Ketopinic acid derived bis(hydroxy amides) as cheap, chiral ligands for the enantioselective ethylation of aromatic aldehydes

Article abstract of DOI:10.1002/ejoc.200901391

Readily accessible, C_2- and pseudo-C₂-symmetric bis(hydroxy amides), derived from, commercially available (+)-ketopinic acid and protic diamines, are promising, cheap, chiral ligands for the synthetically valuable, enantioselective a

Full text of DOI:10.1002/ejoc.200901391

FULL PAPER
DOI: 10.1002/ejoc.200901391
Ketopinic Acid Derived Bis(hydroxy amides) as Cheap, Chiral Ligands for the
Enantioselective Ethylation of Aromatic Aldehydes
Tomás de las Casas Engel,[a] Beatriz Lora Maroto,[a] and Santiago de la Moya Cerero*[a]
Keywords: Asymmetric catalysis / Zinc / Hydroxy amides / Chiral pool / Structure–activity relationships
Readily accessible, C2- and pseudo-C2-symmetric bis(hy-
droxy amides), derived from commercially available (+)-keto-
pinic acid and protic diamines, are promising, cheap, chiral
ligands for the synthetically valuable, enantioselective ad-
dition of organozinc reagents to carbon electrophiles. A
series of ligands of this type, having key structural differ-
ences, has been synthesized and tested in the enantioselec-
tive ethylation of benzaldehydes and (E)-cinnamaldehyde, in
order to gain information on the origin of ligand efficiency.
The results obtained allow for the definition of a privileged
structural pattern for the design of improved cheap ligands
and support interesting models proposed for both the acting
catalytic species and the controlling transition states. The
most efficient ligands proved to be less efficient than com-
mercially available (–)-MIB; nevertheless, an impressive effi-
ciency level was obtained, which should sustain interest in
this cheap type of ligands.
Since its discovery by Oguni and Omi[4] and the following
development of its mechanistic rationalization by Noyori
and co-workers,[5] intensive (and also extensive) research on
this useful, asymmetric reaction has been performed.[1,6]
This effort has mainly centered on developing efficient chi-
ral ligands for less favored additions,[7] finding more versa-
tile, chiral ligands,[8] achieving the coupling of a second
asymmetric reaction by using the same chiral ligand (asym-
metric tandem)[9] and, more recently, immobilizing the chi-
ral ligand on an insoluble matrix without the loss of effi-
ciency with the aim of optimizing the recovery and reuse of
the ligand and designing reactors for flow chemistry.[10]
Nowadays, it is possible to accomplish the enantioselec-
tive alkylation, alkenylation, alkynylation and arylation of
different types of aldehydes and α-oxo esters (the most elec-
trophilic carbonyl compounds) by using different types of
chiral ligands,[1] among which the most significant are the
N,N-dialkylated β-amino alcohols [e.g. Nuggent’s (–)-MIB
(1)[8a] in Figure 1]. It is also possible to achieve enantiose-
lective addition to some activated ketones and imines as
well as addition to some activated Michael acceptors.[1]
However, in these cases, more specific, chiral ligands and
the employment of additional metals are generally required
to gain efficiency. Thus, enantioselective addition to ketones
can be realized by using TiIV and certain chiral hydroxy-
sulfonamides such as C2-symmetric 2 (Figure 1),[1g,8b,8c]
whereas the 1,4-conjugate addition can mainly be ac-
complished with the use of CuII[7b] and certain tridentate,
chiral ligands such as the hydroxy-imino-phosphane 3, de-
veloped recently by Gau and co-workers (Figure 1).[7d]
With the interesting objective of improving both effi-
ciency and versatility, new types of chiral ligand architec-
tures (chiral carbon frameworks and/or involved functional
Introduction
Enantioselective Addition of Organozinc Reagents to
Carbon Electrophiles
The enantioselective addition of organozinc reagents to
carbon electrophiles (both 1,2- and conjugate additions),
promoted by substoichiometric amounts of an enantioen-
riched, chiral ligand, is one of the most important reactions
in asymmetric catalysis (Scheme 1).[1] In fact, the reaction
can be considered a basic tool for the asymmetric construc-
tion of C–C bonds and, therefore, for the preparation of
valuable, chiral, organic molecules.[2] The reaction presents
the additional advantage of the easy preparation, stability,
functional-group compatibility and structural diversity of
the required organozinc reagents.[3]
Scheme 1. Enantioselective addition of organozinc reagents to car-
bon electrophiles.
[a] Departamento de Química Orgánica I, Facultad de Ciencias
Químicas, Universidad Complutense de Madrid,
Ciudad Universitaria s/n, 28040 Madrid, Spain
Fax: +34-91-3944103
E-mail: santmoya@quim.ucm.es
Eur. J. Org. Chem. 2010, 1717–1727
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1717

T. de las Casas Engel, B. Lora Maroto, S. de la Moya Cerero
FULL PAPER
Figure 1. Some chiral ligands for the enantioselective addition of
organozinc reagents to carbon electrophiles.
groupings) appear continuously in the literature.[6] In most
cases, these ligands are extremely elaborate from a synthetic
point of view, either due to their own structural complexity
or because asymmetric syntheses or optical-resolution pro-
cesses are required for their preparation (e.g. Wang’s very
versatile 4 and 5 in Figure 1).[11] This fact makes these new
ligands expensive and, therefore, reduces the possibility of
transferring them to industrial processes, as evidenced by
the practically nonexistent industrial use of catalyzed, enan-
tioselective addition of organozinc reagents to carbon elec-
trophiles. In this sense, we strongly believe that one of the
most important future research interests in this valuable re-
action will be the establishment of procedures for its trans-
fer to industry, including the development of cheap, ef-
ficient, chiral ligands.
Figure 2. Some hydroxy amide based, chiral ligands having protic
amides.
Two years ago, we became interested in testing N,N-di-
substituted hydroxy amides 10a (Scheme 2) as chiral ligands
for the enantioselective addition of organozinc reagents to
aldehydes in the absence of titanium. We then believed that
the hydroxy-aprotic amide grouping should mimic some
functional groupings, which have been successful in the tita-
nium-free process described above [e.g. hydroxyamino
(10b), hydroxyimino (10c) or hydroxyoxazolino (10d)].[1c] In
these cases, the catalytic species is proposed to be a zinc
alkoxide, which is chelated by the lone electron pair of the
coordinating N (see 11b–d in Scheme 2).[15] Therefore,
whereas ligand types 10b–d act as O/N ligands, we proposed
that 10a should act as an O/O ligand (compare 11b–d with
Hydroxy Amides as Chiral Ligands for the Enantioselective
Addition of Organozinc Reagents to Aldehydes
Some chiral, C1-, C2- and C3-symmetric N-monosubsti-
tuted hydroxy amides (protic amide group) have been
tested, generally in combination with titanium tetraisoprop-
oxide, in the enantioselective alkylation and alkynylation
of different aldehydes. These amides are based on either
achiral carboxylic acids and chiral amino alcohols derived
from α-amino acids (e.g. 6 and 7 in Figure 2),[12] chiral hy-
droxy acids and achiral amines (e.g. 8)[13] or chiral hydroxy
acids and chiral amines, including amino alcohols (e.g.
9).[14]
In most cases, the enantioselectivities achieved are poor,
with the exception of the results of Hui and co-workers in
phenylethynylation with C2-symmetric 6 (up to 95% ee),[12a]
Du and co-workers in phenylethynylation with C3-symmet-
ric 7 (up to 92% ee)[12b] and Blay and co-workers with C1-
symmetric ligands 8 for methylation (up to 90% ee),[13] eth-
ylation (up to 88% ee)[13] and alkynylation (up to 92%
ee).[14a] In all these cases, the chiral hydroxy amide acts as
an O/N ligand, coordinating a metal center in the catalytic
organometallic species.[13]
Scheme 2. Different functional groupings in chiral ligands (10) and
the corresponding Zn-chelate catalysts (11). Coordinating atoms
are in bold.
