Nickel nanoparticles as racemization catalysts for primary amines

Article abstract of DOI:10.1002/ejic.201300074

By combining bases that are known to racemize benzylic amines with a nickel(II) salt, active nickel nanoparticles were obtained that can be used as catalysts in the racemization of both aliphatic and benzylic primary amines. The nanoparticles are stable in the ionic liquid tetrabutylammonium bromide and can complete most racemizations within a few hours with excellent selectivity. The problem of the incompatibility of the strongly reducing racemization catalyst and the enzymatic amine resolution catalyst was overcome by using a two-pot system with a biphasic racemization step. Consecutive contact of a nonane layer that contained the amine with the acylating enzyme and with the racemizing Ni nanoparticles in the ionic liquid allowed the 50 % amide yield limit of a kinetic resolution to be successfully surpassed. Copyright

Full text of DOI:10.1002/ejic.201300074

FULL PAPER
DOI:10.1002/ejic.201300074
Nickel Nanoparticles as Racemization Catalysts for
Primary Amines
Inge Geukens,[a] Eva Plessers,[a] Jin Won Seo,[b] and
Dirk E. De Vos*[a]
Keywords: Nanoparticles / Ionic liquids / Nickel / Racemization / Amines
By combining bases that are known to racemize benzylic
amines with a nickel(II) salt, active nickel nanoparticles were
obtained that can be used as catalysts in the racemization of
both aliphatic and benzylic primary amines. The nanopar-
ticles are stable in the ionic liquid tetrabutylammonium
bromide and can complete most racemizations within a few
hours with excellent selectivity. The problem of the incom-
patibility of the strongly reducing racemization catalyst and
the enzymatic amine resolution catalyst was overcome by
using a two-pot system with a biphasic racemization step.
Consecutive contact of a nonane layer that contained the
amine with the acylating enzyme and with the racemizing
Ni nanoparticles in the ionic liquid allowed the 50% amide
yield limit of a kinetic resolution to be successfully surpassed.
Introduction
The combination of an enzymatic acylation and a cata-
lytic racemization is an effective method to produce a wide
range of highly enantiomerically pure and, hence, indus-
trially relevant chiral amines. When the two processes are
performed simultaneously, this process is called dynamic ki-
netic resolution (DKR). However, finding an active and se-
lective racemization catalyst for both benzylic and aliphatic
primary amines still remains a challenge. In their extensive
1997 review, Ebbers et al. discussed the different types of
racemization mechanisms that are possible for various com-
pounds, including primary amines.[1] Considering also more
recent work on the topic, roughly four mechanisms can be
identified for racemization of chiral amines: enzymatic
racemizations, redox racemizations, radical racemizations,
and base-catalyzed racemizations (Figure 1).
The enzymatic racemization of primary chiral amines has Figure 1. The different mechanisms that can be used to racemize
primary amines.
only recently been achieved by Koszelewski and co-workers
using ω-transaminases (ω-TA).[2] Both aliphatic and benz-
ylic amines could be racemized, but no DKR experiments
were conducted, and relatively long reaction times were
ture is the redox system. It requires the use of a transition
needed to reach full racemization. Another disadvantage is
metal, like homogeneous Ir and Ru complexes, which have
the necessity of co-factors, such as pyridoxal-5Ј-phosphate.
been used successfully in DKR; but the catalysts are quite
expensive and often difficult to recycle.[3–8] Pd is another
extensively examined transition metal, and several examples
exist of Pd immobilized on different support types that gave
good results for the DKR of amines.[9–16] However, these
quite costly Pd catalysts usually react more slowly with ali-
phatic amines, and most Pd systems are therefore only ap-
plicable for benzylic amines. Furthermore, these catalysts
have the tendency to form rather large amounts of second-
ary amine condensation products, thus leading to low selec-
tivities. Raney Ni was used by our group as a cheap and
The most widely applied racemization concept in the litera-
[a] Department of Microbial and Molecular Systems, KU Leuven,
Centre for Surface Chemistry and Catalysis,
Kasteelpark Arenberg 23, 3001 Leuven, Belgium
Fax: +32-16-321998
E-mail: dirk.devos@biw.kuleuven.be
Homepage: www.biw.kuleuven.be/m2s/cok
[b] Department of Metallurgy and Materials Engineering, KU
Leuven, Functional Materials Section,
Kasteelpark Arenberg 44, bus 02450, 3001 Leuven, Belgium
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201300074.
