demand.
Although the majority of crude acrolein goes to the manufacture of acrylic acid, the major percentage (e.g., approximately 80 % in 2002[1]) of refined acrolein is consumed in the
synthesis of methionine. Methionine is a sulfur-containing
amino acid that is required by animals, but is not readily found
ChemSusChem 2012, 5, 1162 1180

based route from propylene and the bio-based route from


glycerol, an abundantly available and relatively inexpensive
raw material available from biodiesel production. This review
provides technical details and incentives for industrial proyduction that justify a transition toward bio-based acrolein
production.

in natural food sources. Methionine production capacity has


been expanding steadily since 1977. Currently, the worlds largest methionine facility, built in 2006, is in Antwerp, Belgium,
with an annual capacity of 120 000 tons. The global demand
for methionine is approximately 450 000500 000 tons per year,
with a projected annual growth rate of over 5 %, especially in
Asia.[3]
The remainder of manufactured acrolein is used in the production of glutaraldehyde, 1,2,6-hexanetriol, quinoline, pentaerythritol, cycloaliphatic epoxy resins, oil-well derivatives, and
water treatment chemicals.[4] Also, acrolein can be directly
used as an effective aquatic biocide by injection into water to
control the growth of undesired
microbial material and aquatic
weeds.[5]
Given the growing importance
of this chemical, relatively few
reviews of acrolein production
have appeared.[5, 6] Both of these
reviews focused solely on one
synthesis pathway. This review
provides information associated
with all the synthesis methods
and uses the history of the industrial development itself to rationalize the bio-based acrolein
production. The discussion will
be focused on the fully developed manufacturing (petroleumbased) method and the development of bio-based method. An
attempt of economic evaluation
is made for the first time.

[a] Dr. L. Liu, Dr. X. P. Ye


Department of Biosystems Engineering & Soil Science
University of Tennessee-Knoxville
2506 E.J. Chapman Drive, Knoxville, Tennessee (USA)
Fax: (+ 1) 865-974-4514
E-mail: lliu11@utk.edu
xye2@utk.edu
[b] Dr. J. J. Bozell
Center for Renewable Carbon, Biomass Chemistry Laboratories
University of Tennessee-Knoxville
2506 Jacob Drive, Knoxville, Tennessee (USA)

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemsuschem.org

1163

L. Liu, X. P. Ye, and J. Bozell


1.2. Overview of acrolein synthesis methods
There has always been a demand for acrolein in industry.
Efforts and improvements in the production of acrolein have
been made continuously over several decades (Figure 2). Acrolein can be synthesized by catalytic aldol condensation of acetaldehyde and formaldehyde, which was also the first industrial

Lu Liu recently completed her Ph.D. at


the University of Tennessee with a dissertation entitled Roles of non-thermal plasma in gas-phase glycerol dehydration catalyzed by supported silicotungstic acid. She believes in the
importance of building a supportive
chemical platform to make the renewable energy industry economically
more viable, and seeks to apply her
chemical engineering background to
this effort. Her research interests lie in
developing efficient catalytic routes for value-added production
from renewable resources and transitioning catalytic processes
from the laboratory to the industrial scale.
X. Philip Ye received his Ph.D. from the
University of Minnesota, USA, in 2004.
Prior to that, he worked as an engineer
in a chemical company in China for six
years. Since 2004, he has been an assistant professor and then associate
professor the University of Tennessee.
Working with an interdisciplinary team
with a background in Biosystems/Agricultural Engineering, Chemical Engineering, Applied Chemistry, and Biotechnology, his current research focuses on the interface between biological systems and engineering
processes to fill the void of engineering science in the transformation of biorenewables into fuels, chemicals, and materials, and to
protect our environment.

process.[7] Partial oxidation of propylene catalyzed by multicomponent metal catalysts later replaced the aldol condensation and remains to be the primary manufacturing method.
Acrolein can be produced from partial oxidation of propane,
although it is more difficult to control selectivity and the yield
has not yet reached a commercially viable level. Several other
petroleum-based synthetic routes have been used successfully
to synthesize acrolein. Ai et al. synthesized acrolein through reaction of formaldehyde and ethanol at 240320 8C using nickel
phosphate or silica-supported metal (tungsten, zinc, nickel, or
magnesium) oxides.[8] Acrolein can also be obtained by the oxidation of allyl alcohol,[9, 10] decomposition of allyl ether[11] or
partial oxidation of ethane.[12] However, these routes have significant drawbacks for application on a larger scale, including
low availability and high cost of the reactants, high energy
consumption, or low selectivity.
Acrolein can also be produced through biological routes. As
early as the 1900s, acrolein formation was encountered during
alcoholic fermentation from grains with certain bacterial
strains, such as Bacillus amaracrylus, Clostridium welchii, and
Lactobacillus.[1315] In the case of alcoholic fermentation, acrolein was an undesirable compound, and the target of these studies had always been to minimize or prevent the associated
acrolein production. Only recently has research started to investigate the possibility of using this undesired route for possible chemical production. Vollenweider et al. gave a clear
statement of the potential of the solely biological route of
acrolein production.[16] However, bacterial methods have not
been viable at a production level.
The appearance of acrolein during glycerol distillation,[11]
where glycerol pyrolysis occurs, was probably the prototype of
the double dehydration process for producing acrolein from
glycerol. Acrolein formation from glycerol becomes faster and
considerably more selective in the presence of an acid, compared to pyrolysis alone at the same temperature. Acrolein
produced from glycerol is now categorized as bio-based acrolein, because glycerol is largely available as a coproduct of biodiesel production. With the advent of improved acid catalysts,
the production of bio-based acrolein is increasingly appealing
for industrial manufacture.
1.3. Going toward the bio-based route

Joseph J. Bozell joined the Tennessee


Center for Renewable Carbon in 2006.
Prof. Bozell received his B.Sc. in
chemistry from South Dakota State
University, and a Ph.D. in organic synthesis and organometallic chemistry
from Colorado State University. Before
joining the University of Tennessee, he
did postdoctoral work at the DOEs National Renewable Energy Laboratory.
Prof. Bozells primary interests are in
using the tools of organic synthesis for
converting renewable materials into
chemical products.

1164

www.chemsuschem.org

Before justifying the transition towards the bio-based route, it


is necessary to examine the history of the development of
acrolein production. Degussa launched the first industrial acrolein production in 1942, producing acrolein through aldol condensation of acetaldehyde and formaldehyde [Eq. (1)].[7] Aldol
condensation remained to be the standard industrial practice
into the late 1950s, but problems with incomplete conversion
complicated the downstream separation, as the product and
reactants had similar physical properties. As inexpensive propylene became available in the middle of the 20th century,
partial oxidation of propylene to produce acrolein [Eq. (2)] attracted the attention of industry and research. In 1958, Shell
started the first pilot plant operation of gas-phase propylene
partial oxidation. Later, SOHIO (Standard Oil of Ohio) discov 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemSusChem 2012, 5, 1162 1180

Comparison of Petroleum-Based and Bio-Based Acrolein Production

Figure 2. Summary of the synthesis methods of acrolein.

ered bismuth molybdate catalysts for high-output and cost-effective acrolein production. The partial oxidation of propylene
has been the predominant manufacturing method worldwide
ever since. In parallel, acrolein production from propane continues to be of interest [Eq. (3)] because of propanes considerably lower cost.[17] Propane is the byproduct of both natural
gas production and oil refining, but is only consumed commercially as fuel. In contrast, the carboncarbon double bond of
propylene makes it an important intermediate for the production of a number of value-added chemicals, such as polypropylene, acrylonitrile, propylene oxide, and isopropyl alcohol. This
continual demand in many processes has driven up the propylene price, making development of the route from propane
more attractive. However, despite decades of work on acrolein
synthesis from propane, the yield of acrolein prepared through
partial oxidation of propane has not reached a point of commercial viability.
CH2 O CH3 CHO ! CH2 CHCHO H2 O

CH2 CH3 CHO O2 ! CH2 CHCHO H2 O

H3 CCH2 CH3 3=2 O2 ! H2 CCHCHO 2 H2 O

It is not difficult to find that an inexpensive source of raw


materials usually provides the primary momentum for chemical
process development. The availability and sustainability of the
raw materials are an important consideration. With the continuously growing problem of the non-sustainability of petroleum, a synthesis pathway from a renewable resource is considerably more desirable.
The worlds energy and raw material demands keep increasing, imposing a large burden on the fossil fuel resources. Unfortunately, fossil fuel resources are not sustainable and are
very likely to be mostly depleted in a few decades. Also, the
significant amount of CO2 emission from fossil fuel usage is
also increasing steadily, which might cause irreversible damage
to the environment and climate. One solution to these problems is to utilize renewable resources and gradually transfer
ChemSusChem 2012, 5, 1162 1180

the global economy toward


a more sustainable future. Biodiesel is a good example of such
a transformation; due to the
simple operation, commercial
production has already been
launched on a small scale all
over the world. Biodiesel production and production capacity
worldwide are increasing every
year owing to the regulatory
and socioeconomic pressure of
renewable energy. One mol of
glycerol (an acrolein precursor) is
generated along with each mole
of triglyceride converted to biodiesel. As the result of the
growth of biodiesel, large
amounts of glycerol are produced and are, therefore, available to the marketplace
(Figure 3). Various sources[1517] have reported that the price of
glycerol has been lowered by its large availability, and even
credits are given for reselling the crude glycerol.

Figure 3. Annual worldwide glycerol production resulting from biodiesel


production (19912009). The values were calculated based on the stoichiometric relation between glycerol and biodiesel, the annual production of
which were taken from various sources.[1821] The density of biodiesel used
was 0.88 kg L1.

Because of the excess of glycerol, acrolein, one of the valueadded chemical derivatives of glycerol [Eq. (4)], has attracted
a great deal of attention in recent years. In a recent development of an acid-catalyzed dehydration, a good acrolein yield
was achieved. Hence, there is considerable potential for using
this method as a substitutive or complementary method to
the propylene oxidation route in the future.

H2 COHCHOHCH2 OH ! CH2 CHCHO 2 H2 O

www.chemsuschem.org

1165

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

L. Liu, X. P. Ye, and J. Bozell

2. Technical Aspects of Acrolein Production


2.1. Partial oxidation of propylene
Prior to reviewing the development of a bio-based acrolein
synthesis, it is necessary to briefly review the technical aspects
of the current manufacturing method: partial oxidation of propylene to acrolein [Eq. (2)].
This reaction is usually operated at 250400 8C in a multitubular reactor composed of tubes with diameters of about
2.5 cm and lengths of 35 m containing packed beds of multicomponent metal catalysts.[22] The multi-component metal catalysts for the partial oxidation of propylene usually can be expressed in a general formula, MoBiXIIXIII (XIYZ)O, where
MoBiXIIXIII are the basic structures containing four necessary
metal components: Mo, Bi, XII (Co, Ni, Me, Mg, Pb), and XIII (Fe,
Cr, Ce, Al). Bismuth molybdates are located on the surface of
the catalyst particles and serve as the major active sites; however, the highest activity occurs only with the simultaneous
presence of divalent (e.g., Co2 + ) and trivalent (e.g., Fe3 + ) metal
components. Optionally, the (XIYZ)O component contains
XI as one of the alkaline metals; Y as Sb, Te, W, V, or Nb; and Z
as P or B. Steam is commonly cofed with the reactant gases in
the industrial process due to its benefits to the acrolein production. Water competitively occupies the strong oxidation
sites, preventing over-oxidation of acrolein to CO2.[23] The rate
of acrolein formation increases as a result.[24, 25] Also, water possibly creates new sites for propylene adsorption on the catalyst
surface, thus enhancing the partial oxidation. The service life of

these industrial catalysts is usually around two to five years. As


one of the classical petrochemical processes, partial oxidation
of propylene has undergone continuing industrial improvement in the catalysts and processes since 1960. Table 1 summarizes the most promising disclosures in chronological order.
It is generally agreed that partial oxidization of propylene to
acrolein is initiated by the coordination of the propylene C=C
double-bond to the BiMo site on the catalyst, leading to the
formation of a Mo-bound allyl intermediate surface species
(Figure 4). This is the rate-determining step, which can occur at
either the carbonyl site on the Bi atom (Figure 4 A), or at the
bridging O atom (Figure 4 B), depending on the nature of the
specific catalysts involved.[50] The CO bond is then formed
through O insertion, followed by a second hydrogen abstraction. Generally, the lattice oxygen atom on the catalyst surface
contributes to the specific selectivity, whereas the adsorbed
oxygen radical/molecule contributes to the total oxidation,
leading to COx.[51] The gaseous molecular oxygen supplied to
the reaction serves to reoxidize the catalyst and assures sufficient availability of lattice oxygen atoms.[52] COx could also be
formed from both propylene and acrolein precursors due to
the rather large mobility of the lattice oxygen atom. Therefore,
the mobility of the lattice oxygen, determined by the structure
of catalyst, is critical to both propylene conversion and acrolein
selectivity.[53]
Gas-phase partial oxidation of propylene to acrolein is highly
exothermic.[25] As a result, one of the major issues for the reactor design of acrolein production is how to efficiently release
the heat generated during the reaction. Heat release is essen-

Table 1. Summary of acrolein production by partial oxidation of propylene.