1718
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 1717–1727

Ligands for Enantioselective Ethylation of Aromatic Aldehydes
11a in Scheme 2), since the amide oxygen atom (carbonyl (+)-ketopinic acid in only three synthetic steps,[18] whereas
group) should coordinate the Zn ion better than the amide synthetic access to more efficient ligands 13 is much more
N atom.
complex.[19a] These two facts probably account for the early
Chiral, N,N-disubstituted hydroxy amides 10a have the abandonment of studies on the catalytic use of ligands of
additional advantage of being easily prepared by the amid- type 10a.
ation of simple starting materials (e.g. hydroxy acids and
amines or acids and amino alcohols), which can be ob-
tained in enantiopure form from the chiral pool (renewable,
enantiopure, starting materials).[16] In other words, simple
syntheses and cheap, readily accessible, enantiopure, start-
ing materials make 10a a cheap, potential, chiral ligand.
Once we established the idea that N,N-disubstituted
hydroxy amide 10a could constitute an interesting, new,
cheap, chiral ligand for the enantioselective addition of or-
ganozinc reagents to carbon electrophiles under titanium-
free conditions, we carefully looked for literature precedent
on this utility. Only two precedents were found:[17] the semi-
nal work of Oppolzer and co-workers using ketopinic acid
derived 12 (Figure 3) for the ethylation of benzaldehyde
(68% yield and 91% ee)[18] and the use of BINOL-derived
Figure 3. Previous ligands based on N,N-disubstituted hydroxy
amides (type-10a ligands).
Gratifyingly, we have recently demonstrated in two pre-
13 (Figure 3), reported by Katsuki and co-workers, for the liminary communications a noticeable efficiency improve-
ethylation of different aldehydes (up to 88% yield and 99% ment in the ketopinic acid derived, hydroxy amide based
ee for benzaldehyde).[19] Less efficient ligand 12 is readily ligand 12, through the design of C2-symmetric bis(hydroxy
accessible from commercially available, camphor-derived amides) derived from symmetric, secondary diamines hav-
Figure 4. Selected library of symmetric, ketopinic acid derived bis(hydroxy amides).
Eur. J. Org. Chem. 2010, 1717–1727
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1719

T. de las Casas Engel, B. Lora Maroto, S. de la Moya Cerero
FULL PAPER
ing low steric hindrance around the N atoms. Thus, piper-
azine-based ligand 14a (Figure 4), our first, efficient, cheap,
chiral ligand based on a hydroxy amide, is able to promote
the titanium-free, enantioselective addition of diethylzinc to
benzaldehyde in 97% yield and 90% ee.[20] The participa-
tion of the amide carbonyl groups in the coordination of a
C2-symmetric, zinc dialkoxide was proposed to be impor-
tant for the high enantioselectivity obtained.[20]
In order to gain information on the key structural factors
controlling the efficiency of these interesting, C2-symmetric,
bis(hydroxy amide)-based ligands, this paper gives insight
into the importance of the bis(amide) chelation to the cata-
lytic activity of these ligands as well as the effects of struc-
tural changes in the diamine spacer (length, flexibility and
presence of protic N atoms or additional stereogenic centers)
on ligand efficiency (yield and ee). Moreover, an interesting
comparison between the catalytic activity of our most ef-
ficient ligands with well-known (–)-MIB (1) is included. All
these studies have been realized by using the enantioselective
ethylation of benzaldehyde as the test reaction.
Scheme 3. Synthetic route to bis(hydroxy amides) 14.
Table 1. Reaction yields for the preparation of ligands 14.
Entry
Structure
type
Chemical yields [%]
16
14
Overall
1
2
3
4
5
6
7
8
9
a
b
c
d
e
f
g
h
i
90
88
78
50
86
83
85
98
85
91
90
96
97
88
87
87
85
89
82
79
75
48
76
72
74
83
76
Results and Discussion
The small library of ketopinic acid derived bis(hydroxy
amides) shown in Figure 4, which comprises the above-
mentioned structural variables for the diamine spacer, was
chosen for the study. We then considered the following pa-
rameters for the library design: (1) the use of commercially
available, symmetric diamines for ligand synthesis (short
syntheses for cheap ligands) and (2) the use of reactive, non-
aromatic amines (high conversions for cheap ligands). Pre-
viously described and tested 14a and 14b,[20] as well as the
pseudo-C2-symmetric 14i, have been included in this study
as referable comparison elements. Protic-amide-based li-
gands 14g and 14h have been previously synthesized by
Uang and co-workers and tested in the enantioselective
silylcyanation of aldehydes.[21]
Table 2. Screening of ligands 14 for the enantioselective diethylzinc
addition to benzaldehyde.[a]
Entry
Ligand
1-Phenylpropan-1-ol
Yield [%][b] ee [%][c]
1
2
3
4
5
6
7
8
9
14a
14b
14c
14d
14e
14f
14g
14h
14i
97
92
25
62
95
99
43
45
65
90 (R)
73 (R)
40 (R)
52 (R)
94 (R)
86 (R)
24 (R)
24 (R)
10 (R)
Ligands 14a–i were obtained through a simple, two-step
synthetic process: the amidation of commercially available
(+)-ketopinic acid (15) with the corresponding, commer-
cially available diamine by standard acid activation with N-
[3-(dimethylamino)propyl]-NЈ-ethylcarbodiimide
(EDC),[20b,22] followed by the reduction of the intermediate
bis(norbornan-2-one) 16 with NaBH4.[20,23] This synthetic
route is shown in Scheme 3. Table 1 reports the obtained
yields.
[a] 5 mol-% of ligand, 2 equiv. of Et2Zn (1.0  in hexanes), 1 mL
of additional hexane, 5 h at room temp. [b] Determined by GC with
a capillary silicon-gum (SGL-1) column. [c] Determined by chiral
HPLC (Chiralpak IC). Both elution peaks were previously assigned
from a known mixture of stereoisomers, in which the configuration
for the major isomer is known (assigned by the sign of the optical
rotation of the mixture).
In all these cases, high stereoselectivity was achieved in
the reduction of 16 with NaBH4,[24] since the bis(2-exo-
hydroxy) stereoisomer 14 was obtained nearly exclusively
1
(Ͼ 98% de by H NMR spectroscopy of the isolated crude
material and ca. 100% de after purification).
All the synthesized hydroxy amides 14a–i have been carbon spacers (see Figure 4) provided moderate to high
tested in the enantioselective addition of diethylzinc to efficiencies (73–94% ee and 92–99% yield, see Table 2).
benzaldehyde. The obtained results are shown in Table 2.
Nevertheless, a longer distance between the spacer N atoms
Whereas protic-amide-based ligands 14g–i exhibited (e.g. ligands 14c–d in Figure 4) made such ligands lose sub-
poor efficiencies (10–24% ee and 43–65% yield, see stantial efficiency (40–52% ee and 25–62% yield, see
Table 2), aprotic-amide-based 14a,b,e and f, having 2- Table 2).
1720
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 1717–1727

Ligands for Enantioselective Ethylation of Aromatic Aldehydes
All these data agree with the previously proposed hypo-
On the other hand, for C2-symmetric, aprotic-amide-
thesis of the formation of a privileged, diamide-chelated, based bis(hydroxy amide), the presence of diamine spacers
C2-symmetric, zinc dialkoxide catalyst (17 in Figure 5) from allowing the existence of competitive, less-efficient, pseudo-
C2-symmetric, aprotic-amide-based bis(hydroxy amides).[20] C2-symmetric, rotameric catalysts rot-17 must decrease the
Thus, the formation of this privileged catalyst should be ligand efficiency. This is the case for short-spacer ligand 14b
favored for those ligands with short diamine spacers (e.g. 2- (see Figure 4). In this sense, conformationally restricted
carbon spacers), whereas longer spacers should favor the spacers (e.g. note the spacers of ligands 14e and 14f in Fig-
formation of catalytic species without any centered zinc ion ure 4) or spacers with equally substituted N atoms (e.g. the
(e.g. dimetallic complex 18 in Figure 5). Note that the cata- spacer of 14a) should help to maintain the C2 symmetry of
lytic behavior of these decentered-zinc catalysts must be the corresponding zinc dialkoxide catalyst and, therefore, to
very similar to the behavior of the less efficient, C1-symmet- provide a high catalytic efficiency (compare the efficiencies
ric catalysts, derived from C1-symmetric hydroxy amides.[25] of 14a,e and f versus that of 14b in Table 2).
