Application of a Supramolecular-Ligand Library for the Automated Search for Catalysts for the Asymmetric Hydrogenation of Industrially Relevant Substrates

Article abstract of DOI:10.1002/chem.200901110

A procedure is described for the automated screening and lead optimization of a supramolecular-ligand library for the rhodium-catalyzed asymmetric hydrogenation of five challenging substrates relevant to industry. Each catalyst is (self-) assembled from two urea-functionalized ligands and a transition-metal center through hydrogen-bonding interactions. The modular ligand structure consists of three distinctive fragments: the urea binding motif, the spacer, and the ligand backbone, which carries the phosphorus donor atom. The building blocks for the ligand synthesis are widely available on a commercial basis, thus ena-bling access to a large number of ligands of high structural diversity. The simple synthetic steps enabled the scale-up of the ligand synthesis to multigram quantities. For the catalyst screening, a library of twelve new chiral ligands was prepared that comprised substantial variation in electronic and steric properties. The automated procedures employed ensured the fast catalyst assembly, screening, and direct acquisition of samples for analysis. It appeared that the most selective catalyst was different for every substrate investigated and that small variations in the building blocks had a major impact on the catalyst performance. For two substrates, a catalyst was found that provided the product with outstanding enantioselectivity. The subsequent automated optimization of these two leads showed that an increase of catalyst loading, dihydrogen pressure, and temperature had a positive effect on the catalyst activity without affecting the catalyst selectivity.

Full text of DOI:10.1002/chem.200901110

DOI: 10.1002/chem.200901110
Application of a Supramolecular-Ligand Library for the Automated Search
for Catalysts for the Asymmetric Hydrogenation of Industrially Relevant
Substrates
Jurjen Meeuwissen,[a] Mark Kuil,[a, b] Alida M. van der Burg,[a] Albertus J. Sandee,[a, b] and
Joost N. H. Reek*[a]
Abstract: A procedure is described for
the automated screening and lead opti-
mization of a supramolecular-ligand li-
brary for the rhodium-catalyzed asym-
metric hydrogenation of five challeng-
ing substrates relevant to industry.
Each catalyst is (self-) assembled from
two urea-functionalized ligands and a
transition-metal center through hydro-
gen-bonding interactions. The modular
ligand structure consists of three dis-
tinctive fragments: the urea binding
motif, the spacer, and the ligand back-
bone, which carries the phosphorus
donor atom. The building blocks for
the ligand synthesis are widely avail-
able on a commercial basis, thus ena-
bling access to a large number of li-
gands of high structural diversity. The
simple synthetic steps enabled the
scale-up of the ligand synthesis to mul-
tigram quantities. For the catalyst
acquisition of samples for analysis. It
appeared that the most selective cata-
lyst was different for every substrate
investigated and that small variations
in the building blocks had a major
impact on the catalyst performance.
screening,
a library of twelve new
chiral ligands was prepared that com-
prised substantial variation in electron-
ic and steric properties. The automated
procedures employed ensured the fast
catalyst assembly, screening, and direct
For two substrates, a catalyst was
found that provided the product with
outstanding enantioselectivity. The sub-
sequent automated optimization of
these two leads showed that an in-
crease of catalyst loading, dihydrogen
pressure, and temperature had a posi-
tive effect on the catalyst activity with-
out affecting the catalyst selectivity.
Keywords: asymmetric catalysis
·
high-throughput screening · hydro-
genation · rhodium · supramolec-
ular chemistry
Introduction
geneous enantioselective catalysis. The current mechanistic
knowledge, although far advanced, does not provide suffi-
cient tools to design a catalyst ab initio that will display the
required (enantio-) selectivity for a certain envisioned appli-
cation because small differences in the transition-state ener-
gies of the competing pathways (1–2 kcalmolÀ1 similar to
In the search for suitable catalysts for the asymmetric syn-
thesis of fine chemicals in industry, time constraints and cost
efficiency are the decisive considerations.[1] Despite consid-
erable efforts in academia and industry, these demands still
represent a large barrier for the wider application of homo-
À
the energy required for C C bond rotation in ethane) al-
ready leads to huge selectivity changes.[2] Therefore, subtle
variation in ligand structure can cause significant changes in
enantioselectivity. The most successful strategy, consequent-
ly, comprises the design of modular ligands that are syntheti-
cally easily prepared because it enables the preparation of
many analogues of a promising ligand type.[3] Besides smart
modular design, empirical approaches are often guided by
knowledge-based intuition or serendipity. The use of high-
throughput experimentation (HTE) and ligand libraries for
asymmetric transformations tremendously increases the
pace of ligand screening and is therefore highly desirable.[4]
Considering the large potential of HTE, its impact in homo-
[a] J. Meeuwissen, Dr. M. Kuil, A. M. van der Burg, Dr. A. J. Sandee,
Prof. Dr. J. N. H. Reek
Homogeneous and Supramolecular Catalysis
Vanꢀt Hoff Institute for Molecular Sciences Institution
Nieuwe Achtergracht 166, 1018 WV Amsterdam (The Netherlands)
Fax : (+31)20-5255604
E-mail: j.n.h.reek@uva.nl
[b] Dr. M. Kuil, Dr. A. J. Sandee
BASF Nederland B. V. Catalysts
Strijkviertel 67, 3454 PK De Meern (The Netherlands)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200901110.
10272
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10272 – 10279

FULL PAPER
geneous catalysis has been surprisingly modest until recently
mainly because the synthetic procedures for chiral ligands
are often quite tedious and there is only limited technology
available to prepare large catalyst libraries of sufficient size
and diversity.[5] This factor is clearly reflected in the scarce-
ness of reports on automated preparations of phosphorus li-
gands.
