Tuning selectivity of Pt/CaCO3 in glycerol hydrogenolysis - A Design of Experiments approach

Article abstract of DOI:10.1016/j.catcom.2011.06.007

19 commercially available catalysts have been effectively screened in only 30 experiments for the hydrogenolysis of glycerol using a D-optimal design. Pt/CaCO₃ emerged as both an active and selective catalyst and was studied in greater detail using a Central Composite Design. Upon addition of boric acid, the Pt/CaCO₃ product selectivity could be changed from 1,2-propanediol towards lactic acid. The formation of 1,2-propanediol and lactic acid could be optimized by considering the Response Surface Model plots and formation pathways.

Full text of DOI:10.1016/j.catcom.2011.06.007

Catalysis Communications 13 (2011) 1–5
Contents lists available at ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short Communication
Tuning selectivity of Pt/CaCO3 in glycerol hydrogenolysis — A Design of
Experiments approach
Jeroen ten Dam a, Freek Kapteijn b, Kristina Djanashvili a, Ulf Hanefeld a,
⁎
a
Biocatalysis and Organic Chemistry, Department of Biotechnology, Delft University of Technology, Julianalaan 136, 2628 BL Delft, The Netherlands
Catalysis Engineering, Department of Chemical Engineering, Delft University of Technology, Julianalaan 136, 2628 BL Delft, The Netherlands
b
a r t i c l e i n f o
a b s t r a c t
Article history:
19 commercially available catalysts have been effectively screened in only 30 experiments for the hydrogenolysis
of glycerol using a D-optimal design. Pt/CaCO3 emerged as both an active and selective catalyst and was studied in
greater detail using a Central Composite Design. Upon addition of boric acid, the Pt/CaCO3 product selectivity
could bechanged from1,2-propanedioltowardslactic acid. The formation of 1,2-propanedioland lacticacid could
be optimized by considering the Response Surface Model plots and formation pathways.
© 2011 Elsevier B.V. All rights reserved.
Received 18 March 2011
Received in revised form 25 May 2011
Accepted 6 June 2011
Available online 13 June 2011
Keywords:
Design of Experiment
Heterogeneous catalysis
Glycerol
Boric acid
1,2-propanediol
Lactic acid
1. Introduction
borate esters in solution, and thereby, leaving a single hydroxyl group
accessible for dehydrogenation (Scheme 2) [11].
The production of chemicals from biomass is an essential feature of
a sustainable future [1,2]. Polyols and sugars constitute a considerable
fraction of the available biomass and are therefore an attractive
feedstock for bio-based chemicals [3,4]. These oxygen rich starting
materials need to be defunctionalized to produce appropriate building
blocks for the chemical industry. Hydrogenolysis is a key process to
achieve this goal [5].
Glycerol is an attractive starting material for hydrogenolysis, since
it can serve as a model compound for larger polyols and sugars [6]. The
added advantage is the economical relevance of its conversion into
1,2-propanediol (12PD) and 1,3-propanediol (13PD), both used in
polymer synthesis [7]. Interestingly, glycerol can also yield lactic acid
(LA), an oxidation product, under hydrogenolysis conditions [8,9].
This base-induced reaction resembles an intra-molecular Cannizzaro
reaction and needs to be taken into consideration in the overall process
(Scheme 1) [10].
Numerous papers describe the formation of 12PD through
hydrogenolysis of glycerol, employing different catalysts and condi-
tions [12–19]. This led to a general picture of the formation of 12PD.
Still, the effect of individual variables is not completely understood.
This is where a method such as Design of Experiments (DoE) can be
very efficient to readily determine the influence of several variables on
activity and selectivity of the catalysts, and elucidate the effects of BA
[20].
2. Experimental
Full details are given in the Supplementary Data.
2.1. Reaction procedure
Standard reaction mixtures were prepared by dissolving appropriate
amounts of glycerol and BA in demineralized water in the volumetric
flask adjusting the pH with 0.1 M aqueous sodium hydroxide (Tables S1
and S2, Supplementary Data).
The hydrogenolysis of glycerol was performed in a HEL PolyBLOCK
8, a parallel autoclave reactor system consisting of eight 16 mL vessels.
The appropriate amount of catalyst (5 mol% active metal, relative to
glycerol) was added to 5 mL of aqueous standard reaction mixture.
