Evaluation of 5-cis-Substituted Prolinamines as Ligands in Enantioselective, Copper-Catalyzed Henry Reactions

Article abstract of DOI:10.1002/cctc.201600240

The development of a new catalytic system for enantioselective Henry reactions, which permits superb 99 % ee with a broad variety of aldehydes, is presented. In-depth structure-selectivity investigations with 33 5-cis-substituted prolinamines, prepared fr

Full text of DOI:10.1002/cctc.201600240

DOI: 10.1002/cctc.201600240
Full Papers
Evaluation of 5-cis-Substituted Prolinamines as Ligands in
Enantioselective, Copper-Catalyzed Henry Reactions
Johannes Kaldun, Felix Prause, Dagmar Scharnagel, Frederik Freitag, and
Matthias Breuning*[a]
Dedicated to Prof. Dr. Dr. h.c. mult. Gerhard Bringmann on the occasion of his 65th birthday.
The development of a new catalytic system for enantioselec-
tive Henry reactions, which permits superb 99% ee with
a broad variety of aldehydes, is presented. In-depth structure–
selectivity investigations with 33 5-cis-substituted prolinamines,
prepared from methyl Boc-l-pyroglutamate, revealed that an
aromatic or sterically demanding aliphatic substituent in 5-cis
position is crucial for high levels of stereocontrol, while bulkier
substituents at the nitrogen atoms diminish both, enantiose-
lectivities and reaction rates. The scope of the prime catalyst
was expanded to gram-scale and diastereomeric Henry reac-
tions (up to 84:16 dr, 99% ee). In the course of mechanistic
studies, it was proven that the resulting b-nitro alcohols are
configurationally stable under the reaction conditions. In addi-
tion, competition experiments were used to determine the rel-
ative reaction rates of some of the prolinamine-modified cata-
lysts.
Introduction
The enantioselective Henry (nitro aldol) reaction[1] has drawn
much attention as an asymmetric carbon–carbon bond form-
ing reaction,[2] which triggered the development of many effi-
cient catalytic systems based on heterobimetal[3] and transi-
tion-metal[4–6] complexes.[7] Chirally modified copper complexes
received particular interest because of the wide structural vari-
ability of successful ligands, among them diamines, amino al-
cohols, amino imines, amino pyridines, imino pyridines, Schiff
bases, box-type ligands, and salen-type ligands.[5,6,8,9] Examples
of diamines (4–7)[5a,j,p,8a] and ligands containing the proline
motif (7, 8)[5j,m] that permit 99% ee in the addition of nitrome-
thane (2a) to at least one aldehyde substrate 1 are shown in
Scheme 1. Notably, Gong’s ligand 7,[5j] which belongs to the
most potent ones for this reaction, combines both structural
features.
As part of our ongoing work on conformationally rigid di-
amines[8,10] and encouraged by the stereodiscriminating power
of 7, we became interested in prolinamines of general type 9
and 10 (Scheme 2),[11] which possess, as compared to other
proline-derived ligands, an additional substituent R1 in 5-cis po-
sition. Upon chelation of a metal M, a bicyclic complex [M·9/
10] will be formed with the substituent R1 shielding the upper
left face, which might permit enhanced levels of stereocontrol
in asymmetric transformations. This assumption was recently
corroborated by copper-catalyzed, enantioselective Henry reac-
Scheme 1. The enantioselective, copper-catalyzed Henry reaction and a selec-
tion of diamine (4–7)[5a,j,p, 8a] and proline-derived (7,8)[5j,m] ligands that give
99% ee with at least one aldehyde substrate.
[a] J. Kaldun, F. Prause, D. Scharnagel, F. Freitag, Prof. M. Breuning
Organic Chemistry Laboratory
University of Bayreuth
Universitätsstraße 30, 95447 Bayreuth (Germany)
E-mail: Matthias.Breuning@uni-bayreuth.de
Supporting information and the ORCID identification number(s) for the
author(s) of this article can be found under http://dx.doi.org/10.1002/
cctc.201600240.
Scheme 2. The proline-derived diamines 9 and 10, their metal complexes
[M·9/10], and enantioselective, copper-catalyzed Henry reactions in the pres-
ence of the chiral diamine 9a.[9]
ChemCatChem 2016, 8, 1846 – 1856
1846
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
tions of nitromethane (2a) with a series of aromatic, heteroaro-
matic, vinylic, and aliphatic aldehydes 1.[9,12] The CuCl2 and
CuBr2 complexes of the simple prolinamine 9a (R1 =Ph; R2,
R4 =Me; R3 =H) provided the corresponding b-nitro alcohols 3
with superb, as yet unrivalled 99% ee in all cases (36 exam-
ples). Herein we present the development of the catalytic
system CuX2·9a, whose optimization included in-depth struc-
ture-enantioselectivity investigations with more than 30 di-
amines of types 9 and 10. In addition, further studies on the
substrate scope, some mechanistic investigations on the origin
of the excellent enantioselectivities reached, and the prepara-
tion of the new diamines used in this study are described.
12c required a deprotection–reductive cyclization–reprotection
sequence.[11] The cis diastereoselectivity was high in cyclizations
(dr>90:10). Final exhaustive reduction with LAH in refluxing
THF afforded the prolinol intermediates 14a–c in 91–95%
yield.
The alcohols 14a–c thus prepared and the known deriva-
tives 14d (R1 =Me)[11] and 14e (R1 =Ph)[9] were converted into
the prolinamines 9 and 10 by mesylation of the hydroxy func-
tion and subsequent amination with an excess of the respec-
tive amine HNR3R4 (Table 2). The conversions of these reactions
were good,[13] but the high polarity of the resulting diamines
led to, in part, significant losses during column chromato-
graphic purification, thus lowering the isolated yields to 50–
76%.
Results and Discussion
Table 2. Preparation of the new prolinamines 9 and 10 from 14.
Synthesis of the prolinamines
A fast and variable access to prolinamines of type 9 and 10
was essential for the extensive ligand screening planned. We
recently developed several routes to this class of diamines that
all start from commercially available methyl Boc-l-pyrogluta-
mate (11), but differ in the order of introduction of the sub-
stituents R1–R4, thus permitting a maximum of flexibility.[11] The
new prolinamines used in this study were prepared with focus
on a late-stage installation of the exocyclic amino function
NR3R4, which is most easily achieved by hydroxy–amine ex-
change on the stage of the prolinol precursors 14.
The substituent R1 in 5-cis position was attached by chemo-
selective Grignard addition to the pyrrolidine carbonyl group
in 11 and reductive cyclization of the resulting b-amino ke-
tones 12 (Table 1). In accordance with earlier results,[11] the
yield of the initial addition step strongly depended on the
steric hindrance of the Grignard reagent. Good 78% were
reached with 3,5-Me2PhMgBr, whereas just mediocre 37% and
31% were obtained with the more bulky secondary alkyl
Grignards cPentMgBr and cHexMgCl, respectively. The aliphatic
b-amino ketones 12a and 12b were directly cyclized to the
corresponding prolines 13a and 13b by using NaBH(OAc)3 as
the reductant, whereas ring closure of the aromatic derivative
Entry
14
R1
9, 10
NR3R4
Yield [%][a]
1
2
3
4[b]
5[d]
6[d]
7
8
9
10
11
a
b
c
cPent
cHex
3,5-Me2Ph
Me
Ph
Ph
Ph
Ph
cPent
cHex
3,5-Me2Ph
9b
9c
9d
9e
NHMe
NHMe
NHMe
NHMe
NHAc
NHMs
NH(CH2)2OH
NH(CH2)2OMe
NMe2
54
56
76
52[c]
78[c]
80[c]
50
73
57
57
73
d
e
e
e
e
a
b
c
9 f
9g
9h
9i
10a
10b
10c
NMe2
NMe2
[a] Isolated yield. [b] Two-step sequence: 1. MsCl, NEt3, then HN(Me)Bn; 2.
H2, Pd(OH)2/C. [c] Yield over two steps. [d] Two-step sequence: 1. MsCl,
NEt3, then NH3-MeOH (85%)[11] 2. for 9 f: Ac2O, NEt3; for 9g: MsCl, NEt3.
Notably, the direct preparation of 9e (R1 =Me, NR3R4 =
NHMe, Table 2 entry 4) from 14d by using the standard proce-
dure, mesylation and amination with methylamine, failed. The
pronounced volatility of the product made a removal of
a higher boiling solvent such as MeOH, which was required as
co-eluent in the chromatography of 9e, practically impossible.
We circumvented this problem by amination of 14d with ben-
zylmethylamine, giving the less polar and less volatile N-benzyl
derivative of 9e, which could be purified. Hydrogenolytic de-
benzylation under acidic conditions, basic extraction into Et2O,
and careful evaporation delivered 9e in high purity and ac-
ceptable 52% yield over two steps. Finally, the amides 9 f and
9g were synthesized by a two-step sequence (entries 5 and 6).
