A theoretically-guided optimization of a new family of modular P,S-ligands for iridium-catalyzed hydrogenation of minimally functionalized olefins

Article abstract of DOI:10.1002/chem.201402978

A library of modular iridium complexes derived from thioether-phosphite/phosphinite ligands has been evaluated in the asymmetric iridium-catalyzed hydrogenation of minimally functionalized olefins. The modular ligand design has been shown to be crucial in finding highly selective catalysts for each substrate. A DFT study of the transition state responsible for the enantiocontrol in the Ir-catalyzed hydrogenation is also described and used for further optimization of the crucial stereodefining moieties. Excellent enantioselectivities (enantiomeric excess (ee) values up to 99%) have been obtained for a range of substrates, including E- and Z-trisubstituted and disubstituted olefins, α,β-unsaturated enones, tri- and disubstituted alkenylboronic esters, and olefins with trifluoromethyl substituents.

Full text of DOI:10.1002/chem.201402978

DOI: 10.1002/chem.201402978
Full Paper
&
Hydrogenation
A Theoretically-Guided Optimization of a New Family of Modular
P,S-Ligands for Iridium-Catalyzed Hydrogenation of Minimally
Functionalized Olefins
Jꢀssica Margalef,[a] Xisco Caldentey,[b] Erik A. Karlsson,[b] Mercꢀ Coll,[a] Javier Mazuela,[a]
Oscar Pꢁmies,[a] Montserrat Diꢂguez,*[a] and Miquel A. Pericꢁs*[b, c]
Abstract: A library of modular iridium complexes derived
from thioether-phosphite/phosphinite ligands has been eval-
uated in the asymmetric iridium-catalyzed hydrogenation of
minimally functionalized olefins. The modular ligand design
has been shown to be crucial in finding highly selective cat-
alysts for each substrate. A DFT study of the transition state
responsible for the enantiocontrol in the Ir-catalyzed hydro-
genation is also described and used for further optimization
of the crucial stereodefining moieties. Excellent enantioselec-
tivities (enantiomeric excess (ee) values up to 99%) have
been obtained for a range of substrates, including E- and Z-
trisubstituted and disubstituted olefins, a,b-unsaturated
enones, tri- and disubstituted alkenylboronic esters, and ole-
fins with trifluoromethyl substituents.
nylalanine (l-DOPA),[3] the broad-spectrum antibiotic levofloxa-
cin (Daichii-Sankyo Co.),[4] and sitagliptin (Merck),[5] as well as
the synthesis of the pesticide (S)-metolachlor.[6] Whereas today
a notable series of chiral ligands (mostly phosphorus-based)
for the Ru- and Rh-catalyzed hydrogenation of olefins possess-
ing polar functional groups is available to the chemical com-
munity,[2a,b] the reduction of minimally functionalized sub-
strates is by far less well-developed.[2d,7] The use of chiral ana-
logues of Crabtree’s catalyst[8] modified with phosphine-oxazo-
line (PHOX) ligands ([Ir(PHOX)(cod)][BArF]) (cod=1,5-cycloocta-
diene; [BArF]=[B{3,5-(CF3)2C6H3}4]ꢀ) represented the first
breakthrough in the hydrogenation of this type of substrate.[9]
Since then, mixed phosphorus-oxazoline ligands have been
the most popular heterodonor ligands in this process. Many
successful P-oxazoline ligands have been prepared by incorpo-
rating P-donor groups other than phosphines and by modify-
ing the chiral backbone.[10] Although these modifications have
aided the development of new ligands that have considerably
expanded the scope of Ir-catalyzed hydrogenation, most of the
screened catalysts are still highly substrate-dependent, and
their preparation involves long synthetic sequences. The devel-
opment of efficient modular chiral ligands, readily available
from simple starting materials, which tolerate a broad range of
substrates, still remains a challenge. More recently, research
has been expanded to design heterodonor P,X-ligands bearing
more robust X-donor groups than oxazolines (pyridines,[11]
amides,[12] thiazoles,[13] oxazoles,[14] etc.). In this respect, we
have recently described the successful use of non-N-donor
heterodonor ligands, sugar-based thioether-phosphorus li-
gands, for enantioselective Ir-catalyzed reduction of minimally
functionalized olefins.[15] Ir-complexes modified with these P-
thioether ligands efficiently catalyzed the hydrogenation of
a large range of E- and Z-trisubstituted olefins and the more
Introduction
The growing demand for enantiomerically pure products, re-
quired in the preparation of both compounds of technological
interest and compounds possessing biological activity, has
stimulated the search for highly efficient asymmetric catalytic
processes that display high selectivity and activity, minimal
consumption of energy, and minimal generation of byprod-
ucts.[1] Compared with other techniques, asymmetric catalysis
is an attractive strategy because it uses only a small amount of
catalyst to produce an extensive amount of the requested
target compound, thus reducing the formation of byproducts.
It also has the advantage of reducing the number of reaction
steps and synthetic operations, thus bringing down the overall
production cost.[1]
Asymmetric hydrogenation has become a highly useful tool
for preparing enantiomerically pure compounds because of its
high efficiency, low catalyst loadings, operational simplicity,
and perfect atom economy.[1–2] Its uses have been largely ac-
cepted by the chemical community as illustrated by the com-
mercial production of the Parkinson’s drug l-3,4-dihydroxyphe-
[a] J. Margalef, Dr. M. Coll, Dr. J. Mazuela, Dr. O. Pꢀmies, Prof. M. Diꢁguez
Departament de Quꢂmica Fꢂsica i Inorgꢀnica, Universitat Rovira i Virgili
C/Marcel·li Domingo, s/n., 43007 Tarragona (Spain)
E-mail: montserrat.dieguez@urv.cat
[b] Dr. X. Caldentey, Dr. E. A. Karlsson, Prof. M. A. Pericꢀs
Institute of Chemical Research of Catalonia (ICIQ)
Av. Pa¨ısos Catalans, 16, 43007 Tarragona (Spain)
E-mail: mapericas@iciq.es
[c] Prof. M. A. Pericꢀs
Departament de Quꢂmica Orgꢀnica
Universitat de Barcelona, 08028 Barcelona (Spain)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201402978.
Chem. Eur. J. 2014, 20, 12201 – 12214
12201
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
difficult disubstituted olefins. The results are comparable to
the best ones reported in the literature. Apart from this, the
use of other phosphorus-thioether ligands in the same process
remains unexplored, and a systematic study of the scope of
P,S-ligands is still needed. No mechanistic studies have been
made using this type of ligands to enable a priori prediction of
the right ligand needed to obtain high enantioselectivity.
Therefore, more research is
since the starting enantiopure epoxides are prepared through
a catalytic Sharpless epoxidation, both enantiomeric series of
the target P,S-ligands are equally available. The potential ap-
plicability of the Ir-thioether-phosphite/phosphinite catalyst
precursors ([Ir(cod)(L1–L10a–g)][BArF]) was further proved
using propylene carbonate as a green alternative solvent,
which allows catalyst recycling.
needed to discern the role of
ligand parameters in the origin
of enantioselectivity.
To address all these points, in
this study we have prepared and
evaluated a new highly modular
thioether-phosphite/phosphinite
ligand library (Figure 1) in the Ir-
catalyzed
hydrogenation
of
a broad range of minimally func-
tionalized olefins, including ex-
amples with neighboring polar
groups. These ligands are easily
prepared in few steps from read-
ily available enantiopure arylgly-
cidols. They also incorporate the
advantages of the robustness of
the thioether moiety[16] and the
additional control provided by
the flexibility of the chiral pocket
through a highly modular ligand
scaffold. In a simple three step
procedure (Scheme 1), several
ligand parameters could easily
be tuned to maximize the cata-
lyst performance. With this
ligand library, we therefore in-
vestigated the effect of system-
atically changing the thioether
(L1–L6) and alkoxy (L1, L7, and
L9) groups, the nature of the
starting material arylglycidol
(L10), the configuration of the
Figure 1. Thioether-phosphite/phosphinite ligand library L1–L10a–g.
biaryl phosphite moiety (a–c), Scheme 1. Synthesis of new thioether-phosphite/phosphinite ligands L1–L10a--g. i) R1X/NaH/DMF;[18] ii) R2SH/
NaOH/dioxane/H2O;[17] iii) ClP(OR)2/pyridine/toluene/808C or ClPR2/NEt3/toluene.
and the consequences of replac-
ing the phosphite moiety by
a phosphinite group (d–g). In
this paper we have also carried out DFT calculations to explain Results and Discussion
the origin of enantioselectivity. These DFT calculations have
Synthesis of ligands
also been crucial in the optimization of the ligand design. In-
terestingly, we found that the catalytic performance of the
new ligands is excellent and similar to the performance of the
previous furanoside thioether-phosphorus counterparts,[15]
which have recently emerged as some of the most successful
catalysts designed for this process, with two added advantag-
es. First, these new Ir-thioether-P catalytic systems are able to
expand the scope to a larger range of olefins, which includes
a,b-unsaturated enones, tri- and disubstituted alkenylboronic
esters and olefins with trifluoromethyl substituents. Second,
The new thioether-phosphite L1–L10a–c and phosphinite L1–
L10d--g[17] ligands were efficiently synthesized in one step
from the corresponding readily accessible thioether-alcohols
(7–16; Scheme 1). These compounds are easily prepared in
two steps from enantiopure arylglycidols readily available on
a large scale (0.5–1.0 mol)[18] following previously reported pro-
cedures.[17] In the first step, the protection of the free hydroxyl
group enables us to introduce the desired variety in the alkoxy
group (Scheme 1, step i).[18c] In the second step, the regioselec-
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12202
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tive and stereospecific ring opening by thiolates produced the
corresponding thioether-hydroxyls (7–16; Scheme 1, step ii),
thus giving room for additional diversity by performing the
opening with different thiolates.[17] The last step of the ligand
synthesis (Scheme 1, step iii) is the reaction of the correspond-
ing thioether-hydroxyl in the presence of base with one equiv-
alent of either the corresponding biaryl phosphorochloridite
(ClP(OR)2; P(OR)2 =a–c) to provide thioether-phosphite ligands
(L1–L10a–c) or the required chlorophosphine (ClPR2; PR2 =d–
g) to achieve the new thioether-phosphinite ligands (L1–
L10d–g (Scheme 1, step iii).
