Hydrogenation of lactic acid on reduced copper-containing catalysts

Article abstract of DOI:10.1007/s11172-009-0144-x

The copper-containing catalysts were prepared by reduction of various oxide precursor compounds. Their catalytic properties were studied in the hydrogenation of lactic acid to propylene glycol under atmospheric pressure. An efficient catalyst, whose precursor compound is copper hydroxosilicate with the structure close to the natural mineral chrysocolla, was developed. The optimum conditions for the synthesis of propylene glycol were established that make it possible to achieve a conversion of lactic acid of 97% at atmospheric hydrogen pressure and 473 K. ; 2009 Springer Science+Business Media, Inc.

Full text of DOI:10.1007/s11172-009-0144-x

1
114
Russian Chemical Bulletin, International Edition, Vol. 58, No. 6, pp. 1114—1118, June, 2009
Hydrogenation of lactic acid on reduced copperꢀcontaining catalysts
M. N. Simonov, I. L. Simakova, T. P. Minyukova, and A. A. Khassin
G. K. Boreskov Institute of Catalysis, Siberian Branch, Russian Academy of Sciences,
5
prosp. Akad. Lavrent´eva, 630090 Novosibirsk, Russian Federation.
Fax: +7 (383) 330 8056. Eꢀmail: simakova@catalysis.ru
The copperꢀcontaining catalysts were prepared by reduction of various oxide precursor
compounds. Their catalytic properties were studied in the hydrogenation of lactic acid to
propylene glycol under atmospheric pressure. An efficient catalyst, whose precursor compound
is copper hydroxosilicate with the structure close to the natural mineral chrysocolla, was
developed. The optimum conditions for the synthesis of propylene glycol were established that
make it possible to achieve a conversion of lactic acid of 97% at atmospheric hydrogen pressure
and 473 K.
Key words: copperꢀcontaining catalysts, hydrogenation, lactic acid, propylene glycol.
7
Propylene glycol (propaneꢀ1,2ꢀdiol) is a valuable
It is known that alkyl formates are hydrogenated on
the copper chromite catalyst. Similar catalysts also
exhibit high activity and selectivity in the hydrogenation
chemical solvent and is used in the production of unsatꢀ
urated polyether resins (45%), nonionic detergents (7%),
nontoxic antifreezing agents (5%), and plasticizers, in the
production of wines and nonꢀalcoholic beverages, and in
food, cosmetic, and pharmaceutical industries. The world
production of propylene glycol is 1.5—2.0 million tonnes
per year and continuously increases by 5—6% on the
average annually. This is likely due to the fact that
propylene glycol is a promising substitute of ethylene
glycol because of its nonꢀtoxicity.
8
of lactic acid. However, an appropriate yield of the target
product in the case of using catalysts based on copper
chromite is achieved under high pressure, which substanꢀ
tially decreases the efficiency of the process.
Thus, the catalytic hydrogenation of carboxylic acids
and their esters in the presence of the proposed catalysts,
9
such as rhenium and ruthenium blacks (Re/C, Ru/C)
or CuCr O , in most cases requires severe conditions
2
4
The production of propylene glycol is traditionally
based on organic compounds of natural gas, whose
market cost tends to increase. The industrial production
of propylene glycol is based on the hydration of propylene
(Table 1).
Under the mildest conditions, the reduction of lactic
acid to propylene glycol occurs in the presence of the
silicaꢀsupported copper catalysts. The preparation of
the catalyst by impregnating silica gel (Cab—O—Sil
BH—5, Cabot Corporation) with a solution of copper
1
®
oxide in an acidic medium. The synthesis of propylene
®
oxide is a complicated multistage process characterized
by a low yield of the target product and formation of many
byꢀproducts.
nitrate hydrate in ethanol to incipient wetness followed
Therefore, an important task is the development of an
alternative environmentally safe method for the producꢀ
tion of propylene glycol. Its production from the renewꢀ
Table 1. Activity of various catalysts in the hydrogenation of
lactic acid to propylene glycol
2
—4
able plant sources of such raw materials as lactic acid,
Catalyst
T
P
/atm
Yield of propylene Reference
glycol (%)
5
,6
sugar, and sugar alcohols is of special interest.