Amination of 14e with ammonia afforded the corresponding
primary amine,[11] which was converted into 9 f and 9g by N-
acetylation and N-mesylation, respectively.
Table 1. Preparation of the prolinols 14 from 11.
Entry
R1
Yield of 12[a]
Yield of 13[a]
Yield of 14[a]
[%]
[%]
[%]
1
2
3
cPent
cHex
3,5-Me2Ph
37 (12a)
31 (12b)
78 (12c)
90 (13a)
70 (13b)
59 (13c)[b]
92 (14a)
95 (14b)
91 (14c)
Optimization of the catalytic system
All enantioselective Henry reactions were performed under an
argon atmosphere in a well-tempered cooling bath. In the case
[a] Isolated yield. [b] 93:7 mixture of 13c and its C5-epimer.
ChemCatChem 2016, 8, 1846 – 1856
www.chemcatchem.org
1847
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 3. First structure–selectivity investigations: optimization of the substituents R1–R4.[a]
Entry
Diamine
R1
R2
NR3R4
t [h]
Yield [%][b]
ee [%][c]
Configuration[d]
1
2
3
4
5
6
7
8
10d
10e
10 f
10g
10a
10b
10h
10i
10j
10c
10k
10l
10m
9a
9j
9k
9l
9m
9 f
9g
9h
9i
9n
H
Me
Bn
iPr
cPent
cHex
Ph
4-MeOPh
3,5-(CF3)2Ph
3,5-Me2Ph
1-naphthyl
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
H
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
N(Me)tBu
pyrrolidinyl
NHMe
NHEt
NHiPr
NHtBu
NHPh
24
18
24
40
18
18
20
24
24
18
24
48
40
19
48
48
48
48
24
24
40
40
48
113
40
40
24
99
93
99
99
95
93
95
99
93
92
72
0
99
99
70
50
25
0
71
23
13
84
87
88
84
83
88
90
87
–
94
98
98
85
30
–
R
S
S
S
S
S
S
S
S
S
S
–
S
S
S
S
S
–
–
–
S
S
S
S
S
S
–
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
NHAc
0
0
–
–
NHMs
NH(CH2)2OH
NH(CH2)2OMe
NH2
NHMe
NHMe
NHMe
NHMe
35
34
28
23
99
32
0
84
95
93
77
98
90
–
9o
9p
9q
9r
Et
Bn
iPr
Ph
Ph
[a] Performed on a 1 mmol scale in MeOH (600 mL) and MeNO2 (600 mL). [b] Isolated yield. [c] Determined by HPLC on chiral phase and rounded off to
whole numbers. [d] Assigned by comparison with literature data.
of an important or unexpected result, the reaction was repeat-
ed at least twice. The enantiomeric excess of the products 3
was determined by HPLC on chiral phase with an accuracy of
up to Æ0.1 percentage points.
copper atom in the catalyst into account (see complex [M·9/
10] in Scheme 2, with R1 =H, R2–4 =Me).
In a first set of experiments we kept the methyl groups for
R2À4 and varied the 5-cis substituent R1 (entries 2–11), which
was assumed to exert a strong effect on the chirality transfer.
And indeed, its impact is clearly seen on the sense of the
asymmetric induction. Compared to the reaction with 10d
(R1 =H), the enantiomeric product, (S)-3a, was preferentially
formed with all prolinamines carrying such a substituent (R1 ¼
H). The level of stereoinduction rose with an increasing steric
demand of R1. Good enantioselectivities of 83–90% ee were
reached with all diamines that possess an a-branched aliphatic
or an aromatic substituent R1 as in 10a–c,g–k (entries 4–11).
The good chirality transfers with the aliphatic diamines also ex-
clude a decisive role of a p–p-stacking between R1 and the
aromatic substrate benzaldehyde (1a). Among the promising
prolinamines, we chose to continue the ligand optimization
with derivatives possessing a phenyl group as R1, since these
compounds are most easily accessible (for a reinvestigation on
R1 under optimized conditions, see Table 7).
Ligand structure (I)
The initial ligand screening was done on the addition of nitro-
methane (2a) to benzaldehyde (1a) as the model reaction
(Table 3), by using the following protocol: The chiral catalyst
(4 mol%), prepared prior to use from CuCl2 (4.0 mol%) and
a slight excess of the chiral diamine 9 or 10 (4.4 mol%), and
the aldehyde 1a were dissolved in a 1:1 mixture of MeNO2
(ꢀ11 equivalents with respect to 1a) and MeOH. After cooling
to À208C, the reaction was started by addition of the ancillary
base NEt3 (3.0 mol%) and stirred for 18–113 h. Under these
conditions, the most simple diamine, the 5-cis-unsubstituted
prolinamine 10d (R1 =H), which furthermore possesses a pyrro-
lidine N-methyl and an exocyclic dimethylamino group (R2–4
=
Me), provided the R-configured b-nitro alcohol (R)-3a in ac-
ceptable 71% ee and excellent 99% yield after 24 h (Table 3,
entry 1). The level of enantioselection reached was quite re-
markable, taking the low steric differentiation around the
The influence of the substituents R3 and R4 at the exocyclic
aminomethyl group was investigated next (Table 3, entries 12–
23). Increasing the size of one of these substituents as in 10l
(NR3R4 =N(Me)tBu) caused a complete breakdown in reactivity.
ChemCatChem 2016, 8, 1846 – 1856
www.chemcatchem.org
1848
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
With pyrrolidinyl instead of NMe2, improved 94% ee were
reached. Another gain in stereocontrol was observed upon
switching to the prolinamines 9, which carry secondary amino-
methyl groups NHR4 (entries 14–22). Excellent 98% ee were
reached with the diamines 9a (NHMe) and 9j (NHEt), whereas
bulkier substituents R4 as in 9k,l (NHiPr, NHtBu) resulted in di-
minished asymmetric inductions. As a general trend, the cata-
lytic activity significantly dropped with increasing steric
demand of R4, which is clear from the falling yields in the row
9a, 9j, 9k to 9l, even at prolonged reactions times. No prod-
uct formation was observed with the anilinyl derivative 9m
(NHPh) and the amides 9 f (NHAc) and 9g (NHMs). The poten-
tially tridendate diamines 9h (NH(CH2)2OH) and 9i
(NH(CH2)2OMe) and the primary diamine 9n (NH2) provided (S)-
3a in acceptable 84–95% ee, but low 28–35% yield.
After having identified the NHMe group as the optimal
NR3R4 function, we finally turned our attention to the substitu-
ent R2 at the pyrrolidine nitrogen atom (entries 24–27). The
same trend as with the NR3R4 group was observed: Excellent
asymmetric inductions of 98% ee were achieved with small R2
as in 9a and 9p (R2 =Me, Et), while larger substituents R2 as in
9q and 9r (R2 =Bn, iPr) or an NH function as in 9o drastically
reduced the activity of the catalyst.
Table 4. Variation of the copper salt, solvent, and the solvent–MeNO2
ratio.[a]
Entry
Cu Salt
Solvent
Solvent:MeNO2
t
[h]
Yield
[%][b]
ee
[%][c]
1[d]
2
3
4
5
6
7
8
9[e]
10[e]
11[e]
12[e]
13
14
15
16
17[g]
18[h]
CuCl2
CuCl2
CuCl2
CuCl2
CuBr2
CuBr2
CuBr2
CuBr2
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
CuBr2
CuBr2
CuBr2
CuBr2
MeOH
EtOH
THF
MeNO2
MeOH
EtOH
THF
MeNO2
MeOH
EtOH
THF
MeNO2
THF
1:1
1:1
1:1
0:2
1:1
1:1
1:1
0:2
1:1
1:1
1:1
0:2
1:1
1:1
3:1
1:3
1:1
1:1
19
20
18
21
22
21
18
22
70
70
70
70
17
15
16
16
16
16
99
99
99
99
90
99
99
99
7
12
30
20
88
97
99
99
99
99
97.7
98.2
98.3
98.3
98.7
98.9
99.0
98.0
98.2
99.0
98.7
97.8
90.8
91.2
99.1
98.6
99.0
99.1
[f]
[f]
[f]
[f]
[f]
[f]
MeOH
THF
THF
THF
THF
In summary, the best result (98% ee, 99% yield) was ach-
ieved with the prolinamine 9a possessing a phenyl substituent
in 5-cis position, a pyrrolidine N-methyl group, and a 2-(meth-
ylaminomethyl) side chain. All further experiments were there-
fore performed with this diamine.
[a] Performed on a 1 mmol scale in the respective solvent–MeNO2 mixture
(1200 mL total, c(1a)=0.83m). [b] Isolated yield. [c] Determined by HPLC
on chiral phase. [d] See Table 3, entry 14. [e] No NEt3 added. [f] Dihydrate.
[g] Reaction in 600 mL solvent, c(1a)=1.66m. [h] Reaction in 2400 mL sol-
vent, c(1a)=0.42m.
