The role of the aromatic ligand in the asymmetric transfer hydrogenation of the CN bond on Noyori's chiral Ru catalysts

Article abstract of DOI:10.1016/j.tetasy.2014.08.011

Only four types of dimeric precursors [RuCl₂(η⁶-arene)]₂ for the synthesis of Noyori's half sandwich diamine catalysts [RuCl(TsDPEN)(η⁶-arene)] are commercially available, yet so far no study has tried to system

Full text of DOI:10.1016/j.tetasy.2014.08.011

Tetrahedron: Asymmetry 25 (2014) 1346–1351
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
The role of the aromatic ligand in the asymmetric transfer
hydrogenation of the C@N bond on Noyori’s chiral Ru catalysts
a
a
a
a
b
b
a,
⇑
ˇ
ˇ
ˇ
Petr Šot , Beáta Vilhanová , Jan Pechácek , Jirí Václavík , Jakub Zápal , Marek Kuzma , Petr Kacer
a Department of Organic Technology, Institute of Chemical Technology, Technická 5, 166 28 Prague, Czech Republic
Laboratory of Molecular Structure Characterization, Institute of Microbiology, v.v.i., Academy of Sciences of the Czech Republic, Vídenská 1083, 142 20 Prague 4, Czech Republic
b
ˇ
a r t i c l e i n f o
a b s t r a c t
Only four types of dimeric precursors [RuCl2(g
6-arene)]2 for the synthesis of Noyori’s half sandwich
Article history:
Received 29 July 2014
Accepted 20 August 2014
diamine catalysts [RuCl(TsDPEN)(g
6-arene)] are commercially available, yet so far no study has tried to
systematically evaluate how these systems perform during the asymmetric transfer hydrogenation of
various 3,4-dihydroisoquinolines (i.e., the typical substrates for Noyori asymmetric transfer hydrogena-
tion benchmarking). Experiments combined with molecular modeling allowed us to assess their proper-
ties and formulate a hypothesis clarifying the difference in enantioselectivity of these systems.
Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
the corresponding dimeric precursor of the [RuCl2(g
6-arene)]2 type
is necessary, which involves Birch reduction or [4+2] cycloaddition
Asymmetric catalysis represents a method for preparing enanti-
omerically pure compounds, which play a highly important role for
example in the pharmaceutical industry. Noyori’s (also known as
Noyori–Ikariya) catalytic complexes are well established for the
synthesis of chiral amines and alcohols via asymmetric transfer
hydrogenation under relatively mild conditions (room tempera-
ture, atmospheric pressure) and with high asymmetric induction.1,2
Results published by Xue and Deng indicate that these complexes
can hydrogenate even activated olefins.3 The catalytic complex con-
sists of chiral N-(p-toluenesulfonyl)-1,2-diphenylethylenediamine
and subsequent reaction with ruthenium(III) chloride.5
Generally speaking, the enantioselective course of the reaction
can be viewed as a direct consequence of sophisticated interplay
between stabilizing and destabilizing interactions, which results
in a split of the energy levels of the transition states for the Re
and Si pathways. In this Noyori–Ikariya class of half-sandwich cat-
alysts, one of the major contributors to the aforementioned ener-
getic split is the CH/
CH/ interaction) between the molecule of the substrate and the
6-coordinated arene ligand. According to a computational study
p weak hydrogen bond (also known as the
p
g
(TsDPEN) and a
g
6-arene coordinated to the ruthenium central
by Yamakawa and Noyori6 benzene ligand interacts with the aro-
matic part of the benzaldehyde and can provide stabilization up
to 8.6 kJ/mol. If benzene is substituted by a methyl group, stabiliza-
tion can reach a value of 12.3 kJ/mol. With these results in mind it
is reasonable to anticipate different enantioselectivities for differ-
ently substituted benzenes. This hypothesis has been to some
extent supported by experiments of Yamada7 and Takehara8 with
carbonyl-bearing substrates, yet it is difficult to transfer these find-
ings to the asymmetric transfer hydrogenation of the C@N bond
due to many fundamental differences between the hydrogenation
of imines and ketones (e.g., reaction mechanism, structure of tran-
sition states, acidity of reaction medium, structure and electronic
effects of substrates, etc.).
atom. Many structural modifications of the aforementioned
catalyst have been developed, such as ligands differing in the aryl-
sulfonyl group, alkylated amino group, and substituted aromatic
rings within the 1,2-diphenylethylendiamine (DPEN) fragment.4
Herein we have focused on the modification of the
arene ligand.
g
6-coordinated
Simple alkyl-substituted benzene is usually employed as the
g
6-arene. Currently, only two such catalysts are commercially
available, that is, [RuCl(TsDPEN)(
6-mesitylene)]. Several other dimeric precursors are also
available, such as [RuCl2(
6-hexamethylbenzene)]2 and [RuCl2
6-benzene)]2, which must be further reacted with the TsDPEN
g
6-p-cymene)] and [RuCl(TsDPEN)
(g
g
(g
ligand to obtain the catalyst. These four catalysts thus belong to
the most frequently used systems. For the preparation of a catalyst
with an alternatively substituted aromatic ligand, the synthesis of
These structural effects remain unexplored in the case of the
asymmetric transfer hydrogenation of imines. To the best of our
knowledge, there is currently no such study of the [RuCl(TsDPEN)
(g
6-arene)] catalytic complex during imine hydrogenation.