The catalytic behavior of all the studied aprotic-amide-
based ligands (14a–f in Figure 4 and Table 2) demonstrates
that both the length and the conformational flexibility of
the diamine spacer are key structural factors controlling the
catalytic activity, whereas the presence of additional
stereogenic elements in the spacer does not seem to be very
important at the ligand loading used (note the similar be-
havior of diastereomeric ligands 14e and 14f in Table 2).[29]
The stereodifferentiation preference for pro-(R) in all the
studied cases (see Table 2), can be easily explained on the
basis of the most stable diastereomeric transition states
(controlling TS), shown in Figure 7; pro-(R) controlling TS
20, generated from catalyst 17,[20] explains the stereochemi-
cal outcome with the aprotic-amide-based ligands having 2-
carbon spacers (14a,b,e and f in Figure 4). A similar, pro-
(R) controlling TS 21 (Figure 7), generated from catalyst
18,[25] can be proposed for aprotic-amide-based ligands hav-
ing long spacers 14c–d (see Figure 4). Both models are
based on a previous one established by Oppolzer et al. for
ligand 12 (see Figure 3).[18] Analogously to 21, pro-(R) con-
trolling TS 22 (Figure 7), derived from catalyst 19, can also
be proposed for protic-amide-based hydroxy amides 14g–i
(see Figure 4).
All these TSs are based on the seminal model proposed
by Noyori and co-workers for explaining the stereochemical
outcome of the isoborneol-based DAIB [3-exo-(dimeth-
ylamino)isoborneol].[5] This is the endo-anti-anti-type TS, in
which the reactive molecules (benzaldehyde and diethylzinc)
are coordinated to the isoborneol/zinc alkoxide catalyst on
the less hindered endo face, far away from the catalyst hy-
drocarbon core (both lateral oxametallacycles are located
in an anti disposition with respect to the central dioxadime-
tallacycle), and the (alkoxido)zinc moiety and the phenyl
ring are anti (note this endo-anti-anti disposition in models
20–22 in Figure 7).
Figure 5. Proposed zinc-centered (17 and rot-17) and zinc-decen-
tered (18) catalysts for short and long diamine spacers, respectively.
R ϶ H.
On the other hand, in the case of protic-amide-based
bis(hydroxy amides), dimetallic, decentered-zinc catalysts
19 (Figure 6) must also be formed by the deprotonation of
the amide groups,[26,27] which should make each single zinc
center easily coordinated by two close oxide and amide
groups when short diamine spacers are involved in the li-
gand structure (e.g. ligands 14g–i in Figure 4).[28]
The centered-zinc character of the C2-symmetric catalyst
17 (see Figure 5), generated from aprotic-amide-based, C2-
symmetric bis(hydroxy amides) having short spacers, im-
pedes the formation of a competitive, diastereomeric, pro-
(S) exo-anti-anti TS.[20] This undesired competition is pos-
sible for the decentered-zinc, C2-symmetric catalyst 18 (see
Figure 5), derived from aprotic-amide-based, C2-symmetric
bis(hydroxy amides) having long spacers (14c–d in Fig-
ure 4), which explains the lower enantioselectivity obtained
with such ligands (compare 14c,d versus 14a,b,e and f in
Table 2).
Figure 6. Proposed, zinc-decentered catalysts (18) for aprotic-
amide-based ligands with long spacers (another perspective is rep-
resented in Figure 5) and protic-amide-based ligands (19).
Eur. J. Org. Chem. 2010, 1717–1727
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1721

T. de las Casas Engel, B. Lora Maroto, S. de la Moya Cerero
FULL PAPER
Figure 7. Proposed controlling TSs, derived from catalyst 17 (20), 18 (21) and 19 (22 and 23). R = alkyl.
The more naked character of the (alkoxido)zinc group of
Table 3 also shows that a temperature decrease from
catalyst 19, generated from protic-amide-based bis(hydroxy 20 °C to 0 °C results in an important loss of chemical yield,
amides), when compared with the zinc centers of catalysts whereas the enantioselectivity does not practically change
17 and 18 (compare the coordinative groups around the in the temperature interval between 0 °C and 40 °C. Ad-
zinc centers in Figure 5 and Figure 6), makes possible the ditionally, the extra solvent can be removed without any
competition of an undesired, pro-(S), endo-anti-syn TS (23 significant loss of the reaction efficiency (only a slight loss
in Figure 7) and explains the low enantioselectivity ob- of enantioselectivity is detected, see Table 3),[30] and the li-
tained with the protic-amide-based ligands 14g–i (see Fig- gand loading can be decreased from 5 to 2 mol-% without
ure 4 and Table 2).
the results suffering (see Table 3). The reaction time can
Before evaluating the versatility of the three best ligands also be diminished (from 5 h to 4 h, see Table 3) without
(14a,e and f, see Table 2) for the ethylation of other alde- any substantial loss of efficiency. Unfortunately, the amount
hydes, an optimization of the reaction conditions for benz- of Et2Zn cannot be decreased without a significant decrease
aldehyde ethylation was performed by using the cheapest in the yield (see Table 3).
ligand, 14a. Table 3 shows the results of this optimization
Once we established the optimal reaction conditions, we
study.
tested the most efficient ligands 14a,e and f in the ethylation
In contrast to the ligands 13 of Katsuki and co-workers of different stereoelectronically activated benzaldehydes as
(see Figure 3),[19] the efficiency of our N,N-substituted well as cinnamaldehyde under the same conditions
amide-based ligand 14a decreases strongly when a coordi- (Table 4).The obtained results show that the selected hyd-
native solvent (THF) is used as a cosolvent in the enantiose- roxy amides 14a,e and f are able to efficiently promote (71–
lective test addition (see Table 3). This fact supports the 92% ee and 73–97% yield) the room-temperature ethylation
proposed, diamide-chelated, catalyst model for our ligands of different aldehydes by using a low ligand loading
(17, Figure 5). Thus, whereas Katsuki et al. proposed for (0.02 mol-%) and a short reaction time (4 h). Table 4 also
ligands 13 a catalyst model consisting of a solvent-coordi- shows that the different behavior of diastereomeric 14e and
nated zinc dialkoxide without any carbonyl chelation (al- 14f we expected was only detected for the less efficient ethyl-
though the amide carbonyl groups are proposed to be in- ations.
volved in the controlling TSs),[19] the diamide chelation in
Finally, the catalytic activities of the most efficient li-
17 is crucial for its catalytic activity, since the interruption gands for the benzaldehyde ethylation (14a,e and f) have
of such chelation by the competitive THF coordination to been compared with the catalytic activity exerted by well-
the zinc center lowers both the yield and enantioselectivity known (–)-MIB (1) at a low ligand loading (0.005 equiv.).
(see Table 3).
For such a comparison, the turnover number (TON), turn-
1722
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 1717–1727

Ligands for Enantioselective Ethylation of Aromatic Aldehydes
Table 3. Optimization of the reaction conditions.
Entry
14a
[equiv.]
Et2Zn
Cosolvent[b]
T [°C]
t [h]
1-Phenylpropan-1-ol
[equiv.][a]
Yield
ee
[%][c]
[%][d]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
0.05
0.05
0.05
0.05
0.05
0.05
0.10
0.02
0.01
0.01
0.01
0.02
0.02
0.02
0.02
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
1.5
1.0
2.0
2.0
2.0
2.0
hexane
hexane
hexane
THF
CH2Cl2
–
hexane
hexane
hexane
hexane
hexane
hexane
hexane
hexane
–
0
5
5
5
5
5
5
5
5
5
5
5
5
4
3
4
68
97
96
38
90
97
85
95
90
64
30
96
95
90
91
88 (R)
90 (R)
80 (R)
14 (R)
84 (R)
84 (R)
80 (R)
90 (R)
86 (R)
92 (R)
88 (R)
90 (R)
90 (R)
90 (R)
88 (R)
20
40
20
20
20
20
20
20
20
20
20
20
20
20
[a] 1.0  Et2Zn in hexanes. [b] 1 mL of cosolvent was used. [c] Determined by GC (SGL1). [d] Determined by chiral HPLC (see Table 2).