Reactors were purged 3 times with N2 (20 bar) and 3 times with H2
(40 bar). Then they were pressurized to 40 bar H2, stirred at 800 rpm
and heated to the reaction temperature within 20 min. Stirring was
The aim of this research is to identify a selective catalyst for the
hydrogenolysis of glycerol in water, and to achieve a better under-
standing of the variables influencing its activity and selectivity. Further-
more, we envision the influence of boric acid (BA) on selectivity of
catalysts in the hydrogenolysis reaction of glycerol through formation of
⁎
Corresponding author. Tel.: +31 15 278 93 04; fax: +31 15 278 14 15.
E-mail address: u.hanefeld@tudelft.nl (U. Hanefeld).
1566-7367/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.catcom.2011.06.007

2
J. ten Dam et al. / Catalysis Communications 13 (2011) 1–5
Scheme 1. The formation pathway of 12PD and LA from glycerol.
stopped after 18 h and the reactors were allowed to cool down to room
temperature. Then the reaction mixture was filtered over a nylon
micro filter (Rotilabo, 0.2 μm). An HPLC sample was prepared by
diluting 500 μL filtered reaction mixture using 500 μL 5 mM H2SO4. The
liquid products were analyzed by HPLC. Products detected by HPLC
were 12PD, LA, EG and EtOH [21].
includes the experimental error and the effects of any uncontrolled
variable present. This initial result allows for choosing an appropriate
catalyst, which can be optimized in a subsequent, more detailed study.
We envisioned that BA could influence the selectivity of the
hydrogenolysis by forming borate esters in solution. Borate esters are
only formed at alkaline pH [22]. Therefore, the pH values were varied
between 10 and 12. Preliminary experiments showed that hydro-
genolysis occurs at temperatures between 150 and 200 °C. The effect
of sulfolane as an additive was investigated, because earlier research
had demonstrated that this polar, aprotic solvent could affect the
selectivity of several catalysts [11,23].
3. Results and discussion
3.1. D-optimal design
A series of variables (temperature, pH, additive) and 19 hetero-
geneous catalysts, with comparable porosity and surface area within
each support class (Table 1), were screened for hydrogenolysis
activity, using a DoE approach to optimize the experimental effort.
Temperature was chosen for its expected influence on the conversion
of glycerol, while the effect of pH influences the equilibrium reaction
of glycerol with BA and can therefore affect the overall reaction as
well. Sulfolane has been shown to have an influence on reaction
selectivity and was therefore included in this initial screening [11].
Factorial designs or fractional factorial designs can be used in case
of screening of continuous variables, such as pH and temperature.
However, in this case some of the variables are discrete (active metal
and support type), for which a D-optimal design is more efficient. This
is due to the fact that a complete factorial is needed for every com-
bination of discrete variables, which hugely increases the experimen-
tal effort. The aim of the initial D-optimal screenings was to evaluate
the main influence of the individual variables, which results in a
minimal amount of experiments. For this reason the amount of con-
tinuous variables was limited to three. The design will result in a
model by fitting the experimental data in a polynomial, which will
only describe the main variables in the form of a first order polynomial
(Eq. (1)).
The catalysts were divided into two groups and a D-optimal design
was constructed for both. Design 1 was used to screen 15 catalysts,
considering the variables: active metal (Ni, Pd, Pt, Rh, Ru), support
type (Al2O3, C, SiO2), temperature (150–200 °C), pH (9–12) and
additive (0–5% v/v sulfolane). Design 2 was employed to screen the
remaining four catalysts (Ir/C, Pd/BaSO4, Pd/CaCO3, Pt/CaCO3) and the
same three continuous variables. The division of the catalysts in these
two groups was done since not all the combinations of active metal
and support type were available. Thus, using these two designs, 19
catalysts could be effectively screened in only 30, out of a total of 240,
experiments (Table S1, Supplementary Data).
It appeared that the Ru catalysts in design 1 were especially active.
However, they mainly degraded the glycerol and only small amounts
of C3 products were detected. Due to the setup of our parallel
autoclaves (no option for sampling over time) we could not compare
catalyst selectivities at comparable conversions. The reaction products
are prone to further degradation. Therefore, the performance of the Ru
catalysts will probably be improved when the reaction is stopped
before 18 h. Therefore, the ruthenium catalysts were excluded from
the dataset for polynomial fitting. This resulted in a statistical
polynomial model, of which the significant coefficients are plotted
in the coefficient plot regarding the glycerol conversion (Figure S1a).
It shows that only the variables active metal, support type and tem-
perature have a significant influence on glycerol conversion. In this
case the additive and pH had no significant influence on glycerol
conversion and are therefore not included in the coefficient plot.
In design 2, the Pd/CaCO3 catalyst appeared to be not active at all.
By excluding the three Pd/CaCO3 experiments, the total number of
y = β0 + β1x1 + β2x2 + ε
ð1Þ
Here, y is the response (in this case glycerol conversion), xi is
variable i, β0 is the intercept and βi is the coefficient of variable i and ε
Scheme 2. BA as a hypothetical selectivity inducer through the formation of borate esters.

J. ten Dam et al. / Catalysis Communications 13 (2011) 1–5
3
Table 1
3.2. LA formation
Porosity and BET surface area of catalysts.
It is remarkable that an oxidation product such as LA can be formed
under reductive hydrogenolysis conditions. Its hydrothermal forma-
tion from glycerol in the absence of a metal (95% conversion, 90%
selectivity, Scheme 1) was reported by Kishida et al. [24]. However,
300 °C was required to overcome the energy barrier raised by the
hydride elimination. Maris et al. reported the formation of LA at a
relatively reduced temperature (200 °C) in the presence of heteroge-
neous catalysts (Pt/C, Ru/C, AuRu/C or PtRu/C) [8]. Apparently, the
addition of a catalyst could lower the reaction temperature. This can
be explained by the presence of a metallic surface, facilitating the
dehydrogenation of a hydroxyl group. To exclude that boric acid is
sufficient to catalyze the LA formation, a blank experiment was per-
formed in the absence of catalyst. This showed only 0.5% LA formation,
confirming the catalytic role of the metal catalyst.
Catalyst
Ni/Al2O3
Metal loading (wt.%)
Vpore (mL/g)
SBET (m2/g)
a
50
5
5
5
5
75
5
5
5
5
5
66
5
1
5
5
5
5
5
0.52
0.24
0.44
0.40
0.38
0.34
0.70
0.68
0.77
0.54
0.57
0.17
1.00
1.03
0.95
0.99
0.03
0.03
0.04
209
171
212
150
100
803
951
1010
985
839
954
104
221
236
247
324
10
a
Pd/Al2O3
a
Pt/Al2O3
a
Rh/Al2O3
Ru/Al2O3
Ni/Ca
a
Pd/Ca
Pt/Ca
Rh/Ca
Ru/Ca
Ir/Cb
Ni/SiO2
Pd/SiO2
Pt/SiO2
Rh/SiO2
Ru/SiO2
a
a
a
a
a
b
b
Pd/BaSO4
Pd/CaCO3
Pt/CaCO3
3.3. Influence of BA on activity and selectivity
8
11
b
Our initial hypothesis was that BA could influence the selectivity of
the hydrogenolysis through formation of borate esters in solution,
leaving one hydroxyl group available for reduction. As a Lewis acid, BA
can also activate the hydroxyl groups of glycerol and increase the
activity.
a
Used in D-optimal design 1.
Used in D-optimal design 2.
b
catalysts was reduced from four to three and the results could be
correlated by a polynomial model. The coefficients of this model are
plotted in the coefficient plot regarding the glycerol conversion
(Figure S1b). This model showed that in addition to the catalysts,
temperature and the presence of sulfolane influence the conversion of
glycerol, with a large positive and a small negative effect, respectively.
pH had no significant effect on glycerol conversion and was therefore
omitted from the coefficient plot.
The effect of BA on the activity and selectivity of the catalysts was
investigated by repeating a set of six successful experiments from the
two D-optimal designs, this time in the absence of BA (Fig. 1). This
showed that in most cases, the activity is increased upon addition of BA
(Fig. 1a); in the case of Pt/CaCO3 the activity is even more than doubled.
Not only conversion of glycerol was increased (in most cases) in the
presence of BA, also less glycerol was degraded (Fig. 1b). Consequent-
ly, more glycerol is converted into desired products, best expressed for
a
b
100
100
80
60
40
20
0
0 mM BA
100 mM BA
0 mM BA
100 mM BA
80
60
40
20
0
3
2
4
3
3
O
Pt/SiO
2
4
3
O
2
2
Pt/C
Pt/C
Ru/C
Ru/C
Pt/SiO
Rh/Al
Rh/Al
Pt/CaCO
Pd/BaSO
Pt/CaCO
Pd/BaSO
c
d
100
80
60
40
20
0
100
80
60
40
20
0
0 mM BA
100 mM BA
0 mM BA
100 mM BA
3
2
4
3
3
2
4
3
O
O
2
Pt/C
2
Pt/C
Ru/C
Ru/C
Pt/SiO
Pt/SiO
Rh/Al
Rh/Al
Pt/CaCO
Pd/BaSO
Pt/CaCO
Pd/BaSO
Fig. 1. The influence of BA on a conversion; b degradation; c 12PD-selectivity and d LA selectivity. Some reactions were performed in duplicate or triplicate in order to establish the
error margin. Reaction conditions for the respective catalysts are given in Table S1, Supplementary Data.

4
J. ten Dam et al. / Catalysis Communications 13 (2011) 1–5
Rh/Al2O3 and Pt/SiO2, where the degradation of glycerol is almost
completely suppressed.
In addition to BA concentration (0–125 mM), the influence of the
variables temperature (130–220 °C) and pH (8–13) on activity and
selectivity of the catalyst were investigated. These three variables
could be effectively modeled using a CCD consisting of 20 experiments
(8 factorial points, 6 star points and 6 center points) (Fig. 2a). The
center point is repeated six times in order to obtain a good estimate for
the standard deviation of the system. The model for the conversion of
glycerol indicated that the variable temperature and the interaction
pH×temperature have a significant effect on the conversion of
glycerol. These effects are visualized in the RSM plot in Fig. 2b: pH
has an influence on glycerol conversion, but it is mainly controlled by
temperature.
Interestingly, BA's influence on selectivity is not clear-cut. In some
cases (Pt/C, Pt/SiO2) 12PD-selectivity was substantially increased
(Fig. 1c), while in other cases (Rh/Al2O3, Pt/CaCO3) LA formation was
favored upon the addition of BA (Fig. 1d). In case of Pt/CaCO3 the
increase of LA-selectivity was accompanied by a decrease of 12PD-
selectivity, resulting in a selectivity switch from 12PD to LA upon the
addition of BA.
3.4. Response Surface Model design
The selectivity of Pt/CaCO3 towards 12PD and LA was modeled by
omitting the selectivity data at conversions below 5%. These data were
excluded because the low conversion data contain a high uncertainty
in the selectivity. The resulting selectivity models show an influence
of all three variables, best visualized in RSM plots (Fig. 2c–f).
The selectivity, within variable space, towards 12PD is highest at
low BA concentration, low pH and low temperature (Fig. 2c–d).
However, at these conditions the conversion of glycerol is very low,
resulting in the formation of a small amount of 12PD.
A Central Composite Design (CCD) was set up to study the signif-
icant influence of BA on the selectivity of Pt/CaCO3. A CCD consists of a
full factorial design augmented with star points, which increase the
variable space and allow the modeling of nonlinear effects. A CCD
results in a Response Surface Model (RSM), which enables to draw
insightful RSM plots (Fig. 2, raw data in Table S2, Supplementary Data).
The LA-selectivity is the highest at high pH and BA concentration,
while the effect of temperature is less significant (Fig. 2e–f). This
means that a high LA-selectivity is possible at a relatively high glycerol
conversion. Interestingly, it is visualized that upon changing reaction
variables, Pt/CaCO3 can be either selective for 12PD or LA formation.
a
b
3.5. Increasing 12PD-selectivity
Based on the RSM plots in Fig. 2(c and d), it can be expected that
decreasing pH below pH 9 at 200 °C should increase 12PD-selectivity,
while maintaining a high conversion. Therefore, an additional reaction
was run at 200 °C, 100 mM BA concentration and pH 5.6, i.e. without
addition of sodium hydroxide (Fig. 3a). Indeed, this resulted in a
substantial increase of 12PD-selectivity, accompanied by a decimation
of the LA-selectivity. The overall yield of 12PD was increased, even
though the glycerol conversion was decreased. Mechanistically, the
increase in 12PD-selectivity can be explained by the suppression of the
LA formation pathway, since this is base-induced [23], leaving more
glycerol for hydrogenolysis to 12PD. A reason for the decrease of
conversion rate could be that at pH 5.6 only one pathway is opera-
tional, instead of two pathways at pH 12.
T
c
d
3.6. Increasing LA-formation
The RSM plots show that an optimum for LA-selectivity is close to,
or within the chosen parameter space (Fig. 2e–f). Therefore, changing
these variables would not substantially increase the LA-selectivity.
However, the selectivity could be increased by avoiding the 12PD
formation pathway. This can be achieved by running the reaction in
the absence of hydrogen, thereby prohibiting the formation of 12PD
and leaving the dehydrogenated glycerol available for LA formation.
Fig. 3b shows that in the absence of hydrogen pressure, the LA-
selectivity remains the same, while the conversion and, as a conse-
quence, the LA-formation were substantially increased. The 12PD-
selectivity had dropped considerably, but still small amounts of 12PD
were formed. Obviously, hydrogen, necessary for 12PD formation, is
produced during the reaction: one equivalent of hydrogen is formed
for each LA and even more can be formed during the degradation of
glycerol [18].
e
f
Fig. 2. RSM-plots resulting from a Central Composite Design. Relevant cross sections
through the center points were chosen and the value of the remaining parameter is
given in the respective footnote. Central Composite Design, the circles represent
3.7. Catalyst recycling
a
experiments within parameter space, blue=6×center point, red=8×factorial point,
green=6×star point; b RSM-plot for glycerol conversion at 62.5 mM BA; c RSM-plot for
12PD-selectivity at 62.5 mM BA; d RSM-plot for 12PD-selectivity at 175 °C; e RSM-plot
The carrier material of Pt/CaCO3 should be stable under the alkaline
conditions used in this study. However, the elevated temperatures and
stirring might degrade the catalyst to some extent. Therefore, the
f
for LA-selectivity at pH 10.5; RSM-plot for LA-selectivity at 175 °C.

J. ten Dam et al. / Catalysis Communications 13 (2011) 1–5
5
suppressing the base catalyzed LA formation pathway. Performing
the reaction in the absence of hydrogen increased the LA formation
due to interruption of the 12PD formation. Although LA-selectivity
remained the same, conversion was considerably increased, while
12PD-selectivity was diminished. The catalyst could be recycled and
maintained selectivity, but showed some deactivation.
This study shows that DoE was not only highly effective in iden-
tifying active catalysts, it also provided useful insights into the effects
of pH, BA concentration and temperature on the activity and selectivity
of the selected Pt/CaCO3. These insights were successfully used to
direct the selectivity of Pt/CaCO3 to either 12PD or LA.
a
60
50
40
30
20
10
0
pH 12
pH 9
pH 5.6
Acknowledgments
We are grateful to Dr. Silvia Pereira for assistance regarding DoE
and to Dr. Catherine Pinel (IRCELYON) for helping us with the HPLC
analysis. Jeroen ten Dam gratefully acknowledges financial support
from NWO ASPECT (053.62.020).
b
100
80
60
40
20
0
40 bar H2
20 bar N2
Appendix A. Supplementary data
Supplementary data to this article can be found online at doi:10.
1016/j.catcom.2011.06.007.
References
[1] A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 2411–2502.
[2] C.H. Christensen, J. Rass-Hansen, C.C. Marsden, E. Taarning, K. Egeblad,
ChemSusChem 1 (2008) 283–289.
[3] A. Brandner, K. Lehnert, A. Bienholz, M. Lucas, P. Claus, Top. Catal. 52 (2009)
278–287.
[4] Y.G. Zheng, X.L. Chen, Y.C. Shen, Chem. Rev. 108 (2008) 5253–5277.
[5] C.H.C. Zhou, J.N. Beltramini, Y.X. Fan, G.Q.M. Lu, Chem. Soc. Rev. 37 (2008)
527–549.
[6] J.J. Bozell, G.R. Petersen, Green Chem. 12 (2010) 539–554.
[7] J.V. Kurian, J. Polym. Environ. 13 (2005) 159–167.
[8] E.P. Maris, R.J. Davis, J. Catal. 249 (2007) 328–337.
[9] E.P. Maris, W.C. Ketchie, M. Murayama, R.J. Davis, J. Catal. 251 (2007) 281–294.
[10] P.P. Pescarmona, K.P.F. Janssen, C. Delaet, C. Stroobants, K. Houthoofd, A.
Philippaerts, C. De Jonghe, J.S. Paul, P.A. Jacobs, B.F. Sels, Green Chem. 12 (2010)
1083–1089.
[11] J. Chaminand, L. Djakovitch, P. Gallezot, P. Marion, C. Pinel, C. Rosier, Green Chem.
6 (2004) 359–361.
[12] J. Wang, S.H. Shen, B.D. Li, H.Q. Lin, Y.Z. Yuan, Chem. Lett. 38 (2009) 572–573.
[13] A. Shimao, S. Koso, N. Ueda, Y. Shinmi, I. Furikado, K. Tomishige, Chem. Lett. 38
(2009) 540–541.
[14] S. Sato, M. Akiyama, K. Inui, M. Yokota, Chem. Lett. 38 (2009) 560–561.
[15] T. Jiang, Y.X. Zhou, S.G. Liang, H.Z. Liu, B.X. Han, Green Chem. 11 (2009)
1000–1006.
Fig. 3. a Effect of pH on glycerol conversion and selectivity. Reaction conditions: 5 mol%
Pt/CaCO3, 5 mL aqueous reaction mixture (100 mM glycerol, 100 mM BA, variable pH),
b
200 °C, 40 bar H2, 18 h, 800 rpm; Effect of hydrogen pressure vs. inert atmosphere.
Reaction conditions: 5 mol% Pt/CaCO3, 5 mL aqueous reaction mixture (100 mM
glycerol, 100 mM BA, pH 12), 200 °C, 40 bar H2 or 20 bar N2, 18 h, 800 rpm.
stability of Pt/CaCO3 was investigated by means of a catalyst recycling
experiment (Figure S2, Supplemental Data).1
The selectivity of recycled catalyst is comparable with the
selectivity of fresh catalyst, but the activity decreased when using
the recycled catalyst. TEM analysis before and after reaction (Figure S3,
Supplementary Data) shows that the CaCO3 surface is partly eroded
and platinum particles have sintered. The resulting loss of active sites
can explain the lower catalyst activity.
[16] L.Y. Guo, J.X. Zhou, J.B. Mao, X.W. Guo, S.G. Zhang, Appl. Catal. A-Gen. 367 (2009)
93–98.
4. Conclusions
[17] L. Ma, D.H. He, Z.P. Li, Catal. Commun. 9 (2008) 2489–2495.
[18] E. D'Hondt, S. van de Vyver, B.F. Sels, P.A. Jacobs, Chem. Commun. (2008)
6011–6012.
[19] T. Miyazawa, S. Koso, K. Kunimori, K. Tomishige, Appl. Catal. A-Gen. 329 (2007)
30–35.
[20] O.W. Gooding, Curr. Opin. Chem. Biol. 8 (2004) 297–304.
[21] Catalytic hydrogenolysis of glycerol, M. Besson, L. Djakovitch, P. Gallezot, C. Pinel, A.
Salameh and M. Vospernik, Chemical Industries (Boca Raton, FL, United States),
2009, 123 (Catalysis of Organic Reactions), 313–318.
[22] J. Yan, G. Springsteen, S. Deeter, B.H. Wang, Tetrahedron 60 (2004) 11205–11209.
[23] M. Schlaf, P. Ghosh, P.J. Fagan, E. Hauptman, R.M. Bullock, Angew. Chem. Int. Ed. 40
(2001) 3887–3890.
The D-optimal screening of 19 catalysts showed that active metal,
catalyst support and temperature have the highest influence on
glycerol hydrogenolysis activity. A preliminary study regarding the
effect of BA indicated that it can increase glycerol conversion and
influences selectivity of the catalysts. Upon addition of BA, the
selectivity of Pt/CaCO3 was switched from 12PD to LA. Among all
catalysts screened in the D-optimal design, Pt/CaCO3 was found to be
the most active with little glycerol degradation. A Central Composite
Design revealed the strong relation between the pH and 12PD-
selectivity. Decreasing the pH gave higher 12PD-selectivity by
[24] H. Kishida, F.M. Jin, Z.Y. Zhou, T. Moriya, H. Enomoto, Chem. Lett. 34 (2005)
1560–1561.
1
For the recycling study a new batch of catalyst has been used. The new batch had
higher activity and comparable selectivities.