A short base screening (Table 5, entries 1–3) revealed that
the steric demand of the base is not of importance, which is
clear from the excellent 99% yield and 99.0% ee reached with
both, NEt3 and EtNiPr2. A sufficient basicity, however, was re-
Reaction conditions
The copper source (CuCl2, CuBr2, and Cu(OAc)2) and the solvent
(MeOH, EtOH, THF, and MeNO2) were varied first (Table 4, en-
tries 1–12). The influence of both parameters on the chirality
transfer was marginal, which is clear from the excellent 97.7–
99.0% ee obtained in all cases. A distinct difference in reactivi-
ty and, thus, in the yields, was observed between the copper
halide and the copper acetate complexes. In the latter Henry
reactions, no NEt3 was added since the acetate freed from the
catalyst upon coordination of the substrates can act as the
base.[14] The low 7–30% yield obtained after 70 h are presuma-
bly a consequence of the weaker basicity of acetate, which
slows down the deprotonation of nitromethane. Addition of
NEt3 (3 mol%, entries 13 and 14) accelerated the reaction
(ꢁ88% yield after 17 h), but resulted in lower stereocontrol
(91% ee). A closer inspection of the enantioselectivities ach-
ieved with the CuCl2 and CuBr2 complexes revealed the latter
ones as slightly superior (98.0–99.0% ee vs. 97.7–98.3% ee). All
solvents examined permitted similar levels of chirality transfer,
but the reaction with CuBr2 in THF seemed to proceed some-
what faster. Since this will be beneficial for lower-temperature
reactions (see Table 6), we decided to continue with this com-
bination. Changes in the solvent–MeNO2 ratio from 1:1 to 3:1
and 1:3 (entries 15 and 16) as well as in the concentration
from 0.83m to 1.66m and 0.42m (entries 17 and 18) had no
noticeable effect on yield and enantioselectivity.
Table 5. Variation of the base and the catalyst loading.[a]
Entry
1, 3
CuBr2·9a : Base
[mol%/mol%]
Base
t
[h]
Yield
[%][b]
ee
[%][c]
1[d]
2
3
4
5
6
7
8
a
a
a
a
a
a
a
a
a
b
b
b
b
4.0:3.0
4.0:3.0
4.0:3.0
4.0:3.0
4.0:6.0
4.0:1.0
2.0:1.5
1.0:0.75
0.50:0.375
4.0:3.0
2.0:1.5
1.0:0.75
0.50:0.375
NEt3
EtNiPr2
pyridine
–
NEt3
NEt3
NEt3
NEt3
NEt3
NEt3
NEt3
NEt3
NEt3
18
20
20
16
21
16
18
17
41
17
17
17
42
99
99.0
99.0
–
99
traces
0
–
99
13
99
48
7
99
99
64
98.9
99.0
99.1
99.2
98.5
99.0
98.9
99.0
94.4
9
10
11
12
13
13
[a] Performed on a 1 mmol scale in THF (600 mL) and MeNO2 (600 mL),
9a : CuBr2 =1.1:1. [b] Isolated yield. [c] Determined by HPLC on chiral
phase. [d] See Table 4, entry 7.
ChemCatChem 2016, 8, 1846 – 1856
www.chemcatchem.org
1849
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
quired, because only traces of product were formed in the
presence of pyridine. This observation is in good agreement
with the slow reaction rates observed for the Cu(OAc)2 com-
plexes in which acetate served as the base (see Table 4, en-
tries 9–12). As expected, there was no reaction without a base
(Table 5, entry 4).
pected, an increase in stereocontrol was observed by lowering
the temperature to À308C, giving the b-nitro alcohols
(S)-3a–c in excellent 99.1–99.5% ee. The reaction rates, howev-
er, markedly dropped below À258C, which is clear from the
prolonged reaction times required and the incomplete conver-
sion of the least reactive aldehyde 1c. We therefore choose
À258C as a good compromise between yield and chirality
transfer.
Changing the ratio catalyst–NEt3 from standard 4:3 to 4:6 or
4:1 had little to no effect on the enantioselectivity (Table 5, en-
tries 5 and 6). The yield, however, dropped to mere 13% if just
1 mol% of NEt3 was used. The catalyst loading can be reduced
to 2 mol% without any loss in yield and stereocontrol, if the
catalyst–NEt3 ratio is kept constant at 4:3 (entry 7). With just
1 mol% of catalyst and 0.75 mol% of base, the hitherto best
enantioselection of 99.2% ee was achieved (entry 8). Although
the 48% yield reached are just mediocre, the level of conver-
sion is quite surprising as compared to the reaction with
4 mol% CuBr2·9a and 1 mol% NEt3 (see entry 6), which provid-
ed just 13% product within the same time frame, despite of
the higher amounts of base and catalyst. Further lowering of
the catalyst loading to 0.5 mol% resulted in a slight loss of
asymmetric induction (98.5% ee), but a drastically reduced
yield (7% after 41 h, entry 9).
Ligand structure (II)
At this final stage we decided to reinvestigate the influence of
the 5-cis substituent R1, because there had been no clear pref-
erence for a particular group in the initial screening (see
Table 3, entries 1–11). The re-evaluation was performed under
the optimized reaction conditions with the secondary prolin-
amines 9b–e,s,t carrying the better stereo-differentiating exo-
cyclic NHMe group. As seen in Table 7, all derivatives of 9 with
Table 7. Reinvestigation of the influence of the substituent R1.[a]
At this point we checked that our optimization was not too
substrate specific. As electron-deficient aldehydes might more
readily undergo the uncatalyzed background reaction (vide
infra), and, thus, require higher catalyst loadings, the latter ex-
periments were repeated with 2-nitrobenzaldehyde (1b, en-
tries 10–13). In the presence of 2 mol% catalyst, also this sub-
strate provided the corresponding b-nitro alcohol (S)-3b in ex-
cellent 98.9% ee and 99% yield. Further reduction of the
amount of catalyst to 0.5%, however, led to a significantly
stronger depletion in enantioselectivity (94.4% ee), as com-
pared to the analogous reaction with benzaldehyde (1a, see
entry 9).
Entry
9
R1
t [h]
Yield [%][b]
ee [%][c]
1
2
3
4
s
e
t
b
c
a
d
H
Me
iPr
cPent
cHex
Ph
18
18
18
19
18
24
41
95
95
99
99
99
92
99
25.2
96.0
96.9
97.0
97.7
99.3
98.2
5
6[d]
7
3,5-Me2Ph
The last parameter, the temperature, was optimized with
benzaldehyde (1a), 2-nitrobenzaldehyde (1b), and 2-methoxy-
benzaldehyde (1c) as the model substrates (Table 6). As ex-
[a] Performed on a 1 mmol scale in THF (600 mL) and MeNO2 (600 mL).
[b] Isolated yield. [c] Determined by HPLC on chiral phase. [d] See Table 6,
entry 2.
an aliphatic or aromatic substituent R1 provided the b-nitro al-
cohol (S)-3a with high stereocontrol (ꢁ96.0% ee), even the di-
amine 9e, which possesses the small methyl group. The best
stereoselection (99.3% ee) was achieved with the phenyl-sub-
stituted prolinamine 9a (Table 7, entry 6), maybe as a result of
the optimization process done on this compound. The surpris-
ing reversal in the sense of enantioselection, as it had been
found with the tertiary diamine 10d missing the substituent R1
(see Table 3, entry 1), was not observed for the secondary pro-
linamine 9s, which also afforded the S-configured product
(S)-3a, albeit in low 25% ee.
Table 6. Optimization of the temperature.[a]
Entry
1, 3
T [8C]
t [h]
Yield [%][b]
ee [%][c]
1[d]
2[e]
3
4[f]
5[e]
6
a
a
a
b
b
b
c
À20
À25
À30
À20
À25
À30
À20
À25
À30
18
24
66
17
20
66
40
42
67
99
92
99
99
97
99
99
97
55
99.1
99.3
99.5
98.9
99.0
99.1
99.2
99.5
99.5
7
Aliphatic aldehydes
8[e]
9
c
c
When applying the optimized conditions to the Henry reaction
of the aliphatic aldehyde nonanal (1d), the b-nitro alcohol (S)-
3d was produced in disappointing 53% yield and 94.5% ee
(Table 8, entry 1). The yield and the level of enantioselection,
[a] Performed on a 1 mmol scale in THF (600 mL) and MeNO2 (600 mL).
[b] Isolated yield. [c] Determined by HPLC on chiral phase. [d] See Table 5,
entry 7. [e] Data taken from Ref. [9]. [f] See Table 5, entry 11.
ChemCatChem 2016, 8, 1846 – 1856
www.chemcatchem.org
1850
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
forced conditions, the Henry products (S)-3a and (S)-3b were
formed in high 94% yield each. The enantiomeric excess
(99.0% and 98.9% ee, respectively) was as good as in the
small-scale reactions.