Table 4. Screening of ligands 14a, 14e and 14f (L) for the enantiose-
Table 5. Comparison of ligand activities at low ligand loading
(0.5 mol-%).[a]
lective, diethylzinc addition to different aldehydes.[a]
Entry
L
1-Phenylpropan-1-ol TON[d] TOF TOF ϫ ee
Relative
(TOF ϫ ee)[f]
Yield
[%][b]
ee
[%][c]
[h–1][e]
1
2
3
4
14a
14e
14f
1
67
46
61
97
64 (R)
4 (R)
60 (R)
78 (R)
134
92
122
194
27
18
25
39
17.102
72
15.102
30.102
0.57
0.02
0.50
1.00
Aldehyde
L
Corresponding alcohol
Yield (%)[b]
ee (%)[c]
Benzaldehyde
Benzaldehyde
Benzaldehyde
14a
14e
14f
14a
14e
14f
14a
14e
14f
14a
14e
14f
14a
14e
14f
14a
14e
14f
91
91
97
88
87
79
83
94
91
73
77
76
61
83
40
93
83
84
88 (R)
91 (R)
86 (R)
88 (R)
85 (R)
82 (R)
66 (R)
85 (R)
92 (R)
88 (R)
82 (R)
89 (R)
62 (R)
79 (R)
20 (R)
56 (R)
59 (R)
71 (R)
[a] 0.5 mol-% of ligand, 2 equiv. of Et2Zn (1.0  in hexanes), 1 mL
of additional hexane, 5 h at room temp. [b] Determined by GC
(SGL1). [c] Determined by chiral HPLC (see Table 2). [d] TON =
equiv. of 1-phenylpropan-2-ol obtained/0.5. [e] TOF = TON/5. [f]
Relative (TOF ϫ ee) was determined relative to that of 1.
2-Chlorobenzaldehyde
2-Chlorobenzaldehyde
2-Chlorobenzaldehyde
4-Chlorobenzaldehyde
4-Chlorobenzaldehyde
4-Chlorobenzaldehyde
2-Methylbenzaldehyde
2-Methylbenzaldehyde
2-Methylbenzaldehyde
4-Methylbenzaldehyde
4-Methylbenzaldehyde
4-Methylbenzaldehyde
(E)-Cinnamaldehyde
(E)-Cinnamaldehyde
(E)-Cinnamaldehyde
Although Table 5 shows that the most efficient, ketopinic
acid derived bis(hydroxy amides) are less efficient than 1, a
challenging activity for the cheaper piperazine-based 14a is
demonstrated. The different behavior of diastereomeric li-
gands 14e and 14f in the benzaldehyde ethylation is now
highlighted at a low ligand loading, where 14f is matched
(61% yield and 60% ee), and 14e is mismatched (46% yield
and 4% ee).
[a] 2 mol-% of ligand, 2 equiv. of Et2Zn (1.0  in hexanes), 4 h at
room temp. [b] Determined by GC (SGL1). [c] Determined by chi-
ral HPLC (Chiralpak IC or IA). Both elution peaks were pre-
viously assigned from a known mixture of stereoisomers, in which
the configuration for the major isomer is known (assigned by the
sign of the optical rotation of the mixture).
Conclusions
A diversity of enantiopure, ketopinic acid derived, C2-
symmetric bis(hydroxy amides) can be easily obtained from
cheap, commercially available, starting materials. A seminal
screening on a selected library of these ligands by the enan-
over frequency (TOF) and TOF ϫ ee efficiency parameters tioselective ethylation of benzaldehyde demonstrates that:
have been calculated. The obtained results are shown in (1) protic-amide-based bis(hydroxy amides) are less efficient
Table 5.
than aprotic ones (N,N-disubstituted amides), (2) the effi-
Eur. J. Org. Chem. 2010, 1717–1727
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1723

T. de las Casas Engel, B. Lora Maroto, S. de la Moya Cerero
FULL PAPER
(0.63 g, 3.3 mmol), DMAP (0.42 g, 3.3 mmol) and piperazine
(0.13 g, 1.5 mmol) were dissolved in CH2Cl2 (10 mL), and the mix-
ture was stirred at room temp. for 24 h. H2O (5 mL) was then
added, and the resulting layers were separated. The aqueous layer
was extracted with CH2Cl2 (3ϫ5 mL). The combined organic lay-
ers were washed successively with aqueous HCl (10%, 1ϫ5 mL),
H2O (1ϫ5 mL), aqueous NaOH (10%, 2ϫ5 mL), H2O (1ϫ5 mL)
and brine (1ϫ5 mL) and dried with anhydrous MgSO4. After fil-
tration and solvent evaporation, pure 16a (559 mg, 90% yield) was
ciency of aprotic-amide-based bis(hydroxy amides) having
short diamine spacers (2-carbon length) is higher than the
efficiency of related ligands having long spacers, (3) the effi-
ciency of aprotic-amide-based bis(hydroxy amides) having
conformationally restricted diamine spacers is higher than
the efficiency of related ligands having flexible spacers and
(4) additional chiral centers in the diamine spacer are not
very important for ligand efficiency.
isolated as a white solid. M.p. 262–264 °C. [α]2D0 = –18.5 (c = 0.31,
The results obtained are compatible with a C2-symmet-
1
ric, centered-zinc model for the organometallic, catalytic CHCl3). H NMR (CDCl3, 200 MHz): δ = 3.54 (br. s, 8 H, CH2-
N), 2.50 (ddd, J = 18.3, 2.3, 2.3 Hz, 2 H, CHexo-C=O), 2.26–1.94
(m, 8 H) 1.94–1.85 (d, J = 18.3 Hz, 2 H, CHendo-C=O), 1.51–1.37
(m, 2 H), 1.21 (s, 6 H, Me), 1.20 (s, 6 H, Me) ppm. 13C NMR
(CDCl3, 50 MHz): δ = 212.7 (C=O), 167.7 (N-C=O), 67.4 (C), 50.6
(C), 46.8 (N-CH2), 43.7 (CH2), 43.1 (CH), 27.2 (CH2), 27.0 (CH2),
species generated from aprotic-amide-based bis(hydroxy
amides) having short, diamine spacers and with a C2-sym-
metric, decentered-zinc model for catalysts arising from
aprotic-amide-based bis(hydroxy amides) having long, di-
amine spacers. Protic-amide-based ligands, having either
short or long spacers, favored the centered-zinc model
rather than the decentered-zinc one.
21.2 (CH ), 20.9 (CH ) ppm. FTIR: ν = 1740.8 (s), 1621.6 (s) cm–1.
˜
3
3
MS (EI): m/z (%) = 414 (11), 165 (100). HRMS (EI): m/z =
414.2527 (calcd. for C24H34N2O4 414.2519).
Controlling-TS models have been proposed for all the
ligand cases. These models are based on previous ones pro-
posed by the groups of Noyori and Oppolzer for related
ligands. In all cases, the controlling TS is the pro-(R), endo-
anti-anti diastereomer. C2-Symmetric, centered-zinc cata-
lysts prevent the participation of competitive, exo TSs, al-
lowing high efficiencies, whereas C2-symmetric, decentered-
zinc catalysts, derived from protic-amide-based ligands, al-
low more participation by the pro-(S), endo-anti-syn TS,
giving rise to low efficiencies.
The most efficient, ketopinic acid derived bis(hydroxy
amide) 14a turned out to be 1.7 times less efficient than
(–)-MIB (in terms of TOF ϫ ee) in the benzaldehyde ethyl-
ation. Nevertheless, the measured activity of the most ef-
ficient, aprotic-amide-based bis(hydroxy amides) (up to
97% yield and 94% ee) ensures that interest in the improve-
ment and versatility of this interesting type of cheap ligands
will remain strong. Further research on the use of other
zinc reagents and other classes of aldehydes is in progress.
N,NЈ-Dimethyl-N,NЈ-bis{[(1S,4R)-7,7-dimethyl-2-oxonorborn-1-yl]-
carbonyl}ethanediamine (16b): White solid (549 mg, 88 % yield).