this behavior is due to the fast exchange of the biphenyl
moiety on the NMR timescale. This hypothesis is further con-
firmed in the X-ray analysis of [Ir(cod)(L6a)][BArF], which shows
the presence of the two diastereoisomers resulting from the
conformational isomerism of the biphenyl phosphite moiety in
the solid state (see the Supporting Information). All this indi-
cates that the ligand backbone is not able to control the con-
formational isomerism of the biaryl phosphite group. There-
fore, it is not surprising that in catalytic studies the enantiose-
lectivity obtained with [Ir(cod)(L1–L9a)][BArF] precursors was
low (see below). It could thus be concluded from the VT-NMR
experiments that the catalyst precursors are configurationally
stable in solution at the sulfur center, which, however, does
not necessarily imply that the same holds true for the catalyti-
cally active IrIII/IrV complexes during the reaction conditions
(see below).
All of the ligands are stable in air at room temperature and
to hydrolysis. They were isolated in good yields as white solids
or colorless oils after purification on neutral alumina.
Synthesis of the Ir-catalyst precursors
Crystals suitable for X-ray diffraction analysis of [Ir(cod)-
(L1d)][BArF], [Ir(cod)(L4a)][BArF], and [Ir(cod)(L9a)][BArF] com-
plexes were also obtained to determine the coordination
mode of this new ligand class (Figure 2). In contrast to Ir-L6a
complex, the solid-state structure of complexes containing L4a
and L9a indicated that only one of the diastereoisomers crys-
tallized.
The catalyst precursors were prepared by treating
[Ir(m-Cl)(cod)]2 (0.5 equiv) with an equimolar amount of the ap-
propriate P,S-ligand (L1–L10a–g) in dichloromethane at reflux
ꢀ
for 1 h. The Clꢀ/BArF counterion exchange was then per-
formed by reaction with sodium tetrakis[3,5-bis(trifluorome-
thyl)phenyl]borate (NaBArF; 1 equiv) in water (Scheme 2). The
catalyst precursors were obtained in pure form as air-stable
red-orange solids. No further purification was thus needed. It
should be mentioned that all attempts to prepare iridium com-
plexes containing thioether-phosphinite ligands with the ex-
tremely bulky mesityl phosphinite (f) moiety were unsuccess-
In all cases, the six-membered chelate ring adopted a chair
conformation, with the alkoxide group pointing in the oppo-
site direction to the coordination sphere. However, whereas
the crystal structures of [Ir(cod)(L)][BArF] (L=L4a, L6a, and
L9a), containing a phosphite moiety, showed the thioether
substituent in an equatorial position, an axial disposition of the
thioether substituent was observed for [Ir(cod)(L1d)][BArF],
containing a phosphinite group.
Asymmetric hydrogenation
Asymmetric hydrogenation of the minimally functionalized
model olefin E-2-(4-methoxyphenyl)-2-butene (S1): A computa-
tional study for ligand optimization
Scheme 2. Synthesis of Ir precursors [Ir(cod)(PꢀS)][BArF] (PꢀS=L1–L10a–g).
ful.
Initially, we applied phenylglycidol-based ligands L1–L9a–g in
the Ir-catalyzed hydrogenation of the model substrate E-2-(4-
methoxyphenyl)-2-butene (S1). Model substrate S1 has been
successfully reduced by a large number of catalysts, thus ena-
bling a direct comparison of the potential of the new ligands
with the state of the art.[2d,7] The results, which are summarized
in Table 1, indicated that the enantioselectivity is mainly affect-
ed by the thioether substituent and the type of P-donor
group, whereas the effect of the alkoxy substituent is less pro-
nounced. The small effect of the alkoxy substituent on enantio-
selectivity (i.e., Table 1; entries 1, 24, and 32) is not unexpected
since this substituent is located far away from the coordination
sphere as can be seen in the X-ray structures (see above) and
the DFT-calculated transition states (TS; see below).
The HRMS-ESI spectra show the heaviest ions at m/z, which
correspond to the loss of the BArF anion from the molecular
species. The complexes were also characterized by H, 13C, and
31P NMR spectroscopy. The spectral assignments, made using
1H-1H and 13C-1H correlation measurements, were as expected
for these C1-symmetric iridium complexes.
1
Variable-temperature (VT)-NMR spectroscopic experiments in
CD2Cl2 (+35 to ꢀ858C) indicate the presence of a single
isomer in all cases except for [Ir(cod)(L1–L9a)][BArF] com-
pounds. For these latter complexes, the 31P VT-NMR spectra
show that the signals become broader when the temperature
is lowered. This behavior could indicate a rapid exchange of
the possible diastereoisomers formed by conformational iso-
merism of the biphenyl moiety and/or when the thioether co-
ordinates to the metal atom. The fact that the presence of dif-
ferent diastereoisomers in solution is only observed for com-
plexes with ligands containing a conformationally labile bi-
phenyl moiety (a) and not for related complexes with ligands
containing enantiopure biphenyl moieties (b,c), suggests that
We found that the correct choice of the thioether substitu-
ent is crucial to achieve the highest levels of enantioselectivity.
The results showed that the presence of aryl substituents pro-
vided higher enantioselectivities than alkyl thioether substitu-
ents. Among the aryl substituents, enantioselectivities increase
with increasing steric bulk of the thioether substituent (2,6-
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12203
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 1. Results for the Ir-catalyzed hydrogenation of S1 using the P,S-
ligand library L1–L9a–g.[a]
Entry
Ligand
ee[b]
Entry
Ligand
ee[b]
[%]
[%]
1
2
3
4
5
6
7
8
L1a
L1b
L1c
L1d
L2b
L2c
L2e
L3a
L3b
L3c
L3e
L4a
L4b
L4c
L4d
L4e
L5a
L5e
26 (R)
42 (R)
13 (R)
44 (R)
40 (R)
12 (R)
84 (R)
8 (R)
36 (R)
31 (S)
86 (R)
14 (R)
41 (R)
19 (R)
53 (R)
49 (R)
25 (R)
35 (R)
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36[c]
L6a
L6b
L6c
L6d
L6e
L7a
L7b
L7c
L7d
L7e
L7g
L8a
L8e
L9a
L9b
L9c
L9d
L8e
26 (R)
48 (R)
55 (S)
64 (R)
92 (R)
30 (R)
50 (R)
17 (S)
41 (R)
86 (R)
8 (R)
24 (R)
93 (R)
31 (R)
45 (R)
34 (R)
41 (R)
93 (R)
9
10
11
12
13
14
15
16
17
18
[a] Reactions carried out using 0.5 mmol of S1, 2 mol% of Ir-catalyst pre-
cursor, CH2Cl2 as solvent, 100 bar H2, 4 h. Full conversions were achieved
in all cases. [b] Enantiomeric excesses (ee values) determined by chiral
GC. [c] Reaction carried out using 0.25 mol% of Ir-catalyst precursor for
8 h; 99% conversion.
tween the configuration of the ligand backbone and the con-
figuration of the biaryl group that led to a matched combina-
tion for ligands containing an R-biaryl phosphite moiety (b;
Table 1, entries 13 and 14). However, the best enantioselectivi-
ties were obtained with ligands containing a phosphinite
group (ee values up to 93%, Table 1, entry 31). In particular, re-
placing the phosphite moiety by a bulky di-o-tolyl phosphinite
group had a positive effect on enantioselectivity, whereas the
use of a cyclohexyl phosphinite group led to poor enantiose-
lectivities (Table 1, entry 29). This behavior is in contrast with
the negative effect observed when replacing the phosphite
group by a phosphinite moiety in the previous furanoside-
based thioether-P ligands.[15b] These results clearly show the
importance of using a modular scaffold to build new ligand
systems.
Figure 2. X-ray structures of: a) [Ir(cod)(L1d)][BArF] (CCDC-993594); b) [Ir-
(cod)(L4a)][BArF] (CCDC-993595), and c) [Ir(cod)(L9a)][BArF] (CCDC-993597)
(the BArF counterion and solvent molecules have been omitted for clarity).
These data can be obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
ꢀ
We also performed the reaction at low catalyst loading
(0.25 mol%) using ligand L8e. High enantioselectivity (93% ee)
and activity were maintained.
Me2-C6H3 >1-Napth>2-Napth>Ph; Table 1, entries 23, 11, 7,
and 4).
Regarding the effect of the P-donor group on enantioselec-
tivity, we found that the presence of a conformationally labile
biaryl phosphite group (a) provided low enantioselectivities,
because as observed in the VT-NMR spectra and X-ray struc-
tures of the [Ir(cod)(L1–L9a)][BArF] catalyst precursors, the
ligand backbone is not able to control its conformational iso-
merization (Table 1, entries 1, 8, 12, 17, 19, 24, 30, and 32).