Catalyst

Feed composition[a]

Contact time
[sec]

T
[8C]

Conversion
[%]

Selectivity[b]
[%]

Mo5VW1.5Te0.5Sb2Sn/Bi oxide[26]
Te-V-W-As/Sn-(Sb) oxide[27]
Sn-Sb-U oxide[28]
SnSb2VFe0.5Bi0.01 oxide[29]
Cr2Al0.7Mo7Te0.35 oxide[30]
Ni10Co0.3FeBiPMo12O57[31]
Ni4.5Co4FeBiP0.08Mo12O31[32]
NiCo3Fe2BiAs1.5K0.2Mo12O48.35[33]
NiCo3Fe2BiP2K0.2Mo12O49.6[34]
Mo12Bi0.17Mg0.512Fe0.17Mn05O2580[35]
Ni10.5FeBiMo2O54[36]
Mo-Bi-Fe-Mn-X oxide[37][e]
Mo-Co-Fe-Bi-X-O[38][f]
B-W-Co-Bi-Fe-Mo-Si-M-O[39][g]
Mo10Co8Fe2Bi0.9V0.05K0.05 oxide[40]
Mo10BiFe2Co8Zr0.1Ca0.1Ti0.2 oxide[41]
Mo12Te2Sb2Zr0.5Zn0.5V0.5 oxide[41]
Co4BiFeW2Mo10Si1.35K0.06 oxide[42]
Co-K-Mo-W-(P) oxide[43]
Mo12W0.3BiFeZn0.1Co4.5K0.06Al0.1Si5Ox[4446]
Mo12.25Bi1Fe3Co8Cs0.1[47]
Bi-Mo-Fe-Ni-Si oxide[48]
Mo-Bi-Fe-Co-(Ni)-X oxide[49][h]

I/O/steam = 1:1.6:6
I/O/steam = 1:6:3
I/O/N = 5:7.5:87.5
I/O/N/steam = 1:12:67:15
I/air/steam = 4:50:46
I/air/steam = 1:10:4.1
I/air/steam = 1:10:6
I/air/steam = 1:10:5
I/air/steam = 1:10:5
I/air/steam = 4.5:53:42.5
I/air/steam = 1:10:6
I/air/steam = 6:42.8:51.2
I/air/steam = 1:12:7
I/air/steam = 1:14:10
I/air/steam = 1:10:4
I/air/steam = 1:10:4
I/air/steam = 1:10:4
I/O/steam:N = 7:12.6:10:70.4
I/air/steam = 1:10:8
I/O = 1:0.53
I/O/steam = 110:3-20:7096
I/O/N = 1:120:120
I/O/steam:N = 1:1.6:2.3:4.9

3
4.9
3.5
3
4
1.8
4
1.5
1.5
9
4
2.4
2.4
6
1.6
1.6
1.6
2.25
25
N.A.[c]
N.A.[c]
3
N.A.[c]

450
370
430
270
384
360
310
305
305
370
270
340
350
330350
330
310
350
320
350
310
360
360
310

66.5
65.6
89
62.3
95
90
95.5
94
96
94.8
95.80
91.30
90.80
99.60
6392
98.6
98.3
96.20
84.986.9
99.3099.5
96.00
87.30
> 97

N.A.[c]
69.9
81
72.8
83.3
93.1[d]
82.8
93
88
73.7
69.90
80.40
81.50
73.70
7095
91.8
88.3
86.60
83.187.7
90.1092.30
88.75
96.50
87.63

[a] Molar ratio is used (mol/mol); I = propylene. [b] Selectivity is defined as the molar ratio between the carbon in the acrolein produced in the reaction
and the carbon in the converted propylene. [c] Not available. [d] Acrolein +acrylic acid. [e] X = K, Rb, and/or Cs. [f] X = Sn or Sn with Al, Ni, W, Cr, In, Nb.
[g] M = alkali metal. [h] X = Cs and/or K.

1166

www.chemsuschem.org

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemSusChem 2012, 5, 1162 1180

Comparison of Petroleum-Based and Bio-Based Acrolein Production

Figure 4. Mechanisms of the formation of allyl intermediates (initial/rate-determining steps of the partial oxidation
of propylene).[53, 54]

tial to maintain isothermal reaction conditions and to prevent


temperature runoff and explosion. Patience and Mill described
the potential utilization of a circulating fluidized bed reactor
and reported the initial modeling and assessment of its commercial scale application.[55] Leib et al. later used a neural network model to simulate the performance of the fluidized bed
reactor, providing more industrial application flexibility.[56] The
circulating fluidized bed reactor allowed the usage of higher
propylene concentrations with lower restrictions of the upper
explosion limit, gave rise to a higher heat and mass transfer
rate, higher turndown ratio, and simplified catalyst addition
and removal. However, the requirement of a high attrition resistant catalyst and insufficient fundamental kinetic and transport information has hindered its commercialization. Song
et al. improved the multi-tubular reactor by employing the
heat transfer effect of CO2 during the reactor operation and increased the productivity by 14 %. Heat transfer effects caused
by the inert gas were studied both theoretically and experimentally. The addition of CO2 provided a larger safe range of
reactant inlet and coolant temperatures.[57, 58] Redingshofer
et al. introduced a catalytic wall reactor, for which a specific
temperature could be considerably better maintained during
operation.[25, 59] They suggested using this type of wall reactor
to conduct kinetic and mechanistic studies of highly exothermic reactions to achieve more reliable results. ONeil et al. developed a distributive membrane reactor, which could increase
the acrolein yield and suppress COx production by lowering
the partial pressure of oxygen.[60]

2.2. Bio-based production of acrolein through glycerol dehydration


As an alternative to partial oxidation of propylene, a bio-based
route to acrolein using glycerol dehydration has more recently
become a promising approach for acrolein production due to
sufficient availability of glycerol from biodiesel production (4).
ChemSusChem 2012, 5, 1162 1180

Acid catalyzed dehydration of


glycerol is thought to proceed
through the mechanism shown
in Figure 5, and proper choice of
the acid catalyst can greatly improve glycerol conversion and
acrolein selectivity. The added
acid initiates the formation of
carbocations (I and II) as shown
in Figure 5 A and 5 B. The surface/pore structure (if it is solid
acid catalysts) and the acidity
have some influence on the final
ratio of the two carbocations.
Nevertheless, the 28 carbocation
would always be predominant,
as it is thermodynamically much
more stable than the 18 carbo-

Figure 5. Glycerol dehydration to acrolein and acetol.

cation and its formation rate is faster. The 28 carbocation undergoes loss of a proton to form an intermediate enol, which
undergoes subsequent tautomerization to form 3-hydroxypropanal. A second dehydration of 3-hydroxypropanal affords
acrolein (Figure 5 A1). Similarly, the 18 carbocation can form
acetol through formation of the less stable enol (Figure 5 B1).
In addition to intramolecular enol formation, intermolecular
condensation to form glycerol oligomers (Figure 6) can occur
and is often favored at lower reaction temperatures.[61]

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemsuschem.org

1167

L. Liu, X. P. Ye, and J. Bozell


300390 8C and the pressure
range of 2534 MPa, was obtained by reaction of 1wt % glycerol at 360 8C and 25 MPa in the
presence of 470 ppm ZnSO4.
They also found that the residence time was important under
subcritical conditions, but not to
the same degree under supercritical conditions. They believed
that the reaction proceeded
through an ionic pathway (as
shown in Figure 5) under subcritical conditions, and through
a radical pathway (Figure 7)
under supercritical conditions.

Figure 6. Intermolecular condensation of glycerol.

Even with the presence of the acid, the activation energy for
these dehydrations is still considerable. As a result, the studies
of converting glycerol to acrolein conducted in aqueous phase
under atmospheric pressure generally fail to achieve a useful
acrolein yield even in the presence of mineral acids or mineral
salts.[6264] Two different general approaches have been examined to improve the reactivity: 1) increasing the boiling point
of the reactant either by pressurizing the reaction vessel to
conduct the reactions under sub- or supercritical conditions or
by mixing the reactant in a reaction medium with a high boiling point; 2) increasing the reaction temperature and conducting the reaction in the gas phase.
2.2.1. Liquid-phase reaction
In 1987, Ramayya et al. conducted a liquid-phase glycerol dehydration under high pressure using 0.005 m H2SO4 as the catalyst and a 0.5 m glycerol solution as the reactant.[65] Nearly
100 mol % acrolein selectivity was achieved, but glycerol conversion was low (< 40 %). At a fixed pressure of 34.5 MPa, they
found that a higher temperature favored glycerol conversion.
In 2001, Bhler et al. studied the reactions of glycerol under
sub-/supercritical water conditions. This study covered the
temperature range of 349475 8C at 25, 35, and 45 MPa and
residence times ranging from 32165 s.[66] Clearly, high temperature favors the gas-phase products such as H2, CO, and CO2,
and the production was more enhanced at temperatures
above 450 8C. Longer resident times, lower temperature, and
higher pressures seemed preferable for acrolein production;
however, strong trends of the acrolein yield in terms of these
conditional parameters were not observed. Without the addition of any acid, the selectivity to acrolein was low (6.6
37.5 mol %).
Reactor corrosion was found to be a problem under sub-/supercritical water conditions, and was more severe with the
usage of mineral acids or ZnCl2 as catalysts.[66] ZnSO4 is usually
considerably less corrosive, which justified the usage of ZnSO4
as the acid catalyst by Ott et al.[66] They found that under subcritical conditions, glycerol conversion increased as the quantity of ZnSO4 increased. The best result in their study (75 mol %
selectivity and 50 % conversion) over the temperature range of

1168

www.chemsuschem.org

Figure 7. Radical pathway of glycerol conversion to acrolein, acetol, acetaldehyde, and formaldehyde.[68]

As a result, the reaction conducted under subcritical conditions


was more selective for acrolein production. However, Watanabe and co-workers held a different opinion, claiming that supercritical conditions were better for acrolein production.[67]
The glycerol conversion obtained at temperatures above the
critical point (400 8C) was significantly higher than that obtained at the temperatures below the critical point (300 and
350 8C). In their study, results at various temperatures (300,
350, and 400 8C), pressures (25, 30, and 34.5 MPa), glycerol
concentrations (0.25 and 0.05 mol L1), and H2SO4 concentrations (1 and 5 mmol L1), revealed a maximum acrolein selectivity of 81 mol % at 92 % conversion at 400 8C and 30 MPa with
a feed of 0.005 mol L1 H2SO4 and 0.05 mol L1 glycerol solution
and a residence time of 10.23 s. This result showed that the selectivity decreased and the conversion increased as the residence time increased. In the supercritical region, glycerol conversion increased with increasing pressure.
Several drawbacks lie within the sub-/supercritical reaction.
First, there is a constraint associated with the glycerol feed
concentration. The glycerol feed solution should be kept
below 5 wt % because a higher concentration could result in
easier plugging of coke deposits at the reaction wall, pipes, or

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemSusChem 2012, 5, 1162 1180

Comparison of Petroleum-Based and Bio-Based Acrolein Production


the connector. This situation would cause a severe safety problem, as the reaction is operated at high pressure. Second, sub-/
supercritical water itself already imposes a corrosion burden
on the system materials, and in the presence of the acid catalyst, corrosion would be more severe for the reaction process.
As a result, after a period of time, due to corrosion and plugging, the entire reactor and pipeline would need to be replaced, which increases the cost. Third, as it is a homogenous
catalytic reaction, the downstream separation of the acid from
the product and later the recycling or disposal process of the
acid are also significant issues for the practical scale-up of the
liquid-phase reaction.
Using high-boiling solvents was another method to achieve
a liquid-phase reaction. Hoyt and Manninen reported such
a process in 1951. Liquid-phase glycerol dehydration to acrolein was achieved in 72.3 % yield by gradually feeding glycerol
to a suspension of a supported mineral acid catalyst in a highboiling petroleum oil (300400 8C) at a reaction temperature of
250325 8C.[69] A similar process added glycerol to a suspension
of a sulfate salt catalyst in liquid paraffin at 280 8C to yield
acrolein in 57 mol % at 89 % glycerol conversion.[70] Yoshimi
tested a variety of solid acid catalysts within a range of temperatures (250420 8C) and found that the highest acrolein
yield was 79 mol % at 315 8C/328 8C. Katryniok et al. have criticized this process for its complexity and, thus, the difficulty of
scale-up.[6]
2.2.2. Gas-phase reaction
The other approach for conducting glycerol conversion to
acrolein carries out the reaction in the gas phase. Gas-phase
approaches have several advantages compared to the sub-/supercritical water method. First, the gas-phase reaction is simpler. The atmospheric operation reduces corrosions in the reactor and does not need an expensive pressure controlling
system; it may even utilize, to a large degree, existing manufacturing equipment with minimal adjustments to accommodate bio-based acrolein production. Second, the gas-phase reaction is usually accomplished in a PRB using heterogeneous
catalysts, so the downstream separation of acid from the product is minimized. These advantages favor the gas-phase approach over the sub-/supercritical approach with regard to industrialization considerations. Using these advantages, there
are significantly more studies using a gas-phase method in
comparison to those using a sub-/supercritical water method.
The solid catalyst plays an important role in determining
whether good acrolein production can be achieved. Several
catalyst properties taken together determine the catalytic performance: 1) surface area, 2) pore size, 3) acid strength, and
4) the nature of the acid sites. Usually, a larger surface area
provides a larger amount of active sites for the catalytic conversion and leads to a higher yield of the desired catalytic
product. Mesoporosity has been noted as the most appropriate for glycerol dehydration. Tsukuda et al. conducted a parallel
comparison of HSiW supported on silica of different pore sizes
(3, 6, and 10 nm).[71] The fabricated catalyst with the largest
mesopore size exhibited not only higher acrolein selectivity,
ChemSusChem 2012, 5, 1162 1180

but also maintained the space yield for a longer time-onstream (TOS). Chai et al. divided the acidity (using the Hammett acidity function, H0) into three major groups: H0  8.2,
8.2  H0  3.0, and 3.1  H0  +6.8. The solid acids with
strengths in the range 8.2  H0  < M- > 3.0 were most likely
to afford good acrolein production. Outside this range, the catalyst either does not provide sufficient acidity to selectively
catalyze acrolein formation or is too acidic and suffers from
surface coke formation. Brnsted acid sites are superior to
Lewis acid sites for acrolein production; however, several studies show that Lewis acid sites could be converted into Brnsted acid sites in the presence of water vapor.[72, 73]
2.2.2.1. Catalysts
Supported mineral acids
Of the supported mineral acids, phosphoric acid has been reported as the most favorable.[69, 70, 71, 74, 75] The studies have been
conducted on various supporting materials, such as a-Al2O3,
activated bentonite, and activated carbon. Common mineral
acids are usually water soluble. Therefore, they cannot be solid
acids by themselves. Previous studies used them as solid acids
after they were supported on catalyst frameworks. Boronic
acid, sulfuric acid, and phosphoric acid supported by silica,[71]
alumina,[74] or zeolite[69] have been studied in the past. Supported phosphoric acid thus far provided the best result compared to other supported mineral acids. In 1993, Haas et al.[75]
stated that they had produced 1,2- and 1,3-propandiol from
glycerol. The first step described in their patent is the dehydration of glycerol using an alumina-supported phosphoric acid
for the production of acrolein in 70.5 mol % yield. Suzuki patented a semi-batch reaction using H3PO4/aAl2O3, with dropwise addition of glycerol solution, and reported 42 % conversion and 89 mol % selectivity.[70]
Zeolites
Zeolites are crystallized aluminosilicates, consisting of a threedimensional framework of tetrahedral SiO4 or AlO4. The general
formula of a zeolite can be represented as Mx/n [(AlO2)x(SiO2)y]w H2O. The zeolite has well defined yet controllable
pore sizes, good thermal stability, and tunable distribution of
Brnsted and Lewis acidities.[76] Li and co-workers patented the
use of acidic zeolites to catalyze glycerol dehydration to acrolein.[77] The highest acrolein yield was reported as 82.1 mol %
using ZSM-11 as catalyst at 320 8C. Corma and co-workers operated glycerol conversion in a MicroDowner unit in the presence of ZSM-5 zeolites. At the optimal condition, they reported
62.1 mol % acrolein selectivity and 100 % glycerol conversion.[78]
Ghlich et al. screened several zeolites using a microactivity
test unit. They found that a low amount of phosphoric acid
loading boosted the acrolein selectivity.[79] Neher et al. achieved 7175 mol % selectivity using HZSM-5 to catalyze the
liquid-phase reaction (250300 8C) at 7 MPa.[80] However, the
corresponding glycerol conversion was only 1519 %. Chao
et al. disclosed their development of a catalyst using HZSM-5,
with a SiO2/Al2O3 molar ratio of 25360, together with the

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemsuschem.org

1169

L. Liu, X. P. Ye, and J. Bozell


active metal components. Selectivity of 75.2 mol % and a quantitative glycerol conversion was claimed. Kim and co-workers
examined the effect of the SiO2/Al2O3 molar ratio (ranging
from 231000) in ZSM-5, and showed that HZSM-5 with a SiO2/
Al2O3 molar ratio of 150 provided the highest glycerol dehydration activity.[81] Jia and co-workers investigated the performance of a nanocrystalline zeolite. In their examination of the
particle size effect and Al/Si ratio, they concluded that a nanocrystalline zeolite with small pore size and large Al content
was the most promising material.
Metal/mixed-oxide catalysts
In general, for mixed or metal oxides, the preparation procedures (calcination parameters in particular) modify the acid/
base properties, which results in significant differences in their
catalytic performance in the reactions. Chai et al. investigated
Nb2O5, and found that it exhibited the best catalytic performance when calcined at 600 8C.[82] Ulgen and Holderich reported 72.1 mol % acrolein selectivity at 88.7 mol % conversion
using ZrO2-supported WO3 (15.43 wt % WO3 loading on
ZrO2).[61] In 2008, Okuno et al. disclosed that crystalline metallosilicates containing Al, Fe, Ga, In, Co, Ni, or Zn in certain molar
ratios to Si (Si/metal  800) were good catalyst candidates for
glycerol dehydration to acrolein.[83] Arita patented a solid acid
catalyst, that is, a crystal of a rare earth metal salt of phosphoric acid, which was prepared by adding phosphate ions to a solution of water and a hydroxide salt of a rare earth metal and/
or its dehydration condensation product followed by a cofiring procedure.[84] Liu et al. conducted a detailed study of rare
earth (La, Ce, Nd, Sm, Eu, Gd, Tb, Ho, Er, Tm, Yb, Lu) pyrophosphates.[85] Although the unsupported rare earth pyrophosphates did not have large surface areas (< 50 m2 g1), many of
them displayed promising catalytic performance (> 95 % conversion and > 70 % acrolein selectivity at 2 h TOS; > 80 % conversion and > 75 % acrolein selectivity at 8 h TOS). The calcination temperature and the precipitation pH value are two important parameters during catalyst fabrication that provide
a large tunable range in terms of the final catalytic performance in the glycerol dehydration reaction. Also, enlarging the
catalyst surface area with supporting materials provides an opportunity for further potential improvement.
Heteropoly acids
Heteropoly acids (HPAs) feature strong Brnsted acidity, which
approaches the superacid region (defined as an acid with
a strength higher than that of 100 % H2SO4[86]). Compared to
conventional acids, HPAs are economical, environmentally
friendly, and have well-defined structures and tunable acidity
levels.[8688] The most common commercially available HPAs are
H3PW12O40 (denoted as HPW), H4PW11VO40, H4SiW12O40 (denoted
as HSiW), H3PMo12, and H4SiMo12O40. However, they are water
soluble, as are all mineral acids.[8688] Direct usage in a reaction
containing a large amount of polar substances would result in
substantial loss of the HPA during the course of the reaction.
Although they possess high acidity, HPAs usually have low spe-

1170

www.chemsuschem.org

cific surface areas.[89, 90] Thus, in many HPA applications, HPAs


are loaded onto supporting materials.[9193] The high acidity of
HPAs is retained after being loaded on the support, while achieving larger surface areas controlled by the supporting materials.[89, 90, 94] In other words, the supporting materials can modify
and tune HPA reactivity. With appropriate selection, the supported HPAs can be good candidates as solid acid catalysts for
glycerol dehydration to acrolein.
There is an optimal HPA loading for a given support. Underloading necessarily leads to lower acid strength, whereas overloading weakens the proton sites and makes acid distribution
less uniform. Several studies have examined several commercially available HPAs on different supporting materials (Table 2).
Although some studies found that the crystal phase of HSiW
formed on a silica surface for a loading higher than 20 wt %
prior to the crystal phase formation, Tsukuda and co-workers
obtained good acrolein production using silica-supported
HPAs with 30 wt % acid loading.[71] Three different mesoporous
silicas were compared, (3, 6, and 10 nm) with surface areas of
733, 466, and 310 m2 g1, respectively. HSiW was finely dispersed on the silica with 3 nm pores. However, as the HSiW
molecule itself was about 1.2 nm, there was only a very small
opening for the reactants to enter and for the product to
leave the site. Consequently, it was easier for deposited coke
to clog the entrance. HSiW crystalline structures started to aggregate on 6 nm silica pores, and more so on the 10 nm
pores; yet, they still provided appropriate pore size, surface
area, and acid strength to achieve high acrolein yield. All three
catalysts exhibited almost identical acrolein yields initially; the
ones with larger pores showed overall higher yield because
the conversion decreased more slowly, which was attributed to
lower coke formation.[71] A study of HSiW loaded on activated
carbon with a surface area of 1068 m2 g1 showed that
a 10 wt % loading of HSiW provided the best catalytic performance.[95] The acid strength of HSiW was reduced for loadings
on activated carbon of less than 5 wt %,[95] whereas a higher
loading caused more aggregation of HSiW molecules resulting
in a decrease of the pore size and thus more severe deactivation. Dubois et al. patented their work using the salt of a HPA;
the yields were between 49.8 and 93.1 mol %, and the conversion was between 78.6100 %.[96, 97]
Atia et al. conducted a factorial-type of comparison study on
both the HPAs and the catalyst supports.[98] They found that
the same trends occurred as in the case of silica-supported
HSiW:[71] larger mesoporous alumina-supported HSiW showed
a better catalytic performance for acrolein production. They
found that either 10 or 20 wt % loading had no significant
effect on catalytic conversion. HSiW was a better acid than
H3PMo12O4, H3PW12O40(HPW), or (NH4)3PMo12O40 on silica, alumina, and aluminosilicate, which possibly was the combined
effect of acid strength, thermal stability, oxidation potential,
and hydrolytic stability.[89] Alumina- and aluminosilicate-supported acids showed a rather long and stable catalytic life.
Chai et al. used anti-sintering ZrO2 nano-crystals to support
HPW. Optimal loading of HPW was between 1020 wt %.[99] Interaction between ZrO2 and HPW increased the thermal stability of HPW. The ZrO2-supported HPW calcined at 650 8C exhibit-

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemSusChem 2012, 5, 1162 1180

Comparison of Petroleum-Based and Bio-Based Acrolein Production

Table 2. Recent progress of acrolein production through gas-phase glycerol dehydration.


Catalysts

Particle
sizes
[mm]

Surface Glycerol
Acrolein
Feed (gas)
conversion selectivity
area
[mol %]
[mL min1]
[m2 g1] [%]

Nb2O5[d]
ZSM5 (Si/Al = 100)
15 %H3PO4/SiO2[f]
15 %H3BO3/SiO2[f]
30 % H3PW12O40/SiO2
30 % H4SiW12O40/SiO2
30 % H3MoW12O40/SiO2
H4SiW12O40
15.43 % WO3/ZrO2
20 %HPMo/A5
20 %HPW/A5
20 %HSiW/A5
20 %HSiW/AS4
20 %HSiW/AS12
30 %H3PW12O40/SiO2[g]
30 %H3PW12O40/SiO2[i]
15 %H3PW12O40/ZrO2[i]
10 %H4SiW12O40/AC
Nd4(P2O7)3[k]
FePO4 (A)
FePO4 (AH)
FePO4 (P)
FePO4 (H)
HZSM-5[l]
HZSM-5[l]
HZSM-5[l]
HZSM-5[l]
CsPW
Cs2HPW12O40
0.5 %Pd/CsPW
0.5 %Ru/C + CsPW (1:10)
20 %HPW/SiO2
2 %Pd/20 %HPW/SiO2
VOPO4[m]
SAPO-11
SAPO-34
La4(P2O7)3
Ce4(P2O7)3
Nd4(P2O7)3
Sm4(P2O7)3
Eu4(P2O7)3
Gd4(P2O7)3
Tb4(P2O7)3
Ho4(P2O7)3
Er4(P2O7)3
Tm4(P2O7)3
Yb4(P2O7)3
Lu4(P2O7)3
20 %H4SiW12O40/SiO2
20 %H4SiW12O40/Al2O3[n]

N.A.[e]
40120
75650
75650
75650
N.A.[e]
N.A.[e]
N.A.[e]
5001000
315500
315500
315500
315500
315500
350840[h]
350840[h]
350840u[h]
350840[h]
N.A.[e]
125200
125200
125200
125200
2060
2060
2060
2060
45180
45181
45182
45183
45184
45185
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
13
1.22.4

99
70
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
<5
59
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
141
162
133
N.A.[e]
N.A.[e]
43.6
2.6
7.8
8.7
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
130
70
84
117
205
200
10
89
359
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
N.A.[e]
300
140

88
100
69.9
1.5
99.7
98.3
98.4
26.7
88
97.3
99.2
98.4
100
91
25
20
76
92.6
96.4
100.0
97.7
100.0
100.0
100.0
98
83
57
100 (41)
91 (21)
97 (79)
95 (50)
100 (60)
100 (72)
100
88 (65)
59 (42)
76.2
44.8
87.2
89.7
83.1
88.2
87.6
84.4
86.7
87
48.4
58.5
94.4
98.7

51
62.1
48.2
27.3
65.1
86.2
33.4
59
72.1
45
51.7
63.6
75
71
59
54
71
75.1
82.7
78.4
81.5
85.3
92.1
60
63
65
67
98 (94)
97 (94)
83 (96)
94 (92)
97 (95)
85 (94)
64
62 (55)
72 (65)
78.5
42.9
79.9
77.8
78.5
78.9
78.8
77.2
79.7
77.8
63.7
64.4
88.0
84.7

Glycerol
T
TOS[a]
concentration[b] [wt %] feed information[c] [h1]
[8C] [h]

30 (N2)
36.20
20
N.A.[e]
30 (He)
10
30 (He)
10
30 (He)
10
30 (He)
10
30 (He)
10
30 (He)
10
2.5 (O2)
20
30 (He)
10
30 (He)
10
30 (He)
10
30 (He)
10
30 (He)
10
31 (N2)
36.20
32 (N2)
36.20
33 (N2)
36.20
20 (He)
10
30 (N2)
36.2
40
9.6 (N2)
9.6 (N2)
40
40
9.6 (N2)
40
9.6 (N2)
30 (N2)
35 or 50
30 (N2)
35 or 50
30 (N2)
35 or 50
30 (N2)
35 or 50
10
15 (N2)
15 (N2)
10
10
15 (H2)
10
15 (H2)
15 (N2)
10
15 (H2)
10
18 (N2/O2) 20
50 (He)
5
50 (He)
105
30 (N2)
36.2
30 (N2)
36.2
30 (N2)
36.2
30 (N2)
36.2
30 (N2)
36.2
30 (N2)
36.2
30 (N2)
36.2
30 (N2)
36.2
30 (N2)
36.2
30 (N2)
36.2
30 (N2)
36.2
30 (N2)
36.2
60 (Ar)
20
60 (Ar)
20

80
335[e]
0.587[e]
0.587[e]
0.587[e]
0.587[e]
0.587[e]
0.587[e]
330
0.574[e]
0.574[e]
0.574[e]
0.574[e]
0.574[e]
400
400
400
[j]
227
600
600
600
600
155
465
719
1438
227
227
227
227
227
227
227
43 (90)
44 (90)
227
227
227
227
227
227
227
227
227
227
227
227
84.4
84.4

315
350
325
325
325
275
275
275
280
275
275
275
275
275
315
315
315
330
320
280
280
280
280
320
320
320
320
275
275
275
275
275
275
300
280
280
320
320
320
320
320
320
320
320
320
320
320
320
275
275

910
910
5
5
5
5
5
5
8
several
several
several
several
several
910
910
910
5
78
5
5
5
5
910
910
910
910
1 (5)
1 (5)
1 (5)
1 (5)
1 (5)
1 (5)
10
after 1
after 1
78
78
78
78
78
78
78
78
78
78
78
78
7.5

Ref.

[82]
[78]
[71]
[71]
[71]
[71]
[71]
[71]
[61]
[98]
[98]
[98]
[98]
[98]
[101]
[101]
[101]
[95]
[102]
[102]
[102]
[102]
[102]
[103]
[103]
[103]
[103]
[100]
[100]
[100]
[100]
[100]
[100]
[104]
[105]
[105]
[85]
[85]
[85]
[85]
[85]
[85]
[85]
[85]
[85]
[85]
[85]
[85]
[106]
[106]

[a] Time-on-stream. [b] The concentration of glycerol solution = Wglycerol(g)/(Wglycerol+Wwater)(g). [c] The reported data are gas-hourly-space-velocity unless otherwise specified. [d] Calcined @400 8C. [e] Not available. [e] Weight hourly space velocity. [f] 6 nm. [g] Calcined@350 8C. [h] Corresponds to 2040 mesh.
[i] Calcined@650 8C. [j] Unspecified. [k] pH 6, calcined@500 8C. [l] Nanocrystalline. [m] Calcined@800 8C. [n] Together with a non-thermal plasma application.

ed the best catalytic performance and also had the potential


to be regenerated with a conventional oxygen burning
method. Other than using supported HPAs, the usage of the
salt of HPAs plays a similar role in modifying the surface area,
porosity, and thermal stability, which varied greatly as a function of the counter cations.[89] Alhanash et al.[100] conducted
ChemSusChem 2012, 5, 1162 1180

a study of HPA salts with Cs, which had been reported as the
most promising counter cation for acid-catalyzed reactions.
The HPW-Cs salt showed better performance than HSiW-Cs
salt, although both of them showed high acrolein selectivity.
The doping of Pt group metals on the HPW-Cs salt slowed the
deactivation rate without lowering the acrolein selectivity. The

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemsuschem.org

1171

L. Liu, X. P. Ye, and J. Bozell


best catalyst, as they discovered, was 52 wt % Pd-doped
Cs2.5H0.5PW12O40.[100]
Recent achievements in the gas-phase glycerol dehydration
in terms of acrolein production are summarized in Table 2.
2.2.2.2. Reaction conditions
In a diffuse reflectance FTIR study on dehydration of glycerol
on alumina supported HSiW,[107] Cheng and Ye found that
a higher temperature caused formation of polymeric condensation compounds and also facilitated the formation of acetate
and surface-bound carbonates, resulting in smaller carbon
chain products, such as CH3CHO, CH4, CO, and CO2. In Yoda
and Ootawas FTIR study of glycerol dehydration on H-MFI zeolite, the dehydration yielded exclusively acrolein under
87 8C.[108] ZSM-5 zeolite has been used as catalyst for the kinetic
study by Kim and co-workers.[81] They investigated a broad
range of SiO2/Al2O3 ratios, which covered the ratio of the zeolite that Yoda and Ootawa used. However, acetol formation
was not detected at low temperature (80 8C). This finding reinforced the possibility that low temperatures favored acrolein
production. The reason might be that the decomposition of
acrolein into formaldehyde and acetaldehyde occurs much
faster at higher temperatures, causing partial loss of acrolein
when examining the downstream product. Pathak et al. claimed that the condensable products (such as acrolein, acetol, etc)
were dominant at low temperatures, whereas at high temperatures more non-condensable products (such as CO2, H2, etc)
were present.[109] Kuba and Dubois used their catalyst screening program to evaluate various reaction conditions for glycerol conversion to acrolein.[110] The results showed catalysts deactivated considerably faster at higher temperatures. Although
glycerol conversion increased with the increase of the reaction
temperature, glycerol started to pyrolyze at a temperature of
300 8C, reducing the amount of desired dehydration product.
Also, several studies reported that glycerol conversion was
nearly complete before reaching 290 8C (for example, 275 8C in
Atias study[98] and 240 8C in Ulgen and Hoelderichs study[61]),
so a further increase of the reaction temperature would not
necessarily lead to higher acrolein yields. Relatively better acrolein selectivity occurred at 275 8C for all the catalysts investigated by Atia et al.[98] According to Ulgen and Hoelderich, however, lower temperature favors the intermolecular dehydration
leading to glycerol oligomerization. At higher temperatures,
they found that the maximum selectivity occurred at 280 8C.[61]
Water vapor also plays an important role in acrolein production from glycerol. Because the crude glycerol usually contains
water, the initial rationale of using a glycerol solution as the
feed for dehydration was to eliminate any subsequent water
removal. The presence of water requires that the solid catalyst
demonstrates hydrothermal stability. The literature reports several examples of successful glycerol dehydration in the presence of water. Dehydration occurred as a surface-type acid-catalyzed reaction on Cs heteropoly salt catalysts, which were
water-insoluble and hydrophobic, and demonstrated good hydrothermal stability.[100] Supported HPAs, especially supported
HSiW, have a high resistance to water.[71, 100] Wang et al. con-

1172

www.chemsuschem.org

cluded that vanadium phosphate oxides were stable during reaction in the presence of water vapor.[104] The commercial USYbased equilibrium catalyst and ZSM5-based additive used by
Corma et al. did not show noticeable changes in catalyst properties, stability or catalytic activity at different temperatures
with various water fractions in glycerol solution.[78] A slight increase in catalyst stability with the increase of water content in
the glycerol feed was reported for two other zeolites, H-ZSM-5
(150) and H-ZSM-5 (30).[81] As the water faction increased
(within the range of 66.691.3 wt %), the acrolein yield increased with no significant variation in glycerol conversion.
Suprun et al. claimed that water had a positive influence on
the catalysts (Al2O3- and TiO2-supported phosphate ions, SAPO11 and SAPO-3) for glycerol-to-acrolein conversion in that it
created a moderate increase in acidity by the formation of
extra-framework alumina species [e.g., Al(OH) + and Al(OH)2 + ]
and some titania species, which provided weak Lewis acid
sites.[105] Furthermore, water as a dilute agent for glycerol inhibits catalyst deactivation.[78] The surface reaction network of
the chemisorbed glycerol is very likely to be influenced by the
presence of water vapor.[81] The undesired sequential reactions
from acrolein might be suppressed by the water vapor.[78] In
the presence of water, some Lewis acid sites might be converted into Brnsted acid sites, which are preferred for acrolein formation.[98, 107] Therefore, a conclusion may be drawn that, with
a good solid acid catalysts, the presence of water vapor is
a positive factor to the glycerol-to-acrolein conversion.
2.2.2.3. Problems to be solved
Promising progress has been achieved regarding the acrolein
yield, which is already competitive with the current manufacturing methods derived from propylene.[78] To transform this
bio-based process into a completely satisfactory substitution
to the propylene method, one problem needs to be solved:
how to extend the service life of the highly efficient solid acid
catalysts to the timescale of months to years. Catalyst deactivation was reported in almost every study on glycerol dehydration to acrolein.[61, 71, 72, 74, 81, 82, 85, 95, 98101, 103105, 108, 110, 111] The negative influence of catalyst deactivation was usually reflected in
the decrease of glycerol conversion along with reaction time,
which lowered the overall acrolein yield. For example, compared to glycerol conversion at 1 h TOS, the conversion at 5 h
TOS was reduced by 4059 % for the Cs heteropoly salts in Alhanash and co-workers study.[100] In the study by Kim et al.,[81]
the ZSM-5 catalysts with various Si/Al ratios showed a reduction
of 4669 % in glycerol conversion at 1012 h TOS compared to
2 h TOS, and also some reduction in acrolein selectivity.
Catalyst deactivation is caused by coke formation on the catalyst surface, thereby blocking the pores and preventing reactant contact with the active sites. Glycerol conversion would
decrease as a result. Also, part of the coke is formed from acrolein or glycerol, so its formation further reduced the acrolein
yield. Stronger acid sites, or more specifically, strong Brnsted
acid sites, are favored for acrolein formation, but they unfortunately also lead to more severe coking and, therefore, more
severe catalyst deactivation.[100]

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemSusChem 2012, 5, 1162 1180

Comparison of Petroleum-Based and Bio-Based Acrolein Production


Some efforts were made regarding the catalyst deactivation
problem. Although not as catalytically effective as supported
HPAs, one advantage of zeolite or other metal-oxide catalysts
calcined at a higher temperature than 450 8C or 500 8C is that
a conventional coke-removal method using oxygen burning
can be applied. Chai et al. stated that 20 % oxygen blended
with nitrogen at a reaction temperature of 315 8C was sufficient for complete regeneration.[82] However, no data were provided to show this result at such a low temperature, at which
hard coke could not be cleaned according to classical theories.
Some studies claimed that co-feeding small amounts of
oxygen helped to suppress the side products and the coke formation.[97, 104] Arita et al. disclosed a glycerol dehydration process using zeolite as the acid catalyst.[84] In their operation, the
glycerol feed was halted after 12 h of dehydration, and
oxygen-containing gas was flashed through the reactor at
360 8C for 8 h. The regained acrolein yield (63 mol %) could be
seen in the product sampled within 3 h of restarting the reaction. Yan et al.[74] tried to improve on the coking problem by
operating the reaction at a pressure lower than atmospheric to
aid in the evaporation of glycerol and to prevent condensation
of reactants/products on the catalyst surface. The catalysts life
was extended under these conditions. Alhanash et al. tried to
ease the coking problem by doping the solid acid catalyst with
a platinum-group metal and adding hydrogen to the carrier
gas. The addition of hydrogen possibly hydrogenated the coke
precursor, which is the reverse reaction of coke formation.[100]
Katryniok et al. fabricated a catalyst (HSiW supported on zirconia-grafting-modified SBA-15) that showed some potential in
slowing catalyst deactivation.[112] Liu developed a periodical regeneration method using non-thermal oxygen plasma, which
successfully alleviated the catalyst deactivation problem.[106]

3. Comparative Economic Evaluation of BioBased and Conventional Acrolein Production


Glycerol dehydration as a bio-based method for acrolein production has been demonstrated to give yields comparable to
the partial oxidation of propylene, suggesting that it could
offer a promising sustainable alternative technology for the
chemical industry. Continuous efforts are underway to further
improve the acrolein yield and extend catalyst lifetime. To further demonstrate the viability of this process, an economic
evaluation was carried out to demonstrate the benefits of
a bio-based method in comparison to the propylene-based
method. Gas-phase glycerol conversion was used in this evaluation because it has a better opportunity for use in industry
and because it has attracted significantly more attention in
comparison to liquid-phase approaches. The initial economic
assessment was conducted for an annual production rate of
10 000 t of acrolein. This number was based on the annual production of glycerol co-produced from biodiesel and the possible diverse uses of glycerol for other value-added products,
such as 1,3-propanediol,[113] epichlorohydrin,[114] and lactic
acid.[115] Both propylene-based and bio-based processes are
compared on this basis.
The process flow diagram for the partial oxidation of propylene to acrolein is presented in Figure 8.[116118] In this process,
compressed air, propylene, and steam are heated to 250 8C
prior to entering the packed bed reactor (PRB). A typical molar
ratio of these three components is 1:10:5. The reactor is operated at 250400 8C and 150250 kPa, and the gaseous mixture
is quenched with water and partially cooled to remove the byproduct acrylic acid and trace amounts of acetic acid in the
quench tower. The gaseous mixture, including the products,

Figure 8. Process flow diagram for the production of acrolein from propylene.[117, 119]

ChemSusChem 2012, 5, 1162 1180

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemsuschem.org

1173

L. Liu, X. P. Ye, and J. Bozell


remaining reactants, and inert gas, is sent to an absorber,
pounds. Acrolein is absorbed with a large amount of water in
where the noncondensable gases (unreacted propylene,
the absorption tower, and the noncondensable gases (nitrooxygen, nitrogen, and COx) are removed as the off-gas, and
gen, COx, etc) are removed. The diluted acrolein solution is
the dilute acrolein liquid stream is sent to the subsequent
then worked up to crude acrolein in the stripper, and purified
stripping column. In the stripping column, the majority of
by two steps of distillation through the extraction column and
water is condensed at the bottom, whereas the raw acrolein is
the distillation tower.
obtained as distillate. The refined acrolein (> 95 % pure) is obTo analyze the economic feasibility of bio-based acrolein
tained after passage through two subsequent separation
production, the concept of production costs plus profit (PCPP)
towers (an extraction column and distillation tower) to remove
was used. Table 3 lists the elements in the calculation of PCPP,
the lighter organic impurities.
which is a proxy for the market price of the end product. As is
A proposed industrial gas-phase glycerol dehydration prodemonstrated in the table, PCPP is composed of direct operatcess flow diagram is shown in Figure 9. Glycerol is heated to
ing costs, taxes, insurance fees and plant overhead, the allow275 8C, and fed into a mixing and agitation tank, into which
the high-temperature steam is also fed. The output
(  4.7 mol % glycerol/steam) is fed into the PRB
(275 8C). If crude glycerol is used, the pretreatment
Table 3. Components of the production costs plus profits (PCPP, reproduced from
needs to be added to remove any solids, fouled orHermann et al.[120])
ganics, salts, free fatty acids, color, and odor. Glycerolwater vapor, together with the onsitegenerated
Investment costs
Production costs
Other costs
nitrogen, is fed into the reactor packed with alumiinside battery limit
feedstocks
taxes and insurance
nasupported silicotungstic acid operated at 275 8C
outside battery limit auxiliaries/catalysts
plant overhead (80 % of labor costs)
and atmospheric pressure (101 kPa). The gaseous
byproducts credits/debits marketing
utilities
administration
mixture at the reactor outlet includes noncondensawaste
treatment
R&D
ble gases (nitrogen and CO2), water vapor, acrolein,
operating supplies
capital charge (represents the
low-boiling compounds (acetaldehyde and propionalmaintenance supplies
total depreciation and profits)
dehyde), and high-boiling point compounds (acetol,
operating labor
maintenance labor
acetic acid, acrylic acid, phenol, glycerol, etc). From
laboratory labor
this point on, the treatment can be directly adopted
from the propylene-based process. The gaseous mixtotal fixed capital
direct operating costs
ture is washed with a limited amount of water in the
production costs + profit (PCPP)
quench tower to remove the high-boiling com-

Figure 9. Proposed process flow diagram for the production of acrolein from glycerol. The rectangle (dashed line, left hand side) is applied if crude glycerol is
used as starting material.

1174

www.chemsuschem.org

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemSusChem 2012, 5, 1162 1180

Comparison of Petroleum-Based and Bio-Based Acrolein Production

Table 4. Comparative list of reactor and separation columns for the propylene-based and bio-based acrolein production.[a]
Propylene-based process

Bio-based process

pretreatment heating propylene and air


PBR[b] reactor multiple reactors packed with multi-components metal
operating at  320 8C, strong exothermic reaction
operating at 150250 kPa
requires careful temperature control (good cooling
system)
1st column
quenching of the products
removing the high-boiling compounds, mainly acrylic acid
and acetic acid
feed: propylene, COx, steam, acrolein, acetaldehyde, pro2nd column
pionaldehyde
obtaining a diluted acrolein solution devoid of noncondensable gases
feed: acrolein, water, acetaldehyde, propionaldehyde
3rd column
distillate: crude acrolein
feed: acrolein, acetaldehyde, propionaldehyde.
4th column
distillate: acrolein containing less impurity
obtaining refined acrolein; adding stabilizer.
5th column

purification (if crude glycerol is used as starting material)


heating glycerol, mixing glycerol with steam
multiple reactors packed with alumina-supported heteropoly acid
operating at  280 8C, endothermic reaction
operating at atmospheric pressure
good heat isolation, needs heating supply
quenching of the products
removing the high-boiling compounds, including acetol, acetic acid, acrylic acid,
phenol, and glycerol
feed: acrolein, acetaldehyde, propionaldehyde, steam, COx, N2
obtaining a diluted acrolein solution devoid of noncondensable gases

feed: acrolein, water, acetaldehyde, propionaldehyde


distillate: crude acrolein
feed: acrolein, acetaldehyde, propionaldehyde.
distillate: acrolein containing less impurity
obtaining refined acrolein; adding stabilizer

[a] This table corresponds to the schemes presented in Figure 8 and Figure 9. [b] PBR stands for packed-bed reactor.

ance for marketing, administration and research and development, and the capital charge.
To simplify the comparison of the two processes, the following Equations (5) and (6) are used in the calculation. Two parts
comprise the direct operating cost: Cv and the fixed cost, each
of which contains several elements. Cv is represented by:
P l  Y  Cv  Cfix

where P is the profit; l is the market price of the product


acrolein ($ per ton); Y is the annual mass production of acrolein (tons); Cv is the variable cost ($), which includes the cost for
feedstock, catalyst, energy, and labor, etc.; and Cfix is fixed cost
($), which includes all the other costs, such as investment, capital, marketing, administration, etc. The variable cost is represented by:
Cv Ccat Cfeedstock Cen Cother

where Cv is the variable cost, Ccat is the cost for the catalyst,
Cfeedstock is the cost for feedstock, Cen is the cost for energy consumption, Cother includes all the other components in the variable cost, which is assumed to be the same for both processes.
The difference in variable costs between the propylenebased process and bio-based acrolein can thus be expressed in
Equation (7).
DCv DCcat DCfeedstock DCen DCother


g
p
g
p
g
p
Ccat
 Ccat
 Cfeedstock
Cen
 Cen
Cfeedstock

where D denotes the difference between the propylene- and


the bio-based method, the superscript p denotes propyleneChemSusChem 2012, 5, 1162 1180

based process, and the superscript g denotes bio-based


process.
A comparison of the major components with their specifications in the two production processes, based on Figures 8 and
9, is made in Table 4. There are many similarities between
these production schemes. The multiple PBR system is well applicable to both production methods. The partial oxidation of
propylene to acrolein is highly exothermic ( 356 kJ mol1),
whereas glycerol dehydration to acrolein is endothermic
(43 kJ mol1);[121] therefore, an efficient heat exchanger is
needed for the multiple PBRs in the propylene-based production to promptly release the heat, whereas the multiple PBRs
in the bio-based production need a furnace to maintain the reaction at a elevated temperature. The PBR for propylene-based
production usually operates at 150250 kPa, the operation cost
of which is usually higher than that of bio-based acrolein production operated at atmospheric pressure. Considering these
factors, it is reasonable to assume that the fixed costs in terms
of the reactor part would not differ considerably for both processes. The difference in the energy cost will be considered in
Cen as part of Cv. The post-processing steps are essentially the
same for both synthesis methods, so the cost regarding this
part can be considered to be the same. The major differences
exist prior to the reactors, as categorized under Pretreatment
in Table 4, as the bio-based process requires an additional nitrogen generator for acrolein production (23 340 $; quoted by
Gazcon for onsite nitrogen generation at up to 250 L min1,
99 % purity, 0.5 MPa outlet pressure, and 24 h-a-day operation).
Also, a mixing/agitation tank is needed to blend hot glycerol
with steam. If crude glycerol is used, a pretreatment step can
be added to remove any solid, organics, salts, free fatty acids
(FFAs), color, and odor. Several commercial products are available for this application, such as high efficiency electrodialysis
(HEED) or high efficiency electro-pressure membrane (HEEPM)

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemsuschem.org

1175

L. Liu, X. P. Ye, and J. Bozell


sumption of all the other processes are the same for both production methods). The consumption of deionized water used
in the quenching and absorbing processes and the solvent
used in the extraction step (Figure 8 and 9) in the propylenebased and bio-based production is equivalent; therefore, their
costs are not included in Table 5. The quantities of feedstocks
required for both processes are calculated based on a stoichiometric relation assuming 80 mol % acrolein yield, and the
quantity calculated for crude glycerol is based on a glycerol
purity of 80 %. In terms of the last two variations in Cv [Eq. (7)],
the bio-based method using refined glycerol costs 12.95
16.13 million (MM)$, which is
0.821.02 times the cost of these
Table 5. Comparison of propylene-based and bio-based acrolein production (10 000 ton/year) regarding feedparts of the propylene-based
stock and energy consumption.
method; the bio-based method
using crude glycerol costs 2.71
Raw material
Price
Propylene process
Bio-based process [t per year]
5.83 MM$, which is 17.240.5 %
[t per year]
refined glycerol
crude glycerol
[$ t1]
of the propylene-based method.
[a]
[b]
propylene
1664
9 375

The cost for feedstock plays

20 536[d]

refined glycerol
595.25749.57[c]

25 669[e]
crude glycerol
77.16198.42[c]
the most important role comsteam
7.31
20 089[f]
82 142[g]
82 142[g]
pared to the other components
energy
0.05[h]
0
2.67  106[h, i]
2.67  106 kWh[h, i]
in the variable costs (Table 5),
which is consistent with indusTotal price [MM$]
12.9516.13[k]
2.715.83[l]
15.74[j]
trys continuous search for the
least expensive feedstock.
[a] Price reported in February 2011 by ICIS.[121] [b] Calculated based on 10 000 t annual acrolein production and
80 mol % acrolein yield: [10 000 t/(56 g mol1)  (42 g mol1)]/80 % = 9375 t. [c] Price reported in September 2010
The catalyst is the remaining
by ICIS,[122] apart from of the lower end of crude-glycerol price.[123] [d] Calculated based on 10 000 t annual acrocomponent of Cv [Equation (6)]
lein production, 80 mol % acrolein yield, and stoichiometry of glycerol dehydration to acrolein: [10 000 t/
in addition to those listed in
(56 g mol1)  (92 g mol1)]/80 % = 20 536 t. [e] Calculated based on the same assumptions as [d] and the asTable 5. It is rather difficult to acsumption that crude glycerol contains 80 % glycerol: [10 000 t/(56 g mol1)  (92 g mol1)]/80 % = 25 669 t. [f] Calquire accurate price information
culated based on molar ratio of the feed propylene/air/steam = 1:10:5 and calculated propylene amount required: 9375 t/(42 g mol1)  (18 g mol1)  5 = 20 089 t. [g] Calculated based on calculated amount of glycerol
on industrial catalysts. Therefore,
required and the assumption that the feed has a concentration of 20 wt % glycerol: 20 536 t/20 % = 82 142 t.
an indirect approach was used
1
[h] $ per kWh. [i] Calculated for a year based on heat (43 kJ mol ) required for the endothermic dehydration reto evaluate this component. Bisaction[106] and annual working hours of 8000 h: [105 t  (106 g t1)/(56 g mol1)]/80 %  (43 kJ mol1) 
muth molybdate was considered
(0.000278 kWh kJ1) = 2.67  106 kWh. [j] (1664 $ t1)  (9375 t) + (7.31 $ t1)  (20 089 t) = 15.74  106 $ = 15.74 M$.
[k] low: (595.25 $ t1)  (20 536 t) + (7.31 $ t1)  (82 142 t) + (0.05 $ kWh1)  (2.67  106 kWh) = 12.95 MM$. high:
as an example of a multi-compo(749.57 $ t1)  (20 536 t) + (7.31 $ t1)  (82 142 t) + (0.05 $ kWh1)  (2.67  106 kWh) = 16.13 MM$. [l] (77.16 $ t1) 
nent metal catalyst. In general,
1
1
6
1
(25 669 t) + (7.31 $ t )  (82 142 t) + (0.05 $ kWh )  (2.67  10 kWh) = 2.71 MM$ (low) and (198.42 $ t ) 
an acid catalyst is considerably
(25 669 t) + (7.31 $ t1)  (82 142 t) + (0.05 $ kWh1)  (2.67  106 kWh) = 5.83 MM$ (high).
less expensive than a multi-component metal catalyst. For example, the retail price (SigmaAldrich) of active bismuth molybrefer to what is used in the reaction. Also, the energy listed in
date is over twice that of silicotungstic acid (Table 6). At an inthe table only refers to the part of energy required to heat the
dustrial scale, it is reasonable to assume that the relative cost
reactant(s) and for the reaction (assuming the energy condifference remains the same. The average lifetime of a multi-

systems (http://www.eetcorp.com/heepm/glycerinspecs.htm),
the Crown Ion Works Glycerol-Refining-System (http://
www.crowniron.com/technologies/oleo_glycerine.cfm)
and
ROHMIHAASs AMBERSEP BD50 from Dow (http://www.amberlyst.com/glycerol.htm). A conservative assumption is made that
Cfix + Cother of the bio-based acrolein production from crude
glycerol and from refined glycerol is 1.3 and 1.1 times that of
the propylene-based acrolein production.
Several major components in Cv differ greatly between the
propylene-based and bio-based methods for a 10 000-ton/year
acrolein production (Table 5). The quantities for steam only

Table 6. Comparison of catalysts used for the propylene- and glycerol-based acrolein production (10 000 t per year).
Process

Active catalyst

Retail price
[$ g1]

Life time
[year]

Amount
[kg]

Total cost
factor[a] (per year)

Total cost
(106 $ per year)

propylene process
glycerol process

bismuth molybdate
silicotungstic acid

3.53
1.62

1.5
0.5

9657[b]
7083[d]

1
1.01

0.484.83 [d]
0.494.88[e]

[a] Due to the fact that the industrial catalyst price of propylene process is unknown, this factor reflects the relative relationship between the catalysts of
propylene-based process and bio-based process. [b] Calculated for 10 000 t acrolein production. An acrolein yield of 80 mol % was assumed. The density
used was 5670 kg m3, and GHSV used was 800 h1. [c] Calculated based on the assumption that the price of the industrial catalyst for the propylenebased production falls within the range of 50500 $ kg. [d] Calculated for 10 000 t acrolein production. An acrolein yield of 80 mol % was assumed. The
density used was 520 kg m13, and GHSV used was 100 h1. [e] Lower limit: 0.483 MM$  1.01 = 0.488 MM$; upper limit: 4.83 MM$  1.01 = 4.88 MM$.

1176

www.chemsuschem.org

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemSusChem 2012, 5, 1162 1180

Comparison of Petroleum-Based and Bio-Based Acrolein Production


component catalyst is 12 years. According to the estimation
Based on the estimation above, at the current developmenof catalyst lifetime and the regeneration by Atia et al.,[98] it is
tal stage of gas-phase glycerol dehydration, the acrolein price
reasonable to assume that each batch of the catalyst can last
provided by the bio-based route using refined glycerol as the
for about half a year. The catalyst quantity required for an
feedstock is equivalent to the current manufacturing method;
annual production of 10 000 tons of acrolein through the prothe price is significantly reduced (3142 %), if the production
pylene process is 7083 kg, assuming 80 mol % acrolein yield
begins with crude glycerol. It is worth noting that our estimaand 100 h1 gas-hourly-space velocity (GHSV). With these astion was based on the modest assumptions of the current desumptions, the total cost of the acid catalyst is 1.01 times that
velopmental stage of the bio-based method (GHSV 100 h1
of the multi-component catalyst for an annual production of
and 80 mol % acrolein yield). Compared to over a half-century
10 000 tons of acrolein (Table 6), where the factor of 1.01 is caldevelopment of the propylene-based process, extensive reculated through (1.62  7083/0.5)/(3.539657/1.5) total $ per
search on the gas-phase glycerol dehydration began just severyear. Any improvement in GHSV and catalyst lifetime would
al years ago. Research is underway to improve the acrolein
reduce the ratio of catalyst cost of the glycerol process to that
yield and, more importantly, TOS. A longer TOS allows larger
of the propylene process. If an assumption is made that the inGHSV for a given operation period. Any improvement in GHSV,
dustrial catalyst price falls within the range of 50500 $ kg1
the acrolein yield (higher than 80 mol %) and/or catalyst lifetime would further lower the acrolein price produced from
for the propylene process (0.4834.83 MM per year), the cost
glycerol.
of the catalyst for the glycerol-based method is 0.488
The glycerol market has been very volatile in the past few
4.88 MM$ per year. The difference of the catalyst cost between
years. Therefore, a sensitivity analysis was conducted to calcuthe bio-based method and the propylene-based (DCcat) then
late the bio-based acrolein price with a wide price range of
becomes 0.0050.05 MM$.
glycerol sources (both refined and crude, Figure 10). The price
Based on Equation (7), the difference of Cv between the biobased acrolein production from
refined glycerol and the propylene-based method (DCv,) is
(2.78)0.44 MM$, whereas DCv
from crude glycerol is (13.03)
9.86 MM$.
The current acrolein price is
approximately 27002850 $ per
ton.[124, 125] Equivalently, the total
annual market sales of 10 000
tons of acrolein would be 27.00
28.50 MM$ from a propylenebased process. If the profit is assumed to be 20 % of the total
annual market acrolein sales
(lY), the summation of Cv and
Cfix is equal to 21.6022.8 MM$
for current propylene production. If the same profit (5.40
5.70 MM$) is assumed, based on
Equation 5 and the variation of Figure 10. The calculated price of bio-based acrolein as functions of the price of refined/crude glycerol and the
price of the catalyst for propylene oxidation.
Cv and Cfix discussed above, the
total annual market sales of
of bio-based acrolein from crude glycerol is still considerably
acrolein prepared by the bio-based route from refined glycerol
lower than that of propylene-base acrolein even if the price of
should range from 24.81 [(15.742.78) MM$Cv+1.1  (21.60
crude glycerol reaches 250 $ per ton.
15.74) MM$Cfix+5.40 MM$profit) to a value of 29.65 MM$
In addition, it is also worth mentioning that government
[(15.74+0.44) MM$Cv+1.1  (22.815.74) MM$Cfix+$5.70MMprofit).
policies and regulations are usually in favor of the sustainable
Consequently, the market price of acrolein from refined glycerbio-based production. For example, the Renewable Energy and
ol is 24812965 $ per ton. The total annual market sales acroleJob Creation Act (H.R. 6049) that was passed in the USA in
in prepared through the bio-based route from crude glycerol
Sept. 2008 provides approximately 18 billion $ of tax incentives
should be in the range 15.73 [(15.7413.03) MM$Cv+1.3 
for investment in renewable energy, carbon capture and se(21.6015.74) MM$Cfix+5.40MM$profit) to 20.76 MM$ [(15.74
questration demonstration projects, and energy efficiency and
9.86) MM$Cv+1.3  (22.8-15.74) MM$Cfix +5.70 MM$profit). These
conservation. Numerous tax credit programs have long existed
calculations result in a market price of acrolein from crude
supporting biodiesel production. Several proposals are on the
glycerol of 15732076 $ per ton.
way, which provide some possibilities, such as a 30-cents-perChemSusChem 2012, 5, 1162 1180

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemsuschem.org

1177

L. Liu, X. P. Ye, and J. Bozell


pound tax credit for bio-based chemical production in largescale plants, and even a higher rate for small-scale producers.[126] In our calculation, the tax expenses were assumed to
be the same for both processes. However, the fact is that this
part of the expense would be considerably lower for the biobased processes because of the varieties of tax credits. Therefore, the bio-based acrolein production cost could be further
lowered.
Optional strategies are available as alternatives to launching
a completely new bio-based acrolein production plant. As
mentioned before, both gas-phase reactions share some operation similarity. The current manufacturing method uses a PRB
operating in the range of 300400 8C. These conditions should
not require considerable modification to switch to a gas-phase
glycerol dehydration method: the gas pipelines can be readily
connected to a different carrier gas source and, after changing
the heat exchanger to a furnace, the PRB is ready to operate
after changing the oxidation catalyst to the dehydration acid
catalyst. The post-processing layouts may be directly utilized in
the bio-based production with minimal adjustment. Therefore,
the installation cost, including the cost of major equipment,
the installation, engineering and field expenses, and the contractors fee, can be minimized. Consequently, it is a very
tempting scenario for current acrolein manufacturers to launch
bio-based acrolein production on their current plant setting.
Also, instead of completely replacing the propylene method,
an integrated process combining the two methods and operating simultaneously, can fully take advantage of the exothermic
character of one reaction and the endothermic character of
the other, saving some of the production cost. The disclosure
of a combined process of glycerol dehydration and propylene
oxidation indicates that a transitional stage has begun to
launch a glycerol-conversion method industrially.[127]
As mentioned in Section 1.1, acrylic acid production is
a major consumer of acrolein; the production is growing each
year due to the market demand for acrylic acid. It is worth
nothing that there is barely any technical barrier, uncertainty,
or challenge existing in converting bio-based acrolein to acrylic
acid. The gas-phase catalytic oxidation of acrolein to acrylic
acid is an industrially well-developed process. The facility is
usually built on the site of the acrolein plant to minimize the
handling and transportation of acrolein. Usually, the oxidation
of acrolein to acrylic acid is catalyzed by VMo(W) mixed
oxides in a PRB at an elevated temperature (240300 8C). Acrolein conversion is typically around 99 % and selectivity to acrylic acid around 95 mol %.[128131] The current industrial practice
(technology and facilities) can be directly adapted in realizing
this process to produce bio-based acrylic acid.
The potential advantages of establishing bio-based acrolein
production are enormous, including ecological benefits, providing an alternative source of acrolein in the face of a diminishing petroleum resource, and the opportunity to expand the
bio-based industry. The price advantage of glycerol over propylene as a feedstock has already been demonstrated. The
continued growth of the biodiesel industry suggests that this
advantage will continue. Our analyses and conclusions not
only agree with some previous studies,[66, 78] but also are in line

1178

www.chemsuschem.org

with the current industrial activity.[132, 133] Arkema, the worlds


fifth largest acrylics manufacturer, is building a demonstration
plant of bio-based acrolein/acrylic acid production within
3 years.[134] The worlds third largest manufacturer, Nippon Shokubai, received 225 000 $ funding from the Japanese government in 2009, and has been carrying out research and development of a bio-based route from glycerol to acrolein/acrylic
acid ever since. A pilot plant is under development.[135]
There have been recent concerns regarding the diminishing
glycerol glut because of a reduction in the global growth rate
of biodiesel production. However, parallel development of
high-value products from the glycerol byproduct are anticipated to support and enable continued biodiesel production.
Such technology is not yet widely available, but applications
to commodity chemicals, such as acrolein, will drive demand
for more glycerol, and hence, more biodiesel. Potential further
developments in biodiesel production, including improveing
the energy efficiency of production from microalgae or
improving the conversion technology and oil yield from biodiesel crops through genetic modification, will further spur
development.

4. Summary and Outlook


The chemical industry must energize its proven technological
capabilities to change fundamentally how chemical intermediates and products are made. The continuation of a vibrant
chemical industry requires a radical change in strategy akin to
the introduction of the alternative feedstocks that have greater
sustainability and (potentially) lower cost. From aldol condensation, to partial oxidation of propylene, and then to the research interest in propane oxidation, history bears witness to
the incentive of the industrialization of a synthetic approach,
justifying the actions taken toward a bio-based route. The biobased route (producing acrolein from the biodiesel byproduct
glycerol) has already met some promising milestones at this
point; however, challenges remain. Research efforts underway
today focused on the bio-based route might be a necessity for
the survival and growth of acrolein (acrylic acid) manufacturers
tomorrow.

Acknowledgements
This work was supported by the U.S. Department of Agriculture
HATCH project No. TEN00325.
Keywords: biofuel

dehydrogenation
heterogeneous catalysis oxidation

glycerol

[1] W. G. Etzkorn, S. E. Pedersen, T. E. Snead, Acrolein and Derivatives,


2002, http://203.199.213.48/1013/.
[2] Evonik_Industries, Methioninea Success Story.
[3] National_Academies, Formaldehyde and Other Aldehydes., National Research Council (U.S.). Committee on Aldehydes, 1981.
[4] K. H. Bowmer, G. H. Smith, Weed Res. 1984, 24, 201 211.
[5] B. Katryniok, S. Paul, M. Capron, F. Dumeignil, ChemSusChem 2009, 2,
719 730.

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemSusChem 2012, 5, 1162 1180

Comparison of Petroleum-Based and Bio-Based Acrolein Production


[6] B. Katryniok, S. Paul, V. Bellire-Baca, P. Rey, F. Dumeignil, Green Chem.
2010, 12, 2079 2098.
[7] H. H. Szmant, Organic Building Blocks of the Chemical Industry, Wiley,
New York, 1989.
[8] M. Ai, Appl. Catal. 1991, 77, 123 132.
[9] J. L. Solomon, R. J. Madix, J. Phys. Chem. 1987, 91, 6241 6244.
[10] M. Imachi, N. W. Cant, R. L. Kuczkowski, J. Catal. 1982, 75, 404 409.
[11] P. Vitins, K. W. Egger, J. Chem. Soc,. Perkin Trans. 1 1974, 1292 1293.
[12] K. Nakagawa, Y. Teng, Z. Zhao, Y. Yamada, A. Ueda, T. Suzuki, T. Kobayashi, Catal. Lett. 1999, 63, 79 82.
[13] W. C. Serjak, W. H. Day, J. M. V. Lanen, C. S. Boruff, Appl. Environ. Microbiol. 1953, 2, 14 20.
[14] E. Voisenet, Ann. Inst. Pasteur (Paris) 1918, 32, 476 510.
[15] F. B. Humphreys, J. Inf. Dis. 1924, 35, 282 290.
[16] S. Vollenweider, C. Lacroix, Appl. Microbiol. Biotechnol. 2004, 64, 16
27.
[17] J. F. Brazdil, Top. Catal. 2006, 38, 289 294.
[18] N. Rahmat, A. Z. Abdullah, Renewable Sustainable Energy Rev. 2010, 14,
987 1000.
[19] S. Morais, A. A. Martins, T. M. Mata, J. Energy Inst. 2010, 83, 48 55.
[20] F. O. Lichts World Ethanol and Biofuels Report, Vol. 7, on http://www.agra-net.com/portal2/home.jsp?template = productpage&pubid =
ag072, 2008, p. 29.
[21] F. O. Lichts World Ethanol and Biofuels Report, Vol. 7, on http://www.agra-net.com/portal2/home.jsp?template = productpage&pubid =
ag072/, 2009, p. 2009.
[22] R. F. Howard, Handbook of Commercial Catalysts: Heterogeneous Catalysts, CRC Press LLC, Boca Raton, Florida, 2000.
[23] Y. A. Saleh-Alhamed, R. R. Hudgins, P. L. Silveston, Appl. Catal. A: Gen.
1995, 127, 177 199.
[24] E. H. Xie, Q. L. Zhang, K. T. Chuang, Appl. Catal. A: Gen. 2001, 220,
215 221.
[25] H. Redlingshfer, O. Krocher, W. Bock, K. Huthmacher, G. Emig, Ind.
Eng. Chem. Res. 2002, 41, 1445 1453.
[26] H. Kashiwabara, Y. Nakamura (Asahi Electro-Chemical Co., Ltd.), JP
43024645, 1968, p. 8.
[27] Y. Takayama, T. Ikeda (Mitsubishi Rayon Co., Ltd.), JP 44027202, 1969,
p. 4.
[28] A. Tokutaniyama, T. Kato, K. Baison (Asahi Chemical Industry Co., Ltd.),
JP 44013130, 1969, p. 5.
[29] T. Ikeda, H. Ishii, T. Nakano (Mitsubishi Rayon Co., Ltd.), , 1969, JP
44027202, p. 3.
[30] H. Ito, H. Inoue, Y. Nakamura (Toa Gosei Chemical Industry Co., Ltd.),
JP 44025047, 1969, p. 7.
[31] E. Koberstein, T. Luessling, E. Noll, H. Suchsland, W. Weigert (Degussa
AG), Catalyst for the oxidation of alkenes, ZA 6906179, 1970, p. 12.
[32] S. Takenaka, G. Yamaguchi (Nippon Kayaku Co., Ltd.), JP 44006245,
1969, p. 4.
[33] S. Takenaka, Y. Kido, T. Shimabara, M. Ogawa (Nippon Kayaku Co.,
Ltd.), DE 2038749, 1971, p. 13.
[34] S. Takenaka, G. Yamaguchi (Nippon Kayaku Co., Ltd.), US 3959384,
1974, p. 4.
[35] T. Koshikawa (Mitsubishi Petrochemical Co., Ltd.), JP 49030308, 1974,
p. 6.
[36] S. Takenaka, G. Yamaguchi (Nippon Kayaku Co., Ltd.), JP 49003510,
1974, p. 3.
[37] K. Sakakibara, I. Abe, K. Matsuoka (Daicel Ltd.), JP 49075514, 1974,
p. 6.
[38] I. Ono, T. Iiauni, M. Akashi (Rohm and Haas Co), US 3968165, 1974,
p. 5.
[39] I. Nagai, I. Yanagisawa, M. Ninomiya, T. Oohara (Nippon Shokubai
Kagaku Kogyo Co., Ltd.), JP 51004113, 1976, p. 4.
[40] Y. Umemura, K. Oodan, K. Suzuki, Y. Bandou, T. Hisayuki (Ube Industries, Ltd.), JP 55157529, 1980, p. 6.
[41] Acrolein, JP 55160737, Ube Industries, Ltd., Japan, 1980, 6 pp.
[42] M. Takata, R. Aoki, T. Sato (Nippon Shokubai Kagaku Kogyo Co., Ltd.),
DE 3300044, 1983, p. 23.
[43] H. Kieser, K. Ohl, R. Thaetner, R. Feldhaus, K. Anders (Leuna Werke
VEB), DD 204406, 1983, p. 11.
[44] T. Shiotani, T. Kuroda (Mitsubishi Rayon Co, Ltd.), JP 6210183, 1994,
p. 5.

ChemSusChem 2012, 5, 1162 1180

[45] T. Shiotani, T. Kuroda. Manufacture of catalysts for synthesis of unsaturated aldehydes and carboxylic acids, JP 07016464, Mitsubishi Rayon
Co, Japan, 1995, 5 pp.
[46] T. Shiotani, T. Kuroda, Y. Taniguchi (Mitsubishi Rayon Co, Ltd.), JP
08024652, 1996, p. 6.
[47] Y. Tanaka. (Daicel Chem.), JP 8040969, 1996, p. 5.
[48] T. S. Chang, D. H. Cho, D. K. Lee (Korea Res. Inst. Chem. Tech. [US]), US
6410800, 2002, p. 4.
[49] G. Liu, P. Liu, H. Sun, H. Wu. (Hefei Haili Technology Dev. Co., Ltd.), CN
101402044, 2009, p. 7.
[50] G. Creten, D. S. Lafyatis, G. F. Froment, J. Catal. 1995, 154, 151 162.
[51] Y. A. Saleh-Alhamed, R. R. Hudgins, P. L. Silveston, Chem. Eng. Sci. 1992,
47, 2885 2890.
[52] V. M. Bondareva, T. V. Andrushkevich, Y. D. Pankratiev, React. Kinet.
Catal. Lett. 1986, 32, 171 175.
[53] H. Fansuri, G. H. Pham, S. Wibawanta, D. K. Zhang, D. French, Surf. Rev.
Lett. 2003, 10, 549 553.
[54] T. A. Hanna, Coord. Chem. Rev. 2004, 248, 429 440.
[55] G. S. Patience, P. L. Mills in New Developments in Selective Oxidation II,
Vol. 82 (Eds.: V. C. Corberan, S. V. Bellon), Elsevier, Amsterdam, 1994,
pp. 1 18.
[56] T. M. Leib, P. L. Mills, J. J. Lerou, Coord. Chem. Rev. 1996, 51, 2189
2198.
[57] K. H. Song, S. E. Han, K. H. Park, Korean J. Chem. Eng. 2001, 18, 184
189.
[58] S. E. Han, J. Choe, K. H. Song, I. W. Kim, Ind. Eng. Chem. Res. 2003, 9,
301 305.
[59] H. Redlingshfer, A. Fischer, C. Weckbecker, K. Huthmacher, G. Emig,
Ind. Eng. Chem. Res. 2003, 42, 5482 5488.
[60] C. M. ONeill, E. E. Wolf, Ind. Eng. Chem. Res. 2006, 45, 2697 2706.
[61] A. Ulgen, W. Hoelderich, Catal. Lett. 2009, 131, 122 128.
[62] Preparation of acrolein, A. Wohl, B. Mylo, Ber. Dtsch. Chem. Ges. 1912,
45, 2046 2054.
[63] Preparation of acrolein at Stockholm. G. F. Bergh, J. Prakt. Chem. 1909,
79, 351 357.
[64] J. B. Senderens, Comptes Rendus Chimie 1910, 151, 530 532.
[65] S. Ramayya, A. Brittain, C. DeAlmeida, W. Mok, M. J. Antal, Jr., Fuel
1987, 66, 1364 1371.
[66] L. Ott, M. Bicker, H. Vogel, Green Chem. 2006, 8, 214 220.
[67] Acrolein synthesis from glycerol in hot-compressed water: M. Watanabe, T. Iida, Y. Aizawa, T. M. Aida, H. Inomata, Bioresour. Technol. 2006,
98, 1285 1290.
[68] W. Bhler, E. Dinjus, H. J. Ederer, A. Kruse, C. Mas, J. Supercrit. Fluids
2002, 22, 37 53.
[69] H. E. Hoyt, T. H. Manninen (U.S. Industrial Chemicals, Inc.), US 2558520,
1951, p. 2.
[70] N. Suzuki, M. Takahashi (Kao Corp.), JP 2006290815, 2006, p. 7.
[71] E. Tsukuda, S. Sato, R. Takahashi, T. Sodesawa, Catal. Commun. 2007, 8,
1349 1353.
[72] S.-H. Chai, H.-P. Wang, Y. Liang, B.-Q. Xu, Green Chem. 2007, 9, 1130
1136.
[73] M. Craus in Handbook of Heterogeneous Catalysis (Eds.: G. Ertl, H. Knzinger, J. Weitkamp), Wiley-VCH, Weinheim, 1997.
[74] W. Yan, G. J. Suppes, Ind. Eng. Chem. Res. 2009, 48, 3279 3283.
[75] T. Haas, A. Neher, D. Arntz, H. Klenk, W. Girke (Degussa AG), US
5426249, 1993, p. 3.
[76] K. Tanable, M. Misono, M. Ono, H. Hattori, New solid acids and bases:
their catalytic properties, Kadansha, Tokyo, and Elsevier, Amsterdam,
1989.
[77] X. Li, C. Zhang, C. Qin, C. Chen, J. Shao (Shanghai Huayi Acrylic Acid
Co., Ltd.), CN 101070276, 2007, pp. 10.
[78] A. Corma, G. W. Huber, L. Sauvanaud, P. OConnor, J. Catal. 2008, 257,
163 171.
[79] M. Ghlich, K. Ruchle, H. Toufar, W. Reschetilowski, DGMK Tagungsber.
2008, 3, 209 215.
[80] A. Neher, T. Haas, D. Arntz, H. Klenk, W. Girke (Degussa AG), US
5387720, 1993, p. 3.
[81] Y. T. Kim, K.-D. Jung, E. D. Park, Microporous Mesoporous Mater. 2010,
131, 28 36.
[82] S.-H. Chai, H.-P. Wang, Y. Liang, B.-Q. Xu, J. Catal. 2007, 250, 342 349.

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemsuschem.org

1179

L. Liu, X. P. Ye, and J. Bozell


[83] M. Okuno, E. Matsunami, T. Takahashi, H. Kasuga, M. Okada, M. Kirishiki
(Nippon Shokubai Co., Ltd.), WO 2007132926, 2007, p. 55.
[84] Y. Arita, H. Tsuneki, H. Kasuga, M. Okada, M. Kirishiki (Nippon Shokubai
Co., Ltd.), WO 2008066082, 2008, p. 46.
[85] Q. Liu, Z. Zhang, Y. Du, J. Li, X. Yang, Catal. Lett. 2009, 419 428.
[86] M. N. Timofeeva, Appl. Cat. A: Gen. 2003, 256, 19 35.
[87] Y. Wu, X. K. Ye, X. G. Yang, X. P. Wang, W. L. Chu, Y. C. Hu, Ind. Eng.
Chem. Res. 1996, 35, 2546 2560.
[88] I. V. Kozhevnikov, J. Mol. Catal. A: Chem. 2007, 262, 86 92.
[89] I. V. Kozhevnikov, Chem. Rev. 1998, 98, 171 198.
[90] B. B. Bardin, R. J. Davis, Appl. Catal. A: Gen. 2000, 200, 219 231.
[91] W. L. Chu, X. G. Yang, Y. K. Shan, X. K. Ye, Y. Wu, Catal. Lett. 1996, 42,
201 208.
[92] M. J. Verhoef, P. J. Kooyman, J. A. Peters, H. van Bekkum, Microporous
Mesoporous Mater. 1999, 27, 365 371.
[93] Y. Izumi, R. Hasebe, K. Urabe, J. Catal. 1983, 84, 402 409.
[94] K. Nowinska, R. Field, J. Adamiec, J. Chem. Soc. Faraday Trans. 1991,
87, 749 753.
[95] L. Ning, Y. Ding, W. Chen, L. Gong, R. Lin, L. Yuan, Q. Xin, Chin. J. Catal.
2008, 29, 212 214.
[96] J.-L. Dubois, Y. Magatani, K. Okumura (Arkema France), WO
2009128555, 2009, p. 26.
[97] J.-L. Dubois, Y. Magatani, K. Okumura (Arkema France), WO
2009127889, 2009, p. 25.
[98] H. Atia, U. Armbruster, A. Martin, J. Catal. 2008, 258, 71 82.
[99] S.-H. Chai, H.-P. Wang, Y. Liang, B.-Q. Xu, Green Chem. 2008, 10, 1087
1093.
[100] A. Alhanash, E. F. Kozhevnikova, I. V. Kozhevnikov, Appl. Catal. A: Gen.
2010, 378, 11 18.
[101] S.-H. Chai, H.-P. Wang, Y. Liang, B.-Q. Xu, Appl. Catal. A: Chem. 2009,
353, 213 222.
[102] Production of acrolein and acrylic acid through dehydration and oxydehydration of glycerol with mixed oxide catalysts: J. Deleplanque,
J. L. Dubois, J. F. Devaux, W. Ueda, Catal. Today 2010, 157, 351 358.
[103] C.-J. Jia, Y. Liu, W. Schmidt, A.-H. Lu, J. Catal. 2010, 269, 71 79.
[104] F. Wang, J.-L. Dubois, W. Ueda, J. Catal. 2009, 268, 260 267.
[105] W. Suprun, M. Lutecki, T. Haber, H. Papp, J. Mol. Catal. A: Chem. 2009,
309, 71 78.
[106] L. Liu, PhD thesis, University of Tennessee (Knoxville), 2011.
[107] L. Cheng, X. P. Ye, Catal. Lett. 2009, 130, 100 107.
[108] E. Yoda, A. Ootawa, Appl. Catal. A: Gen. 2009, 360, 66 70.
[109] K. Pathak, K. M. Reddy, N. N. Bakhshi, A. K. Dalai, Appl. Catal. A: Gen.
2010, 372, 224 238.
[110] M. Kuba, J.-L. Dubois, Spec. Chem. Mag. 2009, 29, 42 43.
[111] X. Li, C. Qin, C. Chen, J. Shao (Shanghai Huayi Acrylic Acid Co., Ltd.),
CN 101580461, 2009, p. 10.
[112] B. Katryniok, S. Paul, M. Capron, C. Lancelot, V. Bellire-Baca, P. R. a. F.
Dumeignil, Green Chem. 2010, 12, 1922 1925.

1180

www.chemsuschem.org

[113] D. C. Cameron, J. A. Koutsky, Conversion of Glycerol from Soydiesel Production to 1,3-Propanediol; Final Report, National Biodiesel Development
Board, UW-Madison, Madison, Madison, WI, 1994.
[114] E. Santacesaria, R. Tesser, M. D. Serio, L. Casale, D. Verde, Ind. Eng.
Chem. Res. 2010, 49, 964 970.
[115] C. A. Ramrez-Lpez, J. R. Ochoa-Gmez, M. Fern ndez-Santos, O.
Gmez-Jim
nez-Aberasturi, A. Aonso-Vicario, J. Torrecilla-Soria, Ind.
Eng. Chem. Res. 2010, 49, 6270 6278.
[116] T. Ohara, H. Sata, N. Shimizu, G. Prescher, H. Schwind, O. Weiberg in
Ullmanns Encyclopedia of Industrial Chemistry (Ed.: W. Gerhartz), VCH,
Weinheim, Germany, 1987.
[117] S. Matar, L. F. Hatch, Chemistry of Petrochemical Processes, Gulf Pub. Co,
Houston, TX, 1994.
[118] W. G. Etzkorn in Kirk-Othmer Encyclopedia of Chemical Technology, 5th
ed., Wiley, New York, 2006.
[119] K. Weissermel, H. J. Arpe, Industrial Organic Chemistry. Important Raw
Materials and Intermediates, VCH, Weinheim, 1978.
[120] B. G. Hermann, M. Patel, Appl. Biochem. Biotechnol. 2007, 136, 361
388.
[121] ICIS, Propylene Prices and Pricing Information, Vol. 2011, ICIS pricing,
2011.
[122] ICIS, Sample report: Glycerin (US Gulf), Vol. 2011, ICIS Pricing, 2010.
[123] D. T. Johnson, K. A. Taconi, Environ. Prog. 2007, 26, 338 348.
[124] Kunming Jiaxinde Chemicals Corporation, http://www.alibaba.com/
product-gs/297209147/Acrolein.html, accessed Aug. 01, 2011.
[125] Laohekou Jinghong Chemical Co., Ltd., http://www.diytrade.com/
china/4/products/3517179/Acrolein.html, last accessed February 2012.
[126] E. Voegele. Legislating Incentive in Biorefining, November 03, 2010,
http://www.biorefiningmagazine.com/articles/192/legislating-an-incentive.
[127] J.-L. Dubois (Arkema France), US 12162691, 2007.
[128] N. Bertolini, N. Ferlazzo (Euteco Impianti SpA), US 4289654, 1981, p. 6.
[129] U. Hammon, K. Herzog, H.-P. Neumann (BASF AG), US 5264625, 1993,
p. 4.
[130] W. Ruppel, U. Wegerle, A. Tenten, U. Hammon (BASF AG), US 5739391,
1998, p. 4.
[131] S. Unverricht, H. Arnold, A. Tenten, U. Hammon (BASF AG), US
6403829, 2002, p. 9.
[132] M. S. Reisch, Chem. Eng. News 2010, 88 (11), 12.
[133] M. S. Reisch, Chem. Eng. News 2010, 88 (21), 21 22.
[134] R. Coons, Arkema Finds Glycerol-to-Acrylic Acid Catalysts; Demo-Scale
Plant Possible on http://www.chemweek.com, April 10, 2009.
[135] N. Alperowicz, Shokubai Advances Glycerine-to-Acrylic Acid Process on
http://www.chemweek.com, December 1, 2009.
Received: August 10, 2011
Revised: December 12, 2011
Published online on April 11, 2012

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemSusChem 2012, 5, 1162 1180

Вам также может понравиться