The revival of easy-to-synthesize monodentate ligands has
provided new opportunities for the application of HTE tech-
nology in this field.[6,7] De Vries et al. have shown very ele-
gantly that monodentate phosphoramidites ligands can be
prepared in a parallel fashion and subsequently screened in
catalysis without difficult purification steps in between.[8]
These so called “instant libraries” provide a very efficient
tool to prepare and screen a series of 96 catalysts in only
two days.[9] This protocol has already resulted in the discov-
ery of a new catalyst that currently is applied on a ton-scale
commercial process.[10]
blocks and to expand the substrate scope in rhodium-cata-
lyzed asymmetric hydrogenation.
Herein, we report the development of a library of twelve
new structurally diverse UREAphos ligands that were ap-
plied as ligands in the rhodium-catalyzed asymmetric hydro-
genation of five industrially relevant substrates. Because the
easy catalyst-assembly process is potentially a valuable tool
for the fast lead discovery in industry, we intended to per-
form the catalyst screening by means of HTE in an automat-
ed environment. Interesting leads from the screening phase
were subsequently investigated for optimal reaction condi-
tions in an automated environment. The strategy was ac-
cordingly set up in three consecutive stages: 1) preparation
of the ligand library, 2) automated catalyst assembly and
high-throughput screening, and 3) automated lead optimiza-
tion.
Bidentate ligands represent a different important class of
ligand, but automated procedures to make large libraries of
this type of ligand are unknown to date.[11] The recent break-
through of supramolecular bidentate ligands simplifies the
preparation of bidentate ligands through the self-assembly
of two monodentate ligand building blocks.[12] The bidentate
ligands are formed by simply mixing the proper monoden-
tate ligand building blocks with complementary binding
motifs, such as hydrogen-bonding,[13] ionic,[14] or metal–
ligand interactions.[15] It has been shown that this is a very
powerful tool to construct ligand libraries because of the
modular nature of the assembling components. We recently
reported UREAphos ligands as a new class of supramolec-
ular bidentate ligand that form through urea hydrogen
Results and Discussion
Choice of substrates: The investigated substrates are precur-
sors for the synthesis of high-value compounds of interest
for industry. The scope of the substrates is varied notably in
Scheme 2. Substrates A–E.
the structure and substitution pattern of the prochiral olefin
(Scheme 2). The least-substituted substrate methyl 2-(hy-
droxymethyl)acrylate (A; its product is also known as the
Roche ester)[17] is used as a building block in the synthesis
of the antitumor agents tedanolide and discodermolide.[18]
The trisubstituted (unnatural) b-amino acid precursors (Z)-
methyl 3-acetamidobut-2-enoate (B)[19] and (E)-methyl 2-
(acetamidomethyl)-3-phenylacrylate (C)[7j] are interesting
building blocks for the synthesis of biologically active b-pep-
tide chains, which are stable to proteolytic degradation.[20]
Currently no efficient catalyst is reported to obtain C with
high enantioselectivity. The asymmetric hydrogenation of
the trisubstituted cyclic enamide N-(3,4-dihydro-2-naphtha-
lenyl)acetamide (D)[15b,21] is interesting for obtaining chiral
amines to synthesize many biologically active compounds,
such as the melanin-concentrating hormone antagonist as a
potential therapeutic agent for obesity.[22] Tetrasubstituted
substrates, such as the cyclic enamide N-(1-benzyl-3,4-dihy-
dronaphthalen-2-yl)acetamide (E) are notoriously hard to
hydrogenate with high conversion and enantioselectivity.[23]
This compound is used as a building block for the synthesis
of b-aminotetralin compounds, which potential TRPV1 an-
Scheme 1. Self-assembly of the supramolecular catalyst through the hy-
drogen-bonding interactions of urea. M=metal center.
bonding (Scheme 1).[16] These urea-functionalized ligands
are made by connecting three building blocks that are ame-
nable for ligand variation, that is, ligand backbone, spacer,
and urea motif, thus favoring the development of large and
diverse ligand libraries. Two of these ligands form a supra-
molecular bidentate in the presence of a transition-metal
center through self-assembly. An introductory series of six
structurally related ligands was explored for rhodium-cata-
lyzed asymmetric hydrogenation and showed high selectivi-
ties. These results encouraged us to make use of the poten-
tial to enlarge and diversify the number of ligand building
Chem. Eur. J. 2009, 15, 10272 – 10279
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10273

J. N. H. Reek et al.
tagonists for the treatment of inflammatory pain and other
painful conditions in mammals.[24] There is also no efficient
catalyst reported to obtain the hydrogenation product of E
with high enantioselectivity.
Preparation of the ligand library: Three types of urea-func-
tionalized phosphorus ligands, with varying electronic prop-
Scheme 4. UREAphos ligands L1–L12.
phoramidite L5 all contain a flexible pro-atropoisomeric
(tropos) backbone, which is amenable for the transfer of
chirality from the stereogenic groups in the spacer.[25] Phos-
phite L3 has a relatively bulky stereogenic isopropyl group
at the R2 position of the spacer, whereas phosphite L4 con-
tains a less sterically demanding stereogenic methyl group at
the same position. The spacer of L5 is based on (1S,2S)-cy-
clohexane-1,2-diamine. Phosphites L6–L10 all have the
same (R)-BINOL-derived (BINOL=1,1’-bi-2-naphthol)
backbone and vary only in the spacer and urea motif.
Ligand L6 contains a benzyl urea motif and no substituents
at all in the spacer. Ligands L7–L9 contain only an addition-
al stereogenic methyl group at the R2 position of the spacer
and vary by the benzyl, phenyl, and n-butyl urea motifs, re-
spectively. Ligand L10 carries a stereogenic phenyl group at
the R1 position close to the phosphorus donor atom and a
stereogenic methyl group at the R2 position in combination
with a phenyl urea motif. The phosphoramidites L11 and
L12 are diastereomers, which consist of a cyclohexane-1,2-
diamine spacer (1S,2S and 1R,2R isomers for L11 and L12,
Scheme 3. Synthesis of different types of UREAphos ligands. hex=
hexane, tol=toluene.
erties, were prepared: phosphites, phosphoramidites, and
phosphanes (Scheme 3). The steric properties were varied
by changing the substitutional groups at defined positions in
the backbone, spacer, and urea motif. The synthesis of the
phosphite ligands starts with the coupling of an isocyanate
with an amino alcohol followed by a condensation reaction
with a phosphorochloridite. The synthesis of phosphorami-
dite ligands is virtually analogous to the synthesis of phos-
phites. In the first step, the urea amine is formed by cou-
pling an isocyanate with a twofold excess of the diamine fol-
lowed by a condensation reaction with a phosphorochlori-
dite. The synthesis of phosphane ligands is accomplished by
a single reaction of an amino phosphane with an isocyanate.