Enantioselectivities therefore increased by using enantiopure
biaryl phosphite groups (b,c; that is, Table 1, entries 13 and 14
vs. 12). We also found that there is a cooperative effect be-
With the aim to find which ligand parameters should be fur-
ther modified to increase enantioselectivity, we performed
a DFT computational study of the transition states involved in
the enantiocontrol of the iridium-catalyzed hydrogenation of
substrate S1. Several DFT studies using P,N- and carbene-N li-
gands have indicated that the hydrogenation of minimally
functionalized alkenes proceeds via IrIII/IrV tetrahydride interme-
diates.[10p,19] Recent studies by Hopmann and co-workers using
a phosphine-oxazoline (PHOX)-based iridium catalyst,[19e] and
by our group, in conjunction with the groups of Norrby and
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12204
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Andersson, using Ir-phosphite-oxazoline ligands,[10p] strongly
support that the hydrogenation of minimally functionalized
olefins using P,N-ligands follows a mechanism involving an IrIII/
IrV migratory-insertion/reductive-elimination pathway (labeled
3/5-MI in Scheme 3). In these studies, two catalytic pathways
were contemplated. The already mentioned 3/5-MI pathway
and the mechanism involving an IrIII/IrV s-metathesis/reductive-
elimination pathway (labeled 3/5-Meta in Scheme 3). It has
also been shown that the transition states for the migratory-in-
sertion in the 3/5-MI pathway (TSMI) and the s-metathesis in
the 3/5-Meta pathway (TSMETA) are responsible for the selectivi-
ty in the Ir-catalyzed hydrogenation, and that the enantioselec-
tivity therefore could be reliably calculated from the relative
energies of these transition states.[19d]
that the most stable transition state (TSA1MI) matches the
major product obtained experimentally (R product, Table 1, en-
tries 4 and 22), whereas the most stable transition state with
the re face coordinated (TSA8MI) is expected to be responsible
for the formation of the minor S product. The energy differen-
ces between the most stable transition states giving rise to the
major and minor products are 4.5 and 8.5 kJmolꢀ1, respective-
ly, for L1d and L6d. We also found that the hydrogenation
products are formed through the 3/5-MI mechanism, since the
TS energies for the 3/5-Meta pathway, in both the major and
minor configuration, are at least 13 kJmolꢀ1 higher than those
for the 3/5-MI pathway (see Table 2). Nevertheless, since the
energetic difference between the two pathways is relatively
small, both have to be taken into consideration for further cal-
culations. It should be pointed out that the fact that the calcu-
lations indicate that the minor S product is formed through
a transition state in which the configuration at the sulfur
center is S, whereas the major R product results from a transi-
tion state with R configuration at the sulfur center raised some
concerns regarding the validity of the theoretical model. In
general, a model of this kind, which only takes into account
the relative energies of the transition states through which the
various isomeric intermediates are transformed into their corre-
sponding products, to calculate the product distribution, re-
quires that the Curtin–Hammet principle be applicable, that is,
that the interconversion of the said intermediates be faster
than their evolution into the corresponding products.
However, the VT-NMR studies, in combination with X-ray analy-
sis of [Ir(cod)(L1d)][BArF], suggest that, at least at the level of
the IrI catalyst precursors, the R configuration is maintained at
the sulfur center in solution. To address these concerns, the
transition states for the interconversion of the intermediates
A7 and A8 were calculated. The results clearly show that the
barrier for pyramidal inversion at the sulfur center is considera-
bly lower than the barriers leading to product formation, thus
confirming the applicability of the Curtin–Hammet principle
and the validity of the theoretical model (Table SI.3 in the Sup-
porting Information).
Scheme 3. 3/5-MI and 3/5-Meta catalytic cycles for the Ir-catalyzed hydroge-
nation.
On the basis of these previous studies we therefore per-
formed a computational study of the TSMI and TSMETA transition
states. To accelerate the DFT calculations, we initially studied li-
gands L1d and L6d, containing the simple unsubstituted di-
phenyl phosphinite moiety. In addition, these ligands contain
two types of thioether groups that will help us to understand
the already observed key role of introducing a bulky 2,6-dime-
thylphenyl thioether substituent on enantioselectivity. The
transition states using S1 as substrate for the stereochemistry
determining migratory insertion (TSMI) or s-bond metathesis
(TSMETA) were calculated by using the B3LYP functional,[20] the
6-31G*/LANL2DZ basis set,[21] and the PCM solvent model with
Figure 3 shows the most stable calculated transition states
(TS) for the major and the minor pathway with both ligands. In
these key transition states we can see, on the one hand, the
proximity of the phenyl moiety in the ligand backbone group
to the thioether substituent and, on the other hand, that the
hydrogen at the ortho position of the phenyl group in the
ligand skeleton is pointing towards the metal center. All these
findings indicate that the aromatic substituent in the ligand
backbone could have an important influence on the enantiose-
lectivity.
[22]
parameters for CH2Cl2, as implemented in Gaussian 09.[23] The
energies were further refined by performing single-point calcu-
lations at the 6-311+G** level,[24] and by dispersion correction
with the DFT-D3 model.[25]
Table 2 shows the calculated energies for the most stable
isomers of the transition states (TSMI and TSMETA). These key iso-
mers are the result of varying between the two possible con-
figurations at the sulfur center, coordinating to the two enan-
tiotopic faces (re and si) of the olefin, and changing the relative
position of the hydride (up or down).[26] It should be men-
tioned that olefins coordinated through the si face are reduced
to the R product, whereas those coordinated through the re
face give access to the S product. The results in Table 2 shows
These features prompted us to recalculate the relevant tran-
sition states (from A1 for the major pathway and A8 for the
minor pathway) by replacing the phenyl group by a mesityl
group (ligand L10d; Figure 1). The results, which are summar-
ized in Table 3, showed that the energy difference between
the two transition states was unrealistically large
(30.9 kJmolꢀ1). In Figure 4, it can be seen that in the transition
state giving the S product there is a great steric interaction be-
tween the thioether substituent and the mesityl group, essen-
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12205
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 1, entries 27 and 28), we
also performed the calculations
of the relevant transition states
with the mesityl-based ligand
L10e (Figure 1), with tolyl
groups at the phosphinite
moiety. However, the calculated
energy difference between the
most stable transition states
Table 2. Calculated energies for the transition states TSMI and TSMETA with substrate S1 using ligands L1d and
L6d.[a]
Starting
TSMI
Starting
TSMETA
geometry
L1d
L6d
geometry
L1d
L6d
0
0
17.0
20.2
thus obtained was 13.7 kJmolꢀ1
,
very similar to that achieved
with ligand L10d (Table 3 and
Figure 5). So, in contrast to that
observed for ligands L1–L9, con-
taining a phenyl group in the
backbone (see above), the steric
bulk of the phosphinite group
should have little impact on
enantioselectivity for the mesi-
tyl-based ligands.
20.0
25.5
19.1
9.8
30.5
36.7
30.3
19.0
35.1
20.5
8.5
20.2
31.8
32.7
31.7
34.6
13.1
26.2
26.0
34.3
44.1
36.3
42.9
21.1
31.0
With these latter theoretical
results in hand, a decision was
made to prepare and screen thi-
oether-phosphinite ligands L10d
and L10e, with a mesityl group,
in the asymmetric hydrogena-
tion of substrate S1. The experi-
mental results are shown in
Table 4 (entries 3 and 4). As pre-
dicted by the theoretical calcula-
tions, both mesityl-based ligands
afforded similar higher enantio-
selectivities than ligands L1–L9.
If we compare the calculated
22.6
18.2
4.5
and
experimental
values
(Table 4), we can conclude that,
despite the fact that the calcu-
lated free-energy differences are
systematically higher than the
experimental values, the general
trend is reproduced well. The ro-
[a] Energies in kJmolꢀ1; R=4-MeO-C6H4.
tially locking the configuration at the sulfur center to R. We
therefore switched from an S configuration at the sulfur center
to an R configuration choosing again the most stable isomers
previously calculated for ligands L1d and L6d (TS from A5,
A7, and A15; Table 3). Thus, the obtained energy difference
between the two most stable transition states responsible for
the formation of both enantiomers of the hydrogenated prod-
bustness of the theoretical model is demonstrated with the
prediction of the new improved ligands L10d and L10e con-
taining a mesityl group.
Asymmetric hydrogenation of other minimally functionalized
olefins: Scope and limitations
uct was 14.2 kJmolꢀ1 (ligand L10d) surpassing the DDG+
To establish the scope of the new family of ligands in the Ir-
catalyzed hydrogenation, we selected a representative family
of substrates. We first studied the asymmetric hydrogenation
of other E- and Z-trisubstituted olefins (S2–S18), including ex-
amples containing neighboring polar groups, by using the P,S-
ligand library L1–L10a–g. The most noteworthy results are
shown in Table 5 (see the Supporting Information for a com-
plete set of results). We found again that the correct choice of
the ligand parameters is crucial to achieve the highest levels of
calcd
with ligands L1d and L6d (4.5 kJmolꢀ1 and 8.5 kJmolꢀ1, re-
spectively), indicating that this new modification should pro-
vide higher enantioselectivities than the Ir-L1d and Ir-L6d cat-
alysts.
Encouraged by this result and having in mind that the cata-
lytic experiments using phenyl glycidol-based ligands (L1–L9)
showed that replacing the diphenyl phosphinite moiety by o-
tolyl groups has a positive effect on enantioselectivity (i.e.,
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12206
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 5. Calculated transition states (TS) for the major and the minor path-
ways with ligand L10e.
to >99%; Table 5, entries 2, 3, 5, and 6). The result followed
the same trends as those observed for substrate S1. Enantiose-
lectivities were thus best with the optimized ligands L10d and
L10e.
To assess the potential of the new ligand library for Z-trisub-
stituted isomers, which are usually hydrogenated less enantio-
selectively than the corresponding E-isomers, we chose sub-
strates S4 and S5 (Table 5, en-
Figure 3. Calculated transition states (TS) for the major and the minor path-
ways with ligands L1d and L6d.
Table 3. Calculated energies for the relevant transition states with substrate S1 using ligands L10d and
L10e.[a]
tries 7–12). The reduction of the
model Z substrate S4 proceeded
with moderate enantiocontrol
and followed a different trend
than that observed with E sub-
strates S1–S3. The enantioselec-
tivities were thus best with li-
gands L6a and L6c (Table 5, en-
tries 7 and 8). The moderate
enantioselectivity can be ex-
plained by a competition be-
tween direct hydrogenation
versus Z/E-isomerization of the
substrate. The hydrogenation of
the E isomer produces the oppo-
site configuration of the hydro-
genated product than when the
Z isomer is hydrogenated, which
results in low enantioselectivi-
ty.[2d] Accordingly, the reduction
Starting
L10d
L10e
Starting
L10d
L10e
geometry
geometry
0
0
14.2
13.7
30.9
19.9
36.7
37.5
28.4
33.2
[a] Energies in kJmolꢀ1; R=4-MeO-C6H4.
of dehydronaphthalene S5, which has a Z configuration and
for which Z/E-isomerization is not possible, produces higher
Table 4. Comparison between experimental and theoretical results.[a]
ee[a]
DDGþexptl
DDGþcalcd
[b]
[b]
Entry
Ligand
Figure 4. Calculated transition states (TS) for the major and the minor path-
ways with ligand L10d.
[%]
1
2
3
4
L1d
L6d
L10d
L10e
44 (R)
64 (R)
94 (R)
95 (R)
2.3
3.8
8.6
9.1
4.5
8.5
14.2
13.7
enantioselectivity. We initially studied the hydrogenation of E
substrates S2 and S3, related to S1, which differ in the sub-
stituents of both the aryl ring and the substituents trans to the
aryl group. Excellent enantioselectivities, even higher than with
the model substrate S1, were obtained (ee values between 98
[a] Reaction conditions: 0.5 mmol of S1, 2 mol% catalyst precursor,
CH2Cl2 as solvent, 100 bar H2, 4 h. Full conversions were achieved in all
cases. The ee values were measured by GC. [b] Energies in kJmolꢀ1
.
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12207
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 5. Selected results for the Ir-catalyzed hydrogenation of S2–S18 using the P,S-ligand library L1–L10a–g.[a]
Entry
Substrate
Product
L
ee[b]
Entry
Substrate
Product
L
ee[b]
[%]
[%]
1
2
3
4
5
6
L8e
99 (R)
28
29
30
31
32
33
L7e
L8e
L10e
L7e
L8e
78 (R)
79 (R)
81 (R)
81 (R)
80 (R)
85 (R)
L10d
L10e
L8e
L10d
L10e
99 (R)
>99 (R)
97 (R)
98 (R)
99 (R)
L10e
7
8
9
L6a
L6c
L10e
62 (S)
62 (S)
58 (S)
34
35
36
L6d
L10d
L10e
94 (S)
98 (S)
99 (S)
10
22
12
L8c
L8e
L10e
36 (R)
78 (R)
82 (R)
37
38
39
L6d
L10d
L10e
96 (S)
97 (S)
99 (S)
13
14
15
L8e
L10d
L10e
>99 (R)
99 (R)
>99 (R)
40
41
42
L6d
L10d
L10e
95 (S)
98 (S)
98 (S)
16
17
18
L8e
L10d
L10e
99 (R)
99 (R)
99 (R)
43
44
45
L8e
L10d
L10e
70 (S)
69 (S)
72 (S)
19
20
21
22
23
24
L8e
98 (R)
98 (R)
99 (R)
99 (R)
99 (R)
99 (R)
16
47
48
49
50
51
L9d
43 (R)
44 (R)
45 (R)
94 (+)
93 (+)
83 (+)
L10d
L10e
L8e
L10d
L10e
L10d
L10e
L7a
L7c
L10e
25
26
27
L7e
L8e
L10e
60 (R)
61 (R)
68 (R)
[a] Reactions carried out using 0.5 mmol of substrate, 2 mol% of Ir-catalyst precursor, CH2Cl2 as solvent, 100 bar H2, 4 h. Full conversions were achieved in
all cases. [b] Enantiomeric excesses determined by chiral GC or HPLC.
enantioselectivities (ee values up to 82%; Table 5, entry 12).
Moreover in contrast to S4, the best enantioselectivities were
achieved with the optimized mesityl-based ligands L10d and
L10e.
values up to 85%, Table 5, entry 33). The use of the optimized
mesityl-based ligands L10d and L10e was essential to achieve
the highest levels of enantioselectivity in the reduction of sev-
eral a,b-unsaturated ketones S13–S15 (ee values ranging from
98 to 99%; Table 5, entries 34–42), for which the previous fura-
noside PꢀS ligands proved to be unsuccessful.[27] This repre-
sents an important entry point to the formation of ketones
with stereogenic centers in the a position to the carbonyl
group. Despite this, they have been less studied than other tri-
substituted olefins with a neighboring polar group.[2d] Other
challenging substrate types that have been less investigated
are the a,b-unsaturated amides (S16)[28] and alkenylboronic
esters (S17 and S18).[29] Amides with stereogenic centers in the
a position are an important class of compounds since this
motif is present in several natural products and they can be
easily transformed into other useful compounds (i.e.,
amines).[30] The hydrogenation of alkenylboronic esters pro-
vides easy access to chiral borane compounds, which are val-
uable organic intermediates since the CꢀB bond can be easily
transformed to CꢀO, CꢀN and CꢀC bonds with retention of
the chirality.[31] The hydrogenation of a,b-unsaturated amide
S16 followed the same trend as substrate S1. Enantioselectivi-
ties up to 72% were thus achieved with ligand L10e. The re-
duction of alkenylboronic esters followed a different trend
than S1. Whereas for the more studied substrate S17 moderate
We next studied the reduction of a wide range of trisubsti-
tuted olefins containing several types of neighboring polar
groups S6–S18 (Table 5, entries 13–51). The hydrogenation of
this type of substrate is especially relevant because they allow
for further functionalization and could therefore be important
intermediates for the synthesis of more complex chiral mole-
cules. We were pleased to find that enantioselectivities are
among the best observed in most of the examples. A range of
a,b-unsaturated esters (S6–S9) were thus efficiently hydrogen-
ated (ee values ranging from 98% to >99%). It should be
noted that the ee values are highly independent of the nature
of the alkyl substituent and the electronic nature of the sub-
strate phenyl ring. Although enantioselectivities follow the
same trend regarding the effect of the thioether, alkoxy, and
the P-donor group, the nature of the aryl group in the ligand
backbone is less pronounced. Enantioselectivities were thus
best with ligands L6e, L8e, L10d, and L10e. On the other
hand, the presence of a trimethylsilyl group in the substrate
(S10) has a negative effect on enantioselectivity (Table 5, en-
tries 25–27), whereas the reduction of allylic alcohol and ace-
tate S11 and S12 provided higher enantioselectivities (ee
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12208
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
enantioselectivities were achieved, for the less studied sub-
strate S18 high enantioselectivities up to 94% were reached
using phosphite-thioether ligands L7a and L7c. These results
again showed the importance of having a modular ligand
design.
ing trisubstituted ones, we next chose to hydrogenate sub-
strate S19 as a model. The lower enantioselectivity obtained
with 1,1-disubstituted terminal olefins than that obtained with
trisubstituted olefins has been attributed to two main mo-
tives.[2d,7a,e] The first is that enantiofacial olefin coordination is
difficult to control due to the comparable steric size of the
alkyl and aryl substituent at the olefinic C atom. The second
reason is that the terminal double bond can undergo isomeri-
zation under hydrogenation conditions to produce the more
stable internal trans-alkene, whose hydrogenation leads to the
predominant formation of the opposite enantiomer of the
product. The results under optimized conditions are shown in
Table 6.
The stereochemical outcome in the reduction of these tri-
substituted olefins can be easily rationalized by using a quad-
rant diagram based on the optimized DFT calculated structures
of the transition states (Figure 6). In this quadrant model we
found that the thioether substituent blocks the upper-left
quadrant and one of the P-aryl groups partly occupies the
lower-right quadrant making it semi-hindered. The other two
quadrants, which are free from bulky groups, are open. The
DFT structures thus show that the IrꢀPS catalysts generate
a pocket that is well suited to olefins with large trans substitu-
ents (E-olefins; Figure 6a). This fully explains the high enantio-
selectivities obtained with the DFT-optimized thioether-phos-
phinite ligands in the reductions of olefins S1–S3, S6–S9, S11,
and S12. However, the reduction of substrates S13–S16 gives
products with the opposite absolute configuration to what is
suggested by the quadrant model as previously observed for
a-substituted-a,b-unsaturated esters.[2d,32] On the other hand,
in the reduction of alkenylboronic ester S18, the bulky pinaco-
lato boron group (Bpin) faces the steric bulk of the ligand in
the semi-hindered lower-right quadrant. Thus the need to
switch to phosphite ligands L7a and L7c to obtain high enan-
tioselectivity could be justified by the flexibility of the biphenyl
phosphite moiety,[33] which could tune the steric hindrance of
this lower-right quadrant so that it can accommodate the pina-
colato boron substituent of the substrate.
Table 6. Ir-catalyzed hydrogenation of S19 using the P,S-ligand library
L1-L10a–g.[a]
Entry
Ligand
ee[b]
Entry
Ligand
ee[b]
[%]
[%]
1
2
3
4
5
6
7
8
L1a
L1b
L1c
L1d
L2b
L2c
L2e
L3a
L3b
L3c
L3e
L4a
L4b
L4c
L4d
L4e
L5a
L5e
L6a
46 (S)
70 (S)
82 (R)
64 (S)
88 (S)
74 (R)
81 (S)
31 (S)
93 (S)
93 (R)
75 (S)
62 (S)
60 (S)
92 (S)
87 (S)
80 (S)
64 (S)
76 (S)
94 (S)
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38[c]
L6b
L6c
95 (S)
94 (R)
93 (S)
96 (S)
30 (S)
78 (S)
82 (R)
76 (S)
72 (S)
54 (S)
66 (S)
90 (S)
30 (S)
64 (S)
76 (R)
69 (S)
97 (S)
97 (S)
97 (S)
L6d
L6e
L7a
L7b
L7c
L7d
L7e
L7 f
L8a
L8e
L9a
L9b
L9c
L9d
L10d
L10e
L10e
9
10
11
12
13
14
15
16
17
18
19
[a] Reactions carried out using 0.5 mmol of S19, 2 mol% of Ir-catalyst pre-
cursor, CH2Cl2 as solvent, 1 bar H2, 4 h. Full conversions were achieved in
all cases except for entries 9 and 10 (86 and 96% conversion, respective-
ly). [b] The ee values were determined by chiral GC. [c] Reaction carried
out using 0.25 mol% of Ir-catalyst precursor for 8 h.
Figure 6. Quadrant diagram describing the substrate–ligand interactions.
By using this quadrant model, we can also explain the
change in the sense of enantioselectivity observed experimen-
tally when using Z-trisubstituted olefins instead to E-olefins.
The Z-olefin must coordinate preferentially through the re face,
with the aryl substituent in the semi-hindered lower-right
quadrant and the hydrogen atom positioned in the hindered
upper-left quadrant (Figure 6b). This model also explains the
lower enantioselectivities when the optimized ligands were
used in the reduction of Z-olefins. The favorable chiral pocket
for E-olefins generated by our IrꢀPS catalysts, which can ac-
commodate large trans substituents, fails to perfectly control
the face coordination preference of the Z-olefins.
We were again able to fine-tune the ligand parameters to
achieve high activities and enantioselectivities (ee values up to
97%) in the reduction of this substrate at low catalyst loadings
(0.25 mol%) and hydrogen pressures (1 bar).
The results showed that the effect on enantioselectivity of
the thioether and the alkoxy substituents and the aryl-glycidol
group follow the same trend as for S1. However, in contrast to
S1, enantioselectivities for substrate S19 are similar for ligands
containing either an enantiopure biaryl phosphite moiety (b,c)
or a diaryl phosphinite group (d,e; i.e., Table 6, entries 20–23).
Interestingly, we found that the sense of enantioselectivity is
controlled by the configuration of the biaryl phosphite group
(Table 6, entries 20 and 21), and this represents an additional
To assess the potential of the ligand library L1–L10a–g for
the more challenging 1,1-disubstitued olefins, which generally
are hydrogenated less enantioselectively than the correspond-
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12209
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
possibility for the control of the absolute configuration of the
products through ligand modification.[34] As observed for S1,
the tropoisomerism in the fluxional biaryl phosphite group a is
not controlled by the ligand backbone, except for ligand back-
bone L6, containing an 2,6-dimethylphenyl thioether substitu-
ent, which provided similar high enantioselectivities as the
enantiopure phosphite counterparts (Table 6, entries 19 vs. 20
and 21).
We next turned our attention to study substrates with
neighboring polar groups (S22–S29), due to their importance
in the preparation of chiral synthons. The reduction of sub-
strates S22 and S23, containing trimethylsilyl and acetate
groups, respectively, provided moderate enantioselectivities
(up to 71%; Table 7, entries 7–12). To study whether these
enantioselectivities could be due again to their isomerization
to the trisubstituted internal olefins under the reaction condi-
tions, a decision was made to hydrogenate olefins containing
trifluoromethyl and boronate neighboring groups, which
cannot undergo isomerization (substrates S24 and S25). The
hydrogenation of substrate S24 proceeded with excellent
enantiocontrol (ee values up to 99%; Table 7, entries 13–15).[35]
These results are of interest because enantioenriched a-tri-
fluoromethyl chiral molecules are relevant building blocks for
the development of agrochemicals, pharmaceuticals, and ma-
terials owing to the unique properties of the fluorine atom.[36]
Interestingly, the reduction of alkenylboronic ester S25 also
provided high enantioselectivi-
We then investigated the scope of the new ligand library in
the asymmetric hydrogenation of other 1,1-disubstituted sub-
strates (Table 7). The results with substrates S19–S21 indicated
that enantioselectivity is affected by the alkyl chain substituent
(ee values ranging from 27 to 97%; Table 6, entries 36 and 37;
and Table 7, entries 3 and 6). This can be explained by the
competition between isomerization versus direct hydrogena-
tion for substrates S20 and S21. Accordingly, high amounts of
isomerized internal olefins were observed as byproducts in the
hydrogenation of S20 and S21.
ties (up to 91%, Table 7,
entry 18). Encouraged by this
Table 7. Selected results for the Ir-catalyzed hydrogenation of S20–S30 using the P,S-ligand library L1–L10a–
g.[a]
latter result we also tested other
challenging terminal boronic
esters S26–S29 (Table 7, en-
tries 19–30). Although these sub-
strates are also prone to isomeri-
zation they can be reduced with
acceptable values of enantiose-
lectivity (up to 84%). If we com-
pare these latter results with
those achieved by the only suc-
cessful report on this substrate
class using Ir-phosphinite-imida-
zoline ligands,[29b] we can con-
clude that the new P,S catalytic
systems overcome the limitation
of the Pfaltz ligands in the hy-
drogenation of S25 and S29, for
which poor enantioselectivities
were reported (ee values up to
4% for S25 and 33% for S29 at
ꢀ208C).[29b]
Entry
Substrate
Product
L
ee[b]
[%]
1
2
3
L6a
L6e
L10e
L6a
L6e
L10e
L7a
L7c
L10e
L7a
L7c
L10e
L1a
L1c
L8e
L3c
L10d
L3c
L1c
L3c
34 (S)[c]
54 (S)[d]
62 (S)[e]
16 (S)[f]
21 (S)[g]
27 (S)[h]
29 (R)
58 (R)
71 (R)
43 (R)
52 (R)
68 (R)
99 (ꢀ)
99 (ꢀ)
99 (ꢀ)
74 (S)
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
55 (S)
91 (S)[i]
72 (R)
74 (R)
28 (R)
L10d
Finally, we could also obtain
excellent enantioselectivity in
the hydrogenation of heteroaro-
matic alkene S30 (ee values up
to 96%, Table 7, entry 33). Sub-
strates containing heteroarao-
matic groups are popular in fine-
chemistry industries since the
heterocyclic part allows for fur-
ther functionalization.
22
23
24
25
26
27
28
29
30
31
32
33
L1c
L3c
L10d
L1c
L6c
L10e
L1c
L10d
L10e
L6b
L6c
76 (R)
81 (R)
19 (R)
76 (R)
77 (R)
62 (R)
76 (S)
75 (R)
84 (R)
95 (+)
94 (ꢀ)
96 (+)
L10e
[a] Reactions carried out using 0.5 mmol of substrate, 2 mol% of Ir-catalyst precursor, CH2Cl2 as solvent, 1 bar
H2, 4 h. Full conversions were achieved in all cases. [b] The ee values were determined by chiral GC or HPLC.
[c] 38% of isomerized S1 and 2% of S2. [d] 32% of isomerized S1. [e] 35% of isomerized S1. [f] 41% of tetra-
substituted olefin. [g] 32% of tetrasubstituted olefin. [h] 29% of tetrasubstituted olefin. [i] Reaction carried out
at ꢀ208C.
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12210
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Asymmetric hydrogenation using propylene carbonate as en-
vironmentally benign solvent: Recycling experiments
Conclusion
A modular ligand design, with the help of DFT studies, has
been shown to be highly successful in the identification and
tuning of the crucial stereodefining groups to generate more
selective catalysts. Following this approach, a library of modu-
larly constructed thioether-phosphinite/phosphite ligands de-
rived from the ring opening of enantiopure epoxides has been
evaluated in the asymmetric iridium-catalyzed hydrogenation
of a wide range of olefins. An extensive study on the influence
of the different structural parameters has been performed,
demonstrating the highly modular nature of these ligands.
Computations gave an understanding of the enantiocontrol in
the reaction allowing rationalization of the modifications re-
quired for improving selectivity. The computations moreover
indicated that the diastereoisomers resulting from coordination
of the thioether to the metal center interconvert rapidly under
the reaction conditions through pyramidal inversion, thus al-
lowing for the use of the Curtin–Hammet principle in predict-
ing the outcome of the reaction. In general, enantioselectivities
are mainly controlled by the nature of the thioether, the aryl
moieties and the type of P-donor group. However, the effect
of changing these modules depends on the substrate class.
