Asymmetric Hydrogenation of 3-Substituted Pyridinium Salts

Article abstract of DOI:10.1002/chem.201601501

The use of an equivalent amount of an organic base leads to high enantiomeric excess in the asymmetric hydrogenation of N-benzylated 3-substituted pyridinium salts into the corresponding piperidines. Indeed, in the presence of Et₃N, a Rh-JosiPhos catalyst reduced a range of pyridinium salts with ee values up to 90 %. The role of the base was elucidated with a mechanistic study involving the isolation of the various reaction intermediates and isotopic labeling experiments. Additionally, this study provided some evidence for an enantiodetermining step involving a dihydropyridine intermediate.

Full text of DOI:10.1002/chem.201601501

DOI: 10.1002/chem.201601501
Communication
&
Catalysis
Asymmetric Hydrogenation of 3-Substituted Pyridinium Salts
Marc Renom-Carrasco,[a, b] Piotr Gajewski,[a, b] Luca Pignataro,[b] Johannes G. de Vries,[c]
Umberto Piarulli,[d] Cesare Gennari,[b] and Laurent Lefort*[a]
prevent substrate coordination to the catalyst.[5] Unfortunately,
the translation of these methods to 3-substituted pyridines is
Abstract: The use of an equivalent amount of an organic
not straightforward. All reports dealing with this class of sub-
base leads to high enantiomeric excess in the asymmetric
strates are restricted to nicotinic acid or its esters,[6] and the
hydrogenation of N-benzylated 3-substituted pyridinium
obtained enantiomeric excesses were overall low. Only indirect
salts into the corresponding piperidines. Indeed, in the
protocols relying on the use of a chiral auxiliary[6d] or a two-
presence of Et3N, a Rh-JosiPhos catalyst reduced a range
step hydrogenation procedure[6e] allowed to obtain high ee
of pyridinium salts with ee values up to 90%. The role of
values.
the base was elucidated with a mechanistic study involv-
Although there is no general agreement on the mechanism
ing the isolation of the various reaction intermediates and
of the AH of N-heteroarenes,[7,8] the current data led us to
isotopic labeling experiments. Additionally, this study pro-
speculate that the poor enantiomeric excesses obtained with
vided some evidence for an enantiodetermining step in-
3-substituted pyridines in comparison to their 2-substituted
volving a dihydropyridine intermediate.
analogues were due to a non-enantioselective enaminium–imi-
nium isomerization of a partially hydrogenated intermediate
(Scheme 1).[9] Based on this hypothesis, we surmised that the
addition of a base in the AH of N-benzylated pyridines[5] could
The asymmetric hydrogenation (AH) of double bonds is one of
the most efficient methods to prepare enantiopure organic
compounds.[1] Although AH has been applied to a wide range
of substrates, the development of catalysts for the AH of sub-
stituted N-heteroaromatic substrates is still in its early stages.[2]
Such substrates are particularly challenging due to their high
resonance stabilization energy and strong ability to deactivate
the catalyst by irreversible coordination. Among the N-hetero-
aromatic substrates, substituted pyridines are of special inter-
est, since the products of their hydrogenation, that is, chiral pi-
peridines, are subunits of many biologically active compounds.
Moreover, hundreds of pyridines are commercially available
and countless methods exist for their preparation. During the
last ten years, several successful methods have been disclosed
for the homogeneous AH of 2-substituted pyridines,[3] and
more recently of di- and trisubstituted pyridines.[4] In most of
these contributions, a quaternization of the pyridine nitrogen
atom is used to lower the resonance energy of the ring and
retard the tautomerization by scavenging HBr produced
during the reaction. As a consequence, the formation and sub-
sequent hydrogenation of a racemic iminium salt would be
prevented (Scheme 1).
Scheme 1. Proposed mechanism for the hydrogenation of 3-substituted pyr-
idinium salts and possible effect of the addition of base.
For our initial trials, we selected Rh-JosiPhos as AH catalyst.