M.p. 165–167 °C. [α]2D0 = –48.9 (c = 0.62, CHCl3). 1H NMR
(CDCl3, 300 MHz): δ = 3.82–3.35 (m, 4 H, CH2-N), 3.06 (br. s, 6
H, Me-N), 2.48 (ddd, J = 18.4, 4.7, 2.8 Hz, 2 H, CHexo-C=O),
2.35–1.95 (m, 8 H), 1.90 (d, J = 18.4 Hz, 2 H, CHendo-C=O), 1.55–
1.34 (m, 2 H), 1.20 (s, 6 H, Me), 1.18 (s, 6 H, Me) ppm. 13C NMR
(CDCl3, 75 MHz): δ = 212.2 (C=O), 169.1 (N-C=O), 67.5 (C), 50.5
(C), 47.0 (N-CH2), 43.7 (CH2), 43.2 (CH), 37.3 (N-CH3), 26.9
(CH ), 21.5 (CH ), 20.8 (CH ) ppm. FTIR: ν = 1738.9 (s), 1624.5
˜
2
3
3
(s) cm–1. MS (EI): m/z (%) = 416 (1), 165 (100). HRMS (EI): m/z
= 416.2666 (calcd. for C24H36N2O4 416.2675).
N,NЈ-Bis{[(1S,4R)-7,7-dimethyl-2-oxonorborn-1-yl]carbonyl}-4,4Ј-bi-
piperidine (16c): White solid (580 mg, 78% yield). Decomposes at
250 °C. [α]2D0 = –17.0 (c = 0.10, CHCl3). 1H NMR (CDCl3,
300 MHz): δ = 4.68 (br. s, 2 H, CH-N), 3.83 (br. s, 2 H, CH-N),
3.37–2.37 (br. s, 4 H, CH-N), 2.50 (ddd, J = 18.3, 4.9, 2.4 Hz, 2 H,
CHexo-C=O), 2.37–2.15 (br. s, 2 H), 2.15–2.01 (m, 4 H), 1.98 (dd,
J = 4.4, 4.4 Hz, 2 H), 1.91 (d, J = 18.3 Hz, 2 H, CHendo-C=O),
1.83–1.65 (m, 4 H), 1.60 (m, 2 H), 1.53–1.26 (m, 4 H), 1.23 (s, 6
H, Me), 1.21 (s, 6 H, Me), 1.26–0.96 (br. s, 2 H) ppm. 13C NMR
(CDCl3, 75 MHz): δ = 212.6 (C=O), 167.2 (N-C=O), 67.4 (C), 50.6
(C), 46.8 (N-CH2), 43.7 (CH2), 43.1 (CH), 41.0 (N-CH2CH2CH),
29.6 (N-CHCH2), 27.5 (CH2), 27.0 (CH2), 21.3 (CH3), 21.0 (CH3)
Experimental Section
General Considerations: Common solvents were dried and distilled
by standard procedures. All starting materials and reagents were
obtained from commercial sources and used without further purifi-
cation. Flash chromatography purifications were performed on sil-
ica gel (230–400 mesh ASTM). NMR spectra were recorded at
20 °C, and the residual solvent peaks were used as internal stan-
dards. FTIR spectra were obtained by using the thin-layer tech-
nique. GC analyses were realized at 150 or 120 °C in a chromato-
graph equipped with an SGL-1 column, a flame ionization detector
(FID), and N2 as the mobile phase. Chiral-HPLC analyses were
realized at room temp. in a chromatograph equipped with a capil-
lary Chiralpak-IC or Chiralpack-IA column, a diode array detector
(DAD), and hexane/2-propanol as the mobile phase. Mass spectra
were recorded by using the EI (70 eV) or ESI techniques. HR mass
spectra were obtained by using the peak-matching method (EI) or
FTMS (ESI).
ppm. FTIR: ν = 1739.7 (s), 1619.9 (s) cm–1. MS (EI): m/z (%) =
˜
496 (16), 331 (100). HRMS (EI): m/z = 496.3306 (calcd. for
C30H44N2O4 496.3301).
N,NЈ-Bis{[(1S,4R)-7,7-dimethyl-2-oxonorborn-1-yl]carbonyl}-1,7-di-
aza-12-crown-4 (16d): White solid (376 mg, 50% yield). M.p. 197–
198 °C. [α]2D0 = –7.9 (c 0.21, CHCl3). 1H NMR (CDCl3, 300 MHz):
δ = 4.10–2.90 (m, 16 H, O-CH2CH2-N), 2.47 (ddd, J = 18.4, 4.9,
2.4 Hz, 2 H, CHexo-C=O), 2.23–1.93 (m, 6 H), 1.89 (d, J = 18.4 Hz,
2 H, CHendo-C=O), 1.42 (m, 2 H), 1.24 (m, 2 H), 1.18 (s, 12 H,
Me) ppm. 13C NMR (CDCl3, 75 MHz): δ = 212.7 (C=O), 169.9
(N-C=O), 69.6 (CH2-O), 67.9 (C), 50.7 (C), 48.7 (N-CH2), 43.7
(CH2), 43.0 (CH), 27.9 (CH2), 27.0 (CH2), 21.5 (CH3), 20.7 (CH3)
ppm. FTIR: ν = 1737.9 (s), 1623.6 (s) cm–1. MS (ESI): m/z (%) =
˜
503 (1), 525 (100). HRMS (ESI): m/z = 503.3116 (calcd. for
C28H43N2O6 503.3116).
Synthesis of Bis(oxo amides) 16. Typical Procedure: Preparation of
(1R,2R)-N,NЈ-Dimethyl-N,NЈ-bis{[(1S,4R)-7,7-dimethyl-2-oxonor-
N,NЈ-Bis{[(1S,4R)-7,7-dimethyl-2-oxonorborn-1-yl]carbonyl}piper- born-1-yl]carbonyl}cyclohexane-1,2-diamine (16e): White solid
azine (16a): In a round-bottom flask, equipped with a magnetic
stirrer, (1S)-ketopinic acid (0.55 g, 3.0 mmol), EDC hydrochloride
(606 mg, 86% yield). M.p. 215–216 °C. [α]2D0 = +6.0 (c = 0.10,
CHCl3). 1H NMR (CDCl3, 300 MHz): δ = 4.63 (dd, J = 4.9,
1724
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 1717–1727

Ligands for Enantioselective Ethylation of Aromatic Aldehydes
2.8 Hz, 2 H, CH-N), 2.72 (s, 6 H, Me-N), 2.49 (ddd, J = 18.4, 4.6,
CH2), 44.6 (CH), 40.7 (CH2), 29.9 (CH2), 27.3 (CH2), 22.4 (CH3),
2.2 Hz, 2 H, CHexo-C=O), 2.35–2.17 (m, 2 H), 2.15–1.94 (m, 6 H) 21.4 (CH ) ppm. FTIR: ν = 3429.2 (br., w), 1622.3 (s) cm–1. MS
˜
3
1.90 (d, J = 18.4 Hz, 2 H, CHendo-C=O), 1.83–1.66 (m, 4 H), 1.62–
1.29 (m, 6 H), 1.21 (s, 6 H, Me), 1.13 (s, 6 H, Me) ppm. 13C NMR
(CDCl3, 75 MHz): δ = 212.0 (C=O), 169.1 (N-C=O), 68.1 (C), 53.2
(N-CH), 50.8 (C), 44.0 (CH2), 43.4 (CH), 30.4 (N-CH3), 29.0 (N-
CHCH2), 27.4 (CH2), 26.8 (CH2), 25.0 (N-CHCH2CH2), 22.0
(ESI): m/z (%) = 507 (1), 529 (100). HRMS (ESI): m/z = 507.3423
(calcd. for C28H47N2O6 507.3429).
N,NЈ-Bis{[(1S,2R,4R)-7,7-dimethyl-2-hydroxynorborn-1-yl]carb-
onyl}-4,4Ј-bipiperidine (14c): White solid (239 mg, 96% yield). De-
composes at 250 °C. [α]2D0 = –4.0 (c = 0.10, CHCl3). 1H NMR
(CDCl3, 700 MHz): δ = 4.47 (m, 4 H, CH-N and CH-O), 4.18 (m,
2 H, CH-N), 2.72 (m, 4 H, CH-N), 2.00 (ddd, J = 12.3, 12.3,
4.4 Hz, 2 H), 1.96–1.88 (m, 4 H), 1.87–1.77 (m, 4 H), 1.75 (m, 4
H), 1.63 (dd, J = 4.4, 4.4 Hz, 2 H), 1.54–1.50 (m, 2 H), 1.38 (s, 6
H, Me), 1.36–1.08 (m, 8 H) 1.15 (s, 6 H, Me) ppm. 13C NMR
(CDCl3, 175 MHz): δ = 171.4 (N-C=O), 78.0 (CH-O), 60.8 (C),
50.6 (C), 44.8 (CH), 44.7 (CH2), 41.23 (N-CH2CH2CH), 41.22 (N-
CH2), 30.0 (N-CH2CH2), 29.8 (CH2), 27.1 (CH2), 22.3 (CH3), 21.6
(CH ), 20.6 (CH ) ppm. FTIR: ν = 1738.9 (s), 1619.2 (s) cm–1. MS
˜
3
3
(ESI): m/z (%) = 471 (2), 493 (100). HRMS (ESI): m/z = 471.3223
(calcd. for C28H43N2O4 471.3217).
(1S,2S)-N,NЈ-Dimethyl-N,NЈ-bis{[(1S,4R)-7,7-dimethyl-2-oxonor-
born-1-yl]carbonyl}cyclohexane-1,2-diamine (16f): White solid
(585 mg, 83% yield). M.p. 198–200 °C. [α]2D0 = –77.0 (c = 0.12,
CHCl3). 1H NMR (CDCl3, 300 MHz): δ = 4.67 (br. s, 2 H, CH-
N), 2.77 (s, 6 H, Me-N), 2.47 (dd, J = 18.4, 2.6 Hz, 2 H, CHexo
-
(CH ) ppm. FTIR: ν = 3389.1 (br., w), 1596.0 (s) cm–1. MS (ESI):
˜
3
C=O), 2.28 (m, 2 H), 2.11–1.86 (m, 6 H), 1.86 (d, J = 18.4 Hz, 2
H, CHendo-C=O), 1.68 (m, 4 H), 1.57–1.28 (m, 6 H), 1.20 (s, 6 H,
Me), 1.15 (s, 6 H, Me) ppm. 13C NMR (CDCl3, 75 MHz): δ =
211.2 (C=O), 169.0 (N-C=O), 68.2 (C), 53.3 (N-CH), 50.6 (C), 43.8
(CH2), 43.6 (CH), 30.6 (N-CH3), 29.7 (N-CHCH2), 27.9 (CH2),
26.7 (CH2), 25.2 (N-CHCH2CH2), 21.6 (CH3), 20.9 (CH3) ppm.
m/z (%) = 501 (15), 236 (100). HRMS (ESI): m/z = 501.3684 (calcd.
for C30H49N2O4 501.3687).
(1R,2R)-N,NЈ-Dimethyl-N,NЈ-bis{[(1S,2R,4R)-7,7-dimethyl-2-hy-
droxynorborn-1-yl]carbonyl}cyclohexane-1,2-diamine (14e): White
solid (208 mg, 88% yield). M.p. 222–223 °C. [α]2D0 = +22.0 (c = 0.15,
CHCl3). 1H NMR (CDCl3, 300 MHz): δ = 4.72 (br. s, 2 H, CH-
O), 4.20 (br. s, 2 H, CH-N), 2.92 (s, 6 H, Me-N), 2.01–1.69 (m, 14
H), 1.68–1.56 (m, 6 H), 1.40 (s, 6 H, Me), 1.43–1.30 (m, 2 H), 1.18–
1.03 (m, 2 H), 1.09 (s, 6 H, Me) ppm. 13C NMR (CDCl3, 75 MHz):
δ = 172.8 (N-C=O), 77.1 (CH-O), 61.2 (C), 53.2 (N-CH), 50.8 (C),
44.7 (CH), 41.8 (CH2), 29.7 (N-CH3), 29.4 (CH2), 26.9 (CH2), 25.1
FTIR: ν = 1739.7 (s), 1618.8 (s) cm–1. MS (ESI): m/z (%) = 471 (3),
˜
493 (100). HRMS (ESI): m/z = 471.3231 (calcd. for C28H43N2O4
471.3217).
(1R,2S)-N,NЈ-Bis{[(1S,4R)-7,7-dimethyl-2-oxonorborn-1-yl]carb-
onyl}cyclohexane-1,2-diamine (16i): White solid (563 mg, 85 %
yield). M.p. 96–98 °C. [α]2D0 = +63.8 (c = 0.40, CHCl3). H NMR
1
(CH ), 22.2 (CH ), 22.0 (CH ) ppm. FTIR: ν = 3434.7 (br., w),
˜
2
3
3
(CDCl3, 300 MHz): δ = 7.92 (m, 2 H, NH), 4.22 (m, 2 H, CH-N),
2.64–2.43 (m, 4 H), 2.22–2.02 (m, 4 H), 1.97 (d, J = 18.6 Hz, 1 H,
CHendo-C=O), 1.95 (d, J = 18.6 Hz, 1 H, CHendo-C=O), 1.88–1.35
(m, 12 H), 1.25 (s, 3 H, Me), 1.24 (s, 3 H, Me), 0.99 (s, 3 H, Me),
0.98 (s, 3 H, Me) ppm. 13C NMR (CDCl3, 75 MHz): δ = 217.4
(C=O), 217.1 (C=O), 168.7 (N-C=O), 168.6 (N-C=O), 64.6 (C),
64.5 (C), 50.1 (C), 49.8 (C), 48.33 (N-CH), 48.25 (N-CH), 43.73
(CH2), 43.69 (CH2), 43.2 (CH), 43.1 (CH), 29.3 (N-CHCH2), 28.7
(N-CHCH2), 28.4 (CH2), 28.2 (CH2), 27.67 (CH2), 27.65 (CH2),
22.5 (N-CHCH2CH2), 21.9 (N-CHCH2CH2), 20.9 (CH3), 20.8
1601.4 (s) cm–1. MS (ESI): m/z (%) = 475 (3), 497 (100). HRMS
(ESI): m/z = 475.3525 (calcd. for C28H47N2O4 475.3530).
(1S,2S)-N,NЈ-Dimethyl-N,NЈ-bis{[(1S,2R,4R)-7,7-dimethyl-2-hy-
droxynorborn-1-yl]carbonyl}cyclohexane-1,2-diamine (14f): White
solid (205 mg, 87% yield). M.p. 130–131 °C. [α]2D0 = –63.0 (c = 0.17,
1
CHCl3). H NMR (CDCl3, 300 MHz): δ = 4.72 (m, 2 H, CH-O),
4.17 (m, 2 H, CH-N), 3.24 (d, J = 6.6 Hz, 2 H), 2.91 (s, 6 H, Me-
N), 2.86 (d, J = 2.0 Hz, 2 H), 1.92–1.29 (m, 18 H), 1.36 (s, 6 H,
Me), 1.14 (s, 6 H, Me), 1.10–1.05 (m, 2 H) ppm. 13C NMR (CDCl3,
75 MHz): δ = 172.6 (N-C=O), 76.6 (CH-O), 61.2 (C), 53.4 (N-CH),
50.8 (C), 44.7 (CH), 41.6 (CH2), 29.6 (N-CHCH2), 29.5 (N-CH3),
29.3 (CH2), 27.0 (CH2), 25.2 (N-CHCH2CH2), 22.5 (CH3), 21.8
(CH ), 20.4 (CH ), 20.3 (CH ) ppm. FTIR: ν = 1725.9 (s), 1665.9
˜
3
3
3
(s) cm–1. MS (EI): m/z (%) = 442 (1), 261 (100). HRMS (EI): m/z
= 442.2845 (calcd. for C26H38N2O4 442.2832).
(CH ) ppm. FTIR: ν = 3384.1 (br., w), 1608.7 (s) cm–1. MS (ESI):
˜
3
Synthesis of Bis(hydroxy amides) 14. Typical Procedure: Preparation
of N,NЈ-Bis{[(1S,2R,4R)-7,7-dimethyl-2-hydroxynorborn-1-yl]-
carbonyl}-1,7-diaza-12-crown-4 (14d): A two-necked, round-bottom
flask, equipped with a magnetic stirrer and a H2O condenser, was
charged with bis(oxo amide) 16d (0.25 g, 0.5 mmol), methanol
(10 mL) and NaBH4 (0.08 g, 2.0 mmol). The mixture was refluxed
under argon for 24 h and cooled to room temp. H2O (5 mL) was
then added, and the resulting mixture was concentrated under re-
duced pressure (methanol elimination). The obtained residue was
diluted with ethyl acetate (5 mL). H2O (3 mL) was added, and the
resulting layers were separated. The aqueous layer was extracted
with ethyl acetate (3ϫ3 mL). The combined organic layers were
washed with brine (1ϫ5 mL) and dried with anhydrous MgSO4.
After filtration and solvent evaporation under reduced pressure, the
residue was purified by flash column chromatography (silica gel,
CHCl3) to obtain 14d (244 mg, 97% yield) as a white solid. Decom-
poses at 250 °C. [α]2D0 = –72.0 (c = 0.14, CHCl3). 1H NMR (CDCl3,
300 MHz): δ = 4.10 (dd, J = 8.2, 3.7 Hz, 2 H, CH-O) 4.27–2.60
(m, 18 H, N-CH2CH2-O and OH), 2.03–1.85 (m, 4 H), 1.84–1.69
m/z (%) = 475 (4), 497 (100). HRMS (ESI): m/z = 475.3532 (calcd.
for C28H47N2O4 475.3530).
(1R,2S)-N,NЈ-Bis{[(1S,2R,4R)-7,7-dimethyl-2-hydroxynorborn-1-
yl]carbonyl}cyclohexane-1,2-diamine (14i): White solid (197 mg,
89% yield). M.p. 227–229 °C. [α]2D0 = –14.8 (c = 0.21, CH2Cl2). H
1
NMR (CD3OD, 200 MHz): δ = 4.29 (br. s, 1 H, N-CH), 4.03 (dd,
J = 7.4, 3.6 Hz, 2 H, N-CH and CH-O), 3.96 (dd, J = 7.4, 3.8 Hz,
1 H, CH-O), 2.41–2.15 (m, 2 H) 1.96–1.01 (m, 20 H), 1.20 (s, 3 H),
1.19 (s, 3 H), 0.98 (s, 3 H, Me), 0.96 (s, 3 H, Me) ppm. 13C NMR
(CD3OD, 50 MHz): δ = 174.96 (N-C=O), 174.92 (N-C=O), 78.3
(CH-O), 77.8 (CH-O), 59.1 (C), 58.8 (C), 51.6 (C), 50.74 (C), 50.68
(N-CH), 49.6 (N-CH), 47.0 (CH), 46.9 (CH), 43.3 (CH2), 42.9
(CH2), 31.2 (N-CHCH2), 31.0 (N-CHCH2), 29.4 (CH2), 29.3
(CH2), 28.3 (CH2), 28.0 (CH2), 24.6 (N-CHCH2CH2), 22.2 (N-
CHCH2CH2), 22.1 (CH3), 22.0 (CH3), 21.9 (CH3), 21.8 (CH3)
ppm. FTIR: ν = 3250.3 (br., w), 1626.7 (s) cm–1. MS (EI): m/z
˜
(%) = 446 (1), 97 (100). HRMS (EI): m/z = 446.3133 (calcd. for
C26H42N2O4 446.3145).
(m, 4 H), 1.59–1.41 (m, 4 H), 1.36 (s, 6 H, Me), 1.18–1.01 (m, 2 General Procedure for the Enantioselective Ethylation of Aldehydes:
H), 1.12 (s, 6 H, Me) ppm. 13C NMR (CDCl3, 75 MHz): δ = 174.1 Under argon, into a 10 mL, round-bottomed flask, equipped with
(N-C=O), 76.8 (CH-O), 69.8 (CH2-O), 61.1 (C), 50.8 (C), 49.0 (N- a magnetic stirrer and containing the corresponding bis(hydroxy
Eur. J. Org. Chem. 2010, 1717–1727
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1725

T. de las Casas Engel, B. Lora Maroto, S. de la Moya Cerero
FULL PAPER
[6]
[7]
SciFinder (copyright 2009, CAS; licensed to UCM) finds ca.
1500 references by crossing the topics “diethylzinc” and “enan-
tioselective”.
Some recent examples are: a) Y. Yang, S.-F. Zhu, C.-Y. Zhou,
Q.-L. Zhou, J. Am. Chem. Soc. 2008, 130, 14052–14053; b) T.
Thaler, T. Knochel, Angew. Chem. Int. Ed. 2009, 48, 645–648;
c) G. Blay, L. Cardona, E. Climent, J. R. Pedro, Angew. Chem.
Int. Ed. 2008, 47, 5593–5596; d) D. B. Biradar, H.-M. Gau,
Tetrahedron: Asymmetry 2008, 19, 733–738.
Some of the most versatile ligands are 1 and 2. On 1, see: a)
Y. K. Chen, S. J. Jeon, P. J. Walsh, W. A. Nuggent, Org. Synth.
2005, 82, 87; on 2, see: b) J. W. Patrick in e-ROS Encycl. Reag.
Org. Synth. 2001; c) D. J. Ramón, M. Yus, Angew. Chem. Int.
Ed. 2004, 43, 284–287; see also ref.[1g]
amide) (0.02 mmol), was added diethylzinc (2.00 mmol, 1.0  in
hexanes) at room temp. The mixture was stirred at room temp. for
5 min. The corresponding aldehyde (1.0 mmol) was then added,
and the reaction mixture was stirred at room temp. for 4 h. The
reaction was then quenched by the addition of aqueous NH4Cl
(1 , 3 mL). The resulting mixture was extracted with diethyl ether
(3ϫ3 mL). The combined organic layers were submitted to Celite
filtration and solvent evaporation. The obtained residue was dis-
solved in HPLC-grade hexanes and submitted to analysis by GC
and chiral HPLC.
[8]
1-Phenylpropan-1-ol: Chiralpak IC, 260 nm, 2-propanol/hexanes
(2:98), 1.3 mL/min. tR = 7.3 (R), 7.8 (S) min.
[9]
For example, see: a) Z. Chai, Z.-F. Xie, X.-Y. Liu, G. Zhao, J.-
D. Wang, J. Org. Chem. 2008, 73, 2947–2950; b) M. M. Hus-
sain, P. J. Walsh, Acc. Chem. Res. 2008, 41, 883–893; c) K. Li,
A. Alexakis, Chem. Eur. J. 2007, 13, 3765–3771.
Some recent examples are: a) V. Forrat, D. J. Ramon, M. Yus,
Tetrahedron: Asymmetry 2009, 20, 65–67; b) V. Forrat, D. J.
Ramón, M. Yus, Tetrahedron: Asymmetry 2008, 19, 537–541;
c) M. A. Pericàs, C. I. Herrerías, L. Solà, Adv. Synth. Catal.
2007, 350, 927–932.
For instance: a) M.-C. Wang, Q.-J. Zhang, G.-W. Li, Z.-K. Liu,
Tetrahedron: Asymmetry 2009, 20, 288–292; b) M.-C. Wang,
Q.-J. Zhang, X.-D. Wang, X. Ding, T.-T. Jing, M.-P. Song, J.
Org. Chem. 2008, 73, 168–176; c) S. Ay, M. Nieger, S. Bräse,
Chem. Eur. J. 2008, 14, 11539–11556.
a) X.-P. Hui, C. Yin, Z.-C. Chen, L.-N. Huang, P.-F. Xu, G.-
F. Fan, Tetrahedron 2008, 64, 2553–2558; b) T. Fang, D.-M.
Du, S.-F. Lu, J. Xu, Org. Lett. 2005, 7, 2081–2084; c) G. Blay,
A. Fernández, A. Marco-Aleixandre, J. R. Pedro, Tetrahedron:
Asymmetry 2005, 16, 1953–1958; d) T. Fang, J. Xu, D.-M. Du,
Lett. Org. Chem. 2006, 3, 780–786; e) S.-M. Lu, H.-B. Zhou,
R.-G. Xie, Huazhong Shifan Daxue Xuebao Zirankexueban
2003, 37, 55–58, 64.
G. Blay, I. Fernández, V. Hernández-Olmos, A. Marco-Aleix-
andre, J. R. Pedro, J. Mol. Catal. A 2007, 276, 235–243.
a) G. Blay, A. Marco-Aleixandre, J. R. Pedro, J. Org. Chem.
2006, 71, 6674–6677; b) Z-Q. Zhou, Huaxue Shiji 2006, 28,
521–522, 540; c) Z. Zhou, Y. Guo, J. Zhang, Wuhan Daxue
Xuebao, Lixueban 2007, 53, 641–645.
For example, see: V. Casarotto, Z. Li, J. Boucau, Y.-M. Lin,
Tetrahedron Lett. 2007, 48, 5561–5564; see also ref.[5]
On interest in using renewable starting materials, see the P. T.
Anastas, J. C. Wagner, “7th Principle of the Green Chemistry”
in Green Chemistry: Theory and Practice, Oxford University
Press, Oxford, 1988, p. 30.
1-(2-Chlorophenyl)propan-1-ol: Chiralpak IA, 260 nm, 2-propanol/
hexanes (1:99), 1.3 mL/min. tR = 12.7 (R), 13.4 (S) min.
1-(4-Chlorophenyl)propan-1-ol: Chiralpak IC, 260 nm, 2-propanol/
hexanes (5:95), 1.3 mL/min. tR = 4.3 (R), 4.5 (S) min.
[10]
1-(2-Methylphenyl)propan-1-ol: Chiralpak IA, 260 nm, 2-propanol/
hexanes (2:98), 1.2 mL/min. tR = 8.8 (R), 10.1 (S) min.
1-(4-Methylphenyl)propan-1-ol: Chiralpak IA, 260 nm, 2-propanol/
hexanes (2:98), 1.2 mL/min. tR = 10.4 (R), 11.7 (S) min.
[11]
[12]
1-Phenylpent-1-en-3-ol: Chiralpak IC, 260 nm, 2-propanol/hexanes
(5:95), 1.3 mL/min. tR = 6.4 (R), 7.2 (S) min.
Acknowledgments
We thank the Ministerio de Educación y Ciencia of Spain (research
project CTQ2007-67103-C02) and the Universidad Complutense de
Madrid (research projects PR34/07-15782 and 910107) for support-
ing this work. T. C. E. thanks the Comunidad de Madrid for a
research contract.
[13]
[14]
[1] Some reviews are: a) R. Noyori, M. Kitamura, Angew. Chem.
Int. Ed. Engl. 1991, 30, 49–69; b) K. Soai, S. Niwa, Chem. Rev.
1992, 92, 833–856; c) L. Pu, H.-B. Yu, Chem. Rev. 2001, 101,
757–824; d) L. Pu, Tetrahedron 2003, 59, 9873–9886; e) S. E.
Denmark, J. Fu, Chem. Rev. 2003, 103, 2763–2793; f) H.-J.
Zhu, J.-X. Jiang, J. Ren, Y.-M. Yan, C. U. Pitmann, Curr. Org.
Synth. 2005, 2, 547–587; g) D. J. Ramón, M. Yus, Chem. Rev.
2006, 106, 2126–2208; h) M. Hatano, T. Miyamoto, K. Ishih-
ara, Curr. Org. Chem. 2007, 11, 127–157.
[15]
[16]
[2] Recent examples of the preparation of biologically active com-
pounds are: a) V. J. Forrat, D. J. Ramón, M. Yus, Tetrahedron:
Asymmetry 2007, 18, 400–405; b) H. Y. Kin, A. E. Lurain, P.
García-García, P. J. Carroll, P. J. Walsh, J. Am. Chem. Soc.
2005, 127, 13138–13139; c) M. Yus, J. R. Diego, O. Prieto, Eur.
J. Org. Chem. 2003, 1745–2748; see also ref.[1g] for some ad-
ditional, interesting examples. On liquid-crystal preparation,
see: d) J. D. Armstrong, J. C. McWilliams, Patent No. US
5977371; some recent examples of the synthesis of useful, chi-
ral, building blocks are: e) K. Yeanick, C. Wolf, Org. Lett.
2008, 10, 3015–3918; f) A. Wilsily, E. Fillion, Org. Lett. 2008,
10, 2801–2804; g) J. Sedelmeir, B. Carsten, J. Org. Chem. 2007,
72, 8859–8862.
[17]
A third example describing an efficient chiral imino-hydroxy-
amide (93% ee in the aldehyde ethylation) was previously re-
ported: A. Mori, D. Yu, S. Inoue, Synlett 1992, 427–428; never-
theless, from a functional point of view, the involved ligand
must be considered a hydroxy-imine-based ligand.
W. Oppolzer, R. N. Radinov, Tetrahedron Lett. 1988, 29, 5645–
5648.
a) H. Kitajima, Y. Aoki, T. Katsuki, Bull. Chem. Soc. Jpn.
1997, 70, 207–217; b) H. Kitajima, K. Ito, T. Katsuki, Chem.
Lett. 1996, 343–344.
a) T. de las Casas Engel, B. Lora Maroto, S. de la Moya Cer-
ero, Tetrahedron: Asymmetry 2008, 19, 646–650; b) T.
de las Casas Engel, B. Lora Maroto, S. de la Moya Cerero, Tet-
rahedron: Asymmetry 2008, 19, 2003–2006.
a) B.-J. Uang, I.-P. Fu, C.-D. Hwang, C.-W. Chang, C.-T. Yang,
D.-R. Hwang, Tetrahedron 2004, 60, 10479–10586; b) C.-D.
Hwang, D.-R. Hwang, B.-J. Uang, J. Org. Chem. 1998, 63,
6762–6763.
A cheaper amidation of ketopinic acid can be achieved by
acidic activation with thionyl chloride (ketopinoyl chloride
generation). For example, see ref.[18,20a,21]
[18]
[19]
[20]
[21]
[22]
[3] P. Knochel, N. Millot, A. L. Rodrigues, Org. React. 2001, 58,
417–713.
[4] M. Oguni, T. Omi, Tetrahedron Lett. 1984, 25, 2823–2824.
[5] a) M. Kitamura, S. Suga, K. Kawai, R. Noyori, J. Am. Chem.
Soc. 1986, 108, 6071–6072; b) M. Kitamura, S. Okada, S. Soga,
R. Noyori, J. Am. Chem. Soc. 1989, 111, 4028–4036; c) M.
Yamakawa, R. Noyori, J. Am. Chem. Soc. 1995, 117, 6327–
6335; d) M. Yamakawa, R. Noyori, Organometallics 1999, 18,
128–133.
1726
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 1717–1727

Ligands for Enantioselective Ethylation of Aromatic Aldehydes
[23] Uang and co-workers have used -selectride instead of NaBH4
for the stereoselective reduction of 16g and 16h (see ref.[21]).
[24] On the diastereoselectivity of carbonyl additions in 1-substi-
tuted 2-norbornanones, see: A. García Martínez, E. Teso Vilar,
A. García Fraile, S. de la Moya Cerero, P. Martínez Ruiz, Tet-
rahedron: Asymmetry 1998, 9, 1737–1745.
symmetric catalysts from ketopinic acid derived, protic-amide-
based bis(hydroxy amides) with short diamine spacers (e.g., see
the use of ligands 14g–h in ref.[21]).
[28] Note that the formation of an N-coordinated, zinc-centered
complex (-CO-N-Zn-N-CO-) from a 2-carbon-spacer bis(hy-
droxy amide) should involve a rigid five-membered-ring spacer,
whereas the carbonyl-chelated one (-N-C=O-Zn-O=C-N-) in-
volves a less rigid, eight-membered-ring spacer.
[29] The reduction of the ligand loading results in the different be-
havior of these diastereomeric ligands (see Table 4).
[30] Walsh has demonstrated the utility of highly concentrated con-
ditions for enantioselective ketone alkylations: S.-J. Jeon, H.
Li, P. J. Walsh, J. Am. Chem. Soc. 2005, 127, 16416–16425.
Received: December 1, 2009
[25] Although catalyst 18 possesses two equal catalytic positions,
the probability of a double, synchronous activation (two molec-
ules of benzaldehyde and two molecules of diethylzinc) in a C2-
symmetric fashion must be extremely low.
[26] On the possibility of the protic-amide groups generating a ZnII
complex from a bis(hydroxy amide), see: N. Nishat, M. M.
Haq, T. Ahamad, V. Kumar, J. Coord. Chem. 2007, 60, 85–96.
[27] The use of appropriate metals (e.g. TiIV instead of ZnII) makes
possible the generation of highly efficient, metal-centered, C2-
Published Online: February 3, 2010
Eur. J. Org. Chem. 2010, 1717–1727
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1727