According to these synthetic procedures, a library of twelve
new UREAphos ligands was prepared manually by using
almost exclusively commercially available building blocks
respectively) coupled to
BINOL-derived backbone.
a benzylisocyanate and a R-
A multigram-scale (83 g) batch of phosphite L8 was pre-
pared to demonstrate the possibility of scaling up the ligand
synthesis. Instead of an excess of triethylamine as the base
to trap HCl as a salt, a solid Amberlyst base was used,
which is easily filtered off after completion of the reac-
tion.[26] Remaining HCl salts present in the product would
have a deleterious effect on the activity in the rhodium-cata-
lyzed hydrogenation, and therefore need to be removed.[8]
For automated synthesis and multigram-scale synthesis, the
solid-base method is suitable because of the simple separa-
1
(Scheme 4). All the ligands were fully characterized by H,
13C, and 31P NMR spectroscopic and high-resolution mass-
spectrometric analysis.
The two phosphanes L1 and L2 were synthesized from 2-
(diphenylphosphano)ethanamine coupled to a chiral isocya-
nate and are the only ligands with stereogenic elements ex-
clusively at the urea motif. Phosphites L3 and L4 and phos-
10274
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10272 – 10279

Development of a Ligand Library for Catalytic Automated High-Throughput Experiments
FULL PAPER
tion step involved. This approach ensures that all the em-
ployed synthetic (i.e., coupling of amine and isocyanate
groups and condensation of the PCl and alcohol moieties)
and purification steps (i.e., filtration and recrystallization)
are viable reaction paths in an industrial environment.
Figure 2. 31P{1H} NMR spectrum of [Rh
ACTHNUGTRNE(NUG nbd)(L2)2]BF4.
Complex and dilution studies: Complex and dilution studies
were performed with phosphane ligand L2 as an exemplary
ligand of the library to investigate the coordination behavior
and the nature of the hydrogen bonding in the supramolec-
ular assembly. The coordination study was carried out with
(Figure 2). The urea proton signal of the ligand shows a
strong dependency on the concentration because of the for-
mation of intermolecular hydrogen bonds.[28] The same urea
proton signal in the complex [RhACHTUNTGRNEUNG(nbd)(L2)2]BF4 was shifted
[Rh
(nbd)2]BF4 (nbd=2,5-norbornadiene) as the metal pre-
downfield with respect to L2 by more than Dd=1 ppm and
showed no dependency on concentration because of intra-
molecular hydrogen bonding. These results support the for-
mation of the supramolecular bidentate species through
urea hydrogen bonding. Furthermore, these results are in ac-
cordance with a complex and dilution study performed with
a structurally similar phosphite ligand to L8, which provided
comparable results.[16b]
Automated catalyst assembly and high-throughput screen-
ing: An automated sequence was programmed in an Accel-
erator SLT workstation of Chemspeed Technologies to pre-
pare the catalyst reaction mixtures and perform the subse-
quent catalyst screening (60 experiments in total).[29] Sam-
ples of the product mixtures for GC analysis were prepared
automatically directly after completion of the sequence. The
results of the screening experiments are presented in
Table 1.
In general, these results show that the conversion and en-
antiomeric excess (ee) are strongly influenced by the ligand
type and changes in the ligand structure, which is also evi-
denced by the fact that for every substrate a different cata-
lyst performed best in terms of selectivity. Interestingly, for
defined cases a correlation was found between the catalyst
performance and variations in the ligand structure. These
variations expressed by the ligand backbone, spacer (R1 and
R2 position), and urea moiety (R3 position) allowed us to
Figure 1. Dilution study of ligand L2 and [RhACHTNUTGRNEUNG(nbd)(L2)2]BF4.
(JRhÀP =152 Hz) that is typical for
a
cis geometry
1
(Figure 1).[27] The H NMR spectrum indicated that only one
nbd unit remained coordinated. The general structure of
[Rh
for C61H62O2N4P2Rh: 1047.3403; found: 1047.3395).
dilution study of ligand L2 and complex [Rh-
ACHTUNGTRENNUNG(nbd)(L2)2]BF4 was performed and the signal of one of the
ACHTUNGTRENNUNG(nbd)(L2)2]BF4 was confirmed by using HRMS (calcd
A
À
urea protons (CH2 NH; the position of which was deter-
mined by COSY 1H NMR) was monitored by 1H NMR
spectroscopic analysis in the concentration range 5–40 mm
Table 1. Results of the high-throughput screening of substrates A–E with ligands L1–L12.[a]
Substrate
Ligand
A
B
C
D
E
Conversion
ee[b]
Conversion
ee[b]
Conversion
ee[b]
Conversion
ee[b]
Conversion
ee[b,c]
[%]
[%]
[%]
[%]
[%]
[%]
[%]
[%]
[%]
[%]
L1
L2
L3
L4
L5
L6
L7
L8
49
14
100
100
45
80
100
88
100
100
66
21 (+)
–
4.5 (+)
30 (+)
13 (+)
–
13
–
36
100
–
6.8
21
8.7
29
33
–
–
–
–
–
–
–
–
–
–
–
70
36
34
18
–
18
36
20
2
53 (+)
14 (+)
23 (À)
53 (À)
–
19 (+)
24 (+)
24 (+)
26 (+)
66 (+)
15 (+)
32 (À)
6.4
25
3.0
2.0
–
8.6
4.6
3.0
3.8
5.2
5.4
–
26 (+)
34 (+)
19 (À)
44 (À)
–
20 (+)
71 (+)
71 (+)
72 (+)
69 (+)
13 (+)
–
64 (+)
–
–
27 (À)
15 (À)
45 (À)
31 (À)
38 (À)
–
–
–
34
16
23
18
33
1.6
–
47 (+)
95 (+)
96 (+)
78 (+)
94 (+)
62 (+)
–
–
14 (+)
–
33 (+)
54 (À)
17 (À)
L9
L10
L11
L12
14
3.2
29
54
6.5
76 (À)
[a] Reaction conditions: [Rh]=1 mm [RhACTHNUGRTNE(UNG nbd)2]BF4, [L]=2.2ꢂ[Rh], [S]=50ꢂ[Rh], reaction time=12 h, temperature=RT, pressure=10 bar H2, sol-
vent=CH2Cl2. [b] The signs are reported to emphasize changes in the absolute chirality of the product as measured by GC and HPLC; however, the
signs do not reflect a measured optical rotation. [c] Only the cis-hydrogenated product was observed.
Chem. Eur. J. 2009, 15, 10272 – 10279
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10275

J. N. H. Reek et al.
emphasize relevant structural elements to obtain high selec-
tivity for a specific substrate. Phosphoramidite L11 serves as
an illustrative example of applied changes in the ligand
structure as an ee value of 54% was obtained for substrate
A, whereas the diasteromeric phosphoramidite L12 led to a
low ee value of 17%. Other ligands performed rather poorly
for this substrate and only low-to-modest ee values were ob-
tained (<33%). In the case of substrate B, again the diaste-
reomeric phosphoramidites L11 and L12 differed distinctive-
ly. The highest ee value was accomplished with L12 (76%),
whereas with L11 the complex was completely inactive. In-
terestingly, also the tropos-based phosphite L3 led to an ee
value of 64%, whereas with the analogous tropos-based
phosphite L4 a racemic product was obtained. The other li-
gands of the library performed poorly with ee values lower
tion of the spacer of this phosphite discriminates this ligand
from the other BINOL-derived phosphites. Under the con-
ditions applied, the conversions remained low (<10%) for
the hydrogenation of tetrasubstituted substrate E. The
BINOL-derived phosphites L6–L10 were the most selective
ligands with a ee value of 72% for L9, which is among the
highest reported values for this substrate.[23,31] Only L6 per-
formed relatively poorly (20% ee), which indicates that sub-
stitution with a stereogenic methyl group at the R2 position
of the spacer is an important structural variation to obtain
high enantioselectivity for this substrate.
Automated lead optimization: The high-throughput screen-
ing provided interesting leads in terms of ee values for sub-
strates C and E (96% with L8 and 72% with L9, respective-
ly); however, the conversion of
both substrates (23 and 3.0%
for C and E, respectively) was
rather low under the conditions
applied. To investigate if the ac-
tivity of the catalysts could be
improved, the reaction condi-
tions were optimized using an
automated AMTEC SPR16
robot, which is suited to the ap-
plication of different reaction
conditions in each individual re-
actor.[32] The reaction condi-
tions varied were temperature,
dihydrogen pressure, and, for
E, the rhodium precursor con-
centration.
The obtained results for the
lead optimization of C are pre-
sented in Table 2. The optimiza-
tion experiments showed that
within the span of the investiga-
tion an increase in temperature
Scheme 5. Hydrogenation of E with phosphites L6–L10.
and pressure caused an en-
hancement in the catalyst activ-
than 45%. In the case of C, the phosphites L6–L10 were the
only ligand type that provided active catalysts (Scheme 5).
The ee values obtained with L6 and L9 (47 and 78%, re-
spectively) were only moderate but very high ee values were
obtained with L7, L8, and L10 (95, 96, and 94%, respective-
ly), which are currently the highest reported ee values for
this substrate.[7j,30] These ligands have the BINOL-derived
backbone, the stereogenic methyl group at the R2 position
of the spacer, and a phenyl or benzyl urea motif in common.
Ligands that lack these substitution patterns, such as L6 and
L9, were less successful. In the case of substrate D, all the
ligand types provided active catalysts. Remarkably, phos-
phane L1 which only has a stereogenic element at the urea
moiety reached an ee value of 53%, thus also showing that
variation in the urea motif can be effective in specific cases.
The best result was obtained with BINOL-derived phosphite
L10 (66% ee, 14% conversion). Substitution at the R1 posi-
ity and, therefore, higher conversions. Fortunately, in all
cases this enhancement was accompanied by the retention
of the catalyst selectivity, which becomes evident by compar-
ing the ee value of Entries 1 and 6 (95% ee at 10 bar, 258C
versus 96% ee at 30 bar, 458C, respectively; Table 2). Fur-
Table 2. Lead optimization of substrate C.[a]
Entry
PH [bar]
T [8C]
Conversion [%]
ee [%]
2
1
2
3
4
5
6
10
10
10
30
30
30
25
35
45
25
35
45
29
45
61
45
64
84
95
97
97
93
95
96
[a] Reaction conditions: [Rh]=3.75 mm [Rh
ACHTUGNRTEN(NUGN nbd)2]BF4, [L8]=2.2ꢂ[Rh],
[S]=100ꢂ [Rh], reaction time=4–5 h, solvent=CH2Cl2.
10276
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10272 – 10279

Development of a Ligand Library for Catalytic Automated High-Throughput Experiments
FULL PAPER
thermore, the conversion for these entries increased from 29
to 84%, respectively.
The lead optimization for E was carried out using L8
(Table 3). Interestingly, the optimization experiments result-
ed in even higher ee values (82–87%) in the lead-optimiza-
lectivities were obtained for two inherently difficult sub-
strates: 96 and 87% for substrates C and E, respectively.
These results support the significance of large and diverse
catalyst libraries for lead discovery. Optimization of the re-
action conditions for two successful catalysts showed that
catalyst activity is easily improved by varying temperature,
pressure, and catalyst loading. These results demonstrate the
powerful combination of the application of supramolecular
ligands, based on easily accessible ligand building blocks,
and automated high-throughput screening of their catalysts,
thus leading to new catalyst solutions for industrially rele-
vant substrates.