Rh has indeed been extensively used for the AH of enamines
and enamides.[10] Equally, the chiral bis-phosphine, JosiPhos
J002-2 was shown to be an efficient ligand[11] for the AH of re-
lated substrates, that is, 2,4-disubstituted pyrimidines[12] and
3,4-disubstituted pyridinium salts.[13] Diisopropylamine (DIPEA)
was chosen as the base to prevent the iminium formation.
Indeed, its steric bulkiness was expected to decrease its pro-
pensity to coordinate to the rhodium center. A mixture of THF/
MeOH was selected as solvent to solubilize both the catalyst
and the pyridinium salt. Our first hydrogenation was carried
out with N-benzyl-3-phenylpyridinium bromide (1a) as a sub-
strate at 508C under 50 bar of H2 with and without DIPEA
(Scheme 2). In both cases, N-benzylpiperidine 2a was obtained
in low yield (13 and 16%, respectively) together with some N-
[a] M. Renom-Carrasco, P. Gajewski, Dr. L. Lefort
DSM Ahead R&D B.V.–Innovative Synthesis
P. O. Box 18, 6160 MD Geleen (The Netherlands)
E-mail: laurent.lefort@dsm.com
[b] M. Renom-Carrasco, P. Gajewski, Dr. L. Pignataro, Prof. C. Gennari
Dipartimento di Chimica, Universitꢀ degli Studi di Milano
via C. Golgi 19, 20133 Milan (Italy)
[c] Prof. J. G. de Vries
Leibniz-Institut fꢁr Katalyse e.V. an der Universitꢂt Rostock
Albert-Einstein-Str. 29a, 18059 Rostock (Germany)
[d] Prof. U. Piarulli
Dipartimento di Scienza e Alta Tecnologia
Universitꢀ degli Studi dell’Insubria, via Valleggio 11, 22100 Como (Italy)
Supporting information and the ORCID identification number for the
author of this article can be found under http://dx.doi.org/10.1002/
chem.201601501.
Chem. Eur. J. 2016, 22, 1 – 6
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
benzyl tetrahydropyridine TH-1. In the absence of base, the pi-
peridine was obtained with only 9% ee. In sharp contrast,
when DIPEA (5 equivalents relative to the substrate) was
added to the reaction, the enantiomeric excess rose to 73%,
which was, to our knowledge, the highest enantiomeric excess
ever obtained in the direct AH of a 3-substituted pyridine.
Table 1. Selected results on the optimization of the reaction conditions.[a]
#
Solvent
Base (pKBHþ in H2O)
2a
ee
TH-1
[%][b]
[%][b] [%][b]
1
2
3
4
5
6
7
8
9
THF/MeOH (2:1) DIPEA (11.07)
13
81
5
10
6
27
40
40
5
73
13
34
9
MeOH
DIPEA
THF
DIPEA
ꢀ59[c] 41
THF/EtOH (2:1)
THF/iPrOH (2:1)
THF/TFE (2:1)
THF/TFE (1:1)
THF/TFE (1:2)
THF/TFE (2:1)
DIPEA
DIPEA
DIPEA
DIPEA
DIPEA
–
63
34
83
68
52
23
31
74
84
85
48
90
76
85
78
50
48
17
2
2
0
Scheme 2. First hydrogenation attempts of N-benzyl-3-phenylpyridinium
bromide 1a, with and without base.
10 THF/TFE (2:1)
11 THF/TFE (2:1)
2,6-lutidine (6.65)
N-Me-morpholine (7.38) 40
98
0
6
12 THF/TFE (2:1)
13 THF/TFE (2:1)
14 THF/TFE (2:1)
15 THF/TFE (2:1)
16 THF/TFE (2:1)
17 THF/TFE (2:1)
18 THF/TFE (2:1)
DMAP (9.2)
Et3N (10.75)
Cs2CO3 (10.32)
Et3N
Et3N [0.5 equiv]
Et3N [1 equiv]
Et3N [10 equiv]
23
57
40
50[d]
40
54
58
13
20
32
26
9
Encouraged by this result, we initiated a solvent and base
screening (Table 1) in order to increase both yield and ee of 2a
and to minimize the amount of side product TH-1. When pure
MeOH was used as solvent, a high yield but a low ee was ob-
tained (Table 1, entry 2). In pure THF, the reaction gave the op-
posite enantiomer with very low yield (entry 3). THF/alcohol
mixtures turned out to be optimal, as they allowed to obtain
good yields and high ee values. Among the mixtures tested,
2:1 THF/trifluoroethanol (TFE) gave the best result (entry 6)
and thus was selected for further experiments.