Eur. J. Inorg. Chem. 2013, 2623–2628
2623
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
recyclable catalyst for the DKR of aliphatic amines. The those methods in which hydrogen was evolved during nano-
catalyst could also be used to racemize benzylic amines, but particle formation were found to give active racemization
a two-pot system had to be applied to combine the system catalysts. This is a clear indication that nickel hydride spe-
with an enzymatic acylation owing to incompatibility is- cies are responsible for the catalytic activity of the nanopar-
sues.[17] Although these catalysts showed markedly better ticles, as in the case of Raney Ni, which is a typical hydro-
selectivities than Pd, large quantities of the catalyst genation catalyst. After detailed optimization of the synthe-
(Ͼ200 mol-%) as well as long reaction times (Ͼ24 h) are sis procedure, the reduction of nickel(II) bromide with so-
still necessary to obtain full racemization.
dium hydride in the presence of lithium tert-butoxide
In addition to redox catalysts, radicals can also be em- (LiOtBu) in toluene gave the most active NPs for the race-
ployed as racemization catalysts, and the use of thiyl radi- mization reaction (Scheme 1).
cals resulted in moderate to good yields in a multistep
DKR. As opposed to the case of Pd, however, these radicals
are only known to racemize aliphatic amines, due to signifi-
cant side-product formation when using benzylic amines,
Scheme 1. Reduction reaction for Ni NPs synthesis.
and quite large quantities of catalyst are required to per-
Dubois et al. describe a similar synthesis in THF of Ni–
Al NPs for use in hydrogenation reactions, and claim that
the role of the added metal alkoxide is to activate the metal
hydride, which facilitates the reduction of the metal ions.[28]
Houdayer et al. proposed that the alkoxide also acts as a
stabilizing agent for the nickel nanoparticles.[29]
The as-synthesized nanoparticles were tested for their
capability to racemize three different amines: the benzylic
amine (S)-(–)-1-phenylethylamine [(S)-PEA], and the ali-
phatic amines (S)-(+)-1-cyclohexylethylamine [(S)-CHEA]
and (S)-(+)-2-aminoheptane [(S)-AH]. As can be seen from
Table 1, (S)-PEA is completely racemized within one hour
(Table 1, entry 1), though selectivity is not yet perfect.
Bis(1-phenylethyl)amine was formed as side product by
means of a similar reaction mechanism proposed by Parvu-
lescu et al.[17] The aliphatic amines, on the other hand,
needed longer reaction times (4–5 h), but the catalysts dis-
played excellent selectivities (Table 1, entries 2 and 3). This
form the reaction (120 mol-%).[18–21]
Base-catalyzed racemizations have only scarcely been re-
ported in the literature. In a few patents, alkali-metal alk-
oxides or hydrides were used as fast and cheap racemization
catalysts for benzylic amines with high selectivities.[1,22,23]
The abstraction of the proton at the chiral center can be
facilitated by adding a ketone or aldehyde, preferably with
electron-withdrawing groups, to the mixture prior to ad-
dition of the base. This results in the formation of an imine,
which makes the proton more acidic and hence more liable
to be removed by the base afterwards (Figure 1, b).[1,24,25]
In addition to alkali-metal alkoxides, bases such as diazabi-
cycloundecane (DBU) or diazabicyclooctane (DABCO) can
be used for this racemization. Only benzylic amines seem to
undergo base-catalyzed racemizations with sufficient rates,
clearly because the intermediates become stabilized through
resonance. There are, to the best of our knowledge, no ex-
amples in the literature in which these bases are directly
used as racemization catalysts in a DKR, probably owing
to compatibility issues with the enzymatic acylation.
Table 1. Amine racemization using Ni NPs and Raney Ni in tolu-
ene.
Herein we report an improved system for the racemiza-
tion of primary amines that exploits the beneficial proper-
ties of nickel. In contrast to Pd, Ni has only seldomly been
used in racemization. Just like the activity of Pd can be
enhanced by working with nanoparticles (NPs),[14–16] nickel
nanoparticles were used to increase the activity and thus
lower the amount of catalyst needed for the racemization
reaction relative to the use of Raney Ni.[17] By using ionic
liquids, stable catalysts with high racemization activity and
selectivity for both aliphatic and benzylic amines were ob-
tained. Furthermore, a method was developed to make this
system, which contains bases that are incompatible with the
enzymatic system, combinable with a kinetic resolution
(KR). This leads to enhanced yields of enantiopure amides.
Catalyst
[mmol]
R
t
[h]
Conv.