/
К
The method of preparation of propylene glycol from
lactic acid by the selective reduction of the carboxy group
of lactic acid to the hydroxy group is most promising.
Various methods of reduction of the carboxy to
hydroxy group are described. For example, the Ru and
Ru/C catalysts were used for the synthesis of propylene
glycol from lactic acid. The reaction occurred with a
high selectivity to propylene glycol at 373—543 K and a
high hydrogen pressure of 140 bar.
CuCr O
Re black
Re black
488
378
423
423
210
258
200
145
97
84
85
74
40
88
8
9
2
2
3
3
2
4
5
1
wt.% Ru/C
3 wt.% Cu/SiO2 453
53
а
1.0
а
4
7.1
2
a
Hydrogenation in the gaseous phase, the flow rate of lactic acid
is WHSV = 0.07 h–1
.
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 6, pp. 1086—1090, June, 2009.
066ꢀ5285/09/5806ꢀ1114 © 2009 Springer Science+Business Media, Inc.
1

Hydrogenation of lactic acid
Russ.Chem.Bull., Int.Ed., Vol. 58, No. 6, June, 2009
1115
by reduction in a hydrogen flow (the copper conꢀ
copper atoms to form nitrogen. The reaction stoichiometrically
corresponds to the following equation:
2
2
tent ranged from 9 to 22 wt.%) was described. The
carboxy group of lactic acid was hydrogenated in the
vapor phase at hydrogen pressures of 0.9—27.5 bar at
N O + 2 Cu = Cu O + N .
2
2
2
3
83—548 K. The 100% conversion of lactic acid with a
A weighed sample of the oxide catalyst precursor was placed
into a temperatureꢀcontrolled fixedꢀbed reactor and reduced in
a hydrogen flow at 653 K. After the catalyst was reduced, the
reactor was cooled in a hydrogen flow to 333 K. Then an exactly
measured volume of nitrous oxide in a helium flow was injected
by pulses into the reactor. Quantitative analysis of gas mixtures
at the outlet of the reactor was carried out by GLC on an
LKhMꢀ80 chromatograph (NPO Agropribor) using a heatꢀ
conductivity detector and helium as a carrier gas. The comꢀ
ponents were separated on a Porapack Q capillary column
selectivity to propylene glycol of 88.2% was achieved at
the pressure 7.1 bar, T = 453 K, and the relative flow rate
—
1
of the reactant WHSV = 0.07 h . A drawback of this
catalyst is a low productivity with respect to the target
product.
The purpose of the present work is the search for a
catalytic system with the highest activity under mild conꢀ
ditions compared to the earlier known catalytic systems.
(
71 m, 500 μm, 50 μm). Injection of nitrous oxide pulses was
terminated when evolution of nitrogen no longer occurred.
The specific surface was calculated as follows. The surface
Experimental
area occupied by a surface copper atom located at the faces
The copperꢀcontaining oxide compounds of various comꢀ
positions and structures were chosen as catalyst precursors: a
sample of the Cu—Crꢀcontaining catalyst, viz., copper chromite,
CuCr O , with the structure of tetragonally distorted spinel
2
(
100), (110), and (111) is 0.065, 0.092, and 0.0563 nm ,
20
respectively. If the copper surface is presented by these faces in
an equal ratio, then the average surface area occupied by one
copper atom is 0.0711 nm . According to the stoichiometry of
2
2
4
1
0
(
I4 /amd, a = b = 6.031 Å and c = 7.780 Å) ; a sample of the
1
the reaction, the formation of one nitrogen molecule is accomꢀ
panied by the oxidation of two copper atoms. Then the metal
surface can be determined by the equation
Cu—Zn—Siꢀcontaining catalyst, copper—zinc hydroxosilicate,
Cu,Zn) [Si O ](OH) •nH O, with the structure close to the
(
3
4
10
2
2
1
1
zincsilite natural mineral ; a sample of the Cu—Siꢀcontaining
catalyst, copper hydroxosilicate, with the structure close to the
chrysocolla natural mineral, (CuH) [Si O ](OH) •nH O. A
SCu = 14.2•10–22•νN2•NA,
4
4
10
8
2
sample of the Cu—Crꢀcontaining catalyst was prepared by the
calcination at 900 °С of Cu—Cr hydrocarbonate prepared by
the coprecipitation of the cations from an aqueous solution of
their nitrates. Samples of the Сu—Siꢀ and Сu—Zn—Siꢀcontainꢀ
ing catalysts were prepared by the homogeneous deposition—
where νN2 is the number of moles of nitrogen, and NA is
Avohadro´s number.
Experiments on lactic acid hydrogenation were carried out
in the vaporꢀphase regime in the temperature range from 433 to
493 K (160—220 °С) at atmospheric hydrogen pressure in a glass
fixedꢀbed reactor. After the reductive activation of the catalyst
samples, the initial aqueous solution of lactic acid (17.5 wt.%)
was fed with a syringe pump into the reactor with a rate of 0.22
and 0.31 mL h–1. Hydrogen from a cylinder was preꢀpassed
through a column with the Ni—Cr catalyst for the purification
from oxygen impurities and through a column with the sorbent
for the purification from water admixtures and then fed to the
1
2
precipitation method with different copper contents (Table 2).
The procedure of sample preparation, their structures,
spectral and physicochemical properties, and the structure
evolution in a reductive medium have been described preꢀ
1
3—19
viously.
To activate catalyst precursors, a weighed sample of the
precursor (0.5 g) was placed in a fixedꢀbed reactor and treated
–
1
–1
with hydrogen with a flow rate of 4—10 L h at 653 K.
After activation the catalyst was cooled in a hydrogen flow
to the temperature of the reaction.
reactor with a rate of 4—10 L h . After one hour reactant
feeding the condensed reaction products were collected in a
sampler at specified time intervals and analyzed by GLC analysis.
Measurements were carried out until a constant composition of
the reaction mixture was established.
The procedure of quantitative determination of the specific
surface of metallic copper is based on nitrous oxide with surface
Table 2. Composition and properties of the copperꢀcontaining catalysts
0
2 –1
Sample
Precursor
Taa
Content of
Size of
Specific surface of Cu /m g
Xꢀray diffractionb Titrationc
/
К
copper (wt.%) particles/Å
Cu—Cr
CuCr O
573
653
653
653
28
13
14
45
50×100×100
30—70
30—50
—
∼ 8
∼ 13
∼ 27
—
6.0
7.0
31.0
19.0
2
4
Cu—Zn—Si (Cu0.3Zn ) [Si O ](OH) •nH O
Cu—Si—14 (CuH) [Si O ](OH) •nH O + SiO
2
Cu—Si—45
0.7 3
4
10
2
2
4
4
10
8
2
(CuH) [Si O ](OH) •nH O
4 4 10 8 2
a
Ta is the activation temperature.
The specific copper surface was calculated from the Xꢀray diffraction data.
The specific copper surface was determined by the N O titration.
b
c
2

1
116
Russ.Chem.Bull., Int.Ed., Vol. 58, No. 6, June, 2009
Simonov et al.
Liquid products were identified by the GC/MS method on a
medium, for example, in an atmosphere of unpurified
argon with oxygen, absorbed protons interact with oxygen
with the simultaneous oxidation of metallic copper and
VGꢀ7070 GC/MS instrument (VG Analytical) on a quartz
capillary column 30 m × 0.2 mm (Silicone SEꢀ30). Quantitative
analysis of the reaction mixtures was carried out by GLC on an
LKhMꢀ80 chromatograph (NPO Agropribor) with a flameꢀ
ionization detector using nitrogen as a carrier gas. The compoꢀ
nents were separated on a quartz packing column 2.0 m × 2.0 mm
with the FFAP supported liquid phase. The surface areas of the
chromatographic peaks were determined by the absolute stanꢀ
dardization method. Chromatograms were processed using the
Gepard program package.
+
Cu cations to return copper cations to the tetrahedral
1
5
positions of the chromite structure.
A distinctive feature of copper and zinc hydroxosilicate
is the irreversible character of reduction with the formaꢀ
tion of highly dispersed metallic copper particles on the
surface of the amorphous silicate structure. It is important
that the copper particles are stable and do not suffer sinꢀ
In GLC analysis the product content was determined by the
formula
17,19
tering in a wide temperature interval.
In the present work, we attempted to compare the
properties of the catalysts in the hydrogenation of lactic
acid.
,
According to the GLC analysis results, the major
products of lactic acid hydrogenation are propylene
glycol (PG) and propionic acid (PA). According to the
GLC analysis data, propanol, isopropanol, lactaldehyde,
and products of propylene glycol oligomerization in
amounts not higher than 2 wt.% are formed in the experiꢀ
ments.
where С is the content of the ith substance in the sample, wt.%;
Si is the surface area of the chromatographic peak of the ith
i
substance, mV s; Q is the Q factor for the ith substance; (S •Q )
i
j
j
is the sum of the surface areas of all peaks in the chromatoꢀ
gram, mV s.
The selectivity of the reaction to a product formation was
determined as the ratio of the amount of this product to the
2
1
amount of the converted starting reactant.
The results of testing the copperꢀcontaining catalysts
under comparable conditions are listed in Table 3.
The weight hourly space velocity (WHSV) was determined
as the ratio of the weight flow rate of the reactant (v ) to the
r
catalyst weight (mcat
)
Table 3. Catalytic properties of the copperꢀcontaining samples
(1 atm, T = 473 K, WHSV = 0.08 h–1
)
WHSV = v /m = v •ρ•ω/mcat,
r
cat
s
where v is the volume flow rate of the reactant, cm h–1; ρ is the
3
s
Sample
γ (%)
SPG (%)
SPA (%)
–
3
density of the solution, g cm ; ω is the content of lactic acid in
an aqueous solution, wt.%.
Cu—Cr
3
7
45
97
24
14
50
65
76
80
47
33
The statistical analysis of the determined concentration
values was performed by the leastꢀsquares method using the
Microcal Origin 6.0 program. The relative error of determination
of the starting and final products by GLC is ±5%.
Cu—Zn—Si
Cu—Si—14
Cu—Si—45
Note. The following designations are accepted here and in Table 4: γ
is the conversion of lactic acid, SPG is the selectivity of formaꢀ
tion of PG, and SPA is the selectivity of formation of PA.
Results and Discussion
Lactic acid was hydrogenated in the presence of the
copperꢀcontaining catalysts obtained by the reductive
activation with hydrogen of the copperꢀcontaining oxide
compounds of different composition and structure: copꢀ
per chromite CuCr O , copper—zinc hydroxosilicate
α (%)
1
00
3
2
4
(
Cu Zn ) [Si O ](OH) •nH O, and copper hydroxoꢀ
0.3 0.7 3 4 10 2 2
silicate (CuH) [Si O ](OH) •nH O.
80
20
4
4
10
8
2
These catalyst were chosen due to the difference in the
character of their interaction with hydrogen during
the reduction. It is known that planar particles of metallic
copper epitaxially bound to the structure of reduced
chromite are formed during the reduction of copper
1
3
2
chromite at a temperature not higher than 643 K.
1
A fraction of reacted hydrogen is desorbed in the form
of water, but the major part is absorbed in the lattice of
reduced chromite. It should be mentioned that a portion
of copper is reduced only to the state Cu . When the
obtained reduced system is placed in a mild oxidative
0
5
0
100
150
200
t/min
Fig. 1. Effect of time on the product content (α) in the reaction
mixture: propionic acid (1), propylene glycol (2), and lactic acid
(3); Cu—Cr catalyst, Т = 473 K.
+

Hydrogenation of lactic acid
Russ.Chem.Bull., Int.Ed., Vol. 58, No. 6, June, 2009
1117
The data indicate that a low conversion of lactic acid
is observed when copper chromite is used as a catalyst,
and the catalyst is rapidly deactivated during the hydrogeꢀ
nation (Fig. 1).
α (%)
2
60
Probably, under the reaction conditions (473 K), the
hydrogen ion absorbed in the structure of reduced copper
chromite rapidly interacts with lactic acid, resulting in the
oxidation of copper and deactivation of the catalyst
40
1
20
3
1
00
200
300
400 t/min
0
+
2 Cu2+
+
2
Сu + 4 H +
Fig. 3. Effect of time on the product content (α) in the reaction
mixture: propionic acid (1), propylene glycol (2), and lactic
acid (3); Cu—Si—45 catalyst, Т = 473 K.
The step of reduction of copper cations by molecular
hydrogen with the formation of absorbed hydrogen, which
could close the catalytic cycle, occurs very slowly at such
a low temperature
surface of metallic copper formed during the reduction of
the oxide precursors. In fact, for the Cu—Cr, Cu—Zn—Si,
and Cu—Si—14 catalyst samples the observed difference
in the catalytic activity values (see Table 3) corresponds
H + Cu2
+
2 H + Cu .
+
0
2
0
to the difference in the specific surface values of Cu
measured by the titration with N O and calculated from
Unlike copper chromite, the reduction of copper catꢀ
ions in the silicate composition is irreversible with the
removal of oxygen anions in the form of water. This
explains the rather slow deactivation of the catalyst and
their high stationary activity. Although the weight conꢀ
tents of copper in the starting catalysts are approximately
equal (the copper—silicon catalyst with a copper content
of 14 wt.% (Cu—Si—14) and the copper—zinc—silicon
catalyst (Cu—Zn—Si)), the values of conversion of lactic
acid for them differ strongly (Fig. 2). The highest converꢀ
sion of lactic acid is 97%, and the high yield of propylene
glycol is observed in the presence of the copper—silica
catalyst with the copper content 45 wt.% (Cu—Si—45)
2
the Xꢀray diffraction data. A high activity of the Cu—Si—45
catalyst sample observed in the hydrogenation of lactic
acid indicates, most likely, that the specific surface of
metallic copper is not the single factor determining
the catalytic activity of the reduced copperꢀcontaining
catalysts with a high copper content. It appears that the
composition and structure of the active site of the copperꢀ
containing catalyst in lactic acid hydrogenation need
further investigation.
Thus, a high activity of the Сu—Si catalysts having
the precursor compound with the structure of the chrysoꢀ
colla mineral type can be associated with stable highly
dispersed metallic copper particles that are formed during
(
Fig. 3).
Since all the catalysts studied exhibit catalytic activity
2
2
the reduction of these catalysts. Copper hydroxosilicates
with the structure of the chrysocolla mineral type can be
promising for the industrial production of dihydroxyꢀ
in the hydrogenation of lactic acid only after their reducꢀ
tive activation in a hydrogen atmosphere, it can be asꢀ
sumed that the differences in the catalytic activity of the
catalysts are due to the different values of the accessible
2
3
alkanes (Table 4).
The catalytic properties of the following copperꢀconꢀ
taining catalysts were studied in the present work: copper
chromite, copper—zinc hydroxosilicate, and copper
hydroxosilicate. The efficient catalytic system prepared
by the reduction of copper hydroxosilicate was found for
the hydrogenation of lactic acid. The optimum conditions
for the synthesis of propylene glycol were determined,
which make it possible to achieve a high conversion of
lactic acid (97%) at atmospheric hydrogen pressure.
α (%)
1
00
3
8
6
4
2
0
0
0
0
1
Table 4. Activity of the Cu—Si—45 catalysts23 and 10 wt.%
Cu/SiO (see Ref. 4) in the hydrogenation of lactic acid at 473 K
2
2
and a hydrogen pressure of 1 atm
1
50
100
150
200
t/min
WHSV/h–1
Sample
γ (%)
SPA (%)
Fig. 2. Effect of time on the product content (α) in the reaction
mixture: propionic acid (1), propylene glycol (2), and lactic acid
1
0 wt.% Cu—Si
0.1
0.11
7.3
95
75
65
Cu—Si—45
(
3); Cu—Zn—Si catalyst, Т = 473 K.

1
118
Russ.Chem.Bull., Int.Ed., Vol. 58, No. 6, June, 2009
Simonov et al.
The authors are grateful to the workers of the Analytical
14. A. A. Khassin, G. N. Kustova, H. Jobic, T. M. Yurieva,
Y. A. Chesalov,, G. A. Filonenko, L. M. Plyasova, V. N.
Parmon, Phys. Chem. Chem. Phys., 2009, DOI:10.1039/
B821381J.
Laboratory of the G. K. Boreskov Institute of Catalyꢀ
sis (Siberian Branch, Russian Academy of Sciences)
V. I. Zheivot for the development of the procedure of
GLC analysis of condensed products, V. A. Utkin for
performing GC/MS analysis, M. P. Demeshkina for preꢀ
paring the catalyst samples, and T. M. Yurieva for fruitful
discussions.
1
5. A. A. Khassin, T. M. Yurieva, L. M. Plyasova, G. N. Kustova,
E. Zhobik, A. Ivanov, Yu. A. Chesalov, V. I. Zaikovskii,
A. V. Khassin, L. P. Davydova, V. N. Parmon, Ros. Khim.
Zh., 2008, 32 [Mendeleev Chem. J. (Engl. Transl.), 2008].
16. T. M. Yurieva, G. N. Kustova, T. P. Minyukova, E. K. Poels,
A. Bliek, M. P. Demeshkina, L. M. Plyasova, T. A. Kriger,
V. I. Zaikovskii, Mater. Res. Innov., 5, 2001, 3.
References
1
7. T. M. Yurieva, T. P. Minyukova, G. N. Kustova, L. M.
Plyasova, T. A. Kriger, M. P. Demeshkina, V. I. Zaikovskii,
V. V. Malakhov, L. S. Dovlitova, Mater. Res. Innov., 2001,
1
2
. O. A. Reutov, A. L. Kurts, K. P. Butin, Organicheskaya
khimiya [Organic Chemistry], Ch. 4, Binom, Moscow, 2004,
5
, 74.
7
26 pp. (in Russian).
. Z. Zhang, J. E. Jackson, D. J. Miller, Appl. Catal. A, 2001,
19, 89.
1
1
8. T. P. Minyukova, I. I. Simentsova, A. V. Khassin, N. V.
Shtertser, N. A. Baronskaya, A. A. Khassin, T. M. Yurieva,
Appl. Catal. A: Gen., 2002, 273, 171.
9. T. P. Minyukova, N. V. Shtertser, A. A. Khassin, L. M.
Plyasova, G. N. Kustova, V. I. Zaikovskii, Yu. G. Shvedenkov,
N. A. Baronskaya, Johannes C. van den Heuvel, A. V.
Kuznetsova, L. P. Davydova, T. M. Yurieva, Kinet. Katal.,
2
3
4
. Pat. US 6455742; Chem. Abstr., 2002, 137, 249503.
. R. D. Cortright, M. SanchezꢀCastillo, J. A. Dumesic, Appl.
Catal. B, 2002, 39, 353.
. Pat. US 6479713; Chem. Abstr., 2002, 137, 325152.
. D. C. Cameron, N. E. Altaras, M. L. Hoffman, A. J. Shaw,
Biotechnol. Progr., 1998, 1, 116.
. J. W. Evans, P. S. Casey, M. S. Wainwright, D. L. Trimm,
N. W. Cant, Appl. Catal., 1983, 7, 31.
. E. Bowden, H. Adkins, J. Am. Chem. Soc., 1934, 56, 689.
. H. S. Broadbent, G. C. Campbell, W. J. Bartley, J. H.
Johnson, J. Org. Chem., 1959, 24, 1847.
5
6
2
008, 49, 865 [Kinet. Catal. (Engl. Transl.), 2008, 49, 821].
0. G. L. Kington, J. M. Holmes, Trans. Faraday Soc., 1953,
9, 417.
1. G. K. Boreskov, Geterogennyi kataliz [Heterogeneous Catalyꢀ
sis], Nauka, Moscow, 1986, 304 pp. (in Russian).
2
2
2
2
7
4
8
9
2. C. J. G. Van der Grift, A. Mulder, J. W. Geus, Appl. Catal.,
1
990, 60, 181.
1
1
0. O. Crottaz, F. Kubel, H. Schmid, J. Mater. Chem., 1997, 7,
43.
3. Pat. RF 2290994; Byul. Izobr. [Invention Bulletin], 2007,
1
No. 1 (in Russian).
1. Dana’s New Mineralogy, Eds R. V. Gaines, H. C. Skinner,
E. E. Foord, B. Mason, A. Rosenzweig, John Wiley and
Sons, New York, 1997.
1
1
2. C. J. G. Van der Grift, J. W. Geus, M. J. Kappers, J. H.
Van der Maas, Catal. Lett., 1989, 3, 159.
3. O. V. Makarova, T. M. Yurieva, G. N. Kustova, A. V.
Ziborov, L. M. Plyasova, T. P. Minyukova, L. P. Davydova,
V. I. Zaikovskii, Kinet. Katal., 1993, 34, 681 [Kinet. Catal.
Received December 2, 2008;
in revised form February 25, 2009
(
Engl. Transl.), 1993, 34, 683].