16
16
[a] Reaction conditions: 1a (0.025 mmol), [Rh(cod)2]OTf (2 mol%), Josi-
Phos J002-2 (2.2 mol%), base (5 equiv relative to the substrate unless oth-
erwise indicated), solvent (1.5 mL), 508C, 50 bar H2, 20 h. [b] Determined
by GC analysis with CP-Chirasil-Dex CB column and dodecane as internal
standard. [c] R configuration was obtained in this case. [d] Reaction time:
16 h.
In the absence of base or in the presence of weak bases
(pKBHþ <8, entries 9–11), a lesser amount of TH-1 was detected,
but the obtained ee values were very low. On the contrary, the
use of stronger amines gave the best ee values, together with
a significant amount of TH-1 (entries 12–13). Among them,
Et3N was found to be the best performing (entry 13). When the
reaction time was reduced from 20 to 16 h, a slightly higher ee
was obtained at the expense of the yield (entry 13 vs. 15), sug-
gesting an erosion of ee with time. Varying the amount of base
had little effect as long as at least one equivalent of base rela-
tive to the substrate was used (entries 16–18). Unfortunately,
further catalyst screening involving 8 metal precursors (includ-
ing an iridium source) and 25 chiral ligands failed to give
better results (see the Supporting Information).
Table 2. Substrate scope under the optimized conditions, with and with-
out base.[a]
#
R=
With Et3N
Without Et3N
2 [%][b]
ee [%][b]
2 [%][b]
ee [%][b]
1
2
3
4
5
6
7
8
Ph (1a)
50
90 (S)
84 (S)
83 (ꢀ)
75 (ꢀ)
90 (ꢀ)
86 (ꢀ)
57 (R)
33
55 (R)
41
32 (ꢀ)
5
23 (S)
Ph (1a)[c]
57 (52)[d]
4-CF3C6H4 (1b)
2-MeC6H4 (1c)
4-MeOC6H4 (1d)
2-naphthyl (1e)
Me (1 f)
CO2Et (1g)
NHBoc (1h)
CF3 (1i)
20
50
52
42
36
2
24
2
43
12
7
8
21
<1
3
25
2
14 (ꢀ)
30 (ꢀ)
40 (ꢀ)
20 (ꢀ)
n.d.
ꢀ17[e]
27 (R)
11
Under the optimized conditions, a substrate screening
(Table 2) was performed both in the presence and in the ab-
sence of Et3N. The best results were obtained with N-benzylat-
ed pyridines bearing an aromatic substituent. For these sub-
strates, the ee values of the corresponding N-benzylpiperidines
ranged from 75 to 90% in the presence of Et3N, as compared
with the 14–40% ee obtained without base (Table 2, entries 1–
6). The presence of the base was also beneficial with alkyl-sub-
stituted pyridines, but overall the ee values were lower with
this class of substrates (entries 7 and 11).
9
10
11
nBu (1j)
<1
n.d.
[a] Reaction conditions: 1 (0.025 mmol), [Rh(cod)2]OTf (2 mol%), JosiPhos
J002-2 (2.2 mol%), Et3N (5 equiv), THF/TFE (2:1, 1.5 mL), 508C, 50 bar H2,
16 h. [b] Determined by GC or HPLC analysis with dodecane as internal
standard. Absolute configuration assigned by comparison of the optical
rotation with literature data (see the Supporting Information). [c] Reaction
performed on 500 mg scale (1.53 mmol) for 20 h. [d] Isolated yield.
[e] The opposite enantiomer was obtained.
Finally, lower conversions were obtained for substrates pos-
sessing electron-withdrawing groups (entries 3, 8 and 10).
Since the catalyst was optimized with 3-phenylpyridine as
model substrate, it is not surprising that the best results were
obtained with aryl substituents. However, a positive effect of
the base was observed for all substrates, suggesting a similar
hydrogenation pathway for all substrates.
&
&
Chem. Eur. J. 2016, 22, 1 – 6
www.chemeurj.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
To gain some insight into the mechanism, the reaction was
monitored over time by GC and NMR spectroscopy (Figure 1).
The starting material 1a was consumed relatively quickly (over
90% conversion within the first 4 h), with concomitant forma-
tion of piperidine 2a and of several partially hydrogenated in-
termediates that were identified as N-benzylated tetrahydro-
pyridines.[14] Tetrahydropyridine TH-1, the main side product at
the end of the reaction, was formed at a fast rate during the
first two hours and then its amount very slowly decreased.
This compound was isolated and fully characterized (see the
Supporting Information). A small amount of tetrahydropyridine
TH-2[15] (<3% yield) was also produced, and remained con-
stant during the course of the reaction.
bromide formed at the same rate as that of the pyridinium salt
consumption (see the Supporting Information). This finding
confirms the role of the added base as a scavenger of HBr.
The tetrahydropyridines TH-1 and TH-2 were submitted to
hydrogenation under optimized conditions. After 20 h of reac-
tion, only 7% of TH-1 was hydrogenated to give the racemic
piperidine product. Similarly, the hydrogenation of TH-2 oc-
curred with 7% conversion to 2a (with 15% ee) and 13% iso-
merization to TH-1.[16] These results strongly indicate that TH-
1 and TH-2 are not involved in the enantioselective formation
of 2a. Since they are the only prochiral tetrahydropyridines,
the enantioselective step must occur through the hydrogena-
tion of a dihydropyridine leading to an unidentified tetrahydro-
pyridine derivative (TH-3 or TH-4), the ee of which was mea-
sured and found to be high (Figure 1).
The possible pathways for the hydrogenation of the pyridini-
um salt 1a are shown in Scheme 3. The first hydride addition
can occur at the 2-, the 4- or the 6-position leading to three
different dihydropyridines (DH-1, DH-2 and DH-3, respectively)
that are not observed experimentally due to their instability or
high reactivity. Their enantioselective reduction into TH-3 or
TH-4 involves either the reduction of the conjugated enamine
double bond (DH-3!TH-4 or DH-2!TH-3) or the reduction
of the C3-C4 double bond (DH-1!TH-3).
Figure 1. Evolution of the yield and ee of different reaction species over
time. Reaction conditions: 1a (1.53 mmol), [Rh(cod)2]OTf (2 mol%), JosiPhos
J002-2 (2.2 mol%), Et3N (5 equiv), THF/TFE (2:1, 27 mL), 508C, 50 bar H2. Con-
sumption of 1a determined by NMR spectroscopy with dimethyl terephtha-
late as internal standard. Yield and ee of 2a and the different tetrahydropyri-
dines monitored by GC analysis with CP-Chirasil-Dex CB column with dodec-
ane as internal standard.
Scheme 3. Possible pathways and intermediates for the AH of 1a to 2a.
To identify the main pathway, an isotopic labeling experi-
ment was carried out by conducting the AH of 1a with D2 in-
stead of H2 (Scheme 4). As expected from all the possible inter-
mediates involved in this reaction, the deuteration pattern of
2a was complex but provided some useful information. One of
the most important observations is the syn incorporation of
two deuterium atoms in C3 and C4. According to the postulat-
ed pathways (Scheme 3), this can be due to the hydrogenation
of either TH-2!2a, or DH-1!TH-3. Since the hydrogenation
of isolated TH-2 was shown to occur with very poor stereocon-
trol (see above), it is reasonable to assume that the conversion
DH-1!TH-3 is the one leading to high enantiomeric excess
(93% ee) and to assign the unidentified tetrahydropyridine as
TH-3. A second observation is that the total amount of deuteri-
um incorporated accounts for four deuterium atoms. There-
fore, one proton from the solvent is incorporated, probably in
A third compound, also identified by GC-MS (m/z 249) as an
N-benzyl tetrahydropyridine, was formed in appreciable
amounts at the beginning of the reaction and then was slowly
consumed. Although it was not possible to isolate or synthe-
size it by alternative methods, this compound showed two
peaks when injected in a chiral GC column, suggesting that it
contains a stereocenter and therefore may be either tetrahy-
dropyridine TH-3 or TH-4. Remarkably, its ee was very high
(93%) and remained constant during the entire course of the
reaction. N-Benzylpiperidine 2a was slowly formed with an
enantiomeric excess that increased sharply during the first
10 h of the reaction and then slowly decreased over the next
10 h. Furthermore, the monitoring the reaction by NMR spec-
troscopy revealed that one equivalent of triethylammonium
Chem. Eur. J. 2016, 22, 1 – 6
www.chemeurj.org
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
the least deuterated carbon C5 through an enamine–iminium
tautomerization of TH-3. Finally, it should be noted that, while
the deuterium incorporation at the two diastereotopic posi-
tions of C2 is quite similar (36 vs. 49%), the extent of deutera-
tion of the pro-S hydrogen at C6 is remarkably higher than
that of the pro-R hydrogen (80 vs. 28%). A possible interpreta-
tion of this finding is that the hydride attack at C2 occurs
before the formation of the stereocenter at C3 (and thus is
only influenced by the chiral catalyst), whereas the hydride
attack at C6 occurs after the stereocenter formation, so that
a remarkable substrate control also operates.[17] The analysis of
the deuterium distribution in the isolated intermediates TH-
1 and TH-2 (see the Supporting information) suggests that
these compounds are formed by hydrogenation of DH-3 and
DH-1, respectively (Scheme 3).
In conclusion, we have successfully developed the first
highly enantioselective hydrogenation of 3-substituted pyri-
dines using a Rh-JosiPhos system. The use of a simple base,
like Et3N, remarkably improved the yields and ee values ob-
tained in the AH of such substrates. Furthermore, our mecha-
nistic studies shed some light on the complex reaction net-
work involved in the AH of pyridines. All experimental data
point towards an enantioselective step taking place during the
hydrogenation of one of the dihydropyridine intermediates.
The beneficial role of the base was also identified as prevent-
ing the erosion of ee of the piperidine by retarding the non-
enantioselective hydrogenation of one of the prochiral enam-
ines. We hope that our results will contribute to the rational
design of more efficient catalysts towards an efficient produc-
tion of 3-substituted chiral piperidines.
Experimental Section
Inside a glovebox, [Rh(cod)2]OTf (0.23 mg, 2 mol%) and JosiPhos
J002-2 (0.30 mg, 2.2 mol%) were stirred for 1 h at 408C in 0.5 mL
of THF. The catalyst solution was then transferred into a vial con-
taining a mixture of 1 (0.025 mmol), Et3N (17.4 mL, 0.125 mmol) and
dodecane (10 mg) in 0.5 mL of THF and 0.5 mL of TFE. The vial was
capped and placed into a Premex A96 hydrogenation reactor. After
being flushed five times with N2 and five times with H2, the vial
was pressurized to 50 bar of H2 and stirred at 508C for 16 h. The
yield and ee of the reaction were determined by GC or HPLC using
dodecane as internal standard. Alternatively, the reaction crude
could be concentrated and purified by flash column chromatogra-
phy with hexane/AcOEt.
Scheme 4. Isotopic labeling experiment for the AH of N-benzyl-3-phenylpyri-
dinium bromide 1a.
On the basis of these experimental data, a pathway for the
highly enantioselective formation of piperidine 2a can be pro-
posed (Scheme 5). Initially, a 1,2-hydride addition at the C2 po-
sition takes place. The formed dihydropyridine DH-1 is then
enantioselectively hydrogenated to TH-3 by reduction of the
C3=C4 double bond. Finally, TH-3 slowly tautomerizes to the
iminium form and a final 1,2-hydride addition at the C6 posi-
tion affords the enantioenriched piperidine 2a. At the same
time, at least one other pathway operates, involving the pro-
chiral enamine TH-1 (which is observed as a side-product) as
a key intermediate. As the hydrogenation of TH-1 occurs with
no enantioselectivity, the reaction path(s) involving this inter-
mediate lead to erosion of the final observed ee. A plausible
role of the added base is to scavenge HBr generated by the
first hydride addition, thus slowing down the tautomerization
of TH-1 to the corresponding iminium form and its non-enan-
tioselective hydrogenation to 2a. The other enamine TH-3 is
more readily hydrogenated than TH-1 because it is more
prone to tautomerization due to the absence of an adjacent
aromatic group leading to stabilization through conjugation.