Herein we report the effects of four alkyl-substituted
g
6-arene
⇑
Corresponding author. Tel.: +420 220 444 158; fax: +420 220 444 340.
ligands on the asymmetric transfer hydrogenation of six
ˇ
E-mail address: petr.kacer@vscht.cz (P. Kacer).
http://dx.doi.org/10.1016/j.tetasy.2014.08.011
0957-4166/Ó 2014 Elsevier Ltd. All rights reserved.

P. Šot et al. / Tetrahedron: Asymmetry 25 (2014) 1346–1351
1347
3,4-dihydroisoquinolines. A better understanding of the ligand
effects should allow us both to select a suitable commercially
available catalytic system and to optimize the structure of the
catalyst for each of the tested substrates to achieve optimal
performances in asymmetric transfer hydrogenations.
according to Scheme 1 and all parameters (temperature, substrate-
to-catalyst ratio, HCOOH-to-triethylamine ratio, etc.) were kept
constant (see Scheme 2).
Over the course of the reaction, the conversion was measured in
order to calculate the reaction rates. Substrates 1a and 1b represent
the most frequently used substrates with a strong divergence in
their electron densities; in their case we carried out a detailed mon-
itoring (see Graph 1 for kinetic curves and Table 1 for turnover fre-
quencies). For all alkyl-substituted substrates the conversions after
70 min are available (see Graph 2). Phenyl-substituted molecules
1e and 1f are quite non-reactive and in their case the conversion
was measured after 24 h.17 Very fast progression of the reaction
was observed in the case of 2b; the reaction with 2c proceeded
slightly slower (for both substrates). The slowest progress of the
reaction was apparent in the case of catalyst 2d; after 24 h, rela-
tively low conversion was observed (53% for 1b and 38% for 1a),
but fortunately high enough to determine enantioselectivity (vide
infra). Catalyst 2a was slightly less active than 2b and 2c. Generally
speaking, there is only little difference between the reaction rates of
the asymmetric transfer hydrogenation of substrates 1a and 1b on
2b and 2c. Substrate 1b is more electron rich due to the presence of
two methoxy groups, and thus it can be attracted to the proximity
of the active site (via hydrogen bonds between the oxygen atom of
the methoxy group and the hydrogen atom within the amino group
of the catalyst). This could explain its slightly higher reactivity over
1a. A mathematical model for the description of a kinetic profile of
analogous reaction for ketones (on the tethered variant of the
catalyst) was described in 2009 by Wills et al.18
2. Results and discussion
In order to systematically evaluate the effects of the aromatic
ligand, we performed kinetic measurements of the asymmetric
transfer hydrogenation of 3,4-dihydroisoquinolines and subse-
quently determined the enantiomeric purity of the products, that
is, chiral 1,2,3,4-tetrahydroisoquinolines. We used six imine
substrates: 1-methyl-3,4-dihydroisoquinoline 1a, 6,7-dimethoxy-
1-methyl-3,4-dihydroisoquinoline 1b, 6-methoxy-1-methyl-3,
4-dihydroisoquinoline 1c, 7-methoxy-1-methyl-3,4-dihydroiso-
quinoline 1d, 1-phenyl-3,4-dihydroisoquinoline 1e, and 1-(4-tri-
fluoromethylphenyl)-3,4-dihydroisoquinoline
1f,
and
four
catalysts differing by their aromatic ligands (benzene, mesitylene,
p-cymene, and hexamethylbenzene). The reactions were performed
R1
R2
R1
R2
Ru cat.
N
HCOOH/TEA, CH3CN
NH
*
R3
R3
1a: R1 = H, R2 = H, R3 = Me
1d: R1 = H, R2 = OMe, R3 = Me
1b: R1 = OMe, R2 = OMe, R3 = Me
R
1 = H, R2 = H, R3 = Ph
1f: R1 = H, R2 = H, R3 = 4-CF3-Ph
1e:
1 = OMe, R2 = H, R3 = Me
1c:
R
Table 1
Turnover frequencies for performed asymmetric reductionsa
Scheme 1. General scheme of dihydroisoquinoline reduction.
Substrate/catalyst
TOF [hꢀ1
]
1a
1b
2a
2b
2c
2d
107
224
139
115
232
156
Rn
O
O
2a:
arene = benzene
2b: arene = mesitylene
2c:
b
b
S
—
—
Cl
Ru
H2N
N
arene = p-cymene
2d: arene = hexamethylbenzene
a
Based on linear regression of the linear part of the kinetic curve.
Not calculated because of low conversion.
b
Ph
Ph
The enantioselectivity of the reaction is reported as enantio-
meric excess (ee) and was found constant during the reduction
Scheme 2. Structure of the catalysts.
100
75
50
25
0
0
10
20
30
40
50
60
70
80
time [min]
1a-2a
1b-2a
1a-2b
1b-2b
1a-2c
1b-2c
1a-2d
1b-2d
Graph 1. Kinetic profiles of dihydroisoquinolines 1a, 1b reductions using catalysts 2a–2d.

1348
P. Šot et al. / Tetrahedron: Asymmetry 25 (2014) 1346–1351
100
75
50
25
0
substrate-catalyst combination
conversion
enantiomeric excess
Graph 2. Conversions and enantiomeric excesses after 70 min.
chloroform, larger amounts could be dissolved only in DMF or
Table 2
Enantiomeric excesses determined after 70 mina
DMSO) and even in protic solvents (small amounts could be dis-
solved in methanol), contrary to the rest of the tested catalysts
2b, 2c, and 2d, which all exhibited very good solubility in most
polar aprotic and protic solvents. Even the solubility of the dimeric
precursor of 2a was very low in most solvents.
We employed the methods of molecular modeling available
within the Gaussian 09 framework with the primary aim of
describing effects, which could help us to understand the experi-
mentally observed phenomena described above. A detailed compu-
tational study of the hydrogenation mechanism is beyond the
scope of this paper and corresponding investigations have already
been carried out previously.21–23
Substrate/catalyst
ee [%]
1a
1b
1c
1d
1eb
1fb
2a
2b
2c
2d
58
83
81
85
75
92
93
82
74
91
92
72
71
87
88
58
2
9
—
—
—
—
—
—
a
Enantioselectivity was determined by gas chromatography after derivatization
of the samples with (R)-(ꢀ)-menthyl chloroformate, which we described in our
recent paper.20
b
Enantiomeric excesses measured only for catalyst 2a; in the case of 2b–2d the
conversion was <1%.
Optimized geometries of 2a–2d were compared with the aim of
locating significant differences in interatomic distances and angles.
The bond length between ruthenium and amidic nitrogen (Ru-N-
Ts) was the longest within 2d (2.168 Å) and the shortest in 2a
(2.145 Å). Such a difference could be related to the charge on the
ruthenium atom, which is strongly influenced by the structure of
the aromatic ligand.24 Charges on the amino groups of TsDPEN
exhibited only small differences.
(Table 2). In the case of substrate 1a, the highest ee was achieved
with 2d. Catalysts 2b and 2c displayed comparable enantioselec-
tivity. Dramatically lower ee was achieved with 2a. With regard
to substrate 1b, the enantioselectivity was generally higher and a
different trend was observed: the ee decreased in the order
2c > 2b > 2d > 2a. The hydrogenation of substrates 1c and 1d led
to quite similar results: catalysts 2c and 2b displayed the highest
enantioselectivity, catalyst 2a the second highest and catalyst 2d
the lowest. In the case of substrate 1d, catalyst 2d performed sig-
nificantly worse than in the case of 1c. Since the enantioselectivity
is dependent on the difference between the free energies of the
diastereomeric transition states (thermodynamically favored and
disfavored), the strong differences in ee suggest that the alkyl-sub-
stitution of the aromatic ligand can affect the differences of the free
energies via steric and electronic effects.
The extremely low enantiomeric excess in the case of 1e and 1f
is noteworthy. This might be caused by the relatively equal distri-
bution of the electrons between both aromatic rings and a nearly
symmetrical structure (substrate 1e would attain mirror symmetry
if a –CH2–CH2– fragment was present between the imine nitrogen
and the 1-phenyl ring at the ortho-position), which could result in a
very low difference between the Re and Si pathways. The phenyl
ring of substrate 1f is partly deactivated by the CF3 group, thus
the symmetry is broken and a slightly higher ee was achieved.19
The substitution of the aromatic ligand can also dramatically
affect the solubility of the catalyst. Catalytic complex 2a was prac-
tically insoluble in most polar aprotic solvents (nearly insoluble in
acetonitrile, dichloromethane, and ethyl acetate, slightly soluble in
It is known that the coordination of an arene to a metal slightly
distorts its structure. The carbonAcarbon bonds within the aro-
matic ligand are slightly elongated and the whole molecule exhib-
its noticeable deviation from its usual planarity. The acidity of the
hydrogen atoms within the aromatic ligand is increased due to the
g
6-coordination. The aforementioned structural effect was mani-
fested noticeably in the case of 2d, where the g
6-hexamethylben-
zene ligand has an average CAC bond length of 1.510 Å and
deviates from planarity by 8.95°, while non-bonded hexamethyl-
benzene has an average bond length of 1.410 Å and is strictly
planar (maximum deviation is 3.51°). On the other hand, these
effects are the least noticeable in the case of 2a, where the benzene
ligand has an average length of the CAC bond 1.420 Å and deviates
from planarity by up to 6.22° (non-bonded benzene has the CAC
bond length of 1.396 Å and practically no deviation from planar-
ity). Our computations also showed that the more substituted
g
6-arene, the bigger the distance between the ruthenium and the
center of the
g
6-arene: 2a (1.689 Å) < 2b (1.692 Å) < 2c
(1.699 Å) < 2d (1.703 Å).
In the next stage, we investigated the molecular phenomena
within the transition states. Our calculations were primarily
focused on substrate 1a because of its structural simplicity. The

P. Šot et al. / Tetrahedron: Asymmetry 25 (2014) 1346–1351
1349
comparison of favorable transition states revealed that in the case
protonated (N–H distance: 1.070 Å) by formic acid, substrate 1e
merely forms a hydrogen bond (N–H distance: 1.568 Å). Naturally,
this raises the question as to why catalyst 2a can hydrogenate the
aforementioned substrates. Over the course of our experiments, we
discovered that part of catalyst 2a slowly decomposes (i.e., decoor-
dinated benzene appears after some time in 1H NMR spectra,
corresponding spectra can be provided upon request). In theory,
the degradation products could facilitate substrate activation (by
providing free sites for the coordination of imine to ruthenium)
and thus allow the imine to undergo reduction; however, this deg-
radation is very slow (2% after 12 days), so its influence remains
arguable. Other reasonable possibilities include structure-caused
differences in hydride donating power, participation of molecules
of solvent in the reaction mechanism, or influence of solvent on
the acidic/basic properties of the principal reaction components.
of 1a–2b, two CH/p interactions can be found; one from a hydro-
gen connected to the sp2 carbon (2.740 Å 76C–19H, 2.914 Å 73C–
19H) and a second one from a hydrogen within the methyl group
(2.646 Å 61H–74C, 3.005 Å 61H–77C). This possibility of stabiliza-
tion could directly explain higher enantiomeric excess attained
with 2b, 2c, and 2d (see Fig. 1).
3. Conclusion
In conclusion, we have clearly demonstrated that the structure
of the aromatic ligand strongly influences the reaction rate and
enantioselectivity of the asymmetric transfer hydrogenation of
imines. The nonlinear dependence of the reaction rate and enanti-
oselectivity upon increasing substitution of the aromatic ligand
suggests the existence of a multiparametric relation, although it
is difficult to identify all of those parameters since the mechanism
is not yet fully understood. Thus, the role of the parameters con-
nected with the regeneration or stabilization of the cationic species
requires further investigation. Our DFT study showed that the
number of possible stabilizing interactions (which might be higher
in the case of catalysts 2b, 2c, and 2d) and the length of hydride
transfer trajectory (which strongly increases with increasing sub-
stitution of the aromatic ligand) most likely represent two such
determining parameters.
4. Experimental
4.1. General
Figure 1. Favorable transition state 1a–2b with highlighted important non-
covalent interactions and DFT single-point energy. Unimportant hydrogen atoms
are omitted for clarity.
All substances with the exception of 1a–1f, 2a, and 2d
employed in this work were obtained from commercial sources
and used without further purification. Compounds 1a–1f were pre-
pared according to previously reported procedures.9,10 The cited
articles also contain the corresponding characterization data (all
substrates belong to the batches characterized in these papers).
Compounds 2a and 2d were prepared according to a reported pro-
cedure,11 which was slightly modified.
High resolution mass spectrometry measurements were carried
out on a LTQ Orbitrap Velos (Thermo Fisher Scientific, USA) spec-
trometer. Heated electrospray ionization in positive (HESI+) was
used as an ion source. Full-scan mass spectra (range of m/
z = 100–2000 Da at a resolution of 30,000) were acquired using a
Our approximate model lacks a counterion in order to allow for
faster convergence of the computations. On the other hand this
partly complicates further efforts to provide more detailed investi-
gations of the electronic effects, because the catalyst then under-
goes transformation from a neutral species to a cationic one. The
distance between the ruthenium and the C@N carbon increases
with increasing bulkiness of the aromatic ligand; 2a (3.246 Å) <
2c (3.277 Å) < 2b (3.297 Å) < 2d (3.339 Å). The unexpectedly lower
ee value in the case of the 1b–2d combination might be due to the
possible repulsion between the methoxy groups of the substrate
and methyl groups of the hexamethylbenzene ligand.
The structure of the aromatic ligand also affects its rotational
freedom. The rotation of the p-cymene ligand is strongly restricted
by the bulky isopropyl group, which causes significant steric hin-
drance in the proximity of the sulfonyl group. In contrast, highly
symmetric ligands such as benzene, mesitylene, and hexamethyl-
benzene can rotate freely and this rotation can allow the formation
continual infusion of the sample solutions (concentration: 1
lg/
mL, flow: 2 L/min). The conditions on the mass spectrometer
l
were optimized to the following values: spray voltage 3000 V
(HESI+), vaporizer temperature 40 °C, sheath gas pressure (nitro-
gen) 5 psi, aux gas pressure (nitrogen) 2 ArbU and capillary
temperature 270 °C. The data were acquired and processed using
the Thermo Xcalibur software, version 2.1.0 (Thermo Fisher Scien-
tific, USA).
NMR spectra were measured on a Bruker AVANCE III 600 MHz
spectrometer (600.23 MHz for 1H and 150.93 MHz for 13C) in,
DMSO-d6 at 303 K. The residual signal of the solvent was used as
an internal standard (DMSO-d6: dH 2.500 ppm, dC 39.60 ppm). 1H
NMR, 13C NMR, COSY, gHSQCAD, and gHMBCAD spectra were mea-
sured using standard manufacturers’ software (Varian Inc., Palo
of the CH/p interaction from all positions within the aromatic
ligand. The rotation is completely restricted in the case of
Wills’s25–27 and Ikariya’s28 tethered complexes, in which the arene
is connected (e.g., via an alkyl chain) to the amino group of the
TsDPEN ligand.
The sluggish reactivity of 1-aryl substituted dihydroisoquino-
lines (herein represented by 1e and 1f) is a well-known phenome-
non and constitutes a wide topic per se. We hypothesize that the
hydrogenation of 1-aryl dihydroisoquinolines is strongly hindered
by their weak activation. While substrate 1b appears to be fully

1350
P. Šot et al. / Tetrahedron: Asymmetry 25 (2014) 1346–1351
Alto, U.S.A. and Topspin 2.1, Bruker Biospin GmbH, Rheinstetten,
Germany). 1D NMR spectra were zero filled to fourfold data points
and multiplied by window function (for 1H NMR with two-param-
eter double-exponential Lorentz–Gauss function and for 13C NMR
line broadening 1 Hz was applied) before Fourier transformation
to improve resolution. Chemical shifts are given in d scale (ppm)
and coupling constants in Hz. Digital resolution enabled us to
report the chemical shifts of protons and carbons to 2 and coupling
constants to 1 decimal place.
Chromatographic analyses were carried out using a Varian CP-
3800 with a flame ionization detector (FID11) and injector 1177.
Specifications of the column: column name: Varian VF-1 (nonpo-
lar); column length: 60 m; inner diameter: 0.25 mm; stationary
4.4. Synthesis of compound 2d
6-hexamethylbenzene)]2 (0.21 mmol,
A
mixture of [RuCl2(g
139 mg), (S,S)-TsDPEN (0.42 mmol, 152 mg), and triethylamine
(0.83 mmol, 84 mg) was dissolved in 30 mL of dry dichloromethane
(25 °C) and stirred for 1 h under an Ar atmosphere. The reaction
mixture was washed with cold water (30 mL) and dried over anhy-
drous Na2SO4. The desiccant was filtered off using sintered glass and
dichloromethane was evaporated via a rotary vacuum evaporator.
The crude product was obtained as a yellow-orange solid (155 mg,
56%). The product can be further purified by crystallization from
boiling MTBE (small amount of hexane was added in order to
decrease the solubility of the product and to facilitate precipitation).
Mp: 158–159 °C. 1H NMR (600 MHz, DMSO-d6, 303.2 K) d 2.19 (18H,
s, CH3), 2.19 (3H, s, C-CH3), 3.15 (m), 3.50 (1H, m, N-CH), 3.52 (1H, m,
NH2-CH), 5.94 (2H, d, NH2, JHH = 10.3 Hz), 6.51 (2H, d, C6H5,
JHH = 7.3 Hz), 6.74 (2H, dd, C6H5, JHH = 7.3 Hz, JHH = 7.4 Hz), 6.76
(2H, m, o0), 6.80 (1H, t, p, JHH = 7.4 Hz), 6.85 (2H, d, C6H4, JHH = 8.0),
7.09 (2H, m, C6H4), 7.09 (1H, m, C6H5), 7.23 (2H, d, C6H4, JHH = 8.0 -
Hz). 13C NMR (150 MHz, DMSO-d6, 303.2 K) d 15.75 (6CH3, q, CH3),
20.74 (CH3, q, C6H4-CH3), 69.46 (CH, d, N-CH), 70.83 (CH, m,
NH2-CH), 90.85 (6C, s, C6), 125.90 (CH, d, C6H5), 126.53 (2CH, d,
C6H5), 127.08 (2CH, d, C6H5), 127.36 (2CH, d, C6H5), 127.53 (CH, d,
C6H5), 127.69 (2CH, d, C6H5), 128.04 (2CH, d, C6H5), 128.88 (2CH,
d, C6H5), 138.38 (C, s, C6H4), 139.59 (C, s, C6H5), 139.87 (C, s, C6H5),
142.37 (C, s, C6H4); HRMS: resolution 30,000, C33H39O2N2RuS—
[MꢀCl]+, error: ꢀ1.95 ppm, m/z (rel. int): 623.1795 (18.45%),
624.1824 (5.87%), 625.1761 (7.45%), 626.1764 (39.08%), 627.1758
(35.86%), 627.1845 (10.54%), 628.1763 (60.47%), 629.1758
(100.00%), 629.1830 (5.57%), 630.1794 (29.45%), 631.1785
(59.04%), 632.1804 (20.08%), 633.1832 (5.47%).
phase film thickness: 0.25 lm; stationary phase: poly(dimethylsi-
loxane); carrier gas: nitrogen.
4.2. Standard hydrogenation procedure
The asymmetric transfer hydrogenation reactions were per-
formed according to a previously reported procedure.12 A round-
bottom flask was equipped with a magnetic stirrer bar and was
pre-heated on a water bath (30 °C). Stock solutions of the sub-
strates and catalyst were prepared. The amounts of reaction com-
ponents were calculated in order to fulfill the following ratios: S/C
ratio = 100, HCOOH/triethylamine ratio = 2.5, concentration = 7.0%
(defined as: (mass of substrate + mass of catalyst + mass of formic
acid + mass of triethylamine)/mass of solvent), hydrogenation mix-
ture/substrate ratio = 8.83, total volume of reaction mixture = 2 mL
(all ratios are molar). The components were transferred into the
flask in the following order: acetonitrile, formic acid, triethyl-
amine, solution of the catalyst. After five minutes, the calculated
amount of the substrate solution containing 0.15 mmol of sub-
strate was added into the reaction mixture. The samples were
taken in defined time intervals.
4.5. Computational study
The samples were treated with a saturated solution of sodium
carbonate (1 mL) and extracted three times with diethyl ether
(3 ꢁ 1 mL). The extract was dried over sodium sulfate, filtered,
and stripped in a stream of argon. The residue was dissolved in
For molecular modeling, the Gaussian 09 framework was
used.13 Density functional theory (DFT) and Møller–Plesset pertur-
bation theory of the second order (MP2) were employed as the
main computational methods. DFT was used for the localization
of transition states, frequency analyses, and computation of sin-
gle-point energies. MP2 was used only for the computation of sin-
gle-point energies. DFT calculations were conducted with the
600
20
l
L of acetonitrile and analyzed via GC. After the addition of
l
L triethylamine and 10
lL of (ꢀ)-(R)-menthyl chloroformate,
the enantioselectivity could be determined.
usage of Head-Gordon’s and Chai’s
xB97X-D functional which
includes an empirical correction for dispersion forces.14 Ahlrichs’s
split valence def2-SVP basis set was used for all atoms.15 Calcula-
tions were carried out without any structural simplifications, only
the counterion (formate) was neglected in most calculations.
Computations involving the counterion were carried out on the
def2-SVPD basis set. The transition state geometries were obtained
by using the QST3 algorithm. All transition states were verified to
have only one imaginary vibration (which corresponded to the
transition state). Frequency analyses were carried out at 298.15 K
and 1.00 atm. Solvation was simulated using the integral equation
formalism (IEFPCM) model with universal force field (UFF) topol-
4.3. Synthesis of compound 2a
A mixture of [RuCl2(g
6-benzene)]2 (0.28 mmol, 142 mg), (S,S)-
TsDPEN (0.57 mmol, 208 mg), and triethylamine (1.13 mmol,
114 mg) was dissolved in 35 mL of warm methanol (40 °C) and
sonicated for 10 min. The mixture was then cooled down (5 °C)
and the brown precipitate was filtered on sintered glass and
washed with cold water (200 mL), ethanol (20 mL), and dichloro-
methane (60 mL). The crude product was obtained as red-orange
crystals (101.5 mg, 27%). Mp: decomposes at >170 °C. 1H NMR
(600 MHz, DMSO-d6, 303.2 K) d 2.22 (s), 3.18 (s), 3.73 (br s), 3.84
(br s), 4.05 (br s), 4.11 (br s), 5.70 (br s), 5.80 (s), 6.21 (br s), 6.30
(br s), 7.33–6.49 (m), 7.66 (s). 13C NMR (150 MHz, DMSO-d6,
303.2 K) d 20.78, 65.54, 68.13, 69.54, 71.64, 83.19, 83.41, 125.29,
125.87, 126.09, 126.39, 126.51, 127.10, 127.62, 127.90, 128.14,
129.09, 129.48, 138.04, 138.40, 139.24, 139.91, 140.66, 143.89,
144.18. Due to extensive overlap of broadened signals and result-
ing lack of appropriate correlations, it was impossible to unambig-
uously assign all of the signals; HRMS: resolution 30,000,
ogy (acetonitrile, e
= 35.688).16 For CHELPG charges, the ruthenium
atomic radius was set to 2.07 Å. Coordinates of all relevant struc-
tures can be provided upon request.
Acknowledgments
This work has been financially supported by the Grant Agency of
the Czech Republic: P106/12/1276, and by Grant for long-term con-
ceptual development of Institute of Microbiology RVO: 61388971.
The research was conducted within the ‘Prague Infrastructure for
Structure Biology and Metabolomics’ which has been built up by
financial support of the Operational Program Prague—Competitive-
C
27H27O2N2RuS—[MꢀCl]+, error: 1.13 ppm, m/z (rel. int.):
539.0870 (14.37%), 540.0845 (3.45%), 541.0846 (5.87%), 542.0854
(33.11%), 543.0848 (40.24%), 545.0850 (54.90%), 545.0838
(100.0%), 546.0864 (20.59%), 547.0841 (52.48%); 548.0870
(12.44%), 549.0797 (1.81%), 549.0899: (1.76%).

P. Šot et al. / Tetrahedron: Asymmetry 25 (2014) 1346–1351
1351
Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.;
Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas,
Ö.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision
C.01; Gaussian: Wallingford CT, 2009.
ness (Project No.: CZ.2.16/3.1.00/24023). We greatly appreciate the
access to computing and storage facilities owned by parties and
projects contributing to the National Grid Infrastructure MetaCen-
trum, provided under the program ‘Projects of Large Infrastructure
for Research, Development, and Innovations’ (LM2010005), as well
as access to the CERIT-SC computing and storage facilities provided
under the program Center CERIT Scientific Cloud, part of the
Operational Program Research and Development for Innovations,
Reg. No. CZ. 1.05/3.2.00/08.0144.
14. Head-Gordon, M.; Chai, J. Phys. Chem. Chem. Phys. 2008, 10, 6615.
15. Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297.
16. Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105, 2999.
17. 70 min: 1a–2a: 96.0%, 1b–2a: 92.1%, 1c–2a: 89.7%, 1d–2a: 93.3%, 1a–2b:
92.5%, 1b–2b: 91.2%, 1c–2b: 87.8%, 1d–2b: 92.8%, 1a–2c: 92.4%, 1b–2c: 92.5%,
1c–2c: 90.6%, 1d–2c: 94.1%, 1a–2d: 5.5%, 1b–2d: 17.4%, 1c–2d: 19.1%, 1d–2d:
15.3%. 24 h: 1e–2a: 29.2%, 1f–2a: 27.6%, 1e–2b: 0.2%, 1f–2b: 0.3%, 1e–2c: 0.5%,
1f–2c: 0.2%, 1e–2d: 0.9%, 1f–2d: 0.7%.
References
18. Cheung, F. K.; Clarke, A. J.; Clarkson, G. J.; Fox, D. J.; Graham, A. M.; Lin, Ch.;
Crivillé, A. L.; Wills, M. Dalton Trans. 2010, 39, 1395.
19. Noyori showed that 6,7-dimethoxy-1-phenyl-3,4-dihydroisoquinoline can be
reduced with high enantiomeric excess.1 It is obvious that in this case, the
aromatic part of the dihydroisoquinoline backbone is electron rich, while the
1-phenyl substituent is electron poor. This desymmetrization probably leads to
1. Uematsu, N.; Fujii, A.; Hashiguchi, S.; Ikariya, T.; Noyori, R. J. Am. Chem. Soc.
1996, 118, 4916.
2. Hashiguchi, S.; Fujii, A.; Takehara, J.; Ikariya, T.; Noyori, R. J. Am. Chem. Soc.
1995, 117, 7562.
3. Xue, D.; Chen, Y.-C.; Cui, X.; Wang, Q.-W.; Zhu, J.; Deng, J.-G. J. Org. Chem. 2005,
70, 3584.
a preferential formation of a CH/
p interaction from the aromatic part of the
ˇ
ˇ
´
4. Václavík, J.; Kacer, P.; Kuzma, M.; Cerveny, L. Molecules 2011, 16, 5460.
5. Cheung, F. K.; Hayes, A. M.; Morris, D. J.; Wills, M. Org. Biomol. Chem. 2007, 5,
1093.
dihydroisoquinoline fragment to the 1-phenyl substituent, which in turn
results in higher ee.
ˇ
ˇ
ˇ
20. Prech, J.; Matoušek, V.; Václavík, J.; Pechácek, J.; Syslová, K.; Šot, P.; Janušcák, J.;
6. Yamakawa, M.; Yamada, I.; Noyori, R. Angew. Chem. Int. Ed. 2001, 40, 2818.
7. Yamada, I.; Noyori, R. Org. Lett. 2000, 2, 3425.
8. Takehara, J.; Hashiguchi, S.; Fujii, A.; Inoue, S.; Ikariya, T.; Noyori, R. Chem.
Commun. 1996, 233.
ˇ
Vilhanová, B.; Kuzma, M.; Kacer, P. Am. J. Anal. Chem. 2013, 4, 125.
ˇ
ˇ
ˇ
ˇ
21. Šot, P.; Kuzma, M.; Václavík, J.; Pechácek, J.; Prech, J.; Janušcák, J.; Kacer, P.
Organometallics 2012, 31, 6496.
ˇ
ˇ
22. Václavík, J.; Kuzma, M.; Prech, J.; Kacer, P. Organometallics 2011, 30, 4822.
23. Soldevila-Barreda, J. J.; Bruijnincx, P. C. A.; Habtemariam, A.; Clarkson, G. J.;
Deeth, R. J.; Sadler, P. J. Organometallics 2012, 31, 5958.
ˇ
ˇ
ˇ
9. Václavík, J.; Pechácek, J.; Vilhanová, B.; Šot, P.; Janušcák, J.; Matoušek, V.; Prech,
ˇ
J.; Bártová, S.; Kuzma, M.; Kacer, P. Catal. Lett. 2013, 143, 555.
ˇ
ˇ
ˇ
10. Prech, J.; Václavík, J.; Šot, P.; Pechácek, J.; Vilhanová, B.; Janušcák, J.; Syslová, K.;
24. Mulliken: Ru-2a: ꢀ0.293, Ru-2b: ꢀ0.211, Ru-2c: ꢀ0.224, Ru-2d: ꢀ0.126; H-2a:
ꢀ0.124, H-2b: ꢀ0.130, H-2c: ꢀ0.124, H-2d: ꢀ0.141. NPA: Ru-2a: ꢀ0.243, Ru-
2b: ꢀ0.204, Ru-2c: ꢀ0.207, Ru-2d: ꢀ0.149; H-2a: 0.050, H-2b: 0.031, H-2c:
0.046, H-2d: 0.010; APT: Ru-2a: ꢀ0.342, Ru-2b: ꢀ0.432, Ru-2c: ꢀ0.380, Ru-2d:
ꢀ0.461; H-2a: ꢀ0.175, H-2b: ꢀ0.175, H-2c: ꢀ0.184, H-2d: ꢀ0.203; CHELPG:
Ru-2a: ꢀ0.047, Ru-2b: ꢀ0.089, Ru-2c: ꢀ0.110, Ru-2d: 0.126; H-2a: ꢀ0.199, H-
2b: ꢀ0.237, H-2c: ꢀ0.203, H-2d: ꢀ0.272.
ˇ
ˇ
Pazout, R.; Maixner, J.; Zápal, J.; Kuzma, M.; Kacer, P. Catal. Commun. 2013, 36,
67.
11. Wang, T.; Zhuo, L.-G.; Li, Z.; Chen, F.; Ding, Z.; He, Y.; Fan, Q.-H. J. Am. Chem. Soc.
2011, 133, 9878.
ˇ
ˇ
ˇ
ˇ
12. Pechácek, J.; Václavík, J.; Prech, J.; Šot, P.; Janušcák, J.; Vilhanová, B.; Vavrík, J.;
ˇ
Kuzma, M.; Kacer, P. Tetrahedron: Asymmetry 2013, 24, 233.
13. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.;
Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.;
Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven,
T., ; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.;
Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.;
25. Morris, D. J.; Hayes, A. M.; Wills, M. J. Org. Chem. 2006, 71, 7035.
26. Hannedouche, J.; Clarkson, G. J.; Wills, M. J. Am. Chem. Soc. 2004, 126, 986.
27. Hayes, A. M.; Morris, D. J.; Clarkson, G. J.; Wills, M. J. Am. Chem. Soc. 2005, 127,
7318.
28. Touge, T.; Hakamata, T.; Nara, H.; Kobayashi, T.; Sayo, N.; Saito, T.; Kayaki, Y.;
Ikariya, T. J. Am. Chem. Soc. 2011, 133, 14960.