[%][a]
Sel.
ee
[%][b]
[%][c]
Ideal
–
–
Ph
C6H11
nC5H11
–
1
4
5
50
54
47
40
100
87
99
0
0
7
1
2
3
NiNPs (0.1)[d]
NiNPs (0.1)[d]
NiNPs (0.1)[d]
96
24
4
5
6
Raney Ni (0.3)[e] Ph
Raney Ni (0.5)[e] Ph
Raney Ni (0.8)[e] Ph
1
1
1
24
53
64
61
58
49
68
21
7
[a] Conversion of (S)-amine. [b] Selectivity for (R)-amine. [c] Enan-
tiomeric excess of (S)-amine after reaction. [d] Reaction conditions:
NiBr2 (0.1 mmol), LiOtBu (0.1 mmol), NaH (0.95 mmol), amine
(0.33 mmol) in toluene (1 mL), 110 °C, 500 rpm, Ar. [e] Reaction
conditions: Raney Ni, amine (0.33 mmol) in toluene (1 mL),
110 °C, 500 rpm, Ar. Raney Ni was washed with ethanol and tolu-
ene prior to use. All reactions were analyzed by chiral GC in com-
Results and Discussion
As a starting point, different methods were applied to
synthesize Ni nanoparticles. These included the polyol
method,[26] electrochemical NP synthesis,[27] use of stabi-
lized bis(cyclooctadiene) nickel(0), and the use of reducing
agents such as sodium borohydride or sodium hydride. Only bination with GC–MS.
Eur. J. Inorg. Chem. 2013, 2623–2628
2624
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
is a vast improvement over the use of Raney Ni, as the fully applied for the stabilization of Ni NPs in other reac-
amount of nickel required is almost 15 times lower than tion types, such as aromatic amination reactions,[30] but so
that with Raney Ni. Furthermore, the reaction time is dras- far not yet for the racemization of amines. Several types of
tically shortened (from 24–72 h to 1–5 h), and there is no ILs were tested, and of all the ILs, ammonium-based ILs
need to apply hydrogen pressure to the system. For com- with halide anions proved to be the most chemically resist-
parison, the same reaction conditions were applied to ant to the reducing action of NaH (see the Supporting In-
Raney Ni: very poor selectivities were obtained and large formation). This is consistent with a recent patent, which
catalyst concentrations were necessary to obtain full race- reports the use of these ionic liquids as alternative media
mization (Table 1, entries 4–6).
for dissolving hydrides and shows that, despite what is often
Next, the effect of the three components, NiBr2, NaH, thought, the ionic liquid does not get degraded by the hy-
and LiOtBu, on the racemization reaction was examined dride.[31] The highest activity was obtained in tetrabutyl-
(Table 2). This shows that in the case of (S)-PEA, NaH ammonium bromide (TBABr), in which (S)-PEA could be
alone can catalyze the reaction, and the addition of NiBr2 fully racemized within one hour without precipitation of
and LiOtBu is not necessary for complete racemization the nanoparticles (Table 3, entry 1).
(Table 2, entry 3). This is consistent with patent litera-
ture.[23] However, for the two aliphatic amines, the presence
of both NiBr2 and LiOtBu is essential, irrespective of the
NaH concentration (Table 2, entries 4 and 5). These results
seem to imply that a combined base-catalyzed and redox
racemization is taking place (Figure 1), which could explain
why the nanoparticles are so active.
Table 2. Effect of the components of the Ni NP system on the race-
mization activity.[a]
NiBr2 LiOtBu NaH
[mmol] [mmol] [mmol]
R
Time Conv. Sel.
[h] [%][b] [%][c]
ee
[%][d]
1
2
3
4
5
0.1
–
–
–
–
–
0.1
–
–
–
–
–
Ph
Ph
1
1
1
4
5
0
1
50
1
0
0
98
Ͼ99
0
100
100
1
98
100
0.95 Ph
0.95 C6H11
0.95 nC5H11
Figure 2. (a) TEM pictures of the nanoparticles in toluene after
30 min of reaction (scale bar: 10 nm) and (b,c) in TBABr (scale
bar: 20 nm) after two hours of reaction.
0
[a] Reaction conditions: amine (0.33 mmol) in toluene (1 mL),
110 °C, 500 rpm, Ar. The reaction products were determined by
chiral GC in combination with GC–MS. [b] Conversion of (S)-
amine. [c] Selectivity for (R)-amine. [d] Enantiomeric excess of (S)-
amine after reaction.