Table 3. Lead optimization of substrate E.[a]
Entry
PH [bar]
T [8C]
[Rh] [mm]
Conversion [%]
ee [%]
2
1
2
3
4
5
6
7
8
10
10
10
10
40
40
40
40
25
40
25
40
25
40
25
40
2.25
2.25
3.75
3.75
2.25
2.25
3.75
3.75
5.6
26
15
42
28
62
45
84
87
85
86
86
83
82
83
83
Experimental Section
[a] Reaction conditions: [L8]=2.2ꢂ[Rh], [S]=50ꢂ [Rh], reaction time=
28 h, solvent=CH2Cl2.
General procedures and materials: Unless stated otherwise, the reactions
were carried out in an inert atmosphere of nitrogen or argon with stan-
dard Schlenk techniques. THF was distilled from sodium benzophenone
ketyl, dichloromethane from CaH2, and toluene from sodium under ni-
tion experiments, which is explained by the use of more ac-
curate procedures because the reactions carried out by the
AMTEC robot were performed on a larger scale. Further-
more, the same retention in catalyst selectivity on changing
the reaction conditions was observed for substrate E. The ee
values remained almost constant (compare Entries 1 and 8
in Table 3: 87% ee at 10 bar, 258C, 2.25 mm versus 83% ee
at 40 bar, 408C, 3.75 mm, respectively). Because this tetra-
substituted cyclic substrate has a very hindered double
bond, it is hydrogenated at a relatively slow rate. However,
the conversion could be optimized from 5.6 to 84% and the
increase in temperature, pressure, and the additional catalyst
loading all had a positive effect on the catalyst activity.
trogen. NMR spectra were measured on
a
Varian Mercury (1H:
300 MHz; 31P{1H}: 121.5 MHz; 13C{1H}: 75.5 MHz) or a Varian Inova
spectrometer (1H: 500 MHz; 31P{1H}: 202.3 MHz; 13C{1H}: 125.7 MHz) at
room temperature unless stated otherwise. Chemical shifts are reported
in ppm and are given relative to trimethylsilane (TMS; 1H and 13C{1H})
or H3PO4 (31P{1H}) as an external standard. High-resolution mass spectra
were recorded at the Department of Mass Spectrometry at the University
of Amsterdam using FAB+ ionization on a JEOL JMS SX/SX102A four-
sector mass spectrometer with 3-nitrobenzyl alcohol as the matrix. With
exception of the compounds given below, all the reagents were purchased
from commercial suppliers and used without further purification. Trie-
thylamine was distilled from CaH2 under nitrogen. The following ligand
precursor compounds were synthesized according to reported proce-
dures: phosphorochloridite of (R)-2,2’-bisnaphthol,[33] 3,3’,5,5’-tetra-tert-
butyl-2,2’-bisphenol,[34] and phosphorochloridite of 3,3’,5,5’-tetra-tert-
butyl-2,2’-dihydroxybisphenol.[35] Substrates A,[36] B,[19b] C,[7j] D,[21c] and
E[23] were also synthesized according to reported procedures.
General procedure for the preparation of urea alcohols: A solution of
isocyanate (5 mmol, 1 equiv) in dichloromethane (10 mL) was added
dropwise to a vigorously stirred solution of the amino alcohol (5 mmol,
1 equiv) in dichloromethane (20 mL) kept at 08C. The reaction mixture
was allowed to warm to room temperature and was left stirring for 1 h.
The formed precipitate was filtered off and washed twice with hexanes.
The remaining solvents were evaporated in vacuo to obtain the product
almost quantitatively.
Conclusions
Herein, the preparation of a library of twelve new chiral
urea-functionalized ligands that form supramolecular biden-
tate ligands in the presence of a rhodium catalyst precursor
was presented. Although the library is small at this stage,
the modular buildup of these ligands from very diverse,
commercially available, building blocks enables access to
large ligand libraries or rapid optimization of a lead found
during screening experiments. In addition, relatively easy
synthetic steps have been employed, thus allowing the facile
scale-up of the ligands to multigram-scale synthesis, which is
demonstrated by the preparation of 83 g of one of the li-
gands in about one week.
To demonstrate the ability of these ligands to induce se-
lectivity, they were applied in the asymmetric rhodium-cata-
lyzed hydrogenation of five industrially relevant substrates
by using automated procedures for high-throughput screen-
ing and lead optimization. The screening experiments dem-
onstrated that variation of the electronic and structural
properties of the ligands through small changes applied to
the urea motif, spacer, and ligand backbone had strong ef-
fects on catalyst selectivity. Unrivaled excellent enantiose-
General procedure for the preparation of urea amines: A solution of iso-
cyanate (5 mmol, 1 equiv) in toluene/hexanes (10 mL, 1:1) was added
dropwise to a vigorously stirred solution of diamine (10 mmol, 2 equiv)
in toluene/hexanes (20 mL, 1:1) kept at 08C. The reaction mixture was al-
lowed to warm to room temperature and was left stirring for 1 h. The
formed precipitate was filtered off and washed twice with hexanes. The
remaining solvents were evaporated in vacuo to obtain the product. The
starting material diamine was recovered from the filtrate.
General procedure for the preparation of urea phosphites and urea phos-
phoramidites: A solution of phosphorochloridite (1 mmol, 1 equiv) dis-
solved in THF (10 mL) was added dropwise to a vigorously stirred solu-
tion of the urea alcohol or urea amine (1 mmol, 1 equiv; dried by the co-
evaporation with toluene (3ꢂ)) and an excess of triethylamine in THF
(20 mL) kept at 08C. The turbid reaction mixture was subsequently
warmed to room temperature and left stirring for 12 h. After filtration
over a small layer of neutral alumina oxide and evaporation of solvents,
the product was obtained almost quantitatively. If necessary, the product
was purified by column chromatography with dichloromethane/ethyl ace-
tate (4:1) as the eluent. After evaporation of the solvents in vacuo, the
last traces of ethyl acetate were removed by coevaporation with toluene.
Chem. Eur. J. 2009, 15, 10272 – 10279
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10277

J. N. H. Reek et al.
General procedure for the preparation of urea phosphanes: A solution of
isocyanate (5 mmol, 1 equiv) in dichloromethane was added dropwise to
a vigorously stirred solution of amino phosphane (5 mmol, 1 equiv) in di-
chloromethane (20 mL) kept at 08C. The reaction mixture was allowed to
warm to room temperature and was left stirring for 2 h. After evapora-
tion of the solvents in vacuo, the product was obtained almost quantita-
tively.
[1] H.-U. Blaser, Chem. Commun. 2003, 293–296.
[2] W. Knowles, Acc. Chem. Res. 1983, 16, 106–112.
[3] P. E. Goudriaan, P. W. N. M. van Leeuwen, M.-N. Birkholz, J. N. H.
Reek, Eur. J. Inorg. Chem. 2008, 2939–2958.
[4] a) C. Jꢃkel, R. Paciello, Chem. Rev. 2006, 106, 2912–2942; b) K.
Ding, Chem. Commun. 2008, 909–921; c) K. Ding, H. Du, Y. Yuan,
J. Long, Chem. Eur. J. 2004, 10, 2872–2884; d) M. Reetz, Angew.
Chem. 2002, 114, 1391–1394; Angew. Chem. Int. Ed. 2002, 41, 1335–
1338; e) C. Gennari, U. Piarulli, Chem. Rev. 2003, 103, 3071–3100.
[5] J. G. de Vries, A. H. M. de Vries, Eur. J. Org. Chem. 2002, 799–811.
[6] a) M. T. Reetz, G. Mehler, Angew. Chem. 2000, 112, 4047–4049;
Angew. Chem. Int. Ed. 2000, 39, 3889–3890; b) M. van den Berg,
A. J. Minnaard, E. P. Schudde, J. van Esch, A. H. M. de Vries, J. G.
de Vries, B. L. Feringa, J. Am. Chem. Soc. 2000, 122, 11539–11540;
c) C. Claver, E. Fernandez, A. Gillon, K. Heslop, D. J. Hyett, A.
Martorell, A. G. Orpen, P. G. Pringle, Chem. Commun. 2000, 961–
962.
[7] a) M. T. Reetz, T. Sell. A. Meiswinkel, G. Mehler, Angew. Chem.
2003, 115, 814–817; Angew. Chem. Int. Ed. 2003, 42, 790–793;
b) M. T. Reetz, G. Mehler, Tetrahedron Lett. 2003, 44, 4593–4596;
c) M. T. Reetz, T. Sell, G. Mehler, A. Meiswinkel, Tetrahedron:
Asymmetry 2004, 15, 2165–2167; d) M. T. Reetz, X. Li, Tetrahedron
2004, 60, 9709–9714; e) M. T. Reetz, X. Li, Angew. Chem. 2005,
117, 3019–3021; Angew. Chem. Int. Ed. 2005, 44, 2959–2962; f) C.
Monti, C. Gennari, U. Piarulli, Tetrahedron Lett. 2004, 45, 6859–
6862; g) C. Monti, C. Gennari, U. Piarulli, J. G. de Vries, A. H. M.
de Vries, A. H. M. L. Lefort, Chem. Eur. J. 2005, 11, 6701–6717;
h) D. PeÇa, A. J. Minnaard, J. A. F. Boogers, A. H. M. de Vries, J. G.
de Vries, B. L. Feringa, Org. Biomol. Chem. 2003, 1, 1087–1089;
i) R. Hoen, J. A. F. Boogers, H. Bernsmann, A. J. Minnaard, A.
Meetsma, T. D. Tiemersma-Wegman, A. H. M. de Vries, J. G. de V-
ries, B. L. Feringa, Angew. Chem. 2005, 117, 4281–4284; Angew.
Chem. Int. Ed. 2005, 44, 4209–4212; j) R. Hoen, T. Tiemersma-
Wegman, B. Procuranti, L. Lefort, J. G. de Vries, A. J. Minnaard,
B. L. Feringa, Org. Biomol. Chem. 2007, 5, 267–275.
Multigram synthesis of L8: Amberlyst A-21 (430 g) and urea alcohol U4
(0.505 mol, 97.86 g, 1 equiv) were dried by coevaporation with toluene
(3ꢂ) in a 3-L flask, and THF (500 mL) was added as a solvent. The reac-
tion mixture was cooled in an ice bath and a solution of phosphorochlori-
dite of (R)-2,2’-bisnaphthol (0.505 mol, 1 equiv) in toluene was added
slowly. The reaction mixture was stirred overnight at 458C. The reaction
mixture was filtered and the Amberlyst A-21 washed with THF. The sol-
vents were evaporated in vacuo to give the crude product. The product
was purified by recrystallization from dichloromethane/petroleum ether
to give a white powder (33%, 0.16 mol, 83 g).
General procedure for complex and dilution studies: The complex study
was carried out on a sample (0.5 mL) in CDCl3 (20 mm) by mixing one
equivalent of the catalyst precursor [RhACTHNUTRGNEU(NG nbd)2]BF4 (nbd=2,5-norborna-
diene) with two equivalents of L2. The 1H NMR dilution experiments
were carried out by preparing a sample (0.5 mL) at a known concentra-
tion: 5, 10, 15, 20, 30, and 40 mm in CDCl3 for L2 and for [Rh-
ACHTUNGTRENNUNG(nbd)(L2)2]BF4. The position of the solvent signal was used as a refer-
ence for the urea NH signal.
General procedure for catalyst-screening experiments: The high-through-
put screening experiments were carried out in a Chemspeed Accelerator
SLT workstation. All the experiments were conducted in an inert atmos-
phere. For a typical screening experiment (hydrogenation), stock solu-
tions of the ligands (3.85 mm), catalyst precursor [RhACHTUNTGRENNUG(nbd)2]ACHTUNTGERN(NUNG BF4)
(3.5 mm), and substrate (175 mm) in dry dichloromethane were prepared.
The ligands (2ꢂ1.0 mL) and catalyst precursor (1.0 mL) were mixed in
the reactor. The substrate (1.0 mL) was added after vortex mixing for
5 min. This procedure resulted in catalyst and substrate concentrations of
1.0 and 50 mm, respectively, in the reactor. The hydrogenation was car-
ried out at dihydrogen pressure of 10 bar at room temperature for 12 h.
The hydrogenation products of substrates A–D were analyzed with an In-
terscience Trace GC Ultra (FID detector) equipped with a CP Chiralsil
DexCB column. The hydrogenation products of substrate E were ana-
lyzed on a Shimadzu 10 A HPLC chromatograph equipped with a UV
detector (column: Chiralpak AD; eluent: n-hexane/isopropyl alcohol
(90:10); flow rate: 0.6 mLminÀ1).
[8] L. Lefort, J. A. F. Boogers, A. H. M. de Vries, J. G. J. G. de Vries,
Org. Lett. 2004, 6, 1733–1735.
[9] J. G. de Vries, L. Lefort, Chem. Eur. J. 2006, 12, 4722–4734.
[10] L. Lefort, J. A. F. Boogers, A. H. M. de Vries, J. G. de Vries, Top.
Catal. 2006, 40, 185–191.
[11] W. Tang, X. Zhang, Chem. Rev. 2003, 103, 3029–3069.
[12] For reviews on supramolecular bidentate ligands, see: a) M. J. Wil-
kinson, P. W. N. M. van Leeuwen, J. N. H. Reek, Org. Biomol. Chem.
2005, 3, 2371–2383; b) B. Breit, Angew. Chem. 2005, 117, 6976–
6986; Angew. Chem. Int. Ed. 2005, 44, 6816–6825; c) A. J. Sandee,
J. N. H. Reek, Dalton Trans. 2006, 3385–3391.
General procedure for lead-optimization experiments: The optimization
experiments were carried out in the AMTEC SPR16 slurry-phase reactor
and consisted of 16 parallel reactors equipped with temperature and pres-
sure sensors and a mass-flow controller. The apparatus is suited for moni-
toring gas-uptake profiles during the catalytic reactions for each reactor
simultaneously. Four autoclaves were heated to 1108C and flushed with
argon (22 bar) five times. Next, the reactors were cooled to 258C and
flushed again with argon (22 bar) five times. The autoclaves were charged
[13] a) F. W. Patureau, M. Kuil, A. J. Sandee, J. N. H. Reek, Angew.
Chem. 2008, 120, 3224–3227; Angew. Chem. Int. Ed. 2008, 47,
3180–3183; b) B. Breit, W. Seiche, J. Am. Chem. Soc. 2003, 125,
6608–6609; c) B. Breit, W. Seiche, Angew. Chem. 2005, 117, 1666–
1669; Angew. Chem. Int. Ed. 2005, 44, 1640–1643; d) M. Weis, C.
Waloch, W. Seiche, B. Breit, J. Am. Chem. Soc. 2006, 128, 4188–
4189; e) F. Chevallier, B. Breit, Angew. Chem. 2006, 118, 1629–
1632; Angew. Chem. Int. Ed. 2006, 45, 1599–1602; f) C. Waloch, J.
Wieland, M. Keller, B. Breit, Angew. Chem, 2007, 119, 3097–3099;
Angew. Chem. 2007, 119, 3097–3099; Angew. Chem. Int. Ed. 2007,
46, 3037–3039; g) M. de Greef, B. Breit, Angew. Chem. 2009, 121,
559–562; Angew. Chem. Int. Ed. 2009, 48, 551–554; h) Y. Liu, C. A.
Sandoval, Y. Yamaguchi, X. Zhang, Z. Wang, K. Kato, K. Ding, J.
Am. Chem. Soc. 2006, 128, 14212–14213.
with the appropriate amount of catalyst precursor [RhACTHNUTRGNE(UNG nbd)2]BF4 ligand
and substrate in dichloromethane (8.0 mL) under argon. The reactors
were pressurized with dihydrogen and the pressure was kept constant
during the whole reaction. The reaction mixtures were stirred at the ap-
propriate reaction temperature and the hydrogen uptake was monitored
and recorded for every reactor during catalysis. After catalysis, the reac-
tors were cooled down to room temperature and the pressure was re-
duced to 2.0 bar and samples (0.2 mL) were taken.[37]
[14] H. Gulyꢄs, J. Benet-Buchholz E. C. Escudero-Adan, Z. Freixa,
P. W. N. M. van Leeuwen, Chem. Eur. J. 2007, 13, 3424–3430.
[15] a) V. F. Slagt, M. Rçder, P. C. J. Kamer, P. W. N. M. van Leeuwen,
J. N. H. Reek, J. Am. Chem. Soc. 2003, 125, 4056–4057; b) X.-B.
Jiang, L. Lefort, P. E. Goudriaan, A. H. M. de Vries, P. W. N. M. van
Leeuwen, J. G. de Vries, J. N. H. Reek, Angew. Chem. 2006, 118,
1245–1249; Angew. Chem. Int. Ed. 2006, 45, 1223–1227; c) M. Kuil,
P. E. Goudriaan, P. W. N. M. van Leeuwen, J. N. H. Reek, Chem.
Acknowledgements
This research was financially supported by BASF (The Netherlands) and
the Dutch Ministry of Economic Affairs by way of SenterNovem. Dr. H.
Peeters is acknowledged for the mass spectrometric experiments.
10278
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10272 – 10279

Development of a Ligand Library for Catalytic Automated High-Throughput Experiments
FULL PAPER
Commun. 2006, 4679–4681; d) M. Kuil, P. E. Goudriaan, A. W.
Kleij, D. M. Tooke, A. L. Spek, P. W. N. M. van Leeuwen, J. N. H.
Reek, Dalton Trans. 2007, 2311–2320; e) J. M. Takacs, D. S. Reddy,
S. A. Moteki, D. Wu, H. Palencia, J. Am. Chem. Soc. 2004, 126,
4494–4495; f) J. M. Takacs, K. Chaiseeda, S. A. Moteki, D. S. Reddy,
D. Wu, K. Chandra, Pure Appl. Chem. 2006, 78, 501–509; g) S. A.
Moteki, J. M. Takacs, Angew. Chem. 2008, 120, 908–911; Angew.
Chem. Int. Ed. 2008, 47, 894–897.
[23] P. Dupau, C. Bruneau, P. H. Dixneuf, Adv. Synth. Catal. 2001, 343,
331–334.
[24] M. C. Jetter, M. A. Youngman, J. J. McNally, M. E. McDonnell, S.-P.
Zhang, A. E. Dubin, N. Nasser, E. E. Codd, C. M. Flores, S. L. Dax,
Bioorg. Med. Chem. Lett. 2007, 17, 6160–6163.
[25] K. Mikami, K. Aikawa, Y. Yusa, J. J. Jodry, M. Yamanaka, Synlett
2002, 1561–1578.
[26] A. J. Sandee, A. M. van der Burg, J. N. H. Reek, Patent
US2009082595(A1), 2009.
[16] a) L. K. Knight, Z. Freixa, P. W. N. M. van Leeuwen, J. N. H. Reek,
Organometallics 2006, 25, 954–960; b) A. J. Sandee, A. M. van der
Burg, J. N. H. Reek, Chem. Commun. 2007, 864–866.
[27] Handbook of Homogeneous Hydrogenation (Eds.: J. G. De Vries,
C. J. Elsevier), Wiley-VCH, Weinheim, 2006.
[17] a) P.-A. R. Breuil, F. W. Patureau, J. N. H. Reek, Angew. Chem.
2009, 121, 2196–2199; Angew. Chem. Int. Ed. 2009, 48, 2162–2165;
b) J. Wassenaar, M. Kuil, J. N. H. Reek, Adv. Synth. Catal. 2008, 350,
1610–1614; c) S. Jeulin, T. Ayad, J. P. Ratovelomanana-Vidal,
Genet, Adv. Synth. Catal. 2007, 349, 1592–1596; d) J. Holz, B.
Schꢃffner, O. Zayas, A. Spannenberg, A. Bçrner, Adv. Synth. Catal.
2008, 350, 2533–2543; e) H. Shimizu, T. Saito, H. Kumobayashi,
Adv. Synth. Catal. 2003, 345, 185–189.
[18] a) A. B. Smith, C. M. Adams, S. A. L. Barbosa, A. P. Degnan, Proc.
Natl. Acad. Sci. USA 2004, 101, 12042–12047; b) S. J. Mickel, G. H.
Sedelmeier, D. Niederer, R. Daeffler, A. Osmani, K. Schreiner, M.
Seeger-Weibel, B. Bꢅrod, K. Schaer, R. Gamboni, Org. Process Res.
Dev. 2004, 8, 92–100.
[28] For an account on intra- and intermolecular interactions, see: A.
Mulder, J. Huskens, D. N. Reinhoudt, Org. Biomol. Chem. 2004, 2,
3409–3424, and references therein.
[29] http://www.chemspeed.com.
[30] With the phthalimide-protected analogue of E also excellents ee
values were obtained, see: H. Huang, X. Liu, J. Deng, M. Qiu, Z.
Zheng, Org. Lett. 2006, 8, 3359–3362.
[31] F. W. Patureau, S. de Boer, M. Kuil, J. Meeuwissen, P.-A. R. Breuil,
M. A. Siegler, A. L. Spek, A. J. Sandee, B. de Bruin, J. N. H. Reek,
J. Am. Chem. Soc. 2009, 131, 6683–6685.
[32] http://www.amtec-chemnitz.de.
[33] G. J. H. Buisman, L. A. van der Veen, A. Klootwijk, W. G. J. de
Lange, P. C. J. Kamer, P. W. N. M. van Leeuwen, D. Vogt, Organo-
metallics 1997, 16, 2929–2939.
[34] J. I. van der Vlugt, A. C. Hewat, S. Neto, R. Sablong, A. M. Mills,
M. Lutz, A. L. Spek, C. Mꢆller, D. Vogt, Adv. Synth. Catal. 2004,
346, 993–1003.
[19] a) D. PeÇa, A. J. Minnaard, J. G. de Vries, B. L. Feringa, J. Am.
Chem. Soc. 2002, 124, 14552–14553; b) S. Enthaler, G. Erre, K.
Junge, J. Holz, A. Bçrner, E. Alberico, I. Nieddu, S. Gladiali, M.
Beller, Org. Process Res. Dev. 2007, 11, 568–577.
[20] R. P. Cheng, S. H. Gellman, W. F. DeGrado, Chem. Rev. 2001, 101,
3219–3232, and references therein.
[35] G. J. H. Buisman, P. C. J. Kamer, P. W. N. M. van Leeuwen, Tetrahe-
dron: Asymmetry 1993, 4, 1625–1634.
[21] a) W. Tang Y. Chi, X. Zhang, Org. Lett. 2002, 4, 1695–1698; b) Z.
Zhang G. Zhu, Q. Jiang, D. Xiao, X. Zhang, J. Org. Chem. 1999, 64,
1774–1775; c) J. L. Renaud, P. Dupau, A-E. Hay, M. Guingouain,
P. H. Dixneuf, C. Bruneau, Adv. Synth. Catal. 2003, 345, 230–238.
[22] T. Guo, H. Gu, D. W. Hobbs, D. E. Busler L. L. Rokosz, Bioorg.
Med. Chem. Lett. 2007, 17, 1718–1721.
[36] C. Yu, B. Liu, L. Hu, J. Org. Chem. 2001, 66, 5413–5418.
[37] Spectral data of the prepared compounds can be found in the Sup-
porting Information.
Received: April 27, 2009
Published online: September 3, 2009
Chem. Eur. J. 2009, 15, 10272 – 10279
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10279