Table 8. Optimization of the reaction conditions for aliphatic aldehydes
with 1d as the model compound.[a]
Entry
Cu Salt ([mol%])
T [8C]
t [h]
Yield [%][b]
ee [%][c]
Extending the substrate scope
1
2
3
4
5[d]
6
CuBr2 (2)
CuBr2 (4)
CuBr2 (8)
CuCl2 (8)
CuCl2 (8)
CuCl2 (8)
À25
À25
À25
À25
À20
À10
40
60
24
21
60
16
53
75
90
87
97
77
94.5
96.2
97.0
98.2
98.6
97.1
The good performance of the prolinamine 9a in Henry reac-
tions prompted us to further study its scope and limitations. A
tempting substrate is nicotinaldehyde (1e, Scheme 4) because
of its basic and nucleophilic pyridine moiety, which might pro-
mote the uncatalyzed background reaction[15] and, in addition,
might competitively coordinate to the catalyst, thus reducing
the amount of catalytically active species. And indeed, the
Henry reaction of 1e in the presence of the catalyst CuBr2·9a
(2 mol%) proceeded sluggishly and delivered (S)-3e in unsatis-
fying 43% yield after 10 d and with low 82% ee. To accelerate
the catalyzed reaction, we raised the amount of CuBr2·9a to
15 mol%. Under these conditions, (S)-3e was obtained in im-
proved 81% yield and acceptable 90% ee after 5 days.
[a] Performed on a 1 mmol scale in THF (600 mL) and MeNO2 (600 mL).
[b] Isolated yield. [c] Determined by HPLC on chiral phase. [d] Data taken
from ref. [9].
however, were raised by increasing the amount of catalyst to
8 mol% and the temperature to À208C, and by changing the
copper source to CuCl2. Under these conditions, the product
(S)-3d was obtained in excellent 97% yield and high 98.6% ee.
One observation made in this context is noteworthy: The
enantioselectivity of the reaction at À258C with CuCl2·9a as
the catalyst was slightly lower than the one at À208C (entry 4
vs. 5). This unexpected result might have its origin in a begin-
ning aggregation of the catalyst at À258C, as judged from the
increasing turbidity of the reaction mixture, which would
reduce the amount of active catalyst and, thus, favor the non-
stereoselective background reaction. A similar effect was not
observed for the complex CuBr2·9a in the reaction with aro-
matic aldehydes (see Table 6).
Scheme 4. Henry reactions with basic nicotinaldehyde (1e).
Diastereo- and enantioselective Henry reactions[16] with nitro-
alkanes 2 (R’¼ H) were first studied using benzaldehyde (1a)
as the substrate (Table 9). Owing to the lower reactivity of the
nitroalkanes 2b–d (R’=Me, Et, CH2OTBS), the following reac-
tions were performed at À208C and with 8 mol% catalyst
CuBr2·9a. Whereas the syn–anti ratio in the nitroethane (2b)
derived product 3 f was meager (60:40), acceptable ratios of
78:22 were obtained in 3g and 3h prepared from nitropro-
pane (2c) and sterically more demanding 2-TBSO-nitroethane
(2d), respectively. The enantioselectivities were always excel-
lent (ꢁ98% ee) in the major syn products and acceptable to
good (82–93% ee) in the minor anti products. A further gain in
selectivity was reached with the combination cyclohexanecarb-
aldehyde (1 f)-nitropropane (2c), which provided the product
3i in a good 84:16 syn-anti ratio and with 99% ee in both dia-
stereomers. Thus, the catalyst CuBr2·9a is also well suited for
enantio- and diastereoselective Henry reactions.
Under the optimized conditions for aromatic aldehydes
[CuBr2 (2 mol%), 9a (2.2 mol%), NEt3 (1.5 mol%), THF/MeNO2 =
1:1, À258C] and aliphatic aldehydes [CuCl2 (8 mol%), 9a
(8.8 mol%), NEt3 (6.0 mol%), THF/MeNO2 =1:1, À208C], enan-
tioselective Henry reactions with a broad variety of substrates
were performed, providing the excellent results reported earli-
er.[9]
Gram-scale reactions
Finally, we decided to prove the practicability of our new cata-
lytic system in gram-scale reactions (10 mmol aldehyde) with
benzaldehyde (1a) and 2-nitrobenzaldehyde (1b) as the model
substrates (Scheme 3). To further demonstrate its effectiveness,
we cut, compared to the optimized procedure above, the
amount of catalyst CuBr2·9a in half (1 mol%), which was the
minimum amount required to preserve the excellent stereo-
control (see Table 5, entries 8 and 12). Even under these en-
Mechanistic investigations
Origin of enantioselection
Next we put our focus on the origin of the enantioselection.
We wanted to prove that the high levels of stereocontrol
solely arise from a kinetic differentiation in the C,C-coupling
step and that processes involving product species, as, for ex-
ample, an additional resolution on the stage of the primarily
Scheme 3. Gram-scale Henry reactions of 1a and 1b.
ChemCatChem 2016, 8, 1846 – 1856
www.chemcatchem.org
1851
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 9. Enantio- and diastereoselective Henry reactions with nitroalkanes 2b-d.[a]
Entry
1
R
2
R’
t [d]
3
Yield [%][b]
syn:anti[c]
eesyn [%][d]
eeanti [%][d]
1
2
3
4
a
a
a
f
Ph
Ph
Ph
cHex
b
c
d
c
Me
Et
CH2OTBS
Et
4
4
7
7
f
99
99
98
84
60:40
78:22
78:22
84:16
99
98
99
99
93
82
93
99
g
h
i
[a] Performed on a 1 mmol scale in THF (600 mL) and nitroalkane 2 (8 equiv). [b] Isolated yield. [c] Determined by 1H NMR. [d] Determined by HPLC on
chiral phase.
Uncatalyzed background reaction
resulting, diastereomeric product–catalyst complexes, do not
contribute.
A necessary precondition for any dynamic process that influ-
ences the stereochemical outcome on the stage of the prod-
ucts is an equilibrium between the product species and the
starting materials. The existence of such an equilibrium, al-
though more or less fully shifted towards the products, was
demonstrated by treatment of the b-nitro alcohol (S)-3a with
4-nitrobenzaldehyde (1g) under standard conditions
(Scheme 5). The cross product 3j observed in this reaction
From the ꢁ98.9% ee reached with 1a and 1b in the presence
of just 1 mol% of CuBr2·9a (see Scheme 3) it follows that the
catalyzed reaction must proceed at least 99 times faster than
the non-stereoselective background reaction.[17] This is in good
agreement with the observation that the latter one is virtually
non-existing for benzaldehyde (1a: <1% conversion within
24 h, Scheme 7). In contrast to that, the rate of the background
Scheme 7. Uncatalyzed background reactions of 1a and 1b.
Scheme 5. The formation of the cross product (S)-3j from (S)-3a and 1g
proves the reversibility of the Henry reaction under the standard reaction
conditions.
reaction is surprisingly high for the more electrophilic 2-nitro-
benzaldehyde (1b, 59% conversion within 24 h). Although this
does not noticeably affect the enantioselectivity of the cata-
lyzed reaction with 1 mol% of CuBr2·9a, it is most likely the
main reason for the more pronounced loss in stereocontrol in
the reaction of 1b with 0.5 mol% catalyst, as compared to the
analogous reaction of 1a (98.5% vs. 94.4% ee, see Table 5,
entry 9 vs. 13).
must have been formed via a retro-Henry–Henry sequence.
The rate of the back reaction, however, is pretty slow, which is
clear from the low 6% yield obtained after 7 d. The good 94%
ee indicates that at least the formation of (S)-3j must have
been catalyzed by CuBr2·9a.
With the existence of the back reaction proven, the question
remained whether this process induces any changes in the
enantiopurity of the product. This cannot be the case because
stirring of the scalemic b-nitro alcohol (S)-3a (55% ee) under
standard conditions for 7 d did not noticeably alter its optical
purity (Scheme 6). Thus, any scenario that affects the overall
stereochemical outcome and involves product species can be
safely excluded. The excellent stereodifferentiation observed
must have its origin exclusively in the C,C-coupling step.
Relative reactivities of the catalysts derived from 9a and 10h
In parallel to the significantly enhanced stereocontrol reached
with prolinamines of type 9 (secondary exocyclic amino
group), as compared to those of type 10 (tertiary exocyclic
amino group), we also observed a gain in reactivity. We there-
fore decided to measure the relative rates of reactions in the
presence of our prime catalyst CuX2·9a in comparison to those
with CuX2·10h (10h: dimethyl analogue of 9a) under different
conditions. Competition experiments, in which equimolar
amounts of two chiral catalysts are used that deliver enantio-
meric products, offer an experimentally simple method to do
this, without the necessity of extensive kinetic studies. The rel-
ative rate constant krel can be calculated from the enantiomeric
ratios of the single-catalyst and competition experiments using
Equation (1):[18]
Scheme 6. The scalemic b-nitro alcohol (S)-3a (55% ee) is configurationally
stable under standard Henry conditions.
ChemCatChem 2016, 8, 1846 – 1856
www.chemcatchem.org
1852
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
of the catalysts CuX2·9, as compared to CuX2·10. Since the op-
posite effect was observed in the competition experiments,
the existence of such a hydrogen bridge seems unlikely.[19] The
higher enantioselectivities reached with secondary prolin-
amines 9 presumably originate from steric and conformational
factors.
The required pseudo-enantiomeric catalyst ent-9a was pre-
pared in analogy to 9a, but starting from methyl Boc-d-pyro-
glutamate (ent-11).[9,11] The single-catalyst and competition ex-
periments were done with the CuBr2 and CuCl2 catalysts de-
rived from 10h (2S,5R-configuration) and ent-9a (2R,5S-config-
uration) in THF and MeOH at À258C (Table 10). In THF and
with CuBr2·ent-9a, the product (R)-3a was obtained after 18 h
in good 92% yield and excellent 99% ee, while the analogous
reaction with CuBr2·10h proceeded more slowly (58% yield
after 43 h) and delivered the enantiomer (S)-3a with signifi-
cantly lower stereocontrol (84% ee). The competition experi-
ment (Table 10, entry 3) with equimolar amounts of both cata-
lysts provided (R)-3a in 50% ee, clearly proving the higher cat-
alytic activity of CuBr2·ent-9a. By using Equation (1), a relative
rate factor krel (=kentÀ9a k10hÀ1) of 2.73 in favor of CuBr2·ent-9a
was calculated. A similar krel value of 3.54 was observed for the
corresponding chloro complexes CuCl2·ent-9a and CuCl2·10h
(entries 4–6). MeOH as the solvent (entries 7–9) causes a gener-
al decrease in reactivity, which is clear from the prolonged re-
action times required, which might be an effect of its better
coordination abilities favoring a deactivation of intermediate
catalyst species. In addition, the ratio krel of the reaction rates
is higher: The catalytic system CuCl2·ent-9a reacted 7.32 times
faster than CuCl2·10h. This furthermore underlines the exis-
tence of direct solvent-catalyst interactions—if this were not
the case, the same krel values in THF and MeOH would be ex-
pected.
Conclusions
Several new, 5-cis-substituted prolinamines of type 9 and 10
were synthesized in 4–6 steps from methyl Boc-l-pyrogluta-
mate (11). Their potential as the chiral ligands in enantioselec-
tive, copper-catalyzed Henry reactions was evaluated. In-depth
structure–selectivity investigations with more than 30 diamines
9 and 10 revealed that an aromatic or sufficiently bulky ali-
phatic substituent in 5-cis position is crucial for high levels of
stereocontrol, while larger groups at the pyrrolidine nitrogen
atom or at the exocyclic aminomethyl group cause an, in part,
drastic loss in reactivity and enantioselectivity. The prolinamine
9a (R1 =Ph; R2 =Me; NR3R4 =NHMe) was found to be the
chiral ligand of choice. Optimization of other reaction parame-
ters, such as temperature, solvent, concentration and catalyst
loading, led to two highly efficient catalytic systems, CuBr2·9a
for aromatic aldehydes and CuCl2·9a for aliphatic ones. With
just 2 mol% of catalyst (8 mol% in the case of aliphatic alde-
hydes), the superb results reported earlier (99% ee with 36 al-
dehydes) were achieved.[9] In further studies we extended the
scope of CuBr2·9a to gram-scale and diastereoselective Henry
reactions (up to 84:16 dr, 99% ee). It was also proven that the
stereodifferentiation originates solely from the C,C-coupling
step and that the product is configurationally stable under the
reaction conditions. The uncatalyzed background reaction is
virtually non-existing for benzaldehyde (1a), but remarkably
high for the more electrophilic 2-nitrobenzaldehyde (1b). The
relative reactivity of the catalysts derived from the prolin-
amines 9a and 10h was studied by competition experiments.
The complexes with 9a reacted up to 7.32 times faster, de-
pending on the reaction conditions. This, however, does not
explain the significant increase in enantioselectivity observed
In the transition states of the Henry reactions with the sec-
ondary prolinamines 9, there is the possibility of an additional
hydrogen bridge between the NH function of the chiral ligand
and the nitronate bound to the copper atom, which would fur-
ther rigidify the system and, thus, explain the better stereocon-
trol observed.[9] This interaction should reduce the nucleophi-
licity of the nitronate and, in consequence, lower the activities
Table 10. Single-catalyst and competition experiments with ent-9a and 10h as the chiral diamines.[a]
À1[d]
Entry
Diamine
Cu Salt
Solvent
t [h]
Yield [%][b]
ee [%][c]
Configuration
krel =kentÀ9a k10h
1
2
3
4
5
6
7
8
9
ent-9a
10h
ent-9a/10h 1:1[e]
ent-9a
CuBr2
CuBr2
CuBr2
CuCl2
CuCl2
CuCl2
CuCl2
CuCl2
CuCl2
THF
THF
THF
THF
THF
THF
MeOH
MeOH
MeOH
18
43
19
18
39
19
40
39
41
92
58
87
99
81
77
82
47
78
99
84
50
99
87
58
98
85
76
R
S
R
R
S
R
R
S
R
2.73
3.54
7.32
10h
ent-9a/10h 1:1[e]
ent-9a
10h
ent-9a/10h 1:1[e]
[a] Performed on a 1 mmol scale in THF (600 mL) and MeNO2 (600 mL). [b] Isolated yield. [c] Determined by HPLC on chiral phase. [d] Calculated using Equa-
tion (1). [e] The catalysts derived from ent-9a and 10h were prepared separately and mixed shortly before use.
ChemCatChem 2016, 8, 1846 – 1856
www.chemcatchem.org
1853
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
with 9a (NR3R4 =NHMe), as compared to its dimethyl ana-
logue 10h (NR3R4 =NMe2).
(2S,5R)-1-tert-Butyl 2-methyl 5-cyclopentylpyrrolidine-1,2-di-
carboxylate (13a)
NaBH(OAc)3 (5.53 g, 26.1 mmol) was added at 08C to a solution of
the amino ketone 12a (4.20 g, 13.4 mmol) in EtOAc (60 mL). After
10 min, TFA (6.66 mL, 9.85 g, 86.4 mmol) was added dropwise and
the reaction mixture was stirred overnight at RT. Sat. aq. NaHCO3
(200 mL) was added and the reaction mixture was extracted with
EtOAc (3”200 mL). The combined organic layers were dried over
MgSO4 and the solvent was evaporated. Column chromatography
(silica gel, petroleum ether–EtOAc, 15:1–4:1) provided diastereo-
merically pure 13a (3.59 g, 12.1 mmol, 90%) as a colorless oil. Rf =
Experimental Section
All reactions with moisture-sensitive reagents were performed
under an argon atmosphere in anhydrous solvents, prepared using
standard procedures.[20] Commercially available reagents (highest
quality available) were used as received. Reactions were monitored
by thin layer chromatography on precoated silica gel (Macherey–
Nagel, Alugram SIL G/UV254). Spots were visualized by UV light
(254 nm) or by staining with aqueous KMnO4, vanillin, or ceric am-
monium molybdate. Silica gel (Macherey–Nagel, particle size 40–
63 mm) was used for column chromatography. Optical rotations
were recorded on a Jasco P-1020 polarimeter (10 cm cell). NMR
spectra were taken on a Bruker Avance III HD 500 instrument and
calibrated using the residual undeuterated solvent as an internal
0.69 (petroleum ether/EtOAc 3:1); ½aŠ32 =À26.5 (c=1.00 in MeOH);
D
IR (ATR): n˜max =2948, 2869, 1756, 1694, 1455, 1387, 1365, 1167,
1
1139, 1108 cmÀ1; H NMR* (500 MHz, CDCl3): d=1.12 (m, 1H, cPent-
H), 1.36 (s, 5.4H, C(CH3)3), 1.42 (s, 3.6H, C(CH3)3), 1.49 (m, 3H,
cPent-H), 1.62 (m, 3H, cPent-H), 1.78 (m, 3H, 4-H2, cPent-H), 1.86–
2.13 (m, 2H, 3-HH, cPent-H), 2.20 (m, 1H, 3-HH), 3.69 (s, 3H, OCH3),
3.74 (m, 0.4H, 5-H), 3.86 (t, J=7.9 Hz, 0.6H, 5-H), 4.17 (t, J=8.6 Hz,
0.6H, 2-H), 4.29 ppm (t, J=8.3 Hz, 0.4H, 2-H); 13C NMR* (125 MHz,
CDCl3): d=25.0, 25.1, 25.3, 27.9 (C-cPent), 28.3, 28.5 (C(CH3)3), 28.8,
28.9 (C-3, C-4), 29.5, 29.7, 30.3, 30.5, 44.6, 44.8 (C-cPent), 51.9, 52.1
(OCH3), 59.6, 60.1 (C-2), 62.4, 62.7 (C-5), 79.7, 80.0 (C(CH3)3), 154.4,
155.0 (1-CO2), 174.1, 174.3 ppm (2-CO2); HRMS (ESI, pos.) m/z calcd
for C16H27NO4 [M+H]+ 298.20128, found 298.20134. * 60:40 mixture
of rotamers.
reference. The peak assignments in the H and 13C NMR data were
1
performed on basis of 2D NMR methods (COSY, HSQC, HMBC). In-
frared spectra were recorded on a PerkinElmer Spectrum 100 FT-IR
spectrometer, high-resolution mass spectra were recorded on
a ThermoFisher Scientific Q-Exactive (Orbitrap) or a Bruker Dalton-
ics micrOTOF focus mass spectrometer using ESI (electronspray
ionization). The enantiomeric excess and the configuration of the
b-nitro alcohols 3 were determined by HPLC analysis on chiral
phase; the diastereomeric ratios were measured by 1H NMR (for de-
tails see Supporting Information). Prolinols 14d[11] and 14e[9] and
prolinamines 9a,[9] 9j–m,[12] 9n–r,t,[11] 10e–k,m[11] and 10l[12] were
prepared according to literature procedures. Diamines 10d and 9s
are commercially available. The synthesis of the prolinamine 9b
and general procedures for the asymmetric Henry reactions are de-
scribed here. For the preparation of all other new compounds, see
Supporting information.
(2R,5S)-2-Cyclopentyl-5-(hydroxymethyl)-1-methylpyrroli-
dine (14a)
LiAlH4 (732 mg, 19.3 mmol) was added at 08C to a solution of the
pyrrolidine ester 13a (822 mg, 2.76 mmol) in anhydrous THF
(25 mL). The reaction mixture was stirred for 1 h at 08C and then
refluxed for 26 h. The resulting suspension was treated with sat.
aq. Na2SO4 until H2 evolution ceased. The resulting mixture was fil-
tered through a pad of Celite and the filter cake was rinsed with
CH2Cl2-MeOH (9:1, 200 mL). Evaporation of the solvent and column
chromatography (silica gel, CH2Cl2-MeOH-NH3 (aq., 25%), 90:9:1)
(S)-Methyl 2-(tert-butoxycarbonylamino)-5-cyclopentyl-5-
provided amino alcohol 14a (467 mg, 2.55 mmol, 92%) as a color-
oxopentanoate (12a)
31
less oil. Rf =0.23 (CH2Cl2/MeOH/NH3 (aq., 25%) 95:4.5:0.5); ½aŠ
=
D
A solution of the pyroglutamate 11 (10.0 g, 41.1 mmol) in anhy-
drous THF (120 mL) was treated at À408C with cPentMgBr, pre-
pared from bromocyclopentane (5.95 mL, 8.27 g, 55.5 mmol) and
Mg (1.49 mg, 61.1 mmol) in anhydrous THF (50 mL). The reaction
mixture was allowed to warm to RT overnight. Sat. aq. NH4Cl
(20 mL) was added and THF was evaporated in vacuo. The result-
ing aqueous suspension was partitioned between sat. aq. NH4Cl
(200 mL) and CH2Cl2 (200 mL) and the layers were separated. The
aqueous layer was extracted with CH2Cl2 (2”200 mL). The com-
bined organic layers were washed with brine (100 mL) and dried
over MgSO4. Removal of the solvent under reduced pressure and
column chromatography (silica gel, petroleum ether–EtOAc, 5:1) af-
forded amino ketone 12a (4.75 g, 15.2 mmol, 37%) as a colorless
+22.5 (c=1.00 in MeOH); IR (ATR): n˜max =3312, 2948, 2866, 2782,
1771, 1455, 1240, 1034 cmÀ1 1H NMR (500 MHz, CDCl3): d=1.21
;
(m, 2H, cPent-H), 1.43–1.67 (m, 6H, 3-HH, 4-HH, cPent-H), 1.76 (m,
4H, 3-HH, 4-HH, cPent-H), 2.02 (m, 1H, cPent-H), 2.32 (s, 3H, 1-CH3),
2.51 (dd, J=13.6, 6.6 Hz, 1H, 2-H), 2.58 (m, 1H, 5-H), 2.80–3.25 (br
s, 1H, OH), 3.36 (d, J=10.6 Hz, 1H, 5-CHH), 3.63 ppm (dd, J=10.6,
3.5 Hz, 1H, 5-CHH); 13C NMR (125 MHz, CDCl3): d=25.4 (C-cPent),
25.9, 26.3 (C-3, C-4), 26.9, 27.5, 30.7 (C-cPent), 39.9 (1-CH3), 43.5 (C-
cPent), 61.0 (5-CH2), 67.6 (C-5), 70.7 ppm (C-2); HRMS (ESI, pos.)
m/z calcd for C11H21NO [M + H]+ 184.16959, found 184.16908.
(2R,5S)-2-Cyclopentyl-1-methyl-5-((methylamino)methyl)pyr-
rolidine (9b)
oil. Rf =0.27 (petroleum ether/EtOAc 6:1); ½aŠ31 =À18.6 (c=1.00 in
D
MeOH); IR (ATR): n˜max =3375, 2952, 2871, 1745, 1706, 1513, 1366,
1
1146 cmÀ1; H NMR (500 MHz, CDCl3): d=1.42 (s, 9H, C(CH3)3), 1.56
MsCl (27.8 mL, 41.1 mg, 360 mmol) and NEt3 (136 mL, 99.4 mg,
982 mmol) were added at 08C to a solution of the prolinol 14a
(60.0 mg, 327 mmol) in anhydrous CH2Cl2 (4 mL). After 3 d at RT, an
excess of methylamine (aq., 40%, 1.30 mL, 1.16 g, 9.81 mmol) and
MeOH (4.0 mL) was added and stirring was continued for 3 d.
Evaporation of the solvent and column chromatography (silica gel,
CH2Cl2–MeOH–NH3 (aq., 25%), 97:2.7:0.3–95:4.5:0.5) delivered pro-
linamine 9b (34.6 mg, 176 mmol, 54%) as a yellowish oil. Rf =0.31
(m, 2H, cPent-H), 1.59–1.75 (m, 4H, cPent-H), 1.79 (m, 2H, cPent-H),
1.89 (m, 1H, 3-HH), 2.10 (m, 1H, 3-HH), 2.55 (m, 2H, 4-H2), 2.84
(quint., J=7.9 Hz, 1H, cPent-H), 3.72 (s, 3H, OCH3), 4.26 (m, 1H, 2-
H), 5.09 ppm (d, J=8.0 Hz, 1H, NH); 13C NMR (125 MHz, CDCl3): d=
26.1 (C-cPent), 26.6 (C-3), 28.4 (C(CH3)3), 29.0, 29.1 (C-cPent), 37.7
(C-4), 51.5 (C-cPent), 52.5 (OCH3), 53.1 (C-2), 80.1 (C(CH3)3), 155.6
(NCO2), 173.1 (C-1), 212.2 ppm (C-5); HRMS (ESI, pos.) m/z calcd for
C16H27NO5 [M+H]+ 314.19620, found 314.19637.
(CH2Cl2/MeOH/NH3 (aq., 25%) 90:9:1); ½aŠ32 = +4.9 (c=0.20 in
D
ChemCatChem 2016, 8, 1846 – 1856
www.chemcatchem.org
1854
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
MeOH); IR (ATR): n˜max =2951, 2865, 2780, 1450, 1209, 1134 cmÀ1
;
Nitro Group in Organic Synthesis (Ed.: N. Ono), Wiley-VCH, New York,
2001.
1H NMR (500 MHz, CDCl3): d=1.12–1.30 (m, 2H, cPent-H), 1.41–1.68
(m, 8H, 3-HH, 4-HH, cPent-H, NH), 1.68–1.87 (m, 3H, 3-HH, 4-HH,
cPent-H), 1.97 (m, 1H, cPent-H), 2.31 (s, 3H, 1-CH3), 2.33 (m, 1H, 2-
H), 2.45 (s, 3H, NHCH3), 2.47 (m, 1H, 5-H), 2.53 (dd, J=11.2, 6.0 Hz,
1H, 5-CHH), 2.64 ppm (dd, J=11.2, 3.9 Hz, 1H, 5-CHH); 13C NMR
(125 MHz, CDCl3): d=25.4, 26.0 (C-cPent), 27.0 (C-3), 27.92 (C-4),
27.93, 30.9 (C-cPent), 37.2 (NHCH3), 40.9 (1-CH3), 43.8 (C-cPent),
55.7 (5-CH2), 67.0 (C-5), 71.3 ppm (C-2); HRMS (ESI, pos.) m/z calcd
for C12H24N2 [M+H]+ 197.20123, found 197.20218.
[2] Reviews: a) P. Drabina, L. Harmand, M. Sedlak, Curr. Org. Synth. 2014, 11,
879–888; b) R. Ballini, M. Petrini in Science of Synthesis C-1 Building
Blocks in Organic Synthesis (Ed.: P. W. N. M. Van Leeuwen), Thieme, Stutt-
gart, 2014, chap. 1.3.3, pp. 507–524; c) N. Ananthi, S. Velmathi, Indian J.
Chem., Sect. B: Org. Chem. Incl. Med. Chem. 2013, 52, 87–108; d) G. Che-
lucci, Coord. Chem. Rev. 2013, 257, 1887–1932; e) G. Blay, V. Hernµndez-
Olmos, J. R. Pedro, Synlett 2011, 1195–1211; f) Y. Alvarez-Casao, E. Mar-
ques-Lopez, R. P. Herrera, Symmetry 2011, 3, 220–245; g) C. Palomo, M.
Oiarbide, A. Laso, Eur. J. Org. Chem. 2007, 2561–2574; h) J. Boruwa, N.
Gogoi, P. P. Saikia, N. C. Barua, Tetrahedron: Asymmetry 2006, 17, 3315–
3326.
[3] Selected examples: a) K. Hashimoto, N. Kumagai, M. Shibasaki, Org. Lett.
2014, 16, 3496–3499; b) T. Ogawa, N. Kumagai, M. Shibasaki, Angew.
Chem. Int. Ed. 2013, 52, 6196–6201; Angew. Chem. 2013, 125, 6316–
6321; c) D. Sureshkumar, K. Hashimoto, N. Kumagai, M. Shibasaki, J. Org.
Chem. 2013, 78, 11494–11500; d) T. Nitabaru, A. Nojiri, M. Kobayashi, N.
Kumagai, M. Shibasaki, J. Am. Chem. Soc. 2009, 131, 13860–13869.
[4] Selected examples (except Cu-catalysis): a) J. Dimroth, M. Weck, RSC
Adv. 2015, 5, 29108–29113; b) G.-H. Ouyang, Y.-M. He, Q.-H. Fan, Chem.
Eur. J. 2014, 20, 16454–16457; c) Y.-L. Wei, K.-F. Yang, F. Li, Z.-J. Zheng,
Z. Xu, L.-W. Xu, RSC Adv. 2014, 4, 37859–37867; d) Y. Liu, P. Deng, X. Li,
Y. Xiong, H. Zhou, Synlett 2014, 25, 1735–1738; e) S. Wu, J. Tang, J. Han,
D. Mao, X. Liu, X. Gao, J. Yu, L. Wang, Tetrahedron 2014, 70, 5986–5992;
f) K. Lang, J. Park, S. Hong, Angew. Chem. Int. Ed. 2012, 51, 1620–1624;
Angew. Chem. 2012, 124, 1652–1656; g) S. Liu, C. Wolf, Org. Lett. 2008,
10, 1831–1834; h) C. Palomo, M. Oiarbide, A. Laso, Angew. Chem. Int.
Ed. 2005, 44, 3881–3884; Angew. Chem. 2005, 117, 3949–3952.
[5] Cu-catalyzed Henry reactions of nitromethane giving 99% ee with at
General procedure for the enantioselective Henry reactions
of aromatic aldehydes under optimized conditions
A solution of anhydrous CuBr2 (66.7 mm in MeOH, 300 mL, 4.47 mg,
20.0 mmol, 2.0 mol%) was evaporated to dryness in a Schlenk tube.
A solution of diamine 9a (36.7 mm in anhydrous THF, 600 mL,
4.49 mg, 22.0 mmol, 2.2 mol%), MeNO2 (2a, 600 mL, 684 mg,
11.2 mmol, 11.2 equiv), and the aldehyde 1 (1.00 mmol, 1.00 equiv)
were added successively at RT. The mixture was ultrasonicated for
10 min to give a clear, brownish solution, which was cooled to
À258C. NEt3 (1.50m in THF, 10.0 mL, 1.52 mg, 15.0 mmol, 1.5 mol%)
was added and the resulting blue-green solution was stirred until
TLC-control indicated complete consumption of the aldehyde. The
crude reaction mixture was purified by column chromatography
(silica gel, hexanes–EtOAc 8:1–4:1) providing b-nitro alcohol 3. The
enantiomeric excess of 3 was determined by HPLC on chiral
phase.[9] All variations were done on basis of this general proce-
dure and are indicated in the corresponding Tables and Schemes.
´
least one aldehyde: a) R. Cwiek, P. Niedziejko, Z. Kałuz˙a, J. Org. Chem.
2014, 79, 1222–1234; b) Y. Zhou, Y. Zhu, S. Yan, Y. Gong, Angew. Chem.
Int. Ed. 2013, 52, 10265–10269; Angew. Chem. 2013, 125, 10455–
10459; c) T. Deng, C. Cai, J. Fluorine Chem. 2013, 156, 183–186; d) A.
Das, R. I. Kureshy, K. J. Prathap, M. K. Choudhary, G. V. S. Rao, N. H. Khan,
S. H. R. Abdi, H. C. Bajaj, Appl. Catal. A 2013, 459, 97–105; e) D.-D. Qin,
W.-H. Lai, D. Hu, Z. Chen, A.-A. Wu, Y.-P. Ruan, Z.-H. Zhou, H.-B. Chen,
Chem. Eur. J. 2012, 18, 10515–10518; f) L. Yao, Y. Wei, P. Wang, W. He, S.
Zhang, Tetrahedron 2012, 68, 9119–9124; g) Y. Wei, L. Yao, B. Zhang, W.
He, S. Zhang, Tetrahedron 2011, 67, 8552–8558; h) R. I. Kureshy, A. Das,
N. H. Khan, S. H. R. Abdi, H. C. Bajaj, ACS Catal. 2011, 1, 1529–1535; i) W.
Jin, X. Li, B. Wan, J. Org. Chem. 2011, 76, 484–491; j) Y. Zhou, J. Dong, F.
Zhang, Y. Gong, J. Org. Chem. 2011, 76, 588–600; k) B. V. Subba Reddy,
J. George, Tetrahedron: Asymmetry 2011, 22, 1169–1175; l) A. T. Herr-
mann, S. R. Martinez, A. Zakarian, Org. Lett. 2011, 13, 3636–3639; m) G.
Lai, F. Guo, Y. Zheng, Y. Fang, H. Song, K. Xu, S. Wang, Z. Zha, Z. Wang,
Chem. Eur. J. 2011, 17, 1114–1117; n) W. Jin, X. Li, Y. Huang, F. Wu, B.
Wan, Chem. Eur. J. 2010, 16, 8259–8261; o) G. Zhang, E. Yashima, W.-D.
Woggon, Adv. Synth. Catal. 2009, 351, 1255–1262; p) T. Arai, M. Wata-
nabe, A. Yanagisawa, Org. Lett. 2007, 9, 3595–3597; q) M. Bandini, F. Pic-
cinelli, S. Tommasi, A. Umani-Ronchi, C. Ventrici, Chem. Commun. 2007,
616–618.
General procedure for the enantioselective Henry reactions
of aliphatic aldehydes under optimized conditions
A solution of anhydrous CuCl2 (267 mm in MeOH, 300 mL, 10.8 mg,
80.0 mmol, 8.0 mol%) was evaporated to dryness in a Schlenk tube.
A solution of diamine 9a (147 mm in anhydrous THF, 600 mL,
18.0 mg, 88.0 mmol, 8.8 mol%), MeNO2 (2a, 600 mL, 684 mg,
11.2 mmol, 11.2 equiv), and the aldehyde 1 (1.00 mmol, 1.00 equiv)
were added successively at RT. The mixture was ultrasonicated for
10 min to give a clear, brownish solution and then cooled to
À208C. NEt3 (1.50m in THF, 40 mL, 6.08 mg, 60.0 mmol, 6.0 mol%)
was added and the resulting blue-green solution was stirred until
TLC-control indicated complete consumption of the aldehyde. The
crude reaction mixture was purified by column chromatography
(silica gel, pentane–Et2O 8:1–4:1) providing b-nitro alcohol 3. The
enantiomeric excess of 3 was determined by HPLC on chiral
phase.[9] All variations were done on basis of this general proce-
dure and are indicated in the corresponding Tables and Schemes.
[6] Selected recent examples of enantioselective Cu-catalyzed Henry reac-
tions: a) P. Niedziejko, M. Szewczyk, P. Kalicki, Z. Kałuz˙a, Tetrahedron:
Asymmetry 2015, 26, 1083–1094; b) A. Das, M. K. Choudhary, R. I. Kure-
shy, K. Jana, S. Verma, N. H. Khan, S. H. R. Abdi, H. C. Bajaj, B. Ganguly,
Tetrahedron 2015, 71, 5229–5237; c) Y. Shi, Y. Li, J. Sun, Q. Lai, C. Wei, Z.
Gong, Q. Gu, Z. Song, Appl. Organomet. Chem. 2015, 29, 661–667; d) Q.