To check if the nanoparticles are indeed stable during
reaction, TEM pictures were recorded of the nanoparticles
after 2 and 4 h of reaction. These show that the size of the
particles does not increase during reaction (see the Support-
TEM pictures were recorded of the Ni NPs. Very fine ing Information). Also remarkable is the cubical shape of
spherical nanoparticles of about 2–4 nm in size were iden- the particles (Figure 2, c). LaGrow et al. described a similar
tified (Figure 2, a). Despite their high activity, the nanopar- synthesis of Ni nanocubes in the presence of trioctylphos-
ticles were not stable in toluene, and a clear precipitation phane and hexadecylamine in a reducing hydrogen atmo-
was observed during and after reaction: the high tempera- sphere and claim that the cubical particles are formed be-
ture probably causes the particles to precipitate, and Li- cause the surfactants and the hydrogen selectively adsorb
OtBu is not sufficiently stabilizing to avoid this. The ad- on the (100) face, thus making it thermodynamically more
dition of stabilizing polymers such as polyvinylpyrrolidone, stable than the (111) face.[32] A similar hypothesis was put
polyethylene glycol, or polyacrylonitrile did not improve the forward by Li et al., who synthesized PbS NPs in
stability significantly. On the contrary, the activity of the TBABr.[33] Figure 2 shows that the initial Ni particle size in
nanoparticles drastically decreased after addition of the TBABr is considerably larger than the particle size in tolu-
polymers (see the Supporting Information), most likely be- ene. Nevertheless, the overall activity of Ni in the IL is at
cause the functional groups of these polymers adsorb too least as high as in toluene (compare Tables 1 and 3). Al-
strongly on the nickel surface. For this reason, we decided though the NPs in toluene are highly active at the start of
to assess the suitability of ionic liquids (ILs) as solvent for the reaction, they quickly become inactive due to aggrega-
the NP synthesis. Ionic liquids have already been success- tion and precipitation. The cubical shape of the Ni NPs in
Eur. J. Inorg. Chem. 2013, 2623–2628
2625
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Table 3. Racemization of amines using Ni NPs stabilized in the periments had shown that CAL-B was capable of per-
ionic liquid tetrabutylammonium bromide.[a]
forming a successful KR using several acylating agents with
(S)-PEA under the same reaction conditions as the racemi-
zation (see the Supporting Information). After initial test-
ing in a one-pot system, the presence of NaH appeared to
cause serious incompatibility issues, as it reacts with the
acylating agent used during the KR. NaH attacks the pro-
ton in the position α to the carbonyl group of the ester, thus
initiating a Claisen condensation, which results in forma-
tion of a β-keto ester together with a sodium alkoxide salt.
This process results in a drastic reduction of the NaH and
ester concentration in solution.[35] Both the racemization
and KR are thus affected in this manner (see the Support-
ing Information). Moreover, NaH also reacts with the (R)-
amide formed during the KR, in an acid–base reaction in
which an amide salt is formed. Although this reaction is
reversible and the amide can be obtained again after reac-
tion (see the Supporting Information), a significant amount
of NaH is again lost because of this reaction. Finally, NaH
can also react with the acidic protons of the enzyme and
disturb its catalytic action. All this makes a monophasic,
one-pot DKR problematic.
R
t
[h]
Conv.
[%][b]
Sel.
ee
Yield
[%][e]
[%][c]
[%][d]
1
2
3
4
5
6
7
8
9
Ph
C6H11
1
2
5
5
5
2
2
2
2
49
49
38
41
42
42
31
46
35
95
Ͼ 99
97
99
98
99
94
64
Ͼ 99
6
2
94
96
91
89
92
88
88
62
91
nC4H9
nC5H11
nC6H13
4-Me-Ph
4-Cl-Ph
4-F-Ph
2-naphthyl
25
17
17
16
41
29
30
[a] Reaction conditions: NiBr2 (0.1 mmol), LiOtBu (0.1 mmol),
NaH (0.95 mmol), amine (0.33 mmol) in TBABr (1 mL), 110 °C,
500 rpm, Ar. The reaction products were determined by chiral GC
in combination with GC–MS after extraction with diethyl ether.
[b] Conversion of (S)-amine. [c] Selectivity for (R)-amine. [d] Enan-
tiomeric excess of (S)-amine after reaction. [e] Isolated yield.
To overcome these problems, a biphasic system was first
developed, physically separating the KR from the racemiza-
tion reaction. Such a concept has already been applied for
the IL could also be related to the high selectivity: the selec- the DKR of alcohols,[36] but has, to the best of our knowl-
tive adsorption of hydrogen on the (100) face could create edge, never been applied for amines. To form the two
active Ni–H species, which we have already linked to selec- phases, several organic solvents were tested to see if they
tive racemization (compare Table 1, entry 1 with Table 3, would form two clear phases with TBABr at the reaction
entry 1). Similar effects were observed for Pt nanocubes temperature. Long aliphatic alkanes, like n-nonane, proved
synthesized in tetradecyltrimethylammonium bromide for to be excellent in this respect, also because CAL-B is active
the hydrogenation of benzene.[34]
in these solvents to perform a KR. Next, we decided to
Different amines were tested for their ability to be race- work with long aliphatic acylating agents, such a methyl
mized by the NiNPs in TBABr; both benzylic and several nonanoate, to prevent them from dissolving in the TBABr
aliphatic amines could be racemized with excellent selectivi- phase and being inactivated by the racemization system. As
ties. Even halogenated amines could be racemized, although a bonus, the amides that are formed contain long aliphatic
in these cases hydrogenolysis of the C–X bond takes place chains, strongly enhancing their affinity for n-nonane. The
over the nickel catalyst, thus leading to lower selectivities amine can thus freely migrate from one phase to the other,
(Table 3, entries 7 and 8).
as was proven by extraction experiments, whereas the amide
In contrast to the reaction in toluene, NaH itself does is retained in the upper phase owing to its long aliphatic
not show racemization activity in TBABr, and the presence tail. Finally, to avoid contact of the enzyme with the TBABr
of all three components, namely, NiBr2, NaH, and LiOtBu, phase, a stainless-steel basket was designed to retain the
is necessary to obtain racemization activity, even for the enzyme in the n-nonane phase. Despite these efforts, the
benzylic amines. Moreover, even the presence of a small biphasic DKR did not give higher yields than the biphasic
amount of ionic liquid in toluene inhibited the catalytic ac- KR, thus indicating that the racemization reaction does not
tion of NaH (see the Supporting Information). This is sur- take place. The methanol formed during the KR probably
prising, since Ebbers et al. found that base racemization migrates to the TBABr phase and reacts with NaH to form
catalysts tend to be more active in polar solvents, such as a methoxide, thereby inhibiting the racemization reaction.
dimethyl sulfoxide.[1] The lack of activity of NaH in the Even with other acylating agents, such as nonyl nonanoate,
ionic liquid could be due to the charges present in the ionic which form hydrophobic aliphatic alcohols, this problem
liquid, or to the high water sensitivity of NaH in the hygro- occurred.
scopic IL, although the ionic liquid was thoroughly dried
prior to use.
Therefore, eventually a two-pot system with a biphasic
racemization step was successfully developed (Figure 3). Af-
Finally, we aimed to combine the optimized racemization ter the first KR step in n-nonane, the methanol co-product
system with an enzymatic system to perform a DKR. For is quickly withdrawn from the reaction mixture, after which
the kinetic resolution (KR), immobilized Candida antarc- the nonane layer is added on top of the TBABr phase. This
tica lipase B (CAL-B) was used as enzyme because of its starts a biphasic racemization, which is followed by another
broad substrate scope and good thermal stability. Prior ex- transfer of the nonane layer to the enzymatic resolution cat-
Eur. J. Inorg. Chem. 2013, 2623–2628
2626
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
alyst. This system was applied successfully to (S)-PEA and enhanced affinity for the n-nonane phase. This two-pot sys-
resulted after two resolution steps in a 68% yield of (R)- tem is not only applicable for our Ni NPs system, but can
(+)-N-(1-phenylethyl)nonanamide with an enantiomeric ex- also potentially be applied to other incompatible racemiza-
cess of 97% (compare with a maximal theoretical yield of tion systems, thereby opening opportunities to use base-cat-
75%). To illustrate the biphasic behavior of the racemiza- alyzed racemizations in the synthesis of enantiopure benz-
tion system, inductively coupled plasma atomic emission ylic amines.
spectroscopy (ICP-AES) analysis was performed on the up-
per n-nonane phase during reaction, and a negligible
amount of Ni leaching (0.0001%) to the upper phase was
Experimental Section
found (see the Supporting Information). The nanoparticles
also maintain their activity in at least three consecutive bi- out under an inert atmosphere. The reactions were carried out in
General: Unless indicated otherwise, all manipulations were carried
oven-dried glass reaction vials (8 mL) and were magnetically stirred
(500 rpm). The reaction mixtures were analyzed with a Shimadzu
2014 GC equipped with a Chiralsil-DEX CB (CP7503) chiral col-
umn and a flame ionization detector (FID) using n-tetradecane as
internal standard. For some amines, the addition of a few drops of
acetic anhydride and triethylamine to the GC vial was necessary to
obtain a good separation on the chiral GC (see also the Supporting
Information). Mass spectra were obtained with an Agilent 6890N
GC with a FID detector and a HP-5MS column coupled to a 5973
MSD mass spectrometer. TEM measurements were carried out
with a Philips FEG CM200 apparatus (200 kV) equipped with an
EDX spectrometer. To prepare the carbon-coated Cu grids
(200 mesh), diluted mixtures of the NPs in toluene were added
dropwise onto the grid, after which the solvent was allowed to
evaporate. All solvents used were dried prior to use by passing them
over molecular sieves (zeolite 5 Å). All ionic liquids were dried un-
der vacuum at 120 °C prior to use. Amines and acylating agents
with high purity were ordered from Sigma–Aldrich.
phasic racemization steps, with the same yield of (R)-(+)-
N-(1-phenylethyl)nonanamide obtained after the three runs.
General Procedures for the Racemization Reactions
Reactions in Toluene: NiIIBr2 (0.1 mmol, 22 mg), LiOtBu
(0.1 mmol, 8 mg), and NaH (60% dispersion in mineral oil;
0.95 mmol, 23 mg) were added to the reaction vial, after which tol-
uene (1 mL) was added, and the Ni NPs were formed by heating
the mixture to 110 °C under reflux conditions for 1 h. The (S)-
amine [0.33 mmol, 42 μL for (S)-PEA] was then added to the pre-
formed NPs and heated again to 110 °C for 1 h. After reaction, the
mixture was cooled and centrifuged at 3000 rpm. The supernatant
was washed with water to eliminate possible traces of NaH or
LiOtBu. The organic phase was then analyzed by chiral GC and
GC–MS.
Reactions in TBABr: NiIIBr2 (0.1 mmol, 22 mg), LiOtBu (0.1 mmol,
8 mg), NaH (0.95 mmol, 60% dispersion in mineral oil; 23 mg),
and TBABr (1 g, solid at room temp.) were added to the reaction
vial and the Ni NPs were formed by heating the mixture to 110 °C
during 1 h. (S)-Amine [0.33 mmol, 42 μL for (S)-PEA] was then
added to the preformed NPs, and the mixture was heated again to
110 °C for 1 h. After reaction, the mixture was cooled down and
water (2 mL) and diethyl ether (1 mL) were added to the mixture,
after which it was shaken and allowed to form two phases. The
upper organic phase was then analyzed by chiral GC and GC–MS.
To determine the isolated yields, the reactions were performed on
a scale that was three times larger and underwent the same treat-
ment as described above, only the aqueous phase was extracted
three more times with diethyl ether. The organic fractions were col-
lected and subsequently dried with MgSO4, after which diethyl
ether was removed under vacuum. The residue was further purified
using flash chromatography [n-heptane/ethyl acetate (1:1) mixture].
Figure 3. Schematic representation of the two-pot system using a
biphasic racemization step.
Conclusion
Nickel nanoparticles proved to be excellent catalysts for
the racemization of primary amines. Of the different synthe-
sis methods used, the reduction of nickel(II) bromide with
NaH in the presence of lithium tert-butoxide gave highly
active and selective particles for the racemization of pri-
mary amines. The combined action of NaH, which was pre-
viously used in the literature to racemize benzylic amines,
and the nickel metal is probably what causes the high ac-
tivity of the particles. To increase the stability of the par-
ticles, several ionic liquids were tested and good results were
obtained using TBABr. Both benzylic and aliphatic amines
could be racemized with excellent selectivity by the ionic-
liquid-stabilized Ni NPs within a short range of time. To
combine this system with an enzymatic acylation, a special
two-pot system was developed with a biphasic racemization
step using n-nonane as the second phase in addition to
TBABr. This was done to avoid contact between the enzy-
matic system and the NaH-containing racemization system.
After screening several acylating agents, methyl nonanoate
was chosen as acylating agent because it shows no affinity
Two-Pot Synthesis
Kinetic Resolution (KR): n-Nonane (1 mL), the racemic amine
for the TBABr phase and produces amides with a strongly (0.33 mmol), and methyl nonanoate (0.165 mmol, 32 μL) were
Eur. J. Inorg. Chem. 2013, 2623–2628
2627
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
[5] C. E. Hoben, L. Kanupp, J. E. Bäckvall, Tetrahedron Lett.
2008, 49, 977–979.
[6] L. K. Thalén, D. B. Zhao, J. B. Sortais, J. Paetzold, C. Hoben,
J. E. Bäckvall, Chem. Eur. J. 2009, 15, 3403–3410.
[7] M. Stirling, J. Blacker, M. I. Page, Tetrahedron Lett. 2007, 48,
1247–1250.
[8] A. J. Blacker, M. J. Stirling, M. I. Page, Org. Process Res. Dev.
2007, 11, 642–648.
[9] M. T. Reetz, K. Schimossek, Chimia 1996, 50, 668–669.
[10] Y. K. Choi, M. J. Kim, Y. Ahn, M.-J. Kim, Org. Lett. 2001, 3,
4099–4101.
added to Novozyme 435 (100 mg) and reacted for 24 h at 110 °C
to obtain a 50% conversion with an enantiomeric excess of 99%.
Biphasic Racemization (Figure 3): After the KR, the liquid phase
was removed and the solid enzyme was stored for later use in the
second kinetic resolution. The liquid organic phase was dried with
molecular 5 Å sieves to remove the methanol and possible traces
of water. The mixture was then added to TBABr that contained
the preformed Ni NPs (formed similarly as described above), and
a biphasic mixture was formed. A magnetic stirrer (500 rpm) was
used in the TBABr phase to enhance phase transfer between the
two phases. After 6 h at 110 °C, the mixture was cooled down and
water (1 mL) and diethyl ether (0.5 mL) were added to the biphasic
system. The upper organic phase was removed and the TBABr
phase that contained the NPs was extracted three more times with
a 1:1 mixture of n-nonane/diethyl ether. All organic phases were
pooled and dried with MgSO4. Finally, the diethyl ether was evapo-
rated from the mixture and the resulting organic mixture was added
to the previously stored solid enzyme, together with methyl non-
anoate (0.165 mmol, 32 μL), and allowed to perform a second KR.
The product, (R)-(+)-N-(1-phenylethyl)nonanamide, was analyzed
by GC–MS (70 eV, EI): m/z (%) = 261 (39) [M], 176 (17), 163 (72),
148 (7), 120 (44), 150 (100).
[11] A. Parvulescu, D. E. De Vos, P. Jacobs, Chem. Commun. 2005,
5307–5309.
[12] A. N. Parvulescu, E. Van der Eycken, P. A. Jacobs, D. E.
De Vos, J. Catal. 2008, 255, 206–212.
[13] A. N. Parvulescu, P. A. Jacobs, D. E. De Vos, Appl. Catal. A
2009, 368, 9–16.
[14] M. J. Kim, W. H. Kim, K. Han, Y. K. Choi, J. Park, Org. Lett.
2007, 9, 1157–1159.
[15] Y. Kim, J. Park, M. J. Kim, Tetrahedron Lett. 2010, 51, 5581–
5584.
[16] M. Shakeri, C.-w. Tai, E. Göthelid, S. Oscarsson, J.-E.
Bäckvall, Chem. Eur. J. 2011, 17, 13269–13273.
[17] A. N. Parvulescu, P. A. Jacobs, D. E. De Vos, Adv. Synth. Ca-
tal. 2008, 350, 113–121.
[18] S. Escoubet, S. Gastaldi, N. Vanthuyne, G. Gil, D. Siri, M. P.
Bertrand, Eur. J. Org. Chem. 2006, 3242–3250.
[19] S. Gastaldi, S. Escoubet, N. Vanthuyne, G. Gil, M. P. Bertrand,
Org. Lett. 2007, 9, 837–839.
[20] L. El Blidi, N. Vanthuyne, D. Siri, S. Gastaldi, M. P. Bertrand,
G. Gil, Org. Biomol. Chem. 2010, 8, 4165–4168.
[21] F. Poulhès, N. Vanthuyne, M. P. Bertrand, S. Gastaldi, G. Gil,
J. Org. Chem. 2011, 76, 7281–7286.
[22] J. R. Jansen, M. Littmann, Bayer Aktiengesellschaft (Lever-
kusen, DE) 1993, US 5,183,939.
[23] G. Bison, K. Breideneichen, W. Heinzelmann, W. Wolfes, Dyn-
amit Nobel Aktiengesellschaft (Troisdorf, DE) 1981, US
4,252,744.
[24] K. Ditrich, BASF Aktiengesellschaft (Ludwigshafen, DE)
1998, US 5,847,215.
[25] S. Nagata, Y. Yamada, K. Hagiya, H. Goto, Sumitomo Chemi-
cal Company, Limited (Osaka, JP) 1998, EP 0 778 260 B1.
[26] R. Laine, A. Sellinger, Canon Kabushiki Kaisha (Tokyo, JP)
2003, US Patent 6,551,960.
Consecutive Biphasic Racemizations: After the biphasic racemiza-
tion step, the upper n-nonane phase was removed and stored, and
a fresh n-nonane phase, which underwent a KR similar as described
in the two-pot synthesis, was added to the TBABr phase with the
Ni NPs and allowed to react for another 6 h at 110 °C. This was
done two times, so in total three biphasic racemization steps were
performed using the same Ni NPs in TBABr. After the three reac-
tions, all n-nonane phases were pooled, and the TBABr phase was
treated similarly to that described above for the two-pot synthesis.
All organic phases were added together, dried with MgSO4, and the
diethyl ether was evaporated again from the mixtures. The resulting
organic mixture was added to the previously stored solid enzyme,
together with methyl nonanoate (0.495 mmol, 98 μL), and allowed
to perform a second KR.
Supporting Information (see footnote on the first page of this arti-
cle): Stability experiments with polymers and ionic liquids, ad-
ditional TEM pictures, extraction and compatibility experiments,
chiral column information, and ICP-AES measurements.
[27] M. Reetz, W. Helbig, S. Quaiser, Studiengesellschaft Kohle
mbH (Mulheim an der Ruhr, DE) 1997, US Patent 5,620,584.
[28] J.-M. Dubois, Y. Fort, O. Tillement, Centre National de la Re-
cherche Scientifique (Paris, FR) 2002, US 6,847,215 B1.
[29] A. Houdayer, R. Schneider, D. Billaud, J. Ghanbaja, J. Lam-
bert, Synth. Met. 2005, 151, 165–174.
Acknowledgments
I. G. thanks the Research Foundation Flanders (FWO) for finan-
cial support as research assistant. Dr. J. Jammaer is thanked for
help with the TEM measurements. The authors are also grateful to
the Agency for Innovation by Science and Technology (IWT) for
support through the SBO project MAPIL and to the Federal Gov-
ernment of Belgium for support within the IAP-PAI P7/05 project.
J. W. S. thanks the Flemish Hercules Stichting for its support
through the project HER/08/25.
[30] I. Geukens, J. Fransaer, D. E. De Vos, ChemCatChem 2011, 3,
1431–1434.
[31] H. Nöth, U. Wietelmann, CHEMETALL GMBH (Frankfurt,
DE) 2009, WO/2009/101201.
[32] A. P. LaGrow, B. Ingham, S. Cheong, G. V. M. Williams, C.
Dotzler, M. F. Toney, D. A. Jefferson, E. C. Corbos, P. T.
Bishop, J. Cookson, R. D. Tilley, J. Am. Chem. Soc. 2012, 134,
855–858.
[33] G. Li, C. Li, H. Tang, K. Cao, J. Chen, Mater. Res. Bull. 2011,
46, 1072–1079.
[1] E. J. Ebbers, G. J. A. Ariaans, J. P. M. Houbiers, A. Bruggink,
B. Zwanenburg, Tetrahedron 1997, 53, 9417–9476.
[2] D. Koszelewski, B. Grischek, S. M. Glueck, W. Kroutil, K.
Faber, Chem. Eur. J. 2011, 17, 378–383.
[3] O. Pàmies, A. H. Ell, J. S. M. Samec, N. Hermanns, J. E.
Bäckvall, Tetrahedron Lett. 2002, 43, 4699–4702.
[4] J. Paetzold, J. E. Bäckvall, J. Am. Chem. Soc. 2005, 127, 17620–
17621.
[34] K. M. Bratlie, H. Lee, K. Komvopoulos, P. Yang, G. A. Somor-
jai, Nano Lett. 2007, 7, 3097–3101.
[35] F. W. Swamer, C. R. Hauser, J. Am. Chem. Soc. 1950, 72, 1352–
1356.
[36] S. Wuyts, K. De Temmerman, D. E. De Vos, P. A. Jacobs,
Chem. Eur. J. 2005, 11, 386–397.
Received: January 20, 2013
Published Online: March 19, 2013
Eur. J. Inorg. Chem. 2013, 2623–2628
2628
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim