Rapid conversion of glycerol to lactic acid under alkaline hydrothermal conditions, by using a continuous flow reaction system

Article abstract of DOI:10.1246/cl.131160

A rapid conversion of glycerol to lactic acid (lactate) could be successfully achieved under alkaline hydrothermal conditions, by using a continuous flow reaction system. A rapid conversion by a continuous flow-type reaction system made it possible to achieve the reaction under high-temperature conditions without side reactions. The rapid conversion and high reaction yield (reaction: 2min; yield: 90%) resulted from the rapid temperature-shift advantageous for the inhibition of side reactions.

Full text of DOI:10.1246/cl.131160

Received: December 11, 2013 | Accepted: December 19, 2013 | Web Released: December 26, 2013
CL-131160
Rapid Conversion of Glycerol to Lactic Acid under Alkaline Hydrothermal Conditions,
by Using a Continuous Flow Reaction System
Toshinori Shimanouchi, Shouhei Ueno, Kazuki Shidahara, and Yukitaka Kimura*
Graduate School of Environmental and Life Science, Okayama University, 3-1-1 Tsushimanaka, Okayama 700-8530
(
E-mail: yktkkimu@cc.okayama-u.ac.jp)
A rapid conversion of glycerol to lactic acid (lactate) could
be successfully achieved under alkaline hydrothermal condi-
tions, by using a continuous flow reaction system. A rapid
conversion by a continuous flow-type reaction system made
it possible to achieve the reaction under high-temperature
conditions without side reactions. The rapid conversion and high
reaction yield (reaction: 2 min; yield: 90%) resulted from the
rapid temperature-shift advantageous for the inhibition of side
reactions.
Reactor: SUS316
Inner diameter: 0.8 mm
Column length: 2 m
pump
2
0 MPa
(back pressure)
glycerol
NaOH
column oven
573 623K
water bath
293K
sample
Figure 1. Experimental set up of continuous flow reaction
system.
Glycerol is a main candidate to be used as a building block
especially in chemical synthesis using a C3 backbone. As a
biomass derivative, glycerol is currently produced in a large
amount as a by-product in manufacturing biodiesel by trans-
esterification of vegetable oils, in such a way that its annual
solution in the liquid state, the pressure was set to 20 MPa by
the back-pressure valve. The samples obtained were recovered
and the yields of both lactate and other compounds
were estimated by high-performance liquid chromatography
(HPLC): UV detector UV SPD-6A (Shimadzu, Kyoto, Japan),
column COSMOSIL 5C18-AR-II ODS column (4.6 © 150 mm)
(Nakalai tesque), and the column oven (CTO-10ASVP). 10 mM
phosphate buffer (pH 2.6) was used as a mobile phase. The
optical density at 215 nm was monitored. Prior to the experi-
ment, no significant change in yield of lactic acid was observed
in the case of SUS316 alone, indicating that the materials of the
reactor wall have no obvious catalytic role.
production from biodiesel has tripled, from 1995 to 2006, from
1
2
00000 to 600000 tons. Thus, methods for effective utilization
of glycerol to produce value-added chemicals have been
1­11
investigated.
Conversion of glycerol to lactic acid has attracted attention
because the lactic acid is a useful compound to obtain value-
2
­6
added materials such as lactide and poly(lactic acid).
The
clarification of reaction mechanisms is required to establish the
industrial chemistry with respect to lactic acid. At the present,
there are two different reaction mechanisms: (i) glycelaldehyde-
2­5
routed reaction mechanism
mechanism.
and (ii) acetol-routed reaction
In the continuous flow reaction system presented here, lactic
acid, formic acid, acrylic acid, and other compounds were
6
1
6
Lactic acid has been generated by batch reaction under
alkaline hydrothermal conditions.3 This method is a green
process for production of lactic acid, because water is used as
a reaction medium. It has been reported that a continuous
degradation of carbohydrates using a capillary reactor is
detected for 2 min (Figure S1a). By varying the alkaline
,4
16
hydrothermal conditions as shown in Table S1, the best yield
was 90% (Entry 1), as shown in Table 1. The performance of
catalysts for glycerol conversion is summarized in Table 1,
together with representative results previously reported. How-
ever, these previous reports were conducted in a batch reaction,
and at least 60-min was required to achieve 90% yield (Entry 2).
Metal-supported catalysts have recently been applied for the
conversion of glycerol to lactic acid. In the case of catalysts such
as Cu2O/NaOH, Au­Pt/TiO2, and Ru/C, the yield of lactic acid
was reported to be 42 to 73% (Entries 4­6). From these results,
we considered that the rapid conversion of glycerol to lactic acid
could be achieved by using the continuous flow reaction system.
From the reaction rate theorem, the conversion of glycerol
1
2
possible. Furthermore, such reactors have recently been
used successfully in various reaction and flow­multistep
1
3­15
reactions.
It is therefore expected that capillary reactions
will be applied to conversion of glycerol to lactic acid.
In this study, it was demonstrated that an alkaline hydro-
thermal reaction of glycerol to lactic acid could be rapidly
achieved and the reaction time could be hastened by 30-fold
over previous reports, which is the fastest example to the best
of our knowledge. We discuss the mechanism of the rapid
conversion of glycerol based on kinetic study, NMR, and
operational aspects.
Glycerol solution mixed with NaOH was prepared and
injected by pump (LC-20AD, Shimadzu Co., Ltd.), into a
SUS316 reactor incubated at 573­623 K, followed by the
termination of reaction by cooling solution at the cooling bath
1
6
to lactic acid requires one glycerol and one NaOH (Figure S1).
1
The reaction mechanism was also investigated by using H- and
H NMR. Figure 2 shows the H- and H NMR spectra of the
2
1
2
solutions after the hydrothermal reactions of glycerol with
NaOH or NaOD at 623 K. As described before, it was observed
that the gradual consumption of reactant glycerol was accom-
panied by the production of lactic acid and a small amount of
other compounds including formic acid. By comparing spectra
(a) and (b) to (c) in Figure 2, it can be observed that the H on the
(293 K) (Figure 1). The reaction time was set as the residence
time in the reactor (column length: 2 m) by varying the flow rate
¹
1
(
approximately 0.0033­0.033 mL s ). To maintain the reaction
Chem. Lett. 2014, 43, 535–537 | doi:10.1246/cl.131160
© 2014 The Chemical Society of Japan | 535

Table 1. Conversion of glycerol to lactic acid using various catalystsa
O–
OH
HO
HO
OH
catalyst
Water
73 ~ 623 K, NaOH
O
3
Batch (B) or
Flow (F)
Substrate
NaOH/Glycerol
Temp.
/K
Reaction
time/min
Yield
/%
Entry
Reaction condition and catalyst
Ref.
1
2
3
4
5
F
B
B
B
B
Hydrothermal
Hydrothermal
Hydrothermal
Cu2O, NaOH/N2-1.4 MPa
2 M/0.5 M
1.25 M/0.33 M
1.25 M/0.33 M
1.19 M/1.09 M
0.88 M/0.22 M
623
573
573
513
363
2
60
60
360
®
90
90
75
73.1
42
This study
4
5
10
6
b
Au­Pt/TiO2
Au:Pt = 1:1, 2.5 © 10 3 mmol),
¹
(
0.1 MPa O2
6
B
Ru/C, 4 MPa H2
0.8 M/1 wt %
473
180
55
9
a
b
Reaction conditions: conversion and yield of lacitic acid was determined by HPLC. Recalculated from the ref 10.
CH
3
CH(OH)COO–
Benzilic acid
rearrangement
(
C) OH (C)
OH
OH
OH
(
a) NaOH/H
2
O
H
H
C
D
O
-
O
O
-H
O
-H2
H O
2
1H-NMR
HO
OH
–
2
CH CH(OH)COO
3
H
O
O
OH
H
(B)
H
Keto-enol
OH–
(
D)
(D)
tautomerization H C
B
3
3
H C
OH
3
H C
Lactic acid
Lactate)
HCOO–
Glycerol
Acetol
Pyruvaldehyde
(
Scheme 1. Possible reaction mechanism of glycerol conver-
sion to lactic acid.
(
2
b) NaOD/D O
6
1H-NMR
CH
3
CH(OH)COO–
HCOO–
(
Figure S4);16 2) no detection of glyceraldehyde as a key
1
6
intermediate (Figure S1); 3) H2 gas production late in the
reaction process (data not shown).
(
c) NaOD/D O
2
CD
3
CD(OD)COO–
2H-NMR
CD
3
CD(OD)COO–
From the results, we believe that the conversion of glycerol
to lactic acid in the present continuous flow reaction system was
driven by the acetol-routed reaction mechanism as shown in
Scheme 1. In any case, the reaction mechanism in our case was
the same as that in the batch reaction system. Therefore, a rapid
conversion in the flow system was likely to be caused by other
factors different from the reaction mechanism.
DCOO–
8
4
0
ppm
1
2
Figure 2. H NMR and H NMR spectra for solution after the
hydrothermal reaction of 0.5 M glycerol at 623 K with 1.5 M
NaOD in D2O for 30 s.
Previously, a side reaction resulted in a decreased yield of
lactic acid under the alkaline hydrothermal conditions (593 and
3
,11
β-C of lactic acid has been almost completely transformed to D
613 K). As far as the reaction mechanism was the same as that
in the batch reaction, the temperature-shift might be a key factor.
We then investigated the influence of temperature-shift on the
glycerol conversion. Since a capillary reactor was used in the
present study, the rapid elevation and reduction of temperature
between 293 and 623 K could be achieved within 15 s whereas
when in D O, and that any remaining glycerol does not take
2
participate in the H­D exchange reaction. These results are
consistent with a previous report. Furthermore, a ketone
6
1
2
carbonyl group such as R ­CO­R as an intermediate product
might be formed during the production of lactic acid from
glycerol because of H­D exchange on the β-C of lactic acid,
which was supported by MS observations (Figure S2). It is
therefore thought that the reaction in the flow system used herein
is the acetol-routed reaction mechanism as reported by Zhang
et al. In this mechanism, glycerol is converted to acetol by
dehydration, followed by the elimination of H to induce
benzilic acid rearrangement
1
6
15 min was required in the batch reaction (Figure S5). The
influence of rapid temperature-shift on the production of by-
product was then compared between 573 and 623 K (Figure 3).
At 573 K, the yield of lactic acid in the batch system was 90%,
which was comparable with the previous report (Entry 2 in
Table 1). And the yield in the continuous system was at most
35%. By-products including formic acid and acetic acid were
not seen. In contrast, a significant production of by-products was
observed in the batch system at 623 K, which is in agreement
1
6
6
2
4,6,17
(Scheme 1). The obtained lactic
acid was racemic (Figure S3),16 meaning that the conversion of
glycerol to lactic acid appeared to be non-enantioselective.
On the other hand, the following experimental results
suggest that the glycelaldehyde-routed reaction pathway is
unlikely in the present reaction system: 1) high activation
energy of conversion from glycerol to 2-hydroxypropenal
3
with the previous report. Meanwhile, no significant production
of by-product was detected in the continuous flow reaction
system. These observations indicated that the rapid temperature-
shift succeeded in the inhibition of production of by-products.
536 | Chem. Lett. 2014, 43, 535–537 | doi:10.1246/cl.131160
© 2014 The Chemical Society of Japan

integration of Low-Carbon Society and Ensuring Food Safety
and Security for financial support.
8
4
0
0
0
References and Notes
1
2
L. Ott, M. Bicker, H. Vogel, Green Chem. 2006, 8, 214.
C. A. Ramírez-López, J. R. Ochoa-Gómez, M. Fernández-
Santos, O. Gómez-Jiménez-Aberasturi, A. Alonso-Vicario, J.
Torrecilla-Soria, Ind. Eng. Chem. Res. 2010, 49, 6270.
H. Kishida, F. Jin, T. Moriya, H. Enomoto, Kagaku Kogaku
Ronbunshu 2006, 32, 535.
3
4
5
6
7
8
9
H. Kishida, F. Jin, Z. Zhou, T. Moriya, H. Enomoto, Chem.
Lett. 2005, 34, 1560.
G LF A
Batch
G L F A
Flow
(2min)
G L F A G L F A
Y. Shen, S. Zhang, H. Li, Y. Ren, H. Liu, Chem.®Eur. J. 2010,
Batch
(60min)
Flow
(2min)
1
6, 7368.
(
60min)
Y. Zhang, Z. Shen, X. Zhou, M. Zhang, F. Jin, Green Chem.
2012, 14, 3285.
T. Mizugaki, T. Yamakawa, R. Arundhathi, T. Mitsudome, K.
Jitsukawa, K. Kaneda, Chem. Lett. 2012, 41, 1720.
A. Tsuji, K. T. V. Rao, S. Nishimura, A. Takagaki, K. Ebitani,
ChemSusChem 2011, 4, 542.
5
73 K
623 K
Figure 3. Temperature effect on the conversion of glycerol
and yields of lactate, formic acid, and acetic acid. G: glycerol;
L: lactate; F: formic acid; A: acetic acid. In the case of batch
system at 623 K, other peaks (not identified) were observed
although the data was not involved here. For batch system, 6 mL
of solution was loaded into the SUS with 0.013 m in inner
diameter (10 mL of volume) to start the reaction.
E. P. Maris, R. J. Davis, J. Catal. 2007, 249, 328.
10 D. Roy, B. Subramaniam, R. V. Chaudhan, ACS Catal. 2011,
1, 548.
11 Y. D. Long, F. Guo, Z. Fang, X.-F. Tian, L.-Q. Jiang, F. Zhang,
Bioresour. Technol. 2011, 102, 6884.
1
1
1
1
1
2 C. Usuki, Y. Kimura, S. Adachi, Chem. Eng. Technol. 2008,
1, 133.
3 J. Wegner, S. Ceylan, A. Kirschning, Chem. Commun. 2011,
7, 4583.
4 J. C. Pastre, D. L. Browne, S. V. Ley, Chem. Soc. Rev. 2013,
2, 8849.
5 T. Chinnusamy, S. Yudha, M. Hager, P. Kreitmeier, O. Reiser,
ChemSusChem 2012, 5, 247.
6 Supporting Information is available electronically on the CSJ-
Journal Web site, http://www.csj.jp/journals/chem-lett/index.
html.
In conclusion, we succeeded in the rapid conversion of
glycerol to lactic acid (reaction time: 2 min; yield: 90%), which
is the best performance to the best of our knowledge. This
achievement resulted from the rapid temperature-shift rather
than the reaction mechanism. Our conditions afforded the
racemic lactate, not L- or D-lactic acid, indicating the require-
ment of both the neutralization after the reaction process and the
induction of enantioselectivy. These are under investigation.
3
4
4
The authors are also grateful to the Special Project from
MEXT for Environmental and Life Scientific Research on the
17 W. S.-L. Mok, M. J. Antal, Jr., J. Org. Chem. 1989, 54, 4596.
Chem. Lett. 2014, 43, 535–537 | doi:10.1246/cl.131160
© 2014 The Chemical Society of Japan | 537