The degree of activity and stereoinduction achieved with the
lead ligands were amongst the highest with respect to the
ones reported in the literature. The asymmetric hydrogena-
tions were also performed using propylene carbonate as sol-
vent, which allowed the Ir catalyst to be reused.
Finally, we focused our attention to replace the widely used di-
chloromethane solvent with propylene carbonate (PC) as an
environmentally benign solvent.[37] The use of PC as solvent
not only allows the hydrogenation to be performed in a more
sustainable way but also makes possible the recycling of the Ir
catalysts by a simple two-phase extraction.[38] Catalyst recycling
is desirable in large-scale processes due to the high cost of iri-
dium.
To assess whether the new IrꢀP,S catalysts could be em-
ployed by using PC as solvent, we screened the Ir-L10e cata-
lytic system in the hydrogenation of model substrates S1 and
S19 (Table 8). Although the reaction rates are lower in PC than
in dichloromethane, similar high enantioselectivities were
achieved (ee values up to 94% for S1 and 96% for S19). In ad-
dition, we were able to recycle the Ir catalysts up to 3 times
without any drop of enantioselectivity. As previously observed,
the reaction times necessary to achieve high conversions in-
creased.[38] This drop in activity could be attributed to the loss
of iridium catalyst to the hexane phase,[10k,38a] to the formation
of inactive iridium clusters,[39] or to both.
Table 8. Asymmetric hydrogenation using propylene carbonate using
catalyst precursor [Ir(cod)(L10e)][BArF]. Recycling experiments.[a]
Cycle
Substrate
% Conv.[b] (t [h])
ee[c] [%]
1[d]
2[d]
3[d]
1[e]
2[e]
3[e]
1[e]
S1
98 (6)
84 (10)
89 (15)
97 (4)
96 (8)
81 (10)
99 (6)[f]
94 (R)
94 (R)
93 (R)
95 (S)
96 (S)
95 (S)
72 (S)
Experimental Section
S19
General considerations
All reactions were carried out by using standard Schlenk tech-
niques under an argon atmosphere. Solvents were purified and
dried by standard procedures. Phosphorochloridites were easily
prepared in one step from the corresponding biphenols.[40] Inter-
mediate compounds 1--2,[18] 3--8,[17] 10--13,[17] and 15;[17] and thio-
ether-phosphinite ligands L1d,[17] L4d,[17] L6L7d,[17] and L9d[17]
S20
[a] Reactions carried out using 0.5 mmol of substrate and 2 mol% of Ir-
catalyst precursor. [b] Conversion measured from the 1H NMR spectrum
for substrate S1 or by GC for substrates S19 and S20. [c] The ee values
were determined by chiral HPLC (substrate S1) or GC (substrates S19 and
S20). [d] Reaction carried out at 125 bar. [e] Reaction carried out at
50 bar. [f] 18% of S1 observed.
were prepared as previously reported. H, 13C, and 31P NMR spectra
1
were recorded using a 400 MHz spectrometer. Chemical shifts are
relative to that of SiMe4 (1H and 13C) as internal standard or H3PO4
1
(31P) as external standard. H, 13C and 31P assignments were made
1
1
1
on the basis of H-1H gCOSY, H-13C gHSQC and H-31P gHMBC ex-
Another important feature of using PC as a solvent in the
asymmetric hydrogenation using Ir-P/N catalytic systems, ob-
served by Bçrner et al., is that the rate of isomerization of ter-
minal olefins to the corresponding trisubstituted ones dimin-
ishes compared with when dichloromethane is used. This be-
havior was exploited to improve enantioselectivity in the re-
duction of 1-methylene-1,2,3,4-tetrahydronaphthalene, which
easily isomerizes to form the trisubstituted olefin.[10k,38a] We
therefore also performed the asymmetric hydrogenation of
substrate S20 with Ir-L10e using PC as solvent. We were
pleased to find that the amount of isomerized trisubstituted
substrate substantially diminished, and that the enantioselec-
tivity consequently improved (ee values up to 72%, compared
with 62% in dichloromethane).
periments.
Computational details
Geometries of all transition states were optimized using the Gaus-
sian 09 program,[23] employing the B3LYP[20] density functional and
the LANL2DZ[21d] basis set for iridium and the 6–31G*[21a–c] basis set
for all other elements. Solvation correction was applied in the
course of the optimizations using the PCM model with the default
parameters for dichloromethane.[22] The complexes were treated
with charge +1 and in the singlet state. No symmetry constraints
were applied. Normal mode analysis of all transition states revealed
a single imaginary mode corresponding to the expected hydride
transfer or s-bond metathesis. In the case of hydride transfer, con-
comitant cleavage of the dihydrogen ligand was observed. The en-
ergies were further refined by performing single point calculations
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12211
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
using the abovementioned parameters, with the exception that
the 6–311+G**[24] basis set was used for all elements except iridi-
um, and by applying dispersion correction using the DFT-D3[25]
model. All energies reported are Gibbs free energies at 298.15 K
and calculated as Greported =G6-31G* +E6-311+G**ꢀE6-31G* +EDFT-D3
Typical procedure for the hydrogenation of olefins
The alkene (0.5 mmol) and Ir complex (2 mol%) were dissolved in
CH2Cl2 (2 mL) in a high-pressure autoclave, which was purged four
times with hydrogen. Then, it was pressurized at the desired pres-
sure. After the desired reaction time, the autoclave was depressur-
ized and the solvent evaporated off. The residue was dissolved in
Et2O (1.5 mL) and filtered through a short Celite plug. The enantio-
meric excess was determined by chiral GC or chiral HPLC and con-
versions were determined by 1H NMR spectroscopy. The enantio-
meric excesses of hydrogenated products from S1,[14a] S2,[41] S3
and S4,[14a] S5,[42] S6,[14a] S7–S9,[10m] S10,[43] S11 and S12,[14a] S13–
S15,[10j] S16,[28] S17,[29a] S18,[29b] S19,[14a] S20,[10 g] S21,[14a] S22,[43]
S23,[44] S24,[10k] S25–S29,[29b] and S30[14a] were determined by using
the conditions previously described.
General procedure for the preparation of thioether-alcohols
9, 14, and 16
To a suspension of the desired chiral epoxide (1.34 mmol) and
sodium hydroxide (107 mg, 2.68 mmol, 2 equiv) in dioxane/water
(10:1 v/v, 6.6 mL) was added the corresponding thiol (2.68 mmol,
2 equiv). The mixture was heated for 4 h at 908C. The reaction was
monitored by TLC until disappearance of the starting epoxide. The
mixture was left to reach RT and then water (15 mL) was added.
The mixture was extracted with CH2Cl2 (3ꢄ15 mL). The combined
organic extracts were washed with brine, dried with Na2SO4, and
filtered. The solvent was removed under vacuum and the crude
was purified by flash chromatography on SiO2 to produce the de-
sired thioether-alcohol as a white solid.
Typical procedure for reutilization of catalysts by using PC
as a solvent
After each catalytic experiment, the autoclave was depressurized.
We then extracted the colorless propylene carbonate solution with
dry/deoxygenated hexane under argon atmosphere with the aim
to remove the remaining substrate and the hydrogenated olefin.
After the extractions, the corresponding amount of substrate
(0.5 mmol) was then added and a new catalytic experiment was
started.
General procedure for the preparation of the thioether-
phosphite ligands L1–L9a–c
The corresponding phosphorochloridite (0.55 mmol) produced in
situ was dissolved in toluene (2.5 mL), and pyridine (0.15 mL,
2.9 mmol) was added. The corresponding thioether-hydroxyl com-
pound (0.5 mmol) was azeotropically dried with toluene (3ꢄ2 mL)
and then dissolved in toluene (2.5 mL) to which pyridine (0.15 mL,
2.9 mmol) was added. The alcohol solution was then transferred
slowly to the phosphorochloridite solution. The reaction mixture
was stirred at 808C for 90 min, after which the pyridine salts were
removed by filtration. Evaporation of the solvent gave a white
foam, which was purified by flash chromatography on alumina (tol-
uene/NEt3 =100:1) to produce the corresponding ligand as a white
solid.
Acknowledgements
X.C., E.K., and M.A.P. thank MINECO (grants CTQ2008-00947/
BQU and CTQ2012-38594-C02-01), Generalitat de Catalunya
(grant 2009SGR623), EU-ITN network Mag(net)icFun (PITN-GA-
2012-290248), and the ICIQ Foundation for financial support.
J.M., M.C., J.M., O.P., and M.D. thank the Spanish Government
for providing grant CTQ2010-15835, the Catalan Government
for grant 2009SGR116, and the ICREA Foundation for providing
M.D. and O.P. with financial support through the ICREA Aca-
demia awards.
General procedure for the preparation of the thioether-
phosphinite ligands L1–L10d–g
The corresponding thioether-hydroxyl compound (0.5 mmol) and
4-dimethylaminopyridine (DMAP; 6.7 mg, 0.055 mmol) were dis-
solved in toluene (1 mL), and triethylamine was added (0.09 mL,
0.65 mmol) at RT, followed by the addition of the corresponding
chlorophosphine (0.55 mmol) through a syringe. The reaction was
stirred for 20 min at RT. The solvent was removed in vacuo, and
the product was purified by flash chromatography on alumina (tol-
uene/NEt3 =100:1) to produce the corresponding ligand as an oil.
Keywords: asymmetric synthesis · density functional theory ·
hydrogenation · iridium · ligands
[1] a) R. Noyori, Asymmetric Catalysis in Organic Synthesis, Wiley, New York,
1994; b) E. N. Jacobsen, A. Pfaltz, H. Yamamoto, Comprehensive Asym-
metric Catalysis, Vol. 1, Springer, Berlin, 1999; c) I. Ojima, Catalytic Asym-
metric Synthesis, 2ednd edWiley-VCH, Weinheim, 2000; d) H.-U. Blaser, E.
Schmidt, Asymmetric Catalysis on Industrial Scale, Wiley, New York,
2004; e) C. A. Busacca, D. R. Fandrick, J. J. Song, C. H. Senanayakl, Adv.
Synth. Catal. 2011, 353, 1825–1864.
General procedure for the preparation of [Ir(cod)(L)][BArF]
(L=L1–L10a–g)
[2] a) J. M. Brown, in Comprehensive Asymmetric Catalysis, Vol. 1 (Eds.: E. N.
Jacobsen, A. Pfaltz, H. Yamamoto), Springer, Berlin, 1999; b) W. Tang, X.
Zhang, Chem. Rev. 2003, 103, 3029–3069; c) J.-H. Xie, S.-F. Zhu, Q.-L.
Zhou, Chem. Rev. 2011, 111, 1713–1760; d) J. J. Verendel, O. Pꢁmies, M.
Diꢂguez, P. G. Andersson, Chem. Rev. 2014, 114, 2130–2169.
[3] a) W. S. Knowles, M. J. Sabacky, B. D. Vineyard, J. Chem. Soc. Chem.
Commun. 1972, 10–11; b) W. S. Knowles, Angew. Chem. 2002, 114,
2096–2107; Angew. Chem. Int. Ed. 2002, 41, 1998–2007.
[4] R. Noyori, Angew. Chem. 2002, 114, 2108–2123; Angew. Chem. Int. Ed.
2002, 41, 2008–2022.
[5] C. S. Shultz, S. W. Krska, Acc. Chem. Res. 2007, 40, 1320–1326.
[6] H.-U. Blaser, Adv. Synth. Catal. 2002, 344, 17–31.
The corresponding ligand (0.074 mmol) was dissolved in CH2Cl2
(5 mL) and [Ir(m-Cl)(cod)]2 (25.0 mg, 0.037 mmol) was added. The
reaction mixture was heated at reflux at 508C for 1 hour. After
5 min at room temperature, NaBArF (77.2 mg, 0.080 mmol) and
water (5 mL) were added and the reaction mixture was stirred vig-
orously for 30 min at room temperature. The phases were separat-
ed and the aqueous phase was extracted twice with CH2Cl2. The
combined organic phases were dried with MgSO4, filtered through
a plug of Celite and the solvent was evaporated to give the prod-
uct as a red-orange solid.
[7] a) X. Cui, K. Burgess, Chem. Rev. 2005, 105, 3272–3296; b) K. Kꢅllstrçm,
I. Munslow, P. G. Andersson, Chem. Eur. J. 2006, 12, 3194–3200; c) S. J.
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12212
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Roseblade, A. Pfaltz, Acc. Chem. Res. 2007, 40, 1402–1411; d) T. L.
Church, P. G. Andersson, Coord. Chem. Rev. 2008, 252, 513–531; e) O.
Pꢁmies, P. G. Andersson, M. Diꢂguez, Chem. Eur. J. 2010, 16, 14232–
14240; f) D. H. Woodmansee, A. Pfaltz, Chem. Commun. 2011, 47, 7912–
7916; g) Y. Zhu, K. Burgess, Acc. Chem. Res. 2012, 45, 1623–1636.
[8] R. H. Crabtree, Acc. Chem. Res. 1979, 12, 331–337.
2628–2644; b) X. Caldentey, X. C. Cambeiro, M. A. Pericꢁs, Tetrahedron
2011, 67, 4161–4168.
[18] a) Y. Gao, R. M. Hanson, J. M. Klunder, S. Y. Ko, H. Masamune, K. B. Sharp-
less, J. Am. Chem. Soc. 1987, 109, 5765–5780; b) E. Medina, A. Moyano,
M. A. Pericꢁs, A. Riera, Helv. Chim. Acta 2000, 83, 972–988; c) A. Vidal-
Ferran, A. Moyano, M. A. Pericꢁs, A. Riera, J. Org. Chem. 1997, 62, 4970–
4982.
[19] a) P. Brandt, C. Hedberg, P. G. Andersson, Chem. Eur. J. 2003, 9, 339–
347; b) Y. Fan, X. Cui, K. Burgess, M. B. Hall, J. Am. Chem. Soc. 2004, 126,
16688–16689; c) X. Cui, Y. Fan, M. B. Hall, K. Burgess, Chem. Eur. J. 2005,
11, 6859–6868; d) T. L. Church, T. Rasmussen, P. G. Andersson, Organo-
metallics 2010, 29, 6769–6781; e) K. H. Hopmann, A. Bayer, Organome-
tallics 2011, 30, 2483–2497; Pfaltz’s group has recently provided experi-
mental support for a catalytic pathway going through the IrIII/IrV spe-
cies, see: f) S. Gruber, A. Pfaltz, Angew. Chem. 2014, 126, 1927–1931;
Angew. Chem. Int. Ed. 2014, 53, 1896–1900.
[9] A. Lightfoot, P. Schnider, A. Pfaltz, Angew. Chem. 1998, 110, 3047–3050;
Angew. Chem. Int. Ed. 1998, 37, 2897–2899.
[10] a) J. Blankenstein, A. Pfaltz, Angew. Chem. 2001, 113, 4577–4579;
Angew. Chem. Int. Ed. 2001, 40, 4445–4447; b) D.-R. Hou, J. Reibenspies,
T. J. Colacot, K. Burgess, Chem. Eur. J. 2001, 7, 5391–5400; c) F. Menges,
A. Pfaltz, Adv. Synth. Catal. 2002, 344, 40–44; d) M. C. Perry, X. Cui, M. T.
Powell, D.-R. Hou, J. H. Reibenspies, K. Burgess, J. Am. Chem. Soc. 2003,
125, 113–123; e) W. Tang, W. Wang, X. Zhang, Angew. Chem. 2003, 115,
973–976; Angew. Chem. Int. Ed. 2003, 42, 943–946; f) D. Liu, W. Tang, X.
Zhang, Org. Lett. 2004, 6, 513–516; g) S. McIntyre, E. Hçrmann, F.
Menges, S. P. Smidt, A. Pfaltz, Adv. Synth. Catal. 2005, 347, 282–288;
h) M. Diꢂguez, J. Mazuela, O. Pꢁmies, J. J. Verendel, P. G. Andersson,
Chem. Commun. 2008, 33, 3888–3890; i) M. Diꢂguez, J. Mazuela, O.
Pꢁmies, J. J. Verendel, P. G. Andersson, J. Am. Chem. Soc. 2008, 130,
7208–7209; j) S.-M. Lu, C. Bolm, Angew. Chem. 2008, 120, 9052–9055;
Angew. Chem. Int. Ed. 2008, 47, 8920–8923; k) J. Mazuela, J. J. Verendel,
M. Coll, B. Schꢅffner, A. Bçrner, P. G. Andersson, O. Pꢁmies, M. Diꢂguez,
J. Am. Chem. Soc. 2009, 131, 12344–12353; l) J. Zhao, K. Burgess, J. Am.
Chem. Soc. 2009, 131, 13236–13237; m) W.-J. Lu, Y.-W. Chen, X.-L. Hou,
Adv. Synth. Catal. 2010, 352, 103–107; n) Y. Zhang, Z. Han, F. Li, K. Ding,
A. Zhang, Chem. Commun. 2010, 46, 156–158; o) A. Franzke, A. Pfaltz,
Chem. Eur. J. 2011, 17, 4131–4144; p) J. Mazuela, P.-O. Norrby, P. G. An-
dersson, O. Pꢁmies, M. Diꢂguez, J. Am. Chem. Soc. 2011, 133, 13634–
13645; q) J. Shang, Z. Han, Y. Li, Z. Wang, K. Ding, Chem. Commun.
2012, 48, 5172–5174; r) X. Wang, Z. Han, Z. Wang, K. Ding, Angew.
Chem. 2012, 124, 960–964; Angew. Chem. Int. Ed. 2012, 51, 936–940.
[11] a) T. Bunlaksananusorn, K. Polborn, P. Knochel, Angew. Chem. 2003, 115,
4071–4073; Angew. Chem. Int. Ed. 2003, 42, 3941–3943; b) W. J. Drur-
y III, N. Zimmermann, M. Keenan, M. Hayashi, S. Kaiser, R. Goddard, A.
Pfaltz, Angew. Chem. 2004, 116, 72–76; Angew. Chem. Int. Ed. 2004, 43,
70–74; c) S. Bell, B. Wꢆstenberg, S. Kaiser, F. Menges, T. Netscher, A.
Pfaltz, Science 2006, 311, 642–644; d) S. Kaiser, S. P. Smidt, A. Pfaltz,
Angew. Chem. 2006, 118, 5318–5321; Angew. Chem. Int. Ed. 2006, 45,
5194–5197; e) J. Margalef, M. Lega, F. Ruffo, O. Pꢁmies, M. Diꢂguez, Tet-
rahedron: Asymmetry 2012, 23, 945–951; f) D. H. Woodmansee, M.-A.
Mꢆller, L. Trçndlin, E. Hçrmann, A. Pfaltz, Chem. Eur. J. 2012, 18, 13780–
13786; g) J. Mazuela, O. Pꢁmies, M. Diꢂguez, Adv. Synth. Catal. 2013,
355, 2569–2583; h) A. Schumacher, M. Bernasconi, A. Pfaltz, Angew.
Chem. 2013, 125, 7570–7573; Angew. Chem. Int. Ed. 2013, 52, 7422–
7425.
[20] a) C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789; b) A. D.
Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[21] a) W. J. Hehre, R. Ditchfield, J. A. Pople, J. Chem. Phys. 1972, 56, 2257–
2261; b) P. C. Hariharan, J. A. Pople, Theor. Chim. Acta 1973, 28, 213–
222; c) M. M. Francl, W. J. Pietro, W. J. Hehre, J. S. Binkley, M. S. Gordon,
D. J. Defrees, J. A. Pople, J. Chem. Phys. 1982, 77, 3654–3665; d) P. J.
Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 299–310.
[22] a) S. Miertus, J. Tomasi, Chem. Phys. 1982, 65, 239–245; b) B. Mennucci,
J. Tomasi, J. Chem. Phys. 1997, 106, 5151–5158; c) M. Cossi, V. Barone,
B. Menucci, J. Tomasi, Chem. Phys. Lett. 1998, 286, 253–260.
[23] Gaussian 09, RevisionA.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, J. E. Peralta, Jr., F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, ꢉ. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[24] a) R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, J. Chem. Phys. 1980,
72, 650; b) A. D. McLean, G. S. Chandler, J. Chem. Phys. 1980, 72, 5639–
5648.
[25] a) S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys. 2010, 132,
154104; b) S. Grimme, S. Ehrlich, L. Goerigk, J. Comput. Chem. 2011, 32,
1456–1465.
[26] The additional isomers generated considering the hydride attack to the
less-substituted carbon of the olefin have not been considered, since it
has been previously demonstrated that they have substantially higher
energies. See ref. [19d].
[27] Unpublished results. In all cases, enantioselectivities were below 50%
ee.
[12] D. Rageot, D. H. Woodmansee, B. Pugin, A. Pfaltz, Angew. Chem. 2011,
123, 9772–9775; Angew. Chem. Int. Ed. 2011, 50, 9598–9601.
[13] a) C. Hedberg, K. Kꢅllstrçm, P. Brandt, L. K. Hansen, P. G. Andersson, J.
Am. Chem. Soc. 2006, 128, 2995–3001; b) P. Cheruku, A. Paptchikhine,
T. L. Church, P. G. Andersson, J. Am. Chem. Soc. 2009, 131, 8285–8289;
c) P. Tolstoy, M. Engman, A. Paptchikhine, J. Bergquist, T. L. Church,
A. W.-M. Leung, P. G. Andersson, J. Am. Chem. Soc. 2009, 131, 8855–
8860; d) J. J. Verendel, J.-Q. Li, X. Quan, B. Peters, T. Zhou, O. R. Gautun,
T. Govender, P. G. Andersson, Chem. Eur. J. 2012, 18, 6507–6513.
[14] a) K. Kꢅllstrçm, C. Hedberg, P. Brandt, A. Bayer, P. G. Andersson, J. Am.
Chem. Soc. 2004, 126, 14308–14309; b) J. Mazuela, A. Paptchikhine, O.
Pꢁmies, P. G. Andersson, M. Diꢂguez, Chem. Eur. J. 2010, 16, 4567–4576.
[15] a) M. Coll, O. Pꢁmies, M. Diꢂguez, Chem. Commun. 2011, 47, 9215–
9217; b) M. Coll, O. Pꢁmies, M. Diꢂguez, Adv. Synth. Catal. 2013, 355,
143–160.
[28] W.-J. Lu, X.-L. Hou, Adv. Synth. Catal. 2009, 351, 1224–1228.
[29] a) A. Paptchikhine, P. Cheruku, M. Engman, P. G. Andersson, Chem.
´
Commun. 2009, 5996–5998; b) A. Ganic, A. Pfaltz, Chem. Eur. J. 2012,
18, 6724–6728.
[30] a) K. Oertel, G. Zech, H. Kunz, Angew. Chem. 2000, 112, 1489–1491;
Angew. Chem. Int. Ed. 2000, 39, 1431–1433; b) A. Fꢆrstner, M. Bindl, L.
Jean, Angew. Chem. 2007, 119, 9435–9438; Angew. Chem. Int. Ed. 2007,
46, 9275–9278; c) J. Li, J. Dai, X. Chen, P. Zhu, J. Nat. Prod. 2007, 70,
1846–1849; d) A. M. McGhee, J.-C. Kizirian, D. J. Procter, Org. Biomol.
Chem. 2007, 5, 1021–1024.
[16] a) A. M. Masdeu-Bultꢇ, M. Diꢂguez, E. Martin, M. Gꢇmez, Coord. Chem.
Rev. 2003, 242, 159–201; b) E. Martin, M. Diꢂguez, C. R. Chim. 2007, 10,
188–205; c) M. Mellah, A. Voituriez, E. Schulz, Chem. Rev. 2007, 107,
5133–5209; d) H. Pellissier, Tetrahedron 2007, 63, 1297–1330; e) R. Mal-
acea, E. Manoury, in Phosphorus Ligands in Asymmetric Catalysis, Vol. 2
(Ed.: A. Bçrner), Wiley-VCH, Weinheim, 2008, pp. 749–784; f) R. Gꢇme-
z Arrayꢈs, J. C. Carretero, Chem. Commun. 2011, 47, 2207–2211.
[31] a) G. W. Kabalka, J. Organomet. Chem. 1984, 274, 1–29; b) G. W. Kabalka,
R. C. Marks, J. Organomet. Chem. 1993, 457, 25–40; c) J. B. Morgan, J. P.
Morken, J. Am. Chem. Soc. 2004, 126, 15338–15339; d) W. J. Moran, J. P.
Morken, Org. Lett. 2006, 8, 2413–2415.
[32] This behavior has been explained by the strong polarization of the
double bond in these a,b-unsaturated compounds. For a,b-disubstitut-
ed substrates the addition to the b-C is hampered by steric interactions
caused by the olefin tilting toward the ligand bulk, whereas addition to
the a-C is electronically disfavored. Since the quadrant model relies on
[17] Thioether-phosphinite ligands L1d, L4d, L6d, L7d, and L9d have been
already prepared: a) X. Caldentey, M. A. Pericꢁs, J. Org. Chem. 2010, 75,
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12213
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
steric effects, the failure of the model for a,b-disubstituted substrates
indicates that electronic factors dominate. See, also: J.-Q. Li, X. Quan,
P. G. Andersson, Chem. Eur. J. 2012, 18, 10609–10616.
[38] a) J. Bayardon, J. Holz, B. Schꢅffner, V. Andrushko, S. Verevkin, A. Preetz,
A. Bçrner, Angew. Chem. 2007, 119, 6075–6078; Angew. Chem. Int. Ed.
2007, 46, 5971–5974; b) B. Schꢅffner, F. Schꢅffner, S. P. Verevkin, A.
Bçrner, Chem. Rev. 2010, 110, 4554–4581.
[39] S. P. Smidt, A. Pfaltz, E. Martꢊnez-Viviente, P. S. Pregosin, A. Albinati, Or-
ganometallics 2003, 22, 1000–1009.
[40] G. J. H. Buisman, P. C. J. Kamer, P. W. N. M. van Leeuwen, Tetrahedron:
Asymmetry 1993, 4, 1625–1634.
[41] T. Ohta, H. Ikegami, T. Miyake, H. Takaya, J. Organomet. Chem. 1995, 502,
169–176.
[33] The flexibility of the biphenyl moiety has been demonstrated to be cru-
cial in the success of ligands containing this moiety in several catalytic
reactions. For a recent review, see: P. W. N. M. van Leeuwen, P. C. Kamer,
C. Claver, O. Pꢁmies, M. Diꢂguez, Chem. Rev. 2011, 111, 2077–2118.
[34] The same can be achieved with thioether-phosphinite ligands by
changing the absolute configuration of the starting enantiopure epox-
ide.
[35] Unpublished results. The use of previously reported furanoside-based
PꢀS ligands led to moderate enantioselectivities.
[42] D. H. Woodmansee, M.-A. Mꢆller, M. Neuburger, A. Pfaltz, Chem. Sci.
2010, 1, 72–78.
[43] K. Kꢅllstrçm, I. J. Munslow, C. Hedberg, P. G. Andersson, Adv. Synth.
Catal. 2006, 348, 2575–2578.
[44] S. Deerenberg, P. C. J. Kamer, P. W. N. M. van Leeuwen, Organometallics
2000, 19, 2065–2072.
[36] a) M. Schlosser, Angew. Chem. 1998, 110, 1538–1556; Angew. Chem. Int.
Ed. 1998, 37, 1496–1513; b) P. V. Ramachandran, Asymmetric Fluoroor-
ganic Chemistry: Synthesis Application and Future Directions, Washing-
ton, 2000; c) B. E. Smart, J. Fluorine Chem. 2001, 109, 3–11; d) A special
issue has been devoted to Fluorine in the Life Sciences: ChemBioChem
2004, 5, 559–562.
[37] PC is a noncorrosive, nonhygroscopic, high boiling, nontoxic, and bio-
degradable solvent.
Received: April 6, 2014
Published online on August 5, 2014
Chem. Eur. J. 2014, 20, 12201 – 12214
www.chemeurj.org
12214
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim