Biopropionic acid production via molybdenum-catalyzed deoxygenation of lactic acid

Article abstract of DOI:10.1039/c3gc36874b

As the search for non-fossil based building blocks for the chemical industry increases, new methods for the deoxygenation of biomass-derived substrates are required. Here we present the deoxygenation of lactic acid to propionic acid, using a catalyst based on the non-noble and abundant metal molybdenum under relatively mild reactive distillation conditions (200-270°C). Good yields of sodium propionate (up to 41%) can be obtained with either triethylene glycol dimethyl ether or water as the solvent in the presence of one equivalent of sodium hydroxide, without an external reductant, and using an industrially relevant feedstock of 44% lactic acid in water. Mechanistic studies show that several reactions occur simultaneously, including decarboxylation and decarbonylation of lactic acid, dehydration, and deoxygenation. The major pathway of propionic acid formation was determined with isotopic labeling studies to proceed via direct C-O cleavage, and to a lesser extent via the dehydration/hydrogenation pathway involving acrylic acid.

Full text of DOI:10.1039/c3gc36874b

Green Chemistry
View Article Online
View Journal
PAPER
Biopropionic acid production via molybdenum-
catalyzed deoxygenation of lactic acid
Cite this: DOI: 10.1039/c3gc36874b
Ties J. Korstanje, Hendrik Kleijn, Johann T. B. H. Jastrzebski and
Robertus J. M. Klein Gebbink*
As the search for non-fossil based building blocks for the chemical industry increases, new methods for
the deoxygenation of biomass-derived substrates are required. Here we present the deoxygenation of
lactic acid to propionic acid, using a catalyst based on the non-noble and abundant metal molybdenum
under relatively mild reactive distillation conditions (200–270 °C). Good yields of sodium propionate (up
to 41%) can be obtained with either triethylene glycol dimethyl ether or water as the solvent in the pres-
ence of one equivalent of sodium hydroxide, without an external reductant, and using an industrially rele-
vant feedstock of 44% lactic acid in water. Mechanistic studies show that several reactions occur
simultaneously, including decarboxylation and decarbonylation of lactic acid, dehydration, and deoxy-
genation. The major pathway of propionic acid formation was determined with isotopic labeling studies
to proceed via direct C–O cleavage, and to a lesser extent via the dehydration/hydrogenation pathway
involving acrylic acid.
Received 23rd November 2012,
Accepted 5th February 2013
DOI: 10.1039/c3gc36874b
www.rsc.org/greenchem
worldwide production is estimated at 300 kilotons per year
and is growing rapidly due to its use as a monomer for poly-
Introduction
As fossil resources such as oil, coal and natural gas are becom- lactic acid production.6,7 Multiple transformations of lactic acid
ing scarcer and therefore more expensive, the search for to platform chemicals have been reported, varying from decar-
alternative feedstocks for the chemical industry is increasing boxylation and decarbonylation to form acetaldehyde and
progressively. The most important alternative feedstock that respectively CO2 and H2 or CO and water, dehydration to form
could sustainably provide sufficient carbon-based material is acrylic acid and water, and deoxygenation to form propionic
lignocellulosic biomass.1–4 This sustainable feedstock consists acid.8–17
of three major fractions: cellulose, a polymer of hexoses; hemi-
Propionic acid is currently produced in ∼400 kilotons per
cellulose, a polymer of pentoses; and lignin, a highly branched year from fossil feedstocks via oxidation of ethylene, directly
aromatic polymer. All fractions have a high oxygen-to-carbon or in a two-step manner via propional, or as a side-product
ratio of close to 1, while fossil feedstocks usually have a very from naphtha oxidation. Its main use is as a food preservative
low oxygen-to-carbon ratio. The major challenge to enable the for animal feed and grain and in baked products and
use of lignocellulosic biomass as a feedstock for the chemical cheese.18,19 It can also be produced via fermentation of
industry accordingly lies in (partial) removal of oxygen-con- glucose with Propionibacteria, although this is not yet viable
taining moieties. One of the most direct ways of achieving this for large-scale commercial usage.20 Its production from lactic
is via dehydration and/or deoxygenation of bio-derived alco- acid has already been mentioned in the 19th century using
hols, yielding respectively an olefinic or an aliphatic moiety.5
hydriodic acid,21 but the current state of the art is the use of
One of the examples of biomass-derived molecules with a solid acid catalysts. Among these are supported nitrate and
high oxygen-to-carbon ratio is lactic acid. Bearing both a phosphate salts,8–12 tungsten- and molybdenum-based hetero-
hydroxyl and a carboxylic acid functionality, it is a highly func- polyacids,13 rare earth metal-doped zeolites,14 and supported
tionalized molecule, thus making its selective transformation platinum catalysts.15,16 The applied temperature range for
challenging. Lactic acid can be obtained from lignocellulosic these catalysts is usually 300–400 °C, and the highest reported
biomass at a low cost via bacterial fermentation. Current yield for propionic acid is 25% using a Pt/Nb2O5 catalyst at
350 °C and 57 bar H2 pressure. Noteworthy is the use of
homogeneous precious metal catalysts based on Pt, Pd or Ir.
PtH(PEt3)3 can convert lactic acid to propionic acid in up to
Organic Chemistry and Catalysis, Debye Institute for Nanomaterials Science, Utrecht
50% yield at a moderate temperature of 250 °C in water at
low pH.22
University, Universiteitsweg 99, 3584 CG Utrecht, The Netherlands.
E-mail: r.j.m.kleingebbink@uu.nl
This journal is © The Royal Society of Chemistry 2013
Green Chem.

View Article Online
Paper
Green Chemistry
Table 1 Molybdenum-catalyzed lactic acid deoxygenation under continuous distillationa
Entry
Catalyst
Solventb
Additive
Mass balancec (%)
Conversiond (%)
Propionic acid yieldd (%)
1
2
3
4
5
6
7
8
Mo(CO)6
MoO3
MoO2(acac)2
MoO2Cl2
MoO2Cl2
DPE
DPE
DPE
DPE
EC
TEGDME
None
TEGDME
TEGDME
TEGDME
TEGDME
TEGDME
TEGDME
—
—
—
—
—
—
—
—
99
96
96
94
86
93
80
94
83
95
97
92
88
90
72
95
95
97
92
86
92
89e
95
87e
86e
96e
1
1
6
3
2
4
3
4
7
5
8
6
—
MoO2Cl2
MoO2(acac)2
MoO2(acac)2
MoO2(acac)2
MoO2(acac)2
MoO2(acac)2
MoO2 ((tBuCO)2CH)2
None
9
—
10
11
12
13
NaOH (0.2 mol%)
NaOH (0.2 mol%)
NaOH (0.2 mol%)
NaOH (0.2 mol%)
a Reaction conditions: 4.5 mL L-lactic acid (54 mmol, 90% solution in water), 0.67 mmol catalyst, 5 g solvent, 200–220 °C, 1 h, 10 mL min−1 N2
flow, magnetic stirring, continuous distillation of products. b DPE: diphenyl ether, EC: ethylene carbonate, TEGDME: triethylene glycol dimethyl
ether. c Combined weight of distillate and residue relative to starting weight. d Determined by 1H NMR, carbon-based yield. e Slow addition of
L-lactic acid to catalyst solution.
Here, we present the use of non-noble metal complexes alcohols.25 Initially, we tested four different molybdenum com-
based on molybdenum and rhenium for the selective deoxy- plexes and all were found to be active in the deoxygenation of
genation of lactic acid to propionic acid under relatively mild lactic acid and gave propionic acid as the major low-boiling
conditions (200–270 °C). Previously, we have shown that these product, varying from trace amounts for Mo(CO)6 to 6% yield
complexes are active in the dehydration of a wide variety of for MoO2(acac)2 (Table 1, entries 1–4, acac = acetylacetonato).
alcohols, and can surpass common acid catalysts in both de- The high conversions observed here (>80%) can primarily be
hydration activity and olefin product selectivity.23–25
attributed to dimerization and polymerization of the lactic
acid as the major side reaction. We also tested two other sol-
vents, ethylene carbonate (EC) and triethylene glycol dimethyl
ether (TEGDME), and these gave similar results as diphenyl
ether (entries 4–6). We preferred a solvent that could dissolve
Results and discussion
Our initial efforts towards deoxygenation of lactic acid both water and lactic acid, and decided to continue our investi-
focussed on the use of rhenium-based catalysts, due to their gations with TEGDME as the solvent and MoO2(acac)2 as the
proven activity in various dehydration reactions.23,24 Attempts most promising catalyst. We tested slow addition of the lactic
to deoxygenate methyl lactate using rhenium(VII) oxide in a acid solution to the hot catalyst solution (entry 9), which
batch fashion using various reaction conditions yielded only yielded more propionic acid (7%) at a lower conversion, likely
minor amounts of methyl acrylate, while much polymerization due to lowering of lactic acid polymerization. When using a
occurred. We hypothesized that switching to a reactive distilla- small amount of sodium hydroxide (0.2 mol%) the yield
tion setup would lower polymerization, and switching the sub- increased, and upon combining slow addition and the pres-
strate to lactic acid would allow higher reaction temperatures ence of sodium hydroxide, further improvement of the propio-
due to its increased boiling point compared to methyl lactate nic acid yield to 8% was obtained (entries 10 and 11). The use
(248 °C versus 145 °C).26 Using diphenyl ether as the solvent, of a structurally related catalyst, MoO2((tBuCO)2CH)2, under
interestingly, minor amounts of propionic acid (3% yield) were these conditions resulted in a lower propionic acid yield (entry
obtained together with trace amounts of acetaldehyde and 2,5- 12). The mere presence of sodium hydroxide did not yield any
dimethyl-4-dioxolanone, a known condensation product of propionic acid (entry 13).
lactic acid with acetaldehyde.27 As the formation of propionic
Despite the use of reactive distillation conditions, consider-
acid is a reductive process and no external reductant is able amounts of di- and polymerization were observed under
present, one equivalent of lactic acid is likely consumed to all tested conditions. Therefore we tested whether also
produce the required equivalent of reducing agent (vide infra).
DL-lactide and polylactide could act as a substrate to obtain pro-
As rhenium-based catalysts showed poor performance in pionic acid. Using MoO2(acac)2 as the catalyst, TEGDME as the
the lactic acid deoxygenation reaction, we decided to test mol- solvent and 0.2 mol% of NaOH as the additive, DL-lactide was
ybdenum-based complexes as a catalyst in this reaction, as converted in 92% and 5% yield of propionic acid was obtained.
these were also shown to catalyze the dehydration of Under the same conditions, but with the addition of
Green Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
yield dropped to 26% (entry 4). Switching to diphenyl ether or
water as the solvent increased the conversion, but a slightly
lower yield of sodium propionate was obtained as compared to
the case when TEGDME was used as the solvent (entries 5 and
6). When performing the reaction without any additional
solvent, the yield of sodium propionate dropped considerably
to 15% (entry 8). The use of a feed of 44% lactic acid in water
gave somewhat better results than the use of a 90% solution,
resulting in 85% conversion and a sodium propionate yield of
41% (TOF = 69 h−1, TON = 69). The 44% lactic acid solution is
of interest from an industrial viewpoint, as this is a widely
used technical grade lactic acid stream.28 Another raw material
that can be supplied from lactic acid production is calcium
lactate, which is obtained during a purification step towards
concentrated lactic acid. We performed the reaction with
calcium lactate in water without the addition of NaOH, but
due to the very poor solubility of calcium lactate in water the
reaction was sluggish, and a calcium propionate yield of
approximately 20% was obtained.
Table 2 MoO2(acac)2-catalyzed deoxygenation of in situ formed sodium
lactate after complete distillation of solventa
Mass
balancec (%) (%)
Conversiond Sodium propionate
Entry Solventb
yieldd (%)
1
2
3
4
5
6
7
8
9
TEGDME
79
86
74
—
95
84
81
—
85
85
40
2
TEGDMEe 90
TEGDMEf 86
TEGDMEg 94
—
26
33
35
—
15
41
DPE
89
94
98
88
92
H2O
H2Of
None
H2Oh
a Reaction conditions: 0.67 mmol MoO2(acac)2, 55 mmol NaOH, 5 g
solvent, slow addition of 4.5 mL L-lactic acid (54 mmol, 90% solution
in water), 200–270 °C, 10 mL min−1 N2 flow, magnetic stirring,
complete distillation of solvent within 1 hour. b TEGDME: triethylene
glycol dimethyl ether, DPE: diphenyl ether. c Combined weight of
distillate and residue relative to starting weight. d Determined by
1H NMR, carbon-based yield. e No catalyst present. f Reflux instead of
distillation. g Direct addition of lactic acid and no N2 flow. h 44% lactic
acid solution in water used.
We also performed some experiments in a batch fashion in
a Parr autoclave using molybdenum-based catalysts in either
the presence or absence of additional hydrogen. In the
absence of hydrogen and a reaction temperature of 270 °C for
1 hour in TEGDME, we obtained 29% propionic acid yield at
82% conversion. In the presence of 10 bar of hydrogen a very
1 equivalent of water to enhance depolymerization, polylactide similar amount of propionic acid was obtained, with a some-
was converted to propionic acid in only 1% yield. Mass transfer what higher conversion of 92%. As the deoxygenation of lactic
limitations may be a limiting factor when polylactide is used acid to propionic acid requires one equivalent of reducing
as the substrate, as the polylactide does not completely dis- agent, it is surprising that the extra addition of hydrogen in
solve in the solvent mixture, thus hampering reactivity. The these batch reactions did not enhance the formation of propio-
fact, however, that propionic acid is detected indicates that nic acid.
indeed polylactide can be converted to propionic acid, and
thus that dimerization and polymerization of lactic acid are
not a dead end in the reaction, but can act as a lactic acid
reservoir.
Mechanistic studies
As the addition of small amounts of sodium hydroxide was In order to get a better understanding of the observed reacti-
found to be beneficial, we tested the addition of one equi- vities, we performed mechanistic studies on the molybdenum-
valent of sodium hydroxide, combined with the slow addition catalyzed lactic acid deoxygenation. The first relevant obser-
of lactic acid. The in situ formed sodium lactate was found to vation is the formation of several side products. One of the
convert smoothly to sodium propionate. As the salts are non- first side-products to be observed was acetaldehyde and its
volatile, we decided to completely distill off the solvent, raising condensation product with lactic acid, 2,5-dimethyl-4-dioxola-
the temperature at the end of the reaction to 270 °C. This none. The formation of the dioxolanone was only observed
resulted in a mixture of sodium lactate, sodium propionate when using diphenyl ether as the solvent, which does not mix
and small amounts of sodium acetate in the red-brown with water. In the presence of water the condensation reaction
residue, with 86% conversion and 40% yield of sodium propio- is reversible and the dioxolanone decomposes to acetaldehyde
nate (Table 2, entry 1). In the absence of the molybdenum cata- and lactic acid.
lyst only minor amounts of sodium propionate were detected
The formation of acetaldehyde from lactic acid has pre-
(entry 2), and when performing the reaction under reflux con- viously been observed as a product from both decarbonylation
ditions instead of distillation of the solvent, no sodium propio- and decarboxylation.9,10,13 In the first case carbon monoxide
nate was formed and only the starting material was retrieved and water are formed, and in the latter case carbon dioxide
(entry 3). As the temperature during distillation (220–270 °C) is and hydrogen. Both processes could provide an explanation
much higher than the reflux temperature, it is likely that the for the formation of propionic acid, as this requires one equi-
reflux temperature is insufficient for deoxygenation to occur. valent of reducing agent, which could be provided by both
When adding the lactic acid solution directly at the start of the carbon monoxide and hydrogen. To distinguish between these
reaction instead of slowly adding it to the hot catalyst solution, two processes, we performed GC-analysis on the headspace of
the conversion increased to 95%, but the sodium propionate the autoclave after lactic acid deoxygenation reaction in
This journal is © The Royal Society of Chemistry 2013
Green Chem.

View Article Online
Paper
Green Chemistry
TEGDME at 270 °C under autogenic pressure. This reaction equal amounts, as would be expected when acrylic acid is an
produced 11 bar of gas at the end of the reaction, estimated to intermediate. At the α-carbon a 24-fold enrichment compared
be equivalent to 13.5 mmol of gas, which is approximately to natural abundance is observed, while at the β-carbon a
13% yield based on the starting amount of lactic acid and the 9-fold enrichment is observed. This indicates that the predomi-
formation of 2 equivalents of gas from 1 equivalent of lactic nant process for the formation of propionic acid is direct deox-
acid. By GC analysis H2, CO, CO2, as well as methane were ygenation of lactic acid, as this would only incorporate
detected, indicating that both the decarbonylation and the deuterium at the α-carbon. The fact that we also find isotopic
decarboxylation pathway are active under the given reaction enrichment at the β-carbon indicates that the dehydration/
conditions.
By means of 1H and 13C NMR, the formation of small occurring, but to a lesser extent. This is also confirmed by the
amounts of acetic acid in this reaction was detected. The for- observed consumption of acrylic acid upon reacting
hydrogenation pathway with acrylic acid as an intermediate is
a
mation of acetic acid and methane has previously been attri- 1 : 1 mixture of lactic acid and acrylic acid (vide supra). The
buted to secondary reactions of acetaldehyde, either via observed deuterium incorporation at the β-carbon can also be
hydration/dehydrogenation or C–C cleavage, respectively.17,22 caused by hydride transfer reactions. When assuming that the
The former would yield another equivalent of hydrogen, while C–OH cleavage reaction proceeds via a carbenium intermedi-
the latter would release another equivalent of CO, in both ate, a 1,2-hydride shift from the β- to the α-carbon could occur,
cases increasing the amount of reducing agent formed from and subsequent hydride transfer would incorporate deuterium
lactic acid.
at the β-carbon. The relatively low enrichments observed (0.5%
Other products that were detected by NMR analysis after of the total amount of propionic acid formed) can be explained
the reaction are lactide and oligo- or polylactide, both originat- by slow hydrogen–deuterium exchange at a possible molyb-
ing from self-esterification of lactic acid. These compounds denum hydride intermediate (vide infra) compared to hydrogen
have been tested as starting materials in the deoxygenation transfer to the propionic acid. This difference also explains the
and show conversion towards propionic acid, thus showing observation that added hydrogen does not enhance the yield
that neither is a dead end for the reactivity in the deoxygena- significantly.
tion reaction.
We did not observe any acrylic acid in the catalytic experi-
One of the proposed intermediates in the formation of pro- ments, and the deuteration experiments indicated that the rate
pionic acid is acrylic acid, formed via dehydration of lactic of hydrogenation is (much) higher than the rate of de-
acid and subsequent hydrogenation. We have, however, never hydration. Therefore, we have further investigated this by
observed traces of acrylic acid or its sodium salt during any of testing the hydrogenation of acrylic acid in the presence of
our reactive distillation experiments. An alternative route for MoO2(acac)2 under 8 bar of hydrogen at 220 °C for 1 hour. In
the formation of propionic acid is the direct deoxygenation via this experiment acrylic acid was mostly polymerized, while
cleavage of the C–OH bond and subsequent hydride transfer. only a minor amount (2%) was converted to propionic acid.
To distinguish between these two pathways, we have per- This can be attributed to the slow activation of hydrogen by
formed deuterium labeling studies using deuterium gas in an the molybdenum catalyst, as molybdenum is an early tran-
autoclave experiment and monitored the incorporation of deu- sition metal in a d0 configuration and is thus not a good
terium at both the α- and β-carbon of the propionic acid hydrogen activator. It is reported that molybdenum(VI)oxo
product (Scheme 1). From these experiments we found that at complexes are able to activate hydrogen with participation of
both positions deuterium enrichment occurs, although not in the oxo ligand, to form a hydrido–hydroxo species, albeit with
a relatively high activation energy.29 Another possibility would
be the formation of a molybdenum hydride species via a
similar pathway. Molybdenum hydrides, although relatively
rare, are known in the literature, and especially oxomolybde-
num complexes can be protonated to a hydroxomolybdenum
intermediate, which can subsequently rearrange to form an
oxomolybdenum hydride species.30,31
We also tested the involvement of acrylic acid by starting
with a 1 : 1 mixture of lactic acid and acrylic acid under reac-
tive distillation conditions in TEGDME as the solvent and in
the presence of MoO2(acac)2 and NaOH. Under these con-
ditions the acrylic acid is consumed for 71%, and the lactic
acid is consumed for 80%, while propionic acid is formed in
30% yield (carbon-based yield, based on the amount of lactic
acid plus acrylic acid). When acrylic acid would be completely
inactive in this reaction, the maximum theoretical carbon-
based yield would be 25%, thus showing that under the
given reaction conditions acrylic acid is indeed one of the
Scheme 1 Two possible pathways for deuterium incorporation in propionic
acid using D2 gas.
Green Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
Scheme 2 Reaction overview of lactic acid degradation using MoO2(acac)2.
intermediates, confirming the results of the isotope labeling non-noble and relatively abundant metal molybdenum.
studies. With either triethylene glycol dimethyl ether or water as
Regarding the role of sodium hydroxide, it is likely that it is solvent under reactive distillation conditions and in the pres-
required to shift the coordination equilibrium towards the ence of 1 equivalent of sodium hydroxide, good yields of
most reactive molybdenum lactate species. It is known that a sodium propionate (up to 41%) were obtained without the
bisoxomolybdenum(VI) dilactate species can be formed, which addition of an external reductant. An industrially relevant
has its optimum stability at pH 4.7.32 The pH of our solution feedstock of 44% lactic acid in water was found to give the
in water was determined to be 5.5 at room temperature, which best results.
decreases upon heating due to the increasing self-ionization
Mechanistic studies have shown that several reactions occur
constant of water at higher temperature, making it possible simultaneously, e.g. di- and polymerization, decarboxylation
that this dilactate molybdenum species is a relevant intermedi- and decarbonylation, dehydration and deoxygenation reac-
ate in the catalytic cycle.
tions. Secondary reactions such as condensation and hydro-
Scheme 2 summarizes the mechanistic implications of our genation reactions were also found to occur. Isotopic labeling
observations. Lactic acid can be reversibly converted to both studies with deuterium gas showed that the major pathway for
lactide and oligo- or polylactide, can both be decarboxylated the formation of propionic acid proceeds via direct deoxygena-
and decarbonylated to form acetaldehyde and, respectively, tion (C–OH breakage). The dehydration/hydrogenation
CO2 and H2 or CO and H2O, and can also both be dehydrated pathway involving acrylic acid as an intermediate was found to
to acrylic acid and deoxygenated to propionic acid. Secondary be active as a minor reaction pathway.
reactions also occur, such as condensation of lactic acid with
Overall, our system opens up the pathway for the pro-
acetaldehyde to 2,5-dimethyl-4-dioxolanone, and hydrogen- duction of bio-based propionic acid from lactic acid, applying
ation of acrylic acid to propionic acid. Furthermore, secondary an abundant metal as the catalyst under relatively mild
reactions involving acetaldehyde are observed, yielding conditions.
methane and acetic acid.