Overall, the high enantiomeric excess of 2a is in part due to
the inhibition of the hydrogenation of TH-1.[18]
Acknowledgements
We thank the European Commission [ITN-EID “REDUCTO” PITN-
GA-2012-316371] for financial support and for predoctoral fel-
lowships (to M.R.-C. and P.G.).
Keywords: asymmetric catalysis · homogeneous catalysis ·
hydrogenation · pyridines · reaction mechanisms
[1] a) The Handbook of Homogeneous Hydrogenation (Eds.: J. G. de Vries,
C. J. Elsevier), Wiley-VCH, Weinheim, 2007; b) D. J. Ager, A. H. M. de V-
ries, J. G. de Vries, Chem. Soc. Rev. 2012, 41, 3340–3380.
[2] For reviews on the asymmetric hydrogenation N-heteroaromatic com-
pounds see: a) D.-S. Wang, Q.-A. Chen, S.-M. Lu, Y.-G. Zhou, Chem. Rev.
2012, 112, 2557–2590; b) Y.-G. Zhou, Acc. Chem. Res. 2007, 40, 1357–
1366; c) F. Glorius, Org. Biomol. Chem. 2005, 3, 4171–4175; d) Y.-M. He,
F.-T. Song, Q.-H. Fan, Stereoselective Formation of Amines (Eds.: W. Li, X.
Zhang), Springer, Heidelberg, 2014, pp. 145–190.
[3] a) C. Y. Legault, A. B. Charette, J. Am. Chem. Soc. 2005, 127, 8966–8967;
b) X.-B. Wang, W. Zeng, Y.-G. Zhou, Tetrahedron Lett. 2008, 49, 4922–
4924; c) A. Cadu, P. K. Upadhyay, P. G. Andersson, Asian J. Org. Chem.
2013, 2, 1061–1065; d) Z.-S. Ye, M.-W. Chen, Q.-A. Chen, L. Shi, Y. Duan,
Y.-G. Zhou, Angew. Chem. Int. Ed. 2012, 51, 10181–10184; Angew. Chem.
2012, 124, 10328–10331; e) M. Chang, Y. Huang, S. Liu, Y. Chen, S. W.
Krska, I. W. Davies, X. Zhang, Angew. Chem. Int. Ed. 2014, 53, 12761–
12764; Angew. Chem. 2014, 126, 12975–12978.
[4] a) Y. Kita, A. Iimuro, S. Hida, K. Mashima, Chem. Lett. 2014, 43, 284–286;
b) M.-W. Chen, Z.-S. Ye, Z.-P. Chen, B. Wu, Y.-G. Zhou, Org. Chem. Front.
2015, 2, 586–589.
Scheme 5. Proposed mechanism for the AH of N-benzyl-3-phenylpyridinium
bromide 1a to the piperidine 2a.
&
&
Chem. Eur. J. 2016, 22, 1 – 6
www.chemeurj.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
[5] For a review on different substrate activation methods for the catalytic
asymmetric hydrogenation of N-heteroarenes, see: B. Balakrishna, J. L.
NfflÇez-Rico, A. Vidal-Ferran, Eur. J. Org. Chem. 2015, 5293–5303.
[6] a) H.-U. Blaser, H. Hçnig, M. Studer, C. Wedemeyer-Exl, J. Mol. Catal. A
1999, 139, 253–257; b) M. Studer, C. Wedemeyer-Exl, F. Spindler, H.-U.
Blaser, Monatsh. Chem. 2000, 131, 1335–1343; c) S. A. Raynor, J. M.
Thomas, R. Raja, B. F. Johnson, R. G. Bell, M. D. Mantle, Chem. Commun.
Mehler, M. T. Reetz, J. G. de Vries, B. L. Feringa, J. Org. Chem. 2005, 70,
943–951; e) D. PeÇa, A. J. Minnaard, J. G. de Vries, B. L. Feringa, J. Am.
Chem. Soc. 2002, 124, 14552–14553; f) V. I. Tararov, R. Kadyrov, T. H. Rier-
meier, J. Holz, A. Bçrner, Tetrahedron Lett. 2000, 41, 2351–2355; g) Y. J.
Zhang, J. H. Park, S. Lee, Tetrahedron: Asymmetry 2004, 15, 2209–2212;
h) L. A. Oro, D. Carmona, The Handbook of Homogeneous Hydrogenation,
Part I (Eds.: J. G. de Vries, C. J. Elsevier), Wiley-VCH, Weinheim, 2007,
pp. 3–30.
˝
2000, 1925–1926; d) L. Hegedus, V. Hµda, A. Tungler, T. MµthØ, L. Szep-
esy, Appl. Catal. A 2000, 201, 107–114; e) A. Lei, M. Chen, M. He, X.
Zhang, Eur. J. Org. Chem. 2006, 4343–4347.
[11] A. Togni, C. Breutel, A. Schnyder, F. Spindler, H. Landert, A. Tijani, J. Am.
Chem. Soc. 1994, 116, 4062–4066.
[7] Mechanistic investigations on the metal-catalyzed homogeneous asym-
metric hydrogenation of quinolines and isoquinolines: a) D.-W. Wang,
X.-B. Wang, D.-S. Wang, S.-M. Lu, Y.-G. Zhou, Y.-X. Li, J. Org. Chem. 2009,
74, 2780–2787; b) M. Rueping, T. Theissmann, Chem. Sci. 2010, 1, 473–
476; c) G. E. Dobereiner, A. Nova, N. D. Schley, N. Hazari, S. J. Miller, O. Ei-
senstein, R. H. Crabtree, J. Am. Chem. Soc. 2011, 133, 7547–7562; d) T.
Wang, L.-G. Zhuo, Z. Li, F. Chen, Z. Ding, Y. He, Q.-H. Fan, J. Xiang, Z.-X.
[12] R. Kuwano, Y. Hashiguchi, R. Ikeda, K. Ishizuka, Angew. Chem. Int. Ed.
2015, 54, 2393–2396; Angew. Chem. 2015, 127, 2423–2426.
[13] S. G. Ruggeri, J. M. Hawkins, T. M. Makowskim, J. L. Rutherford, F. J.
Urban (Pfizer Prod Inc. Groton, CT (US)), WO 2007/012953A2, 2007.
[14] The observed products do not account for the complete disappearance
of the starting material due to the partial decomposition of the unsta-
ble enamine intermediates during sampling and subsequent analysis
(see the Supporting information where we report the relatively fast de-
composition of the enamines even after isolation).
[15] A reference sample of this compound was separately synthesized and
characterized.
[16] Rh-diphosphine complexes are known to catalyze allylamine-enamine
isomerizations: S. Otsuka, K. Tani, T. Yamagata, S. Akutagawa, H. Kumo-
bayashi, M. Yagi (Takasago Perfumery Co., Ltd., Tokyo (Japan)), EP
0068506A1, 1983.
[17] The preferential deuteration of the pro-S hydrogen can be explained
by the Fürst-Plattner effect (A. Fürst, P. A. Plattner, Helv. Chim. Acta
1949, 32, 275–283), where the hydride attack on the iminium inter-
mediate arising from the tautomerization of TH-3 takes places from the
side of the phenyl substituent (see the Supporting Information for fur-
ther details).
[18] The low ee for 2a at the beginning of the reaction is consistent with
the initial fast formation of TH-1 whose non-enantioselective conver-
sion into 2a can compete at this stage with the enantioselective path-
ways due to the large excess of TH-1 relative to TH-3. As the reaction
proceeds, the concentration of TH-3 increases and the enantioselective
pathway overrules the pathway via TH-1 leading to the observed in-
crease of ee. A similar profile for the ee is obtained from a kinetic
model based on our proposed reaction mechanism (see the Supporting
Information).
˝
Yu, A. S. C. Chan, J. Am. Chem. Soc. 2011, 133, 9878–9891; e) G. Eros, K.
Nagy, H. Mehdi, I. Pµpai, P. Nagy, P. Kirµly, G. Tµrkµnyi, T. Soós, Chem. Eur.
J. 2012, 18, 574–585; f) X.-F. Cai, R.-N. Guo, M.-W. Chen, L. Shi, Y.-G.
Zhou, Chem. Eur. J. 2014, 20, 7245–7248; g) L. Shi, Z.-S. Ye, L.-L. Cao, R.-
N. Guo, Y. Hu, Y.-G. Zhou, Angew. Chem. Int. Ed. 2012, 51, 8286–8289;
Angew. Chem. 2012, 124, 8411–8414; h) Z.-S. Ye, R.-N. Guo, X.-F. Cai, M.-
W. Chen, L. Shi, Y.-G. Zhou, Angew. Chem. Int. Ed. 2013, 52, 3685–3689;
Angew. Chem. 2013, 125, 3773–3777; i) A. Iimuro, K. Yamaji, S. Kandula,
T. Nagano, Y. Kita, K. Mashima, Angew. Chem. Int. Ed. 2013, 52, 2046–
2050; Angew. Chem. 2013, 125, 2100–2104; j) Y. Kita, K. Yamaji, K. Higa-
shida, K. Sathaiah, A. Iimuro, K. Mashima, Chem. Eur. J. 2015, 21, 1915–
1927.
[8] For a mechanistic study on the non-enantioselective transfer hydroge-
nation of pyridines see: J. Wu, W. Tang, A. Pettman, J. Xiao, Adv. Synth.
Catal. 2013, 355, 35–40.
[9] As pointed out by one of the reviewer, high enantioselectivities can be
obtained with 2,3-disubstituted N-heterocycles even if a non-enantiose-
lective enamine-imine tautomerization takes place. This involves a dy-
namic kinetic resolution process. See for example: a) Y. Duan, L. Li, M.-
W. Chen, C.-B. Yu, H.-J. Fan, Y.-G. Zhou, J. Am. Chem. Soc. 2014, 136,
7688–7700; b) L. Shi, Z.-S. Ye, L.-L. Cao, R.-N. Guo, Y. Hu, Y.-G. Zhou,
Angew. Chem. Int. Ed. 2012, 51, 8286–8289; Angew. Chem. 2012, 124,
8411–8414; and ref. [7g].
[10] a) Y. Liu, K. Ding, J. Am. Chem. Soc. 2005, 127, 10488–10489; b) A.
Ohashi, S. Kikuchi, M. Yasutake, T. Imamoto, Eur. J. Org. Chem. 2002,
2535–2546; c) S. Enthaler, G. Erre, K. Junge, J. Holz, A. Bçrner, E. Alberi-
co, I. Nieddu, S. Gladiali, M. Beller, Org. Process Res. Dev. 2007, 11, 568–
577; d) H. Bernsmann, M. van den Berg, R. Hoen, A. J. Minnaard, G.
Received: March 31, 2016
Published online on && &&, 0000
Chem. Eur. J. 2016, 22, 1 – 6
www.chemeurj.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
COMMUNICATION
Ariadne’s thread: The asymmetric hy-
drogenation of 3-substitued pyridines
has been achieved with up to 90% ee.
The use of an equivalent amount of an
organic base turned out to be crucial to
avoid unwanted pathways leading to
non-enantioselective reduction. An in-
depth investigation of the mechanism
revealed a labyrinth of hydrogenation
pathways with many dead ends (see
figure).
&
Catalysis
M. Renom-Carrasco, P. Gajewski,
L. Pignataro, J. G. de Vries, U. Piarulli,
C. Gennari, L. Lefort*
&& – &&
Asymmetric Hydrogenation of 3-
Substituted Pyridinium Salts
&
&
Chem. Eur. J. 2016, 22, 1 – 6
www.chemeurj.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!