Asymmetric Henry reactions catalyzed by metal complexes of chiral boron-bridged bisoxazoline (borabox) ligands

Article abstract of DOI:10.1002/ejoc.200800570

Metal complexes of boron-bridged bisoxazolines (borabox ligands) were evaluated as catalysts for the Henry reaction. Copper(II) complexes induced high enantio- and diastereoselectivity in reactions with nitroethane and nitropropane. The amount of base add

Full text of DOI:10.1002/ejoc.200800570

FULL PAPER
DOI: 10.1002/ejoc.200800570
Asymmetric Henry Reactions Catalyzed by Metal Complexes of Chiral
Boron-Bridged Bisoxazoline (borabox) Ligands
Aurélie Toussaint[a] and Andreas Pfaltz*[a]
Keywords: Henry reaction / Asymmetric catalysis / Borabox ligands / Copper complexes / Nitropropane
Metal complexes of boron-bridged bisoxazolines (borabox li-
gands) were evaluated as catalysts for the Henry reaction.
bisoxazoline and other privileged ligands showed that bora-
box was the most suitable ligand system for the enantio- and
Copper(II) complexes induced high enantio- and diastereo- diastereoselective formation of nitroalcohols derived from ni-
selectivity in reactions with nitroethane and nitropropane.
The amount of base added had a strong influence on the for-
mation of the chiral complex and the enantioselectivity. Com-
parison with the corresponding dimethylmethylene-bridged
tropropane and aliphatic aldehydes.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2008)
parts. The results obtained so far indicate that borabox li-
gands are a valuable addition to the known chiral bisoxa-
zolines, as they may provide a useful alternative when stan-
dard box ligands fail.
In view of the successful application of chiral bisoxa-
zolines in enantioselective copper-catalyzed nitroaldol (or
Henry) reactions,[3,4] we decided to evaluate analogous
borabox ligands in this respect. Here we report the results
of this study, which demonstrate that copper(II)-borabox
complexes are efficient catalysts for Henry reactions, and
can induce high enantio- and diastereoselectivities, espe-
cially in reactions between nitroethane or nitropropane and
aliphatic aldehydes.
Introduction
Chiral C2-symmetric bisoxazolines 1 (box) belong to the
most versatile ligands for asymmetric catalysis (Figure 1).[1]
The success of these ligands is explained by the short, ef-
ficient synthesis from readily available precursors, the ease
of modifying the structure, and the ideal position of the
stereogenic centers near the coordination sphere, which per-
mits efficient, direct chirality transfer to a coordinated sub-
strate.
Results and Discussion
Figure 1. Box and borabox ligands.
Initial Screening
Initially the reactivity and selectivity of various borabox
complexes for the Henry reaction were investigated. The re-
action between benzaldehyde and nitromethane in the pres-
ence of 5 mol-% of metal source, 5.5 mol-% of borabox li-
gand and triethylamine was chosen as a model system. Zin-
c(II) and copper(II) triflate proved to be the most suitable
metal sources as other metal salts such as zinc(II) chloride,
copper(II) chloride and copper(II) acetate induced lower
enantioselectivities. Table 1 summarizes the results of the
screening experiments. The ratio of base to ligand was
found to strongly influence both the reactivity and selectiv-
ity of the reaction (Table 1, Entries 1 to 3). Variation of
the amount of base added after complexation of the ligand
showed that in the presence of zinc(II) triflate and ligand
2a, 10.5 mol-% of base gave the highest enantiomeric excess
(70% yield, 24% ee, Entry 2). It was supposed that 5.5 mol-
% of triethylamine was needed for the deprotonation of the
We have recently introduced a new variant of this privi-
leged ligand structure, anionic boron-bridged bisoxazolines
2 (borabox).[2] Like the box ligands the borabox analogs are
readily synthesized from chiral amino alcohols. Compara-
tive studies of the two ligand systems revealed that the reac-
tivities and enantioselectivities of box- and borabox-derived
metal catalysts often differed substantially. For example, in
copper-catalyzed enantioselective acylations of meso-1,2-di-
ols and kinetic resolution of chiral 1,2-diols and pyridyl
alcohols, borabox ligands outperformed their box counter-
[a] Department of Chemistry, University of Basel,
St. Johanns Ring 19, 4056 Basel, Switzerland
Fax: +41-61-267-1103
E-mail: andreas.pfaltz@unibas.ch
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Eur. J. Org. Chem. 2008, 4591–4597
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4591

A. Toussaint, A. Pfaltz
FULL PAPER
Table 1. Optimization of the catalytic enantioselective Henry reac- protonated borabox ligand in the complexation step and
tion between benzaldehyde and nitromethane in the presence of
5 mol-% for the deprotonation of nitromethane. Thus, in-
stead of adding the base in one batch, first 5.5 mol-% of
triethylamine as base.[a]
base was used to induce complexation, followed by a sec-
ond portion when the reaction was started. For the catalyst
system derived from copper(II) triflate and ligand 2b, this
procedure resulted in increased yield and product 3a was
formed with higher enantiomeric excess of 70% (Table 1,
Entries 6 and 7).
To gain additional information on the complexation re-
action, the formation of boraboxzinc(II) complexes was
studied by NMR and mass spectrometry (Scheme 1). When
equimolar amounts of the borabox 2a and zinc(II) triflate
were dissolved in [D6]acetone without the addition of a
base, the NMR and FAB-MS indicated quantitative forma-
tion of a bis(borabox) complex 4. In contrast, in the pres-
ence of one equivalent of potassium carbonate, only the
mono(borabox) zinc triflate complex 5 was observed. Thus,
the addition of base suppresses the formation of a unreac-
tive homoleptic bis(borabox) complex.
The results obtained with different borabox derivatives
showed that ligands carrying a tert-butyl group at the ste-
reogenic center were most effective (Table 1, Entries 8, 9, 14
and 15). The substituents at the boron atom had only a
minor influence on the reactivity and enantioselectivity.
Based on the data summarized in Table 1, copper(II) triflate
and ligand 2b, which gave the best results, were chosen for
subsequent studies.
Entry
Ligand
Metal salt
Et3N [mol-%] Yield [%][b] ee [%][c]
1
2
3
4
5
6
7
8
2a
2a
2a
2a
2b
2b
2b
2c
2c
2d
2d
2e
2e
2f
Zn(OTf)2
Zn(OTf)2
Zn(OTf)2
Cu(OTf)2
Zn(OTf)2
Cu(OTf)2
Cu(OTf)2
Zn(OTf)2
Cu(OTf)2
Zn(OTf)2
Cu(OTf)2
Zn(OTf)2
Cu(OTf)2
Zn(OTf)2
Cu(OTf)2
5.5
59
70
75
90
86
85
94
50
85
81
41
50
33
56
85
2 (R)
24 (S)
22 (S)
1 (R)
3 (R)
35 (S)
70 (S)
25 (R)
66 (S)
3 (R)
5 (R)
6 (R)
6 (R)
17 (R)
54 (S)
10.5
15.5
10.5
10.5
10.5
10.5[d]
10.5[d]
10.5[d]
10.5[d]
10.5[d]
10.5[d]
10.5[d]
10.5[d]
10.5[d]
9
10
11
12
13
14
15
2f
[a] Reactions were performed on a 0.5 mmol scale: Cu(OTf)2
(5 mol-%) and ligand 2 (5.5 mol-%) were stirred in ethanol
(0.75 mL) under Ar at room temperature. After a stirring period of
3 to 5 hours, benzaldehyde (1 equiv.), nitromethane (10 equiv.) and
triethylamine were added and the reaction mixture was stirred for
24 h. [b] After column chromatography. [c] Determined by chiral
HPLC. The absolute configuration of 3a was assigned by compari-
son with that reported in the literature.[4a] [d] 5.5 mol-% of triethyl-
amine were added initially. After complexation for 3 to 5 hours at
room temp. and the addition of benzaldehyde and nitromethane,
5 mol-% of triethylamine was added.
Screening of different reaction media showed that protic
solvents generally gave the best results (EtOH, iPrOH,
MeOH) followed by aprotic polar solvents (MeCN, acetone,
MeNO2), whereas aprotic apolar solvents (CH2Cl2, THF)
were not suitable for this reaction. Other bases were also
tested, including Hünig’s base and potassium carbonate,
but the enantioselectivities were lower than with triethyl-
amine.
Scheme 1. Complexation studies with Zn(OTf)2.
4592
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 4591–4597

Asymmetric Henry Reactions
Influence of the Base
compare Entries 1/2, 3/4, and 5/6). However, for aromatic
aldehydes like ortho-methoxybenzaldehyde, the effect was
negligible (Entries 7/8).
Subsequently, the influence of the amount of triethyl-
amine on the reactivity and stereoselectivity was studied in
greater detail. The quantity of base used for complex for-
mation was kept constant at 5.5 mol-% (1 equiv. based on
the borabox ligand), while the additional portion of trieth-
ylamine added after the complexation step was varied from
0 to 5.5 mol-% (Table 2).
Table 3. Influence of the amount of triethylamine in Henry reac-
tions of nitromethane.[a]
Table 2. Influence of the amount of triethylamine in the Henry re-
action between cyclohexanecarboxaldehyde and nitropropane.[a]
Entry
R
3
Et3N[b]
[mol-%]
Time
[days]
Yield
[%]
ee
[%][d]
1
2
3
4
5
6
7
8
iPr
iPr
Et
Et
Cy
Cy
3n
3n
3o
3o
3p
3p
3g
3g
5
4.5
5
4.5
5
4.5
5
4.5
5
5
5
5
5
5
1
1
78[c]
83[c]
87[c]
90[c]
85[c]
93[c]
82
38 (S)
56 (S)
38 (S)
55 (S)
71 (S)
81 (S)
65 (S)
68 (S)
Entry
Et3N [b] Conversion syn/anti[c]
[mol-%]
ee (syn)
ee (anti)
2-MeOC6H4
2-MeOC6H4
[%][c]
[%][d,e]
[%][d,f]
80
1
2
3
4
5
6
5.5
5
4.5
3.0
1.5
0
64
52
60
68
62
70
86:14
90:10
92:08
90:10
91:09
92:08
82 (1S,2S)
84 (1S,2S)
91 (1S,2S)
91 (1S,2S)
91 (1S,2S)
75 (1S,2S)
50 (1S,2R)
41 (1S,2R)
33 (1S,2R)
31 (1S,2R)
28 (1S,2R)
30 (1S,2R)
[a] Reactions were performed on a 0.5 mmol scale: Cu(OTf)2
(5 mol-%), ligand 2b (5.5 mol-%) and triethylamine (5