Song, X. An, T. Xia, X. Zhou, T. Shen, C. R. Chim. 2015, 18, 215–222;
e) J.-L. Li, L. Liu, Y.-N. Pei, H.-J. Zhu, Tetrahedron 2014, 70, 9077–9083;
f) M. Holmquist, G. Blay, M. C. MuÇoz, J. R. Pedro, Org. Lett. 2014, 16,
1204–1207; g) S. J. Canipa, A. Stute, P. O’Brien, Tetrahedron 2014, 70,
7395–7403; h) X. Wang, W. Zhao, G. Li, J. Wang, G. Liu, L. Liu, R. Zhao,
Acknowledgements
The financial support of the German research foundation (DFG)
is gratefully acknowledged.
´
M. Wang, Appl. Organomet. Chem. 2014, 28, 892–899; i) P. Kiełbasinski,
´
M. Rachwalski, S. Kaczmarczyk, S. Lesniak, Tetrahedron: Asymmetry 2013,
Keywords: aldol reactions · amines · asymmetric catalysis ·
copper · ligand design
24, 1417–1420; j) Q. Dai, N. K. Rana, J. C.-G. Zhao, Org. Lett. 2013, 15,
2922–2925; k) M. W. Leighty, B. Shen, J. N. Johnston, J. Am. Chem. Soc.
2012, 134, 15233–15236.
[1] Reviews: a) H. Sasai in Comprehensive Organic Synthesis Vol. 2, 2nd ed.
(Eds.: P. Knochel, G. A. Molander), Elsevier, Amsterdam, 2014, chap. 2.13,
pp. 543–570; b) F. A. Luzzio, Tetrahedron 2001, 57, 915–945; c) The
[7] Selected recent examples of organocatalyzed and enzymatic Henry re-
actions: a) M. T. Corbett, J. S. Johnson, Angew. Chem. Int. Ed. 2014, 53,
255–259; Angew. Chem. 2014, 126, 259–263; b) K. Matsumoto, S. Asa-
ChemCatChem 2016, 8, 1846 – 1856
www.chemcatchem.org
1855
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
´
kura, Tetrahedron Lett. 2014, 55, 6919–6921; c) M. Vlatkovic, L. Bernardi,
[16] Selected recent examples of diastereo- and enantioselective Cu-cata-
lyzed Henry reactions: a) H. Mei, X. Xiao, X. Zhao, B. Fang, X. Liu, L. Lin,
X. Feng, J. Org. Chem. 2015, 80, 2272–2280; b) T. Arai, A. Joko, K. Sato,
Synlett 2015, 26, 209–214; c) A. Das, M. K. Choudhary, R. I. Kureshy, T.
Roy, N. H. Khan, S. H. R. Abdi, H. C. Bajaj, ChemPlusChem 2014, 79,
1138–1146; d) L. Zhang, H. Wu, Z. Yang, X. Xu, H. Zhao, Y. Huang, Y.
Wang, Tetrahedron 2013, 69, 10644–10652; e) D.-D. Qin, W. Yu, J.-D.
Zhou, Y.-C. Zhang, Y.-P. Ruan, Z.-H. Zhou, H.-B. Chen, Chem. Eur. J. 2013,
19, 16541–16544; f) J. D. White, S. Shaw, Org. Lett. 2012, 14, 6270–
6273; g) K. Xu, G. Lai, Z. Zha, S. Pan, H. Chen, Z. Wang, Chem. Eur. J.
2012, 18, 12357–12362; h) R. Boobalan, G.-H. Lee, C. Chen, Adv. Synth.
Catal. 2012, 354, 2511–2520; i) A. Chougnet, G. Zhang, K. Liu, D. Häus-
singer, A. Kägi, T. Allmendinger, W.-D. Woggon, Adv. Synth. Catal. 2011,
353, 1797–1806; j) L. Cheng, J. Dong, J. You, G. Gao, J. Lan, Chem. Eur. J.
2010, 16, 6761–6765; k) T. Arai, Y. Taneda, Y. Endo, Chem. Commun.
2010, 46, 7936–7938; l) Ref. 5d,f,g,i–k,n,p; 8a,b.
[17] From erobs =(Scat +Sbg)/(Rcat +Rbg)ꢁ99.5:0.5 and under the assumption
of a perfect catalyst (Scat/Rcat =100:0) follows vcat/vbg ꢁ99:1. In the case
of a nonperfect catalyst (Scat/Rcat <100:0), the ratio vcat/vbg would be
even higher. obs=observed, cat=catalyzed reaction, bg=background
reaction.
[18] This formula (deduction see Supporting Information) can be used
under the following assumptions, which are—most probably—fulfilled
by our reactions: (i) the rate laws for the for the two catalytic cycles are
identical (vcat1/vcat2 ꢀconstant) and do not change in the course of the
reaction, (ii) there is no interaction between the catalysts that unpro-
portionally reduces the concentration of one catalyst (as, for example,
the formation of a 2:1 oligomer), and (iii) the uncatalyzed background
reaction is slow (here: 1% at maximum because of the excellent 99%
ee reached) and, thus, does not measurably lower the enantiomeric
excess.
E. Otten, B. L. Feringa, Chem. Commun. 2014, 50, 7773–7775; d) P. B.
Thorat, S. V. Goswami, W. N. Jadhav, S. R. Bhusare, Aust. J. Chem. 2013,
66, 661–666; e) S. Kitagaki, T. Ueda, C. Mukai, Chem. Commun. 2013, 49,
4030–4032; f) K. Kanagaraj, P. Suresh, K. Pitchumani, Org. Lett. 2010, 12,
4070–4073; g) D. Uraguchi, S. Nakamura, T. Ooi, Angew. Chem. Int. Ed.
2010, 49, 7562–7565; Angew. Chem. 2010, 122, 7724–7727; h) M.
Gruber-Khadjawi, T. Purkarthofer, W. Skranc, H. Griengl, Adv. Synth.
Catal. 2007, 349, 1445–1450; i) T. Mandal, S. Samanta, C.-G. Zhao, Org.
Lett. 2007, 9, 943–945; j) Y. Sohtome, N. Takemura, K. Takada, R. Takagi,
T. Iguchi, K. Nagasawa, Chem. Asian J. 2007, 2, 1150–1160.
[8] a) D. Scharnagel, A. Müller, F. Prause, M. Eck, J. Goller, W. Milius, M.
Breuning, Chem. Eur. J. 2015, 21, 12488–12500; b) M. Breuning, D. Hein,
M. Steiner, V. H. Gessner, C. Strohmann, Chem. Eur. J. 2009, 15, 12764–
12769.
[9] D. Scharnagel, F. Prause, J. Kaldun, R. G. Haase, M. Breuning, Chem.
Commun. 2014, 50, 6623–6625.
[10] a) M. Breuning, D. Hein, Eur. J. Org. Chem. 2013, 7575–7582; b) M.
Breuning, A. Paasche, M. Steiner, S. Dilsky, V. H. Gessner, C. Strohmann,
B. Engels, J. Mol. Struct. 2011, 1005, 178–185; c) M. Breuning, T. Häuser,
C. Mehler, C. Däschlein, C. Strohmann, A. Oechsner, H. Braunschweig,
Beilstein J. Org. Chem. 2009, 5, 81; d) M. Breuning, M. Steiner, C. Mehler,
A. Paasche, D. Hein, J. Org. Chem. 2009, 74, 1407–1410.
[11] F. Prause, J. Kaldun, B. Arensmeyer, B. Wennemann, B. Frçhlich, D. Schar-
nagel, M. Breuning, Synthesis 2015, 47, 575–586.
[12] For a successful application of copper complexes of diamines 9 and 10
in enantioselective, oxidative biaryl coupling reactions, see: F. Prause, B.
Arensmeyer, B. Frçhlich, M. Breuning, Catal. Sci. Technol. 2015, 5, 2215–
2226.
[13] Rearrangements of 14 to b-amino piperidines (via the sequence mesy-
lation-aziridinium formation-nucleophilic attack at C-2 under ring en-
largement) did not occur to a significant degree.
[14] Enantioselective Henry reactions in the presence of chiral, Cu(OAc)2-de-
rived catalysts, a) without an ancillary base: Ref. 5e,h,k,l,n–q; 6d,h–k;
16b,c,e,h; b) with an ancillary base: Ref. 5a,c,g; 6a,b; 16d,i.
[19] The existence of a hydrogen bridge cannot be fully ruled out, because
the lower steric hindrance in 9 might induce an acceleration that over-
compensates any deceleration induced by the hydrogen bounding.
[20] W. L. F. Armarego, D. D. Perrin, Purification of Laboratory Chemicals,
4th ed., Butterworth-Heinemann, Oxford, 2000.
[15] Catalytic amounts of pyridine do only weakly promote the Henry reac-
tion (see Table 5, entry 3). In the case of nicotinaldehyde (1e), however,
the pyridyl moiety is part of the substrate and, thus, present in a 50-
fold excess with respect to the amount of catalyst (2 mol%).
Received: February 29, 2016
Published online on May 9, 2016
ChemCatChem 2016, 8, 1846 – 1856
www.chemcatchem.org
1856
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim