Batch versus flow stereoselective hydrogenation of Α-acetamido-cinnamic acid catalyzed by an Au(I) complex

Article abstract of DOI:10.1016/j.mcat.2019.110420

A chiral gold (I) (2S,4S)-1-tert-butoxycarbonyl-4-diphenylphosphino-2-(diphenylphosphino- methyl) pyrrolidine (BPPM) complex has been prepared using [Au(SMe₂)Cl] as precursor. The heterogenization of the Au-BPPM catalyst onto the CNT support followed two routes, ie (i) the non-covalent immobilization of the gold(I)complex by dry-impregnation, and (b) covalent immobilization of the gold(I)complex on a pre-functionalized CNT. These catalysts afford the stereoselective hydrogenation of α-acetamidocinnamic acid to the (R)-N-acetyl-phenylalanine enantiomer. The nature of the solvent affected both the enantioselectivity and TOFs. Among MeOH, EtOH, and TFE, methanol appeared to be the most efficient one (at 80 °C a TOF of 0.37 h^?1 for a total enantioselectivity to the R-isomer). Transferring the reaction in the flow reactor, under similar conditions (methanol, room temperature) led to a 10 time increase of the TOF with no change in the stereoselectivity. The decrease of the TOF in time for both the reference Rh and the Au catalysts was assigned to their partial modification under the reaction conditions. The heterogenization of the Au-BPPM catalyst onto the CNT support, for the same content of Au-complex, led to a very important increase of the conversion with no change in the selectivity. However, the covalent bonding was more efficient affording a very high increase of the conversion even at room temperature (95% after 24 h), thus demonstrating that the anchoring a support increases the dispersion, and in consequence the efficiency. These CNT-Au-BPPM catalysts preserved the catalytic performances during recycling as also confirmed by the characterization results.

Full text of DOI:10.1016/j.mcat.2019.110420

Molecular Catalysis 474 (2019) 110420
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.elsevier.com/locate/mcat
Batch versus flow stereoselective hydrogenation of α-acetamido-cinnamic
acid catalyzed by an Au(I) complex
a
a
a,⁎
b
b,⁎
Alina Negoi , Bogdan Cojocaru , Vasile I. Parvulescu , Nora Imlyhen , Maryse Gouygou
a
Department of Organic Chemistry, Biochemistry and Catalysis, Faculty of Chemistry, University of Bucharest, Bd. Regina Elisabeta no. 4–12, 030018 Bucharest 030016,
Romania
b
CNRS, LCC (Laboratoire de Chimie de Coordination), Universite de Toulouse, UPS, 205, route de Narbonne, 31077 Toulouse, France
A R T I C L E I N F O
A B S T R A C T
Keywords:
A chiral gold (I) (2S,4S)-1-tert-butoxycarbonyl-4-diphenylphosphino-2-(diphenylphosphino- methyl) pyrrolidine
Au(I) (2S,4S)-1-tert-butoxycarbonyl-4-
diphenylphosphino-2-(diphenylphosphino-
methyl) pyrrolidine (BPPM) complex
Stereoselective hydrogenation of α-
acetamidocinnamic acid to the (R)-N-acetyl-
phenylalanine
Heterogenization of the Au-BPPM catalyst onto
a CNT support
Flow stereo-hydrogenation
(BPPM) complex has been prepared using [Au(SMe )Cl] as precursor. The heterogenization of the Au-BPPM
catalyst onto the CNT support followed two routes, ie (i) the non-covalent immobilization of the gold(I)complex
2
by dry-impregnation, and (b) covalent immobilization of the gold(I)complex on a pre-functionalized CNT. These
catalysts afford the stereoselective hydrogenation of α-acetamidocinnamic acid to the (R)-N-acetyl-phenylala-
nine enantiomer. The nature of the solvent affected both the enantioselectivity and TOFs. Among MeOH, EtOH,
−
1
and TFE, methanol appeared to be the most efficient one (at 80 °C a TOF of 0.37 h for a total enantioselectivity
to the R-isomer). Transferring the reaction in the flow reactor, under similar conditions (methanol, room tem-
perature) led to a 10 time increase of the TOF with no change in the stereoselectivity. The decrease of the TOF in
time for both the reference Rh and the Au catalysts was assigned to their partial modification under the reaction
conditions. The heterogenization of the Au-BPPM catalyst onto the CNT support, for the same content of Au-
complex, led to a very important increase of the conversion with no change in the selectivity. However, the
covalent bonding was more efficient affording a very high increase of the conversion even at room temperature
(95% after 24 h), thus demonstrating that the anchoring a support increases the dispersion, and in consequence
the efficiency. These CNT-Au-BPPM catalysts preserved the catalytic performances during recycling as also
confirmed by the characterization results.
1. Introduction
However, three main systems have been successfully implemented
in the development of homogeneous gold-catalyzed enantioselective
Gold (I) complexes have received considerable attention in the field
reactions [5]. Chiral monodentate phosphine-type ligands were first
used in gold-catalysis asymmetric reactions. Examples were reported
early in 1986 by Ito et al for the enantio- and diastereoselective
synthesis of a 5-alkyl-2-oxazoline-4-carboxylate via aldol reaction be-
tween isocyanoacetate and benzaldehyde [6]. Then, various phosphi-
ne–ligated gold(I) complexes have been designed for enantioselective
gold catalysis [5]. However one of the most widely used strategies is the
use of bimetallic gold complexes in particular with atrop-isomeric di-
phosphines ligands. Diphosphine–ligated gold(I) complexes enable
various asymmetric transformations such as enantioselective alkyne,
allenes and alkenes activation and enantioselective transformation of
diazo compounds [5]. The use of chiral counter-anions to achiral ca-
tionic gold(I) catalysts has emerged as an alternative strategy as in-
itially demonstrated in the intramolecular hydroamination and hydro-
alkoxylation reaction of allenes with high asymmetric induction
of organic synthetic chemistry owing to their capability to catalyze a
broad panel of transformations [1–4]. They are the most effective cat-
alysts for the electrophilic activation of alkynes under homogeneous
conditions, and a broad range of versatile synthetic tools have been
developed for the construction of carbon–carbon or carbon–heteroatom
bonds. The majority of gold (I) complexes employed as catalyst pre-
cursors are phosphine complexes or N-heterocyclic carbene complexes
of the type [AuClL].
The development of efficient asymmetric transformations with
chiral gold catalysts was challenging given the intrinsic structural
characteristics of the gold (I) complexes. Because of their linear geo-
metry, the chiral ligand (L *) and the substrate are on the opposite side
to the metal center. As a result, the substrate is easily displaced from the
chiral pocket.
⁎
Corresponding authors.
E-mail addresses: parvulescu@g.unibuc.ro (V.I. Parvulescu), maryse.gouygou@lcc-toulouse.fr (M. Gouygou).
https://doi.org/10.1016/j.mcat.2019.110420
Received 12 March 2019; Received in revised form 23 April 2019; Accepted 20 May 2019
2468-8231/ © 2019 Elsevier B.V. All rights reserved.

A. Negoi, et al.
Molecular Catalysis 474 (2019) 110420
[
7–10].
In contrast, little work has been done on enantioselective hydro-
CNT (Sigma-Aldrich) at dry. Then, the catalysts were treated under
vacuum at room temperature for 8 h.
genation catalyzed by gold(I). In 2005, Corma and co-workers reported
the first gold (I) catalyst, a dimeric gold(I) complex bearing the [(R,R)-
Me-Duphos] diphosphine ligand, able to perform enantioselective hy-
drogenation of alkenes and imines with high catalytic activities and
selectivities under mild reaction conditions [11]. The reaction me-
chanism studied by density functional theory calculations revealed an
ionic hydrogenation [12]. Other examples reported by the Corma’s
group refer to the hydrogenation of prochiral alkenes catalyzed by a bis
N-heterocyclic gold (I) complex [13].
2.1.2. Covalent immobilization of the gold(I)complex onto CNT, CNT-Au-
BPPM
(2S,4S)-BPPM grafted on CNT was prepared according to a pre-
viously reported procedure [17]. (Scheme 2) The concentration of the
grafted ligand was calculated to afford a 4 wt% Au loading after the
complexation. Then the complexation with [Au(SMe )Cl] followed the
2
same protocol with that described above (Scheme 1). The ICP-OES
analysis confirmed the loading of Au (3.92 wt%).
Heterogeneous stereoselective hydrogenation has also been carried
with gold (III) complexes. Thus, studies carried out with Au(III)−Schiff
base complexes in combination with kinetic experiments and theore-
tical calculations shown that the nature of the solid support (polarity
and proton-donating ability) is important in this reaction. As an effect,
the activity of these catalysts may be enhanced by a simple grafting of
these complexes onto selected surfaces [14].
2
.1.3. Catalyst characterization
The investigated catalysts were characterized by ATR-FTIR, Raman
and DR-UV–vis spectroscopy. Attenuated total reflection Fourier
transformed infrared (ATR-FTIR) spectra were recorded using
PerkinElmer Spectrum Two spectrometer having an ATR cell equipped
with a diamond plate (Pike Technologies, Madison, WI). The spectra
were recorded with a 4 cm
were acquired in the extended spectral region from 150 up 4000 cm
a
−
1
As part of our continuing interest in homogeneous gold catalysis
resolution at 20 scans. Raman spectra
-
1
[
15] and in heterogeneous enantioselective hydrogenation [16] we
report herein the preparation of chiral gold catalysts (Fig. 1), and their
application in enantioselective hydrogenation. The aim of this study
was to compare the behavior of a gold complex as a homogeneous or
heterogeneized catalyst (resulted either via non-covalent or covalent
linkage onto carbon nanotubes) in both batch and flow conditions.
with a Horiba JobinYvon - Labram HR UV–vis-NIR Raman Microscope
Spectrometer (˜0.4 μm resolution on X and Y axes and ˜0.7 μm resolu-
tion on Z axe), at 488 nm and 633 nm. Diffuse reflectance UV–vis
measurements were carried out with a Thermo Electron Specord 250
using an integrating sphere and MgO as reference. The slit was set at
4
nm. The spectra of the catalysts were recorded in reflectance units and
2. Experimental
were transformed in Kubelka–Munk remission function F(R).
Unless otherwise stated, all syntheses were run under Argon using
Schlenk techniques. All reagents: (2S,4S)-1-tert-butoxycarbonyl-4-di-
2.2. Catalytic reactions
phenylphosphino-2-(diphenylphosphino- methyl) pyrrolidine (BPPM),
Hydrogenation of ACA to (R)-N-acetyl-phenylalanine and of MAB
has been carried out in batch and flow reactors, respectively. Batch
reactions were carried out in a stainless steel pressure autoclave from
Parr (50 mL) using 80 mg ACA, 10 mg catalyst, and 25 mL MeOH, EtOH
or TFE. After closing, the autoclave was purged three times with ni-
[
2
Au(SMe )Cl], α-acetamidocinnamic acid (ACA), 98% purity, and me-
thyl-3-aminobutanoate (MAB), carbon nanotube (CNT), were pur-
chased from Sigma-Aldrich and used as received without further pur-
2
ification. Dichloromethane and pentane were dried under N using a
solvent purification system (SPS). Methanol (MeOH, 99.8% purity),
ethanol (MeOH, 96% purity), TFE (2,2,2-trifluoroethanol, 99.5%
purity) were purchased from Sigma-Aldrich. The (R, R)-MeDuPhos-Rh
complex was purchased from STREM Chemicals, Inc.
trogen and then pressurized at 40 atm H
orously stirred (1.000 rpm) for 24 h at temperatures in the range rt-
0 °C.
Flow reactions were carried out in a tubular stainless-steel micro-
reactor system (length of 300 mm and i.d. of 9 mm, equipped with a
thermocouple in the middle) from PID&ENG at a pressure of 40 atm H2.
2
. The mixture was then vig-
8
NMR spectra were recorded at 25 °C on Advance 400 Ultrashield
1
and Advance III Ultrashield Plus 500 MHz Bruker apparatus. H NMR,
1
3
1
31
C{ H} NMR chemical shifts are referenced to the solvent signal.
{1
H} NMR chemical shifts are referenced to an external standard (85%
aqueous H PO ). Multiplicity is indicated as follows: s = singlet, bs =
P
80 mg ACA and 10 mg catalyst were dispersed in methanol in a volume
filling of the entire system (around 60 mL). The microreactor was
placed inside a ceramic resistance furnace where the temperature
control was achieved with a TohoTTM-204 controller (K-type thermo-
couple at the middle-level of the ceramic jacket) and an Elko ELK-38
indicator (K-type thermo-couple in contact with the microreactor
body). Both instruments (calibrated against a high temperature glass
thermometer) were computer interfaced through RS-232 to TTL level
converters based on Maxim MAX-232 IC allowing bidirectional Modbus
3
4
broad singulet, d = doublet, m = multiplet. ESI analyses were per-
formed on a UPLC Xevo G2 Q TOF spectrometer.
2
2
.1. Synthesis of the gold(I) complex, Au-BPPM
.1.1. Non-covalent immobilization of the gold(I)complex onto CNT
−
1
The heterogeneous 4 wt% complex CNT@Au-BPPM (CNT-carbon
communication. The flow control (10 mL min ) was performed for the
liquid phase by a pump (KnauerK-501 HPLC pump, computer con-
trolled by RS-232 serial interface). The liquid output from the reactor
was collected into an open glass vessel. Samples were extracted from
nanotube) has been prepared by dry-impregnation of the resulted
complex. For this purpose, the calculated amount of the Au-BPPM was
dispersed in methanol and the solution was added drop-by-drop to the
Fig. 1. Gold (I) complexes used in this work.
2

A. Negoi, et al.
Molecular Catalysis 474 (2019) 110420
2
Scheme 1. Complexation of (2S,4S)-BPPM with [Au(SMe )Cl].
2
Scheme 2. Complexation of (2S,4S)-BPPM grafted on CNT with [Au(SMe )Cl].
the collector vessel with the peristaltic pump P2 (HeidolphPD 5001,
computer controlled through a parallel port relay inter-face) and dis-
pensed in individual glass tubes (covered with small funnels to reduce
evaporation) by the help of a fraction collector (Teledyne ISCO Foxy Jr,
computer connected by RS-232 serial link). The operational parameters
of the installation (temperature, liquid flow rate, collection of samples)
were controlled by a code developed in C-programming language
running on GNU/Linux operating system (32-bit CentOS 6).
1
(R)-N-acetyl-phenylalanine
H-NMR: (500.13 MHz, DMSO-d6, δ ppm, J
Hz): 12.70 (s, 1H, -COOH), 8.21 (s, 1H, H-
1
H-8 and H-9), 4.42 (m, 1H, H-2, J = 5.0
Hz, J = 3.1 Hz), 3.07-2.81 (m, 2H, H-3,
0, -NH), 7.28-7.21 (m, 5H, H-5, H-6, H-7,
A
B
A
B
J =5.0 Hz, J =8.9 Hz), 1.79 (s, 3H, H-12);
13
C-NMR: (125.77 MHz, DMSO-d6, δ ppm):
73.1 (C-1), 169.2 (C-11), 137.6 (C-4),
29.0 (C-5 and C-9), 128.1 (C-6 and C-8),
26.3 (C-7), 53.4 (C-2), 36.7 (C-3), 22.2 (C-
2).
1
1
1
1
2.3. Analysis of the reaction products
3. Results and discussions
Quantitative analysis of the reaction products was performed using
a high performance liquid chromatography coupled with fluorescent
and diode array detection (HPLC-FLD/DAD) method. The analyses were
carried out with an Agilent 1260 modular system equipped with CHI-
RALPAK QD-AX column (150 × 4.6 mm ID, 5 μm). Working parameters
of the HPLC-FLD/DAD system were: injection volume of 20 μL, mobile
3
.1. Catalysts characterization
The ATR-FTIR spectrum of Au-BPPM (Fig. 2) contains typical bands
of this complex, ie bands assigned to the vibrations of the CeH skeletal
−1
(
1229, 1259 cm ), CeH deformation of the CeCO out-of-plane de-
phase composition: MeOH:CH
3 3 4
COOH:CH COONH = 98:2:0.5, flow
−1
−1
formation (412 cm ), C–H scissor (1482 cm ), asymmetric CeH
stretching (2861 and 2972 cm ), CeNeC deformation (421, 432, 439,
4
4
rate of 1.0 mL/min, temperature of the column 25 °C, 338 nm and
−
1
2
3
62 nm for OPA and FMOC derivatives, respectively. FLD was set up at
40 nm extinction wavelength and 450 emission wave-length. Before
−
1
57 1347, 1359 and 1394 cm ), CeC]O in-plane deformation (474,
−
1
89, 618 cm− ), CeC skeletal (499, 515, 521, 545, 584 815, 855, 906,
analysis, the amino-compounds were derivatized using OPA and FMOC
reagents for primary and secondary amino groups, respectively. Finally,
1
−1
and 929 cm ), CeOeC deformation (691, 810 cm ), CeN stretching
2
00 μL sample were mixed with 200 μL derivatization agent and diluted
in 1 mL borate buffer.
2 4 5 4
The borate buffer solution contained 0.4 mol Na [B O (OH) ] in 1 L
of distilled water. The pH value was adjusted to 10.2 by boric acid. The
OPA (o-phthalaldehyde) solution was prepared by adding 10 mg OPA in
1
mL 0.4 M borate buffer. FMOC (9-fluorenylmethyl chloroformate)
solution contained 2.5 mg FMOC in 1 mL acetonitrile.
1
α-acetamidocinnamic acid
H-NMR (500.13 MHz, DMSO-d6, δ ppm, J
Hz): 12.60 (s, 1H, -COOH), 9.45 (s, 1H, H-
0, -NH), 7.62-7.37 (m, 5H, H-5, H-6, H-7,
H-8 and H-9), 7.23 (s, 1H, H-3), 1.99 (s, 3H,
1
1
3
H-12). C-NMR (125.77 MHz, DMSO-d6, δ
ppm): 169.2 (C-1), 166.4 (C-11), 133.7 (C-
3
(
2
), 131.1 (C-4), 129.7 (C-5 and C-9), 129.1
C-7), 128.5 (C-6 and C-8), 127.3 (C-2),
2.4 (C-9).
Fig. 2. ATR-FTIR spectra of Au-BPPM, CNT@Au-BPPM and spent CNT@Au-
BPPM catalysts.
3

A. Negoi, et al.
Molecular Catalysis 474 (2019) 110420
Table 1
Catalytic results for the hydrogenation of α-acetamidocinnamic acid to (R)-N-
acetyl-phenylalanine with (R, R)-Rh-MeDuPhos under batch homogeneous
conditions (24 h).
Entry Time, min Conversion (%) Selectivity (%) Ee(%)
TON TOF
−
1)
(h
1
30
28
100
> 99
6.6
13.2
(
R)
2
3
4
90
180
360
58
79
100
100
100
100
85 (R)
80 (R)
75 (R)
13.6 9.1
21.0 7.0
23.5 5.6
Reactions conditions: 0.05 g ACA, 0.005 g of Rh-complex, 40 atm H
methanol, room temperature.
2
; 25 mL
deformation (1588 cm ), amide C]O stretching (1677 cm−1), and
−
1
−1
amide NeH stretching (3057 and 3075 cm ) bonds. The immobiliza-
tion of the complex (CNT@Au-BPPM spectrum) led to both a shift and
diminution of the intensity of these bands. Accordingly, only the bands
assigned to the vibrations of the CeC]O in-plane deformation (474
−
1
−1
−1
cm ), PeC stretching (784 cm ), CeC skeletal (850 cm ), P-N-C
−
1
−1
asymmetric stretching (1010 cm ), CeH skeletal (1258 cm ),
CeNeC deformation (1352 cm ) and asymmetric CeH stretching
2960 cm ) respectively were detected in this spectrum. Further, the
−1
−
1
(
spectrum collected for the spent CNT@Au-BPPM shown no other
changes in the intensity or location of the bands compared to fresh
CNT@Au-BPPM.
Raman spectra were collected for Au-BPPM, CNT@Au-BPPM and
spent CNT@Au-BPPM catalysts are presented in Fig. 3. Considering
both irradiation sources (488 and 633 nm respectively), the spectrum of
Au-BPPM contains lines assigned to vibrations of the CeC alicyclic and
−1
aliphatic chain at 760, 783, 998, 1030, 1104, 1160 and 1230 cm , the
−
1
asymmetric CH
1
2
at 1410 and 1143 cm , and the CeC aromatic ring at
−
1
587 cm . The immobilization of the Au-BPPM caused both a reduc-
Fig. 3. Raman spectra of Au-BPPM, CNT@Au-BPPM and spent CNT@Au-BPPM
catalysts.
tion of the intensity of these lines and a shift. Thus, the lines assigned to
vibrations of the CeC alicyclic and aliphatic chain shifted at 1053 and
−1
−1
1
157 cm , and that of the CeC aromatic ring at 1570 cm . Like for
the ATR-FTIR spectra the spectrum collected for the spent CNT@Au-
BPPM shown no other changes in the intensity or location of the bands
compared to fresh CNT@Au-BPPM. Supplementary, fresh CNT@Au-
BPPM and spent CNT@Au-BPPM shows the typical Raman shifts of the
−1
D, G and G’ bands at 1322, 1570 and 2716 cm
respectively.
Fig. 4 shows the DR-UV–vis spectra of Au-BPPM, CNT@Au-BPPM
and spent CNT@Au-BPPM catalysts. An absorption band at ˜230 nm is
characteristic to Au ions, while the band at about 290 nm is char-
acteristic to symmetry-allowed p
complexes [18].
σ
-dσ* transition in binuclear gold
3
.2. Hydrogenation of α-acetamidocinnamic acid to (R)-N-acetyl-
phenylalanine under batch homogeneous catalytic conditions
Scheme 3 In the first set of experiments, the catalytic performances
of the new synthesized Au-complex catalyst, Au-BPPM, were compared
to those of the (R, R)-MeDuPhos-Rh complex under batch homogeneous
catalytic conditions. Indeed, Rh-chiral diphosphine based catalysts are
the most successful ligands developed for the asymmetric hydrogena-
tion of C]C bonds such as unsaturated prochiral acids [19]. Particu-
larly, the Rh-DuPhos catalysts show excellent enantioselectivies (up to
98% ee) for the asymmetric hydrogenation of dehydroamino acid de-
rivates [19,20]. In this study, to evaluate the catalytic performance of
Fig. 4. DR-UV–vis spectra of Au-BPPM, CNT@Au-BPPM and spent CNT@Au-
BPPM catalysts.
(
7
1164, 1207 and 1273 cm−1), PeC stretching (712, 725, 741, 750 and
−1
85 cm ), P-N-C asymmetric stretching (958, 971, 997, 1025, 1075
−1
−1
and 1103 cm ), P-Ph stretching (1113, 1136 and 1455 cm ), P-Ph
ring in-plane stretching (1113, 1136 and 1455 cm ), amide NeH
−1
Scheme 3. Hydrogenation of α-acetamidocinnamic acid to (R)-N-
acetyl-phenylalanine.
4

A. Negoi, et al.
Molecular Catalysis 474 (2019) 110420
Table 2
Catalytic results for the hydrogenation of α-acetamidocinnamic acid to (R)-N-acetyl-phenylalanine with Au-BPPM catalyst under batch homogeneous conditions
(24 h).
Entry
Solvent
Temperature (°C)
Conversion (%)
Selectivity (%)
Ee (%)
TON
TOF (h−1)
1
2
3
4
MeOH
MeOH
EtOH
TFE
RT
80
RT
RT
10.8
27.1
7.2
100
100
100
100
> 99 (R)
> 99 (R)
> 99 (R)
98 (R)
2.5
8.8
2.3
1.0
0.10
0.37
0.10
0.04
3.2
2
Reactions conditions: 0.08 g ACA, 0.01 g of Au-complex, 40 atm H ; 25 mL solvent, room temperature.
Table 3
Catalytic results for the hydrogenation of α-acetamidocinnamic acid to (R)-N-acetyl-phenylalanine under flow homogeneous catalytic conditions onto the Au-BPPM
catalyst.
Entry
Temperature (°C)
RT
Time h
Conversion (%)
Selectivity (%)
Ee (%)
TON
TOF (h−1)
1
2
3
4
5
6
7
8
9
0.5
1
3
5
8
0.5
1
3
5
8
1.2
4.6
100
100
100
100
100
100
100
100
100
100
> 99 (R)
> 99 (R)
> 99 (R)
> 99 (R)
> 99 (R)
> 99 (R)
> 99 (R)
> 99 (R)
> 99 (R)
> 99 (R)
0.39
1.49
3.48
4.74
5.46
1.56
5.29
11.76
15.01
18.23
0.78
1.49
1.16
0.95
0.68
3.12
5.29
3.92
3.00
2.28
10.7
14.6
16.8
4.8
16.3
36.2
46.2
56.1
80
1
0
2
Reactions conditions: 0.08 g ACA, 0.01 g of Au-complex, 40 atm H ; 25 mL methanol.
Table 4
Catalytic results for the hydrogenation of α-acetamidocinnamic acid to (R)-phenylalanine with immobilized Au-catalysts.
Entry
Catalyst
Temperature (°C)
Time (h)
Conversion (%)
Selectivity (%)
ee (%)
83 (R)
71 (R)
63 (R)
85 (R)
70 (R)
71 (R)
98(R)
1
2
3
4
5
6
7
CNT@Au-BPPMa
80
80
80
RT
RT
RT
RT
2
17
69
72
21
95
96
57
100
100
100
100
100
100
100
4
wt%
CNT@Au-BPPM
wt%
CNT@Au-BPPM
wt%
CNT-Au-BPPM
wt%Au
CNT-Au-BPPM
wt%Au
CNT-Au-BPPM
wt%Au
CNT-Au-BPPM
wt%Au
a
24
48
2
4
a
4
a
4
a
24
48
2
4
a
b
4
4
Reaction conditions: 0.05 g ACA, 0.005 g of Au-complex, 40 atm H
2
, methanol.
a
Batch conditions.
b
Flow conditions.
the Au-BPPM catalyst, α-acetamidocinnamic acid has been selected as a
model substrate.
(Table 2, entries 1–2). The higher TON was nearby 9 at 80 °C after 24 h
that corresponded to TOFs smaller than 0.5. The solvent has also an
influence on this reaction. While for methanol and ethanol the perfor-
mances were quite similar (Table 2, entry 1 vs entry 3), working in TFE
led to a decrease in both the conversion and the enantioselectivity. A
negative influence of TFE in asymmetric reactions has also been pre-
viously reported in the literature [22].
Tables 1 and 2 compile catalytic results using (R, R)-MeDuPhos-Rh
and the new synthesized Au-complex catalyst. The experiments carried
out with Rh confirmed the total stereoselectivity to the (R) isomer, but
only for small reaction times (ie 30 min), that correspond to conver-
sions below 30% (Table 1, entry 1). An increase of the reaction time led
to an increase of the conversion reaching 100% after 6 h (Table 1, entry
4
) but with a progressive decrease of the enantioselectivity. The cal-
3.3. Hydrogenation of α-acetamidocinnamic acid to (R)-N-acetyl-
phenylalanine under catalytic flow homogeneous conditions
culated values of TON and TOF reflect this behavior. Thus, the evolu-
tion of the TOF in time reflects a deactivation of the catalysts [21] that
may correspond to a partial degradation of the complex leaving a less
complexed metal.
Working with Au-complex catalyst in the same catalytic conditions,
the reaction was much slowly compared to Rh catalysts. As expected
the increase of the temperature led to an increase of the conversion.
However, the enantioselectivity is better than for Rh-MeDuPhos, de-
monstrating an increased stability of the Au-complex. It has been pre-
served at almost total even after the increase of the temperature at 80 °C
The hydrogenation of α-acetamidocinnamic acid to (R)-N-acetyl-
phenylalanine under catalytic flow homogeneous conditions has been
carried out in methanol as solvent at room temperature and at 80 °C
taking as reference the results collected from the reactions carried out
in batch. In all experiments, excellent chemo- and enantioselectivies
were obtained. The increase of the reaction time affords an increased
conversion irrespective of the temperature (room temperature, Table 3,
entries 1–5, and 80 °C, Table 3, entries 6–10). Very important, these
5

A. Negoi, et al.
Molecular Catalysis 474 (2019) 110420
increases were produced with no change in the chemo- or enantios-
electivity. However, the values of TOFs also shown a depletion for re-
action times longer than 3 h that may also suggest a kind of deactiva-
tion. Since the enantioselectivity has not changed this decrease could be
simply assigned to a more advanced strong chemisorption of either the
substrate or the product onto the Au-BPPM catalyst. This process in
favored by the increase of the temperature (Table 3, see RT versus
MeDuPhos, the enantioselctivity to the R-isomer was total only at room
temperature, however with a TOF fifty times higher. Transferring the
reaction in the flow reactor, under similar conditions (methanol, tem-
perature) led to a 10 time increase of the TOF of the Au-BPPM catalyst
with no change in the stereoselectivity. The decrease of the TOF in time
for both the reference Rh and the Au catalysts was assigned to their
partial modification in the reaction conditions.
8
0 °C).
The heterogenization of the Au-BPPM catalyst onto the CNT sup-
port, for the same content of Au-complex, led to a very important in-
crease of the conversion with no change in the selectivity. However, the
covalent bonding was more efficient affording a very high increase of
the conversion even at room temperature (95% after 24 h). The an-
choring onto the support increased the dispersion, and in consequence
the efficiency. However, under batch conditions, for both catalysts, the
enantioselectivity suffered an important depletion (70% in R). But,
transferring the covalent bonded CNT-Au-BPPM catalyst in the flow
reactor corresponded to an excellent ee (98%) for an acceptable con-
version. Furthermore, recycling the CNT-Au-BPPM catalysts occurred
with no change in the catalytic performances as also confirmed by the
characterization results.
3
.4. Hydrogenation of α-acetamidocinnamic acid to (R)N-acetyl-
phenylalanine with the immobilized Au-BPPM catalysts
Table 4 shows the catalytic results with the two immobilized Au
catalysts, CNT@Au-BPPM and CNT-Au-BPPM. The conversions were
higher than those measured for Au-BPPM under flow conditions but the
enantioselectivities were inferior (63–83 % ee). The increase of the
reaction time affords an increased conversion with the prize of a de-
crease in the enantioselectivity (Table 4 entries 1–3). This decrease is
the effect of both non-covalent (CNT@Au-BPPM) and covalent (CNT-
Au-BPPM) immobilization of the Au-BPPM complex. Most probably, for
CNT@Au-BPPM this interaction occurs via a π−π stacking interaction
between the aromatic entities of the ligand and CNT. For the second
case, the decrease of the enantioselectivity is the effect of the rigidity
induced by the anchoring (Scheme 2). In both cases, a change in the
bite-angle of complex could be responsible for this decrease. On the
other hand, the high dispersion provided by the covalent immobiliza-
tion of the gold(I)complex onto CNT produces an enhanced availability
to the complex and, thus, an increased conversion compared to the
impregnated gold complex (Table 4 entries 4–6 vs entries 1–3). Re-
cycling the CNT-Au-BPPM catalysts for five times occurred with no
change in the catalytic performances thus confirming the character-
ization results.
References
[1] A. Corma, A. Leyva-Pérez, M.J. Sabater, Chem. Rev. 111 (2011) 1657–1712.
[2] R. Dorel, A.M. Echavarren, Chem. Rev. 115 (2015) 9028–9072.
[3] D. Pflästerer, A.S.K. Hashmi, Chem. Soc. Rev. 45 (2016) 1331–1367.
[4] Y. Li, W. Li, J. Zhang, Chem. Eur. J. 23 (2017) 467–512.
[5] W. Zi, F.D. Toste, Chem. Soc. Rev. 45 (2016) 4567–4589 and the references cited
therein.
[
[
6] Y. Ito, M. Sawamura, T. Hayashi, J. Am. Chem. Soc. 108 (1986) 6405–6406.
7] R.A. Widenhoefer, Chem. Eur. J. 14 (2008) 5382–5391.
[8] S. Sengupta, X. Shi, ChemCatChem 2 (2010) 609–619.
[9] A. Pradal, P.Y. Toullec, V. Michelet, Synthesis (2011) 1501–1514.
10] Y.-M. Wang, A.D. Lackner, F.D. Toste, Acc. Chem. Res. 47 (2014) 889–901.
11] C. Gonzalez-Arellano, A. Corma, M. Iglesias, F. Sanchez, Chem. Commun. (2005)
3451–3453.
[
[
Finally, the gold(I) complex covalently immobilized onto CNT, CNT-
Au-BPPM, was evaluated under catalytic flow conditions (Table 4, entry
7
could be continuously produced with a 57% conversion and an ex-
cellent enantioselectivity of 98%.
[12] A. Comas-Vives, G. Ujaque, J. Am. Chem. Soc. 135 (2013) 1295–1305.
[
[
[
[
13] A. Arnanz, C. Gonzalez-Arellano, A. Juan, G. Villaverde, A. Corma, M. Iglesias,
). Under the experimental conditions (at 25 °C, for 2 h), the product
F. Sanchez, Chem. Commun. (2010) 3001–3003.
14] A. Comas-Vives, C. Gonzalez-Arellano, A. Corma, M. Iglesias, F. Sanchez,
G.J. Ujaque, J. Am. Chem. Soc. 128 (2006) 4756–4765.
15] K. Fourmy, S. Mallet-Ladeira, O. Dechy-Cabaret, M. Gouygou, Organometallics 32
(2013) 1571–1574.
4
. Conclusions
16] C.C. Gheorghiu, B.F. Machado, C. Salinas-Martínez de Lecea, M. Gouygou,
M.C. Román-Martínez, P. Serp, Dalton Trans. 43 (2014) 7455–7463.
17] T.P. Dang, H.B. Kagan, Chem. Commun. (1971) 481.
18] D.M. Roundhill, J.P. Fackler Jr., Optoelectronic Properties of Inorganic Compounds,
Springer Science & Business Media, 2013, p. 197.
[
[
In conclusion, this study demonstrated the capability of the Au-
BPPM catalyst to afford the stereoselective hydrogenation of α-acet-
amidocinnamic acid to the (R)-N-acetyl-phenylalanine enantiomer. The
nature of the solvent affected both the enantioselectivity and TOFs. In
the investigated series (MeOH, EtOH, TFE) methanol appeared to be the
most efficient one (at 80 °C a TOF of 0.37 h
antioselctivity to the R-isomer). Using the reference (R, R)-Rh-
[19] M.J. Burk, J.E. Feaster, W.A. Nugent, R.L. Harlow, J. Am. Chem. Soc. 115 (1993)
1
0125–10138.
[
20] M.J. Burk, M.F. Gross, J.P. Martinez, J. Am. Chem. Soc. 117 (1995) 9375–9376.
[
21] S. Kozuch, J.M.L. Martin, ACS Catal. 2 (2012) 2787–2794.
−1
for a total en-
[22] N. Fleury-Bregeot, V. de la Fuente, S. Castillon, C. Claver, ChemCatChem 2 (2010)
1346–1371.
6