Modification of supported Pd catalysts by alkalic salts in the selective racemization and dynamic kinetic resolution of primary amines

Article abstract of DOI:10.1039/c3cy00535f

Supported Pd nanoparticles (PdNPs) were modified with various alkalic salts by incipient wetness impregnation method. The effect of the salts on the catalytic activity and selectivity behavior of Pd catalysts for the racemization of (S)-1-phenylethylamine was investigated. The presence of alkalic salts can greatly enhance the selectivity of Pd catalysts, without significantly decreasing the catalytic activity. This modification method is suitable for PdNPs supported on various supports, such as micro/mesoporous silica and activated carbon. Combined with the immobilized lipases (Novozyme 435), this catalyst system can efficiently catalyze the dynamic kinetic resolution (DKR) of 1-phenylethylamine with yield and enantioselectivity up to 97% and 99%, respectively. The Royal Society of Chemistry.

Full text of DOI:10.1039/c3cy00535f

Catalysis
Science &
Technology
View Article Online
View Journal | View Issue
PAPER
Modification of supported Pd catalysts by alkalic
salts in the selective racemization and dynamic
kinetic resolution of primary amines†
Cite this: Catal. Sci. Technol., 2014,
4, 464
ab
a
a
a
Qianru Jin, Guoqing Jia, Yanmei Zhang and Can Li*
Supported Pd nanoparticles (PdNPs) were modified with various alkalic salts by incipient wetness
impregnation method. The effect of the salts on the catalytic activity and selectivity behavior of Pd
catalysts for the racemization of (S)-1-phenylethylamine was investigated. The presence of alkalic salts can
greatly enhance the selectivity of Pd catalysts, without significantly decreasing the catalytic activity. This
modification method is suitable for PdNPs supported on various supports, such as micro/mesoporous
silica and activated carbon. Combined with the immobilized lipases (Novozyme 435), this catalyst system
can efficiently catalyze the dynamic kinetic resolution (DKR) of 1-phenylethylamine with yield and
enantioselectivity up to 97% and 99%, respectively.
Received 23rd July 2013,
Accepted 23rd October 2013
DOI: 10.1039/c3cy00535f
www.rsc.org/catalysis
11–15
the racemization and DKR of amines were developed.
1
. Introduction
For example, Paetzold and Bäckvall developed an efficient
ruthenium (Ru) catalyst. By combining Ru-catalyzed
racemization and lipase-catalyzed resolution, a wide range of
primary amines were transferred into one enantiomer of
Enantiomerically pure amines are widely used as building
blocks for the synthesis of various biologically active com-
pounds in the pharmaceutical and agrochemical industries,
and therefore it is of considerable importance to develop their
preparation methods. Kinetic resolution (KR) of racemic
amine mixtures by enzyme catalysis has been employed in the
preparation of chiral amines for a long time. However, the
major drawback of KR is that the theoretical maximum yield
cannot exceed 50% for either enantiomer. Dynamic kinetic
resolution (DKR) (Scheme 1) that combines enzyme-catalyzed
KR with in situ racemization of the undesired enantiomer
can overcome the limitation of KR and theoretically increase
1
1,16
amides with good yield and selectivity.
As an alternative
1
to homogeneous racemization catalysts, palladium nano-
particles (PdNPs) supported on alkaline earth salts, amino-
functionalized silica, layered double hydroxide (LDH) and
aluminum hydroxide [AlO(OH)] exhibited at least the same or
better performance than the Ru catalyst in the racemization
2
,3
1
2,13,17
of chiral aromatic amines.
Besides, Parvulescu and
®
co-workers found that RANEY nickel and cobalt catalysts, as
an alternative to noble metals such as Pd or Ru, are able to
racemize various chiral amines with high selectivity.
4
–7
14
the yield up to 100%.
To achieve this goal, an efficient
racemization catalyst compatible with the enzymatic KR is
prerequisite.
Recently, Shakeri and co-workers reported that PdNPs
supported in aminopropyl-modified mesocellular foam
(Amp-MCF) showed higher activity than other racemization
8
Palladium on carbon (Pd/C) was first used as a racemization
9
15
catalyst for the DKR of amines in 1996. Using triethylamine as
catalysts reported so far.
solvent, racemic 1-phenylethylamine was transformed to the
corresponding (R)-amide with 64% yield but a long reaction
time (8 days) was necessary. Since then, no significant develop-
ment has been reported for nearly a decade because racemiza-
tion of amines is difficult and usually needs harsh reaction
Porous carbon and silica with high specific surface area
are ideal supports for the immobilization of PdNPs, but the
10
conditions. Until 2005, several successful catalyst systems for
a
State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics,
Chinese Academy of Sciences, 457 Zhongshan Road, Dalian 116023, China.
E-mail: canli@dicp.ac.cn; Web: http://www.canli.dicp.ac.cn;
Fax: +86 411 84694447; Tel: +86 411 84379070
b
Graduate University of Chinese Academy of Sciences, Beijing 100049, China
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
c3cy00535f
Scheme 1 Dynamic kinetic resolution (DKR) of racemic 1-phenylethylamine.
464 | Catal. Sci. Technol., 2014, 4, 464–471
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
acid property of the supports could promote side reactions
NOVA 4200e and Autosorb-1 at 77 K, respectively. Before mea-
surements, the samples were outgassed at 393 K for 6 h. The
X-ray diffraction (XRD) patterns were recorded on a D/
Max2500V/PC powder diffractometer (Rigaku), using Cu Kα
radiation that operated at 40 kV and 100 mA. Infrared spectra
were recorded on a Thermo-Nicolet Nexus 470 Fourier trans-
form infrared (FT-IR) spectrometer. All the samples, with
KBr, were pressed into self-supporting wafers and mounted
in an IR cell for FT-IR spectroscopy.
The amount of Pd loading was measured using a sequential
inductively coupled plasma atomic emission spectrometer
(ICPS-8100 made by SHIMADZU). The samples were prepared
as follows: appropriate amounts of supported Pd catalysts
were successively dissolved in aqua regia and hydrofluoric
acid, then this solution was diluted with pure water to a
final volume of 100 mL.
1
2
during the racemization of chiral amines. The previous
reports regarding SiO supported PdNPs indicate that amino-
2
functionalization can effectively modify the surface property
of SiO and enhance the selectivity of the racemization reac-
2
tion, which is mainly ascribed to the basicity of the amino
1
5,17
groups.
However, most amino-modified SiO2 cannot
resist high temperature, which may limit its application in
catalysis. Recently, incorporation of metal salts has been
demonstrated to modify the surface acid–base properties
and the active component of catalysts in such ways that
1
8–22
are beneficial to their catalytic activity and selectivity.
1
9
For example, Groppo et al. reported that K-doping on a
Pd/SiO –Al CO catalyst had a great influence on the starting
Pd precursor phase, the fraction of truly available Pd
surface sites and the lifetime of Pd catalysts. Liu et al.
reported that incorporation of K CO into cinder could
2
2
3
2
+
2
1
X-ray photoelectron spectroscopy (XPS) was performed on
an ESCALAB 250Xi (Thermo Scientific) with Al Kα radiation at
2
3
enhance the surface alkalinity and catalytic performance of
2
2
−5
catalysts. Zhu et al. reported that modification of the
Cu/SiO catalyst with alkaline earth metals could catalyze the
7.1 × 10 Pa, calibrated internally by silica Si (2p) binding
2
energy at 103.4 eV. The binding energy of Pd 3d5/2 and Pd 3d3/2
peaks were separated by 5.25 eV with peak ratios of 3/2. The
conversion of 1,4-butanediol to γ-butyrolactone with high
selectivity. Thus it is of great interest to explore the effect of
incorporated salts on the catalytic activity and selectivity
behavior of supported Pd catalysts for the racemization of
primary amines.
0
binding energies of Pd 3d5/2 were set as thus: Pd : 335.2 eV;
3 4 2
PdO: 336 eV; Pd(NH ) Cl : 338.4 eV.
2.2 Catalyst preparation procedure
In the present work, we report that the addition of alkalic
salts to supported PdNPs could eliminate the surface acidity
of the supports to some extent and improve the selectivity
towards racemization of (S)-1-phenylethylamine. Moreover,
their ability to suppress side reactions is related to the basic
strength of inorganic salts. Pd catalysts modified with K CO
show excellent catalytic activity and selectivity for the racemi-
zation and DKR of primary amines, irrespective of the type of
support material.
Support materials such as MCF, SBA-15 and FDU-12 were syn-
thesized according to the procedures reported previously.
Supported Pd catalysts were prepared by employing SiO2
2
3–25
(
800 mg) for the adsorption of Pd(NH
3
)
4
Cl
2
precursor solution
2
6
(
5 mM, 40 mL). Before adsorption, the pH of the precursor
2
3
solution was adjusted to 11.0 with ammonia. After stirring for
h, the light yellow solid was centrifuged from the slurry and
6
2
+
2
then dried at 80 °C overnight. Then, the Pd /SiO complex was
reduced in flowing hydrogen at 200 °C for 3 h. The resultant
Pd catalysts were denoted as Pd/R (R = MCF, SBA-15, FDU-12
and SiO ).
2
A series of salt-modified Pd catalysts were prepared by
the incipient wetness impregnation method using 1 mL of
2
. Experimental section
2
.1 Materials and instruments
−
1
Lipase acrylic resin from Candida antarctica (≥10 000 U g ,
Novozyme 435), pluronic P123 (EO20PO70EO20) and F127
0.5 M salt solution and 200 mg of supported Pd/R catalysts
(
Pd/R = Pd/MCF, Pd/FDU-12, Pd/SBA-15, Pd/SiO and 5% Pd/C).
2
(EO106PO70EO106) copolymer were purchased from Sigma-
After shaking for 6 h, the resultant mixtures were lyophilized
for 10 h on a 2.5 L Labconco Freeze Dry at a condenser temper-
ature of −49 °C and a pressure of 20 Pa. The amounts of incor-
Aldrich; silicon(IV) oxide (amorphous fumed), mesitylene,
tetraethoxysilane, (±)-1-phenylethylamine, (S)-1-phenylethylamine
and tetraamine palladium(II) chloride monohydrates were
obtained from Alfa Aesar. Bromothymol blue, phenolphthalein,
NH F, Cs CO , Rb CO , K CO , Na CO , Li CO , K PO ,
Na PO , CH COOK, CH COONa, KCl and NaCl were obtained
from the Shanghai Chemical Reagent Company by the Chinese
Medicine Group and dried before use.
porated K
CO
R′ = Li CO , Na CO , K CO , Rb CO , Cs CO , K PO , Na PO ,
2 3
CO were adjusted by varying the concentration of
K
2
3
solution. The resultant catalysts were denoted as R′–Pd/R
4
2
3
2
3
2
3
2
3
2
3
3
4
(
2
3
2
3
2
3
2
3
2
3
3
4
3
4
3
4
3
3
3 3
CH COOK, CH COONa, KCl, and NaCl).
2.3 Procedure of racemization and dynamic kinetic
Transmission electron microscopy (TEM) images were
2
resolution of 1-phenylethylamine
recorded on an FEI Tecnai G Spirit at an accelerating voltage
of 120 kV. High-resolution TEM (HRTEM) images and energy
dispersive X-ray (EDX) spectroscopy were taken using an FEI
Tecnai G F30 Spirit microscope operating at an accelerating
Typically, 0.50 mmol of (S)-1-phenylethylamine (65 μL),
0.22 mmol of hexadecane (65 μL) as internal standard, 4 mL
of toluene and 0.01 mmol of supported Pd catalysts (43.4 mg,
2.45 wt.%) were added into a 10 mL Schlenk tube. After freez-
ing with liquid nitrogen and removing air under vacuum, the
2
voltage of 300 kV. N sorption isotherms of mesoporous and
2
microporous materials were measured on a Quantachrome
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 464–471 | 465

View Article Online
Paper
Catalysis Science & Technology
racemization reaction was performed under stirring at 70 °C
with a 5 L hydrogen balloon (10% H and 90% N ). 100 μL of
2
2
the mixture were removed from the reaction system at intervals
of 1 h and acylated by acetic anhydride. After a high speed cen-
trifugation, the supernatant liquor was used to monitor the
enantiomeric excess (ee%) and selectivity on an Agilent 7890
gas chromatograph with a Supelco-β-DEX™ 225 chiral column
(0.25 mm × 30 m × 0.25 μm). The oven temperature was firstly
kept at 160 °C for 25 min and then 180 °C for 30 min. The
injector (split 1:50) temperature was 250 °C and the detector
FID temperature was 280 °C. The flow rates of the carrier gases
nitrogen, hydrogen and air were 10, 40 and 400 mL min ,
respectively.
Fig. 1 XRD pattern of MCF and Pd/MCF with different Pd loading.
−1
0
2+
Dynamic kinetic resolution (DKR) was performed simi-
larly, with 0.50 mmol of racemic 1-phenylethylamine (65 μL),
reveals that the ratio of Pd and Pd is 1 : 1 (Fig. S2†).
2+
Although nearly half of the Pd species exist as Pd , there is no
obvious diffraction peak of PdO in the XRD pattern of Pd/MCF.
The metal dispersion and mean particle diameter estimated
from CO adsorption is 47.5% and 2.4 nm, respectively.
1
0
.0 mmol of ethyl methoxyacetate (120 μL), 4 mL of toluene,
.01 mmol of Pd source and 100 mg of immobilized Candida
antartica lipase B (Novozyme 435) as a resolution catalyst.
After the reaction was completed, the reaction mixture was
evaporated and then separated on a silica gel column to give
an isolated yield of product.
No aggregation of PdNPs is observed in the TEM image of
Pd/MCF (Fig. 2a), indicating that PdNPs are uniformly dis-
persed into the nanopores of MCF. However, it is difficult to
discriminate PdNPs from the disordered pore structure of
MCF. The dark field image was also captured at the same
position (Fig. 2b) and a large number of light spots indicate
the presence of PdNPs. The PdNPs in the specific area
labelled 1 were analysed by EDX and the characteristic peaks
at around 2.84 KeV are ascribed to Pd species. Moreover,
individual PdNPs can be observed in the enlarged TEM
images (Fig. 2c). Indeed the PdNPs shows obvious crystalline
character, which is evidenced by the presence of the lattice
fringes (Fig. 2d). However, it seems that the particle diameter
of PdNPs estimated from TEM images is larger than that
from CO adsorption.
2
.4 Basicity characterization
The basic strength (H ) of inorganic salts and salt-modified
−
Pd/MCF catalysts were determined using the Hammett indi-
cator. Generally, 0.20 mmol of inorganic salts or 0.01 mmol of
salt-modified Pd/MCF were suspended in 2 mL of an indicator
solution (0.1 mg of bromothymol blue or phenolphthalein
per mL of toluene) and then left to equilibrate for 2 h until
no further color changes were observed. According to the
2
1
literature, the basic strength of samples is defined as being
stronger than a weaker indicator that exhibits a color change,
and weaker than a stronger indicator that produces no color
change. The basicity values (mol/mol) of inorganic salts were
determined by the method of Hammett indicator–benzene
3
.2 The effect of salts on the catalytic activity and selectivity
of supported Pd catalysts for racemization of
S)-1-phenylethylamine
Using Pd/MCF as a racemization catalyst, the racemization of
S)-1-phenylethylamine proceeded in toluene at 70 °C and the
−
1
carboxylic acid (0.1 mol L anhydrous toluene solution)
titration until the color changed back to the original color.
(
(
3
. Results and discussion
enantiomer excess (ee) decreased to below 5% in 2 h (Entry 2,
Table 1). However, the selectivity for 1-phenylethylamine was
69% and further decreased upon prolonging the reaction time.
Taking the advantage of salt effect in modulating the basicity
3
.1 Textural properties of supported Pd catalysts
Siliceous mesocellular foam (MCF) with high surface area
and continuous 3-D pore structure was chosen as a model
support to immobilize PdNPs. ICP results indicate that Pd
loading of Pd/MCF increases with increasing the concentra-
tion of Pd(NH ) Cl precursor and reaches the maximum
2
0–22
of catalyst,
various alkalic salts were introduced into pores
of Pd/MCF by the incipient wetness impregnation method and
then the salt-modified Pd catalysts were applied to racemize
(S)-1-phenylethylamine. After the addition of inorganic salts,
there is a slight decrease in the catalytic activity of Pd/MCF,
which is partly ascribed to the pore blockage by inorganic salts.
Despite this slight decrease, the addition of alkalic potassium
salts significantly improves the reaction selectivity, such as
97% selectivity for K PO –Pd/MCF (Entry 4, Table 1), 96% selec-
3
4
2
8
wt.% (Fig. S1†). The XRD patterns of Pd/MCF catalysts are
shown in Fig. 1. Besides the typical diffraction peak due to
MCF, the additional peaks at 2θ = 40.14°, 46.69° and 68.17°
are ascribed to PdNPs (according to the ICDD PDF data-
bases). The intensity of Pd diffraction peaks is enhanced
gradually with increasing the Pd loading, indicating the
growth of PdNPs. Pd/MCF with Pd loading of 2.45 wt.% was
selected for further study.
3
4
tivity for K CO –Pd/MCF (Entry 6, Table 1), and 91% selectivity
2
3
3
for CH COOK–Pd/MCF (Entry 8, Table 1), while the effect of
The oxidation states of Pd on the reduced Pd/MCF
catalysts were investigated by XPS and the fitting result
neutral KCl and NaCl on the selectivity of the racemization
reaction is not obvious. It seems that their ability to
466 | Catal. Sci. Technol., 2014, 4, 464–471
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
−
(
pK : 4.76) > Cl (pK : −7). It is noteworthy that, except
a
a
Na PO , the addition of alkalic sodium salts (Na CO and
3
4
2
3
3
CH COONa) to Pd/MCF cannot result in obvious improve-
ment in the selectivity of the racemization reaction. This
suggests that their ability to suppress side reactions is also
related to the counter cations of inorganic salts.
Then the addition of Li PO to Pd/MCF by mechanical
3
4
mixing was tested and it could indeed effectively suppress
the formation of side products with 91% selectivity (Entry 12,
Table 1). Pd/MCF was also modified with other carbonates of
group I and the catalytic results are summarized in Table 1.
The addition of Li
2 3
CO can not effectively suppress the side
reactions during the racemization and Li CO –Pd/MCF
2
3
exhibits poor selectivity (about 73%). On the contrary, the
presence of Rb CO or Cs CO significantly improves the
performance of Pd/MCF and high selectivity up to 98% is
2
3
2
3
1
2,15,17
obtained. Based on the research result of references,
we speculate that the basicity of inorganic salts is the main
cause of higher selectivity for 1-phenylethylamine.
3.3 Basicity characterization of incorporated inorganic salts
and Pd catalysts
The acid–base properties of salt-modified Pd catalysts
Fig. 2 TEM images and EDX spectrum of Pd/MCF, (a: light field TEM;
b: dark field TEM; c and d: enlarged PdNPs).
−
strongly depend on the basic strength (H ) of inorganic salts
in organic solvent. Using Hammett indicators, the basic
strength of inorganic salts was determined. Bromothymol blue
is a common Hammett indicator for weak bases and its basic
Table
1 Salt-modified Pd/MCF catalyzed the racemization of
a
(S)-1-phenylethylamine
strength (H ) is 7.2. It acts as a weak acid in organic solvent and
−
can thus be in protonated or deprotonated form, appearing yel-
low in acid solution or blue in basic solution, respectively. In
neutral solution it is bluish green.
Fig. 3 shows the color of bromothymol blue in toluene
solution suspended with various alkalic salts. Pristine
bromothymol blue appears yellow in toluene solution (mid-
dle). The addition of inorganic salts leads to the color change
of bromothymol blue solution, which is related to the basicity
b
b
b
ee
Conversion
(%)
Selectivity
(%)
Entry
Catalyst
(%)
1
2
3
4
5
6
7
8
9
Ideal
Pd/MCF
Pd/MCF
0
5
1
25
9
4
1
6
4
14
7
50
64
79
39
50
53
68
51
75
56
68
45
63
48
45
100
69
42
97
91
96
65
91
48
78
60
91
73
98
98
c
of inorganic salts, and indicates that their basic strength (H )
−
K
3
PO
Na PO
CO –Pd/MCF
Na CO –Pd/MCF
4
–Pd/MCF
is above 7.2. According to the depth of blue, the basicity of
3
4
–Pd/MCF
inorganic salts can be expressed as K
CO
≈ K
PO
≈ Na
3 4
PO >
2
3
3
4
K
2
3
CH COOK > Na CO > CH COONa. This basicity trend is in
3
2
3
3
2
3
accordance with that of the selectivity of salt-modified Pd/MCF
in the racemization of (S)-1-phenylethylamine (Table 1),
suggesting that high selectivity for amine is mainly ascribed to
the basicity of inorganic salts. In addition, Li PO , Na PO and
CH
CH
3
COOK–Pd/MCF
COONa–Pd/MCF
3
10
11
12
13
14
15
KCl–Pd/MCF
NaCl–Pd/MCF
d
3
4
3
4
Li
Li
Rb
Cs
3
PO
CO
CO
CO
4
+ Pd/MCF
–Pd/MCF
20
2
5
3 4
K PO can convert bromothymol blue from yellow to dark blue
(Fig. S3†), indicating their similar basicity in toluene solution.
2
3
2
3
–Pd/MCF
–Pd/MCF
2
3
13
a
Standard reaction conditions: (S)-1-phenylethylamine (0.50 mmol),
−1
Pd/MCF (2.45 wt.%, 0.01 mmol of Pd), incorporated salt ( 2.5 mmol g
catalyst), hexadecane (0.22 mmol), toluene (4 mL), 0.01 MPa H , 70 °C,
2
b
c
4
h. Determined as corresponding acetamide using chiral GC. 2 h.
Mechanical mixing.
d
suppress side reactions is related to the dissociation con-
stant of conjugated acids of incorporated salts, with an order
Fig. 3 The color of bromothymol blue in toluene solution suspended
with alkalic salts.
3
−
2−
−
PO
4
(pK
a
:
12.38)
≈
CO
3
a
(pK :
10.33)
>
CH
3
COO
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 464–471 | 467

View Article Online
Paper
Catalysis Science & Technology
However, carbonate salts with different counter ions have
different basicity in toluene solution. This can explain why
the counter ion in the carbonate of group I influences the
selectivity of the racemization reaction.
(S)-1-phenylethylamine, and therefore K CO was selected for
2
3
further study. Single K CO or K CO -impregnated MCF
2
3
2
3
shows no catalytic activity, suggesting K
2
3
CO merely serves as
a modifier to eliminate the surface acid of MCF.
The basic strength of inorganic salts was also determined
The amount of K CO is an important parameter which
2
3
using phenolphthalein (H
−
= 9.8) as Hammett indicator.
affects the activity and selectivity of supported Pd catalysts.
As illustrated in Fig. 4, the catalytic activity decreases with
increasing the amount of K CO and nearly disappears when
Among them, K CO , K PO , Na PO , and CH COOK can
2
3
3
4
3
4
3
convert phenolphthalein from colorless to purple and their
basic strengths (H ) are above 9.8. Na CO and CH COONa fail
to change the color of phenolphthalein and their basic
2
3
−
1
−
2
3
3
K
2
CO
decreased catalytic activity of Pd/MCF is mainly ascribed to
the combined effect of pore blockage and basicity of K CO .
3
content reaches up to 30 mmol g . Here the
strengths (H ) are between 7.2 and 9.8. In view of such results,
−
2
3
as well as the observation on the selectivity of salt-modified
Pd/MCF, it permits us to conclude that inorganic salts with
basic strength above 9.8 can be used as an effective additive to
selectively catalyze the racemization of (S)-1-phenylethylamine.
Furthermore, the basicity values (mol/mol) of inorganic
salts were determined by the common method of Hammett
On the other hand, reaction selectivity is significantly
improved and maintained enduringly at around 99% when
−
1
K2CO3 content is higher than 5 mmol g . Therefore, the
−
1
amount of K
2 3
CO was set at 5 mmol g in the following
discussion.
The effect of adding method of K CO to Pd/MCF was also
2
3
1
8
indicator–benzoic acid titration. As shown in Table 2, the
investigated. From the time courses for racemization of (S)-1-
basicity value of these alkalic salts decreases in the order of
phenylethylamine (Fig. 5), the addition of K CO to Pd/MCF
2
3
K
3
PO
4
> K
2
CO
3
> Na
3
PO
4
> CH
3
COOK > Na
2
CO
3
>
by incipient wetness impregnation or mechanical mixing can
significantly suppress the consumption of amines and a high
selectivity (>98%) is maintained even after complete conver-
sion. However, when K CO is added directly to the reaction
2 3
mixture, side reactions take place and the selectivity of the
racemization reaction decreases to 72%.
CH COONa, which also is in accordance with that of the
3
selectivity of salt-modified Pd/MCF and further confirms that
their ability to suppress side reactions is related to the basic-
ity of inorganic salts.
Using Hammett indicator, the basic strength (H ) of salt-
−
modified Pd/MCF was also determined. Pristine and halite-
modified Pd/MCF are weakly acidic and their basic strengths
The FT-IR and XRD results (Fig. S4 and S5†) indicate that
K CO species can be successfully introduced into Pd/MCF
2 3
(
H ) are below 7.2. After the addition of salts with strong
catalyst by incipient wetness impregnation or mechanical
mixing. The reproducibility of catalyst preparation was evalu-
−
2 3 3 4 3 4
basicity, such as K CO , K PO , and Na PO , the salt-
modified Pd/MCF become alkalescent and can change the
color of bromothymol blue from yellow to blue but fail to
convert phenolphthalein from colorless to purple, so their
2 3
ated by comparing the catalytic performance of K CO –Pd/
MCF catalysts with different batches and scales in the race-
mization of (S)-1-phenylethylamine and the results are sum-
marized in the Tables S1 and S2.† From these results, it
seems that our reported preparation method of Pd catalysts
has high reproducibility.
−
basic strengths (H ) are between 7.2 and 9.8. The addition of
Na CO and CH COONa to Pd/MCF cannot convert the color of
2
3
3
bromothymol blue solution from yellow to blue, so their basic
strengths (H ) are below 7.2. These results indicate that the
addition of Na CO and CH COONa is not enough to effectively
−
The fact that the addition of K
cantly improve the selectivity of the racemization reaction
might be related to the basicity of K CO . As reported previ-
a basic support is necessary in the
Pd-catalyzed racemization of primary amines, which can
2 3
CO to Pd/MCF can signifi-
2
3
3
eliminate the surface acidity of MCF. The basicity values of
salt-modified Pd/MCF cannot be exactly determined by the
Hammett method because the color changes of titration
endpoints are seriously disturbed by black catalyst powder.
2
3
1
2,17
ously in the literature,
3
.4 The effect of K CO on the selective racemization of
2 3
(S)-1-phenylethylamine
Among the salts tested, K PO and K CO are found to be the
3
4
2
3
most effective additives in the selective racemization of
Table 2 The basicity values of various alkalic salts in the toluenea
Total basicity
(mol/mol)
Total basicity
(mol/mol)
Salt
Salt
K
K
CH
2
CO
3
0.85
1.0
0.40
Na
Na
CH
2
CO
PO
COONa
3
0.30
0.60
0.03
3
PO
4
3
4
3
COOK
3
a
Conditions: 0.2 mmol salts in 2 mL of bromothymol blue solution,
.1 M benzoic acid, stirring.
Fig. 4 Effect of K
Pd/MCF in the racemization of (S)-1-phenylethylamine.
2 3
CO content on the activity and selectivity of
0
468 | Catal. Sci. Technol., 2014, 4, 464–471
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Table 3 The effect of K
2
CO
3
on the activity and selectivity of Pd
catalysts immobilized on different supports for the racemization of
(
S)-1-phenylethylaminea
b
b
b
K
2
CO
3
ee
Conversion
(%)
Selectivity
(%)
−1
Entry Catalyst
(mmol g
)
(%)
1
2
3
4
5
Pd/MCF
0
5
0
5
0
5
0
5
0
1
1
2
10
2
6
1
3
3
11
5
79
49
78
45
88
48
88
49
89
48
62
42
99
42
99
25
98
23
99
22
90
74
Pd/SBA-15
Pd/FDU-12
c
2
Pd/SiO
c
Fig. 5 Time courses for the racemization of (S)-1-phenylethylamine
Pd/C
5
10
catalyzed by Pd/MCF in the presence of K
selectivity%: dotted line. Standard reaction conditions: a: the addition
of K CO by incipient wetness impregnation; b: mechanical mixing of
Pd/MCF and K CO ; c: K CO was added into reaction system directly).
2 3
CO , (ee%: solid line;
d
Pd/C
2
3
a
b
Standard reaction condition.
Determined as corresponding
d
2
3
2
3
c
acetamide using chiral GC. Commercial. Mechanical mixing of Pd/C
and K CO
2
3
.
suppress the formation of side products. Moreover, Pd on
amino-functionalized silica is proved to be more selective for
the racemization of (S)-1-phenylethylamine, mainly due to the
basicity of aminopropyl group. K CO has similar base strength
with propylamine, which can create a basic microenvironment
for PdNPs to suppress the acid-promoted side reactions of
imine intermediate with 1-phenylethylamine, resulting in an
enhanced selectivity in the racemization reaction.
K CO only increases the selectivity to 74%. Then we tested
the recyclability of the K CO –Pd/C catalyst (Table S4†). It
2 3
can be reused at least 4 times without obvious loss in reac-
tion selectivity. However, there is a slight decrease in the
2 3
catalytic activity of K CO –Pd/C, which is probably due to the
2
3
2
3
aggregation of PdNPs. These results indicate that the addi-
tion of K CO to supported Pd catalysts by incipient wetness
2
3
impregnation is an efficient way for improving the selectivity
of the racemization reaction.
3
.5 Expanding support scope
To examine the generality of K CO addition for improving the
2
3
selectivity of immobilized PdNPs in the racemization of chiral
amines, three other porous silica materials were also chosen as
support materials to immobilize PdNPs. Among them, MCF
and FDU-12 silica are composed of large spherical cells that are
connected by windows to create a 3D system, while hexagonal
one-dimensional mesoporous SBA-15 and commercial SiO2
3.6 Expanding substrate scope
Besides (S)-1-phenylethylamine, other substrates were also
tested. As shown in Table 4, Pd/MCF can catalyze the racemi-
zation of functionalized chiral aromatic amines, such as (S)-
1
1
-(4-MeO-phenyl)ethylamine, (S)-1-(4-tolyl)ethylamine and (S)-
-(4-F-phenyl)ethylamine, but the selectivity is very poor. After
possess no cagelike pore structure (Fig. S6†). N
2
sorption
−1
the addition of K
2
CO
3
(5 mmol g ) to Pd/MCF, the reaction
isotherms of MCF, FDU-12 and SBA-15 (Fig. S7†) possess
obvious hysteresis loops, which further confirm the presence
selectivity for all the functionalized chiral aromatic amines is
improved. For example, the selectivity for 1-(4-MeO-phenyl)-
ethylamine and 1-(4-tolyl)ethylamine are increased to 99%
and 92%, respectively. Compared with 1-phenylethylamine,
of mesoporous structures. Commercial SiO
2
mainly has micro-
and pore volume of
.61 cm g (Table S3†). All these support materials have dif-
2
−1
pores with surface area of 227 m
1
g
3
−1
the selectivity of
2 3
K CO –Pd/MCF for (S)-1-(4-F-phenyl)
ferent mesoporous structures, surface areas and pore volumes.
As shown in Table 3, all the supported Pd catalysts exhibit
similar high activity and poor selectivity in the racemization
of (S)-1-phenylethylamine. The modification of Pd catalysts
by K CO impregnation can result in a significant enhance-
ethylamine is decreased. It seems that the presence of an
electron withdrawing group in amines can affect the reaction
selectivity. In addition, K CO –Pd/MCF can successfully cata-
2 3
lyze the racemization of (S)-1-phenylpropyl-amine with selec-
tivity up to 99%. Unfortunately, aliphatic chiral amine, such
as (S)-1-cyclohexylethylamine and (S)-2-hexyl-amine cannot be
2
3
ment of the reaction selectivity, irrespective of the kind of
supports. We optimized the amount of K CO for each sup-
port (shown in Fig. S8†) and it turned out that 5 mmol g
2
3
racemized with Pd/MCF or K
2 3
CO –Pd/MCF as a catalyst, even
−
1
at high temperature (100 °C) and long reaction time (3 days).
was prerequisite. It is known that side reactions are serious
using Pd/C as racemization catalyst, and in Jacobs' work no
obvious improvement was found either using triethylamine
3.7 Dynamic kinetic resolution of racemic 1-phenylethylamine
1
2
as solvent or catalyst pretreatment with base. Surprisingly,
the addition of K CO to Pd/C by incipient wetness impregna-
With K CO -modified Pd catalysts in hand, DKR of racemic
2
3
2
3
1-phenylethylamine was attempted using Novozyme 435 as
the resolution catalyst and ethyl methoxyacetate as the acyl
agent (Table 5). When PdNPs are supported on mesoporous
tion method increases the selectivity of the racemization
reaction up to 90%. However, mechanical mixing of Pd/C and
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 464–471 | 469

View Article Online
Paper
Catalysis Science & Technology
Table 4 Racemization of functionalized chiral aromatic amines catalyzed by Pd/MCFa
−1
b
b
b
Substrate
K CO
2 3
(mmol g
)
T (°C)
t (h)
ee (%)
Conversion (%)
Selectivity (%)
0
5
70
90
12
12
3
19
81
44
36
92
0
5
70
90
12
12
7
5
86
49
27
99
0
5
90
90
24
24
56
57
71
42
37
74
0
5
70
70
4
8
0
1
70
50
61
99
a
2
Standard reaction conditions: substrate (65 μL), Pd/MCF (2.45 wt.%, 0.01 mmol of Pd), hexadecane (120 μL), toluene (4 mL), 0.01 MPa H .
Determined as corresponding acetamide using chiral GC.
b
Table 5 Dynamic kinetic resolution of 1-phenylethylaminea
yield and selectivity. Noteworthily, modification of Pd/C with
K CO by multiple incipient wetness impregnations has suc-
2
3
cessfully achieved the high yield (97%) of the product from
racemization and DKR of primary amines.
b
Acknowledgements
Time
(h)
Conversion
(%)
Yield
(%)
ee
(%)
Entry
Catalyst
We acknowledge the financial support from the National
Natural Science Foundation of China (grant NSFC 31000392)
and the Program for Strategic Scientific Alliances between
China and The Netherlands (grant 2008DFB50130).
1
2
3
4
5
6
2
K CO
2
K CO
2
K CO
2
K CO
2
K CO
2
K CO
3
3
3
3
3
3
–Pd/MCF
–Pd/SBA-15
–Pd/FDU-12
–Pd/SiO
–Pd/C
9
9
9
9
6
6
97
97
98
97
98
99
96
96
97
90
82
96
99
99
99
98
98
99
c
2
d
–Pd/C
a
Reaction conditions: 0.50 mmol of racemic 1-phenylethylamine,
Notes and references
1
20 μL of ethylmethoxyacetate, 0.01 mmol of Pd source, 100 mg of
b
Novozyme 435, 4 mL of toluene, 0.01 MPa H
2
, 70 °C. Isolated yield.
1
M. Breuer, K. Ditrich, T. Habicher, B. Hauer, M. Keßeler,
c
d
Commercial. Multiple incipient wetness impregnations.
R. Stürmer and T. Zelinski, Angew. Chem., Int. Ed., 2004, 43,
7
88–824.
M. Höhne and U. T. Bornscheuer, ChemCatChem,
009, 1, 42–51.
silicate, such as MCF, FDU-12 and SBA-15, the yield of
R)-1-phenylethylamide exceeds 95% and the enantioselectivity
of product is 99% ee (Entries 1–3, Table 5). When
2
3
4
5
6
7
8
(
2
S. Arseniyadis, A. Valleix, A. Wagner and C. Mioskowski,
Angew. Chem., Int. Ed., 2004, 43, 3314–3317.
O. Pàmies and J.-E. Bäckvall, Chem. Rev., 2003, 103,
PdNPs are supported on commercial SiO , the yield of
2
(
R)-1-phenylethylamide decreases to 90% (Entry 4, Table 5).
Commercial Pd/C affords the lowest yield of the corresponding
R)-amide (82%). However, the yield can be further enhanced
up to 97% by multiple incipient wetness impregnations of Pd/
C with K CO . To the best of our knowledge, this is the first
3
247–3261.
(
A. Parvulescu, J. Janssens, J. Vanderleyden and D. De Vos,
Top. Catal., 2010, 53, 931–941.
J. H. Lee, K. Han, M.-J. Kim and J. Park, Eur. J. Org. Chem.,
010, 999–1015.
Y. Kim, J. Park and M.-J. Kim, ChemCatChem,
011, 3, 271–277.
2
3
case that DKR of amines can be efficiently performed with the
catalyst based on commercial Pd/C.
2
2
4
. Conclusion
Y. Ahn, S.-B. Ko, M.-J. Kim and J. Park, Coord. Chem. Rev.,
2008, 252, 647–658.
In summary, incorporation of alkalic salts (H > 9.8) into
−
supported Pd catalysts is found to be an efficient method for
improving the selectivity for the racemization of (S)-1-phenylethylamine.
These alkalic salts supported on Pd catalysts can efficiently
suppress side reactions during the racemization and enhance
the yield of chiral amines. Combined with immobilized
lipases (Novozyme 435), these salt-modified Pd catalysts can
catalyze the DKR of racemic 1-phenylethylamine with good
9 M. T. Reetz and K. Schimossek, Chimia, 1996, 50, 668–669.
10 E. J. Ebbers, G. J. A. Ariaans, J. P. M. Houbiers, A. Bruggink
and B. Zwanenburg, Tetrahedron, 1997, 53, 9417–9476.
11 J. Paetzold and J.-E. Bäckvall, J. Am. Chem. Soc., 2005, 127,
17620–17621.
12 A. Parvulescu, D. De Vos and P. Jacobs, Chem. Commun.,
2005, 5307–5309.
470 | Catal. Sci. Technol., 2014, 4, 464–471
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
1
1
1
1
1
1
1
3 M.-J. Kim, W.-H. Kim, K. Han, Y. K. Choi and J. Park,
Org. Lett., 2007, 9, 1157–1159.
4 A. N. Parvulescu, P. A. Jacobs and D. E. De Vos, Adv. Synth.
Catal., 2008, 350, 113–121.
5 M. Shakeri, C. Tai, E. Göthelid, S. Oscarsson and
J.-E. Bäckvall, Chem.–Eur. J., 2011, 17, 13269–13273.
6 L. K. Thalén, D. Zhao, J.-B. Sortais, J. Paetzold, C. Hoben
and J.-E. Bäckvall, Chem.–Eur. J., 2009, 15, 3403–3410.
7 A. N. Parvulescu, P. A. Jacobs and D. E. De Vos, Appl. Catal.,
A, 2009, 368, 9–16.
20 S. L. Yan, M. Kim, S. O. Salley and K. Y. S. Ng, Appl. Catal.,
A, 2009, 360, 163–170.
21 H. Liu, L. Su, F. Liu, C. Li and U. U. Solomon, Appl. Catal.,
B, 2011, 106, 550–558.
22 B. Zhang, Y. Zhu, G. Ding, H. Zheng and Y. Li, Appl. Catal.,
A, 2012, 443–444, 191–201.
23 Y. Han, S. S. Lee and J. Y. Ying, Chem. Mater.,
2006, 18, 643–649.
24 D. Zhao, Q. Huo, J. Feng, B. F. Chmelka and G. D. Stucky,
J. Am. Chem. Soc., 1998, 120, 6024–6036.
25 J. Fan, C. Yu, F. Gao, J. Lei, B. Tian, L. Wang, Q. Luo, B. Tu,
W. Zhou and D. Zhao, Angew. Chem., Int. Ed., 2003, 42,
3146–3150.
8 N. Boz, N. Degirmenbasi and D. M. Kalyon, Appl. Catal., B,
2
009, 89, 590–596.
9 R. Pellegrini, G. Leofanti, G. Agostini, L. Bertinetti,
S. Bertarione, E. Groppo, A. Zecchina and C. Lamberti,
J. Catal., 2009, 267, 40–49.
26 C. Li, Q. Zhang, Y. Wang and H. Wan, Catal. Lett., 2008, 120,
126–136.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 464–471 | 471