Experimental
Conclusions
All chemicals were obtained from Aldrich or Acros and used
In conclusion, we have presented a catalytic system capable without further purification. 1H and 13C{1H} NMR spectra were
of deoxygenating lactic acid, using a catalyst based on the recorded on a Varian 400 MHz spectrometer in D2O or CDCl3
This journal is © The Royal Society of Chemistry 2013
Green Chem.

View Article Online
Paper
Green Chemistry
at 400 and 100 MHz, respectively, and referenced against
residual solvent signal (4.79 and 7.26 ppm for 1H and
77.16 ppm for 13C). 2H NMR spectra were recorded on a Varian
400 MHz spectrometer at 61 MHz in H2O.
4 D. B. Turley, in Introduction to Chemicals from Biomass, ed.
J. H. Clark and F. E. I. Deswarte, John Wiley & Sons,
Chichester, 2008, ch. 2, pp. 26–46.
5 M. Schlaf, Dalton Trans., 2006, 4645.
Batch experiments were performed in a 50 mL Parr auto-
6 L. Nattrass and A. Higson, NNFCC Renewable Chemicals
Factsheet: Lactic Acid, National Non-Food Crops Centre,
York, 2011.
7 R. H. Platel, L. M. Hodgson and C. K. Williams, Polym.
Rev., 2008, 48, 11.
8 D. Wadley, M. Tam, P. Kokitkar, J. Jackson and D. Miller,
J. Catal., 1997, 165, 162.
9 G. C. Gunter, D. J. Miller and J. E. Jackson, J. Catal., 1994,
148, 252.
clave with glass insert, by mixing 0.22
g (0.67 mmol)
MoO2(acac)2, 10 g of solvent, 4.5 mL (54 mmol, 90% solution
in water) L-lactic acid and, where applicable, 2.2 g (55 mmol)
NaOH. The autoclave is, where applicable, flushed three times
with hydrogen or deuterium and pressurized, and the mixture
is heated to the desired temperature under mechanical stirring
(1500 rpm). The desired temperature is maintained for 1 hour,
after which the autoclave is cooled to room temperature using
an ice bath.
10 G. C. Gunter, R. H. Langford, J. E. Jackson and D. J. Miller,
Ind. Eng. Chem. Res., 1995, 34, 974.
A typical reactive distillation reaction was performed by dis-
solving 0.22 g (0.67 mmol) MoO2(acac)2 and 2.2 g (55 mmol) 11 J. H. Hong, J.-M. Lee, H. Kim, Y. K. Hwang, J.-S. Chang,
NaOH in 5 g of triethyleneglycol dimethyl ether in a 50 mL S. B. Halligudi and Y.-H. Han, Appl. Catal., A, 2011, 396, 194.
three-neck round-bottom flask, fitted with a gas inlet, a 12 C. Paparizos, W. G. Shaw and S. R. Dolhyj (The Standard
septum and a distillation setup. The distillation setup is
cooled with water and the distillate is collected in a two-neck
round-bottom flask, connected to an additional cooler, con-
Oil Company), Catalytic Conversion of Lactic Acid and
Ammonium Lactate to Acrylic Acid, EU Pat., EU0181718,
1985.
taining acetone/dry ice. The solution is heated with a heating 13 B. Katryniok, S. Paul and F. Dumeignil, Green Chem., 2010,
mantle to 220 °C under magnetic stirring and a nitrogen flow 12, 1910.
of 10 mL min−1, and 4.5 mL (54 mmol, 90% solution in water) 14 H. Wang, D. Yu, P. Sun, J. Yan, Y. Wang and H. Huang,
L-lactic acid is slowly added by a syringe over the course of Catal. Commun., 2008, 9, 1799.
10 minutes. After complete addition of the lactic acid, the 15 J. C. Serrano-Ruiz and J. A. Dumesic, ChemSusChem, 2009,
temperature is further increased to distill off all volatiles to an
endpoint of ∼270 °C.
2, 581.
16 J. C. Serrano-Ruiz and J. A. Dumesic, Green Chem., 2009,
11, 1101.
Yields and conversions were determined by NMR by dissol-
ving an aliquot from both the distillate and the residue in D2O 17 W. S.-L. Mok, M. J. Antal and M. Jones, J. Org. Chem., 1989,
(CDCl3 when using diphenyl ether as the solvent) and adding
sodium acetate as the external standard.
54, 4596.
18 Product Safety Assessment: Propionic Acid, The DOW Chemi-
cal Company, Midland, MI, 2008.
19 H. Bülow, W. Kohler, H. Görth, U. Pakka and P. Kirsth, in
Propionsäure und –Derivate, Verlag Chemie, Weinheim,
1972, vol. 19.
20 Y. Liu, Y.-G. Zhang, R.-B. Zhang, F. Zhang and J. Zhu, Curr.
Microbiol., 2011, 62, 152.
Acknowledgements
This research has been performed within the framework of the
CatchBio program. The authors gratefully acknowledge 21 E. Lautemann, Justus Liebigs Ann. Chem., 1860, 113, 217.
support from the Smart Mix Program of the Netherlands 22 B. Odell, G. Earlam and D. J. Cole-Hamilton, J. Organomet.
Ministry of Economic Affairs and the Netherlands Ministry
Chem., 1985, 290, 241.
of Education, Culture and Science.
23 T. J. Korstanje, J. T. B. H. Jastrzebski and R. J. M. Klein
Gebbink, ChemSusChem, 2010, 3, 695.
24 T. J. Korstanje, E. F. de Waard, J. T. B. H. Jastrzebski and
R. J. M. Klein Gebbink, ACS Catal., 2012, 2, 2173.
25 T. J. Korstanje, E. Folkertsma, M. Lutz, J. T. B.
H. Jastrzebski and R. J. M. Klein Gebbink, Eur. J. Inorg.
Chem., 2013, DOI: 10.1002/ejic.201201350, in press.
References
1 R. D. Perlack, L. L. Wright, A. F. Turhollow, R. L. Graham,
B. J. Stokes and D. C. Erbach, Biomass as Feedstock for a 26 Physical constants of organic compounds. CRC Handbook
Bioenergy and Bioproducts Industry: The Technical Feasibility
of a Billion-Ton Annual Supply, U.S. Department of Energy
and U.S. Department of Agriculture, 2005.
of Chemistry and Physics, ed. W. M. Haynes, CRC Press/
Taylor and Francis, Boca Raton, FL, 92nd edn (Internet
Version 2012), 2012.
2 J. J. Bozell and G. R. Petersen, Green Chem., 2010, 12, 27 M. Farines and J. Soulier, Bull. Soc. Chim. Fr., 1970, 332.
539.
28 H. Benninga, in A History of Lactic Acid Making: A Chapter in
the History of Biotechnology, Kluwer Academic Publishers,
Dordrecht, 1990, p. 181.
3 P. N. R. Vennestrøm, C. M. Osmundsen, C. H. Christensen
and E. Taarning, Angew. Chem., Int. Ed., 2011, 50, 10502.
Green Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
29 P. Reis, P. Costa, C. Romão, J. Fernandes, M. Calhorda and 31 F. Cimadevilla, M. Esther García, D. García-Vivó,
B. Royo, Dalton Trans., 2008, 1727.
30 E. J. Sundstrom, X. Yang, V. Sara Thoi, H. I. Karunadasa,
M. A. Ruiz, M. Teresa Rueda and S. Halut, J. Organomet.
Chem., 2012, 699, 67.
C. J. Chang, J. R. Long and M. Head-Gordon, J. Am. Chem. 32 A. Beltrán-Porter, A. Cervilla, F. Caturla and M. José Vila,
Soc., 2012, 134, 5233.
Transition Met. Chem., 1983, 8, 324.
This journal is © The Royal Society of Chemistry 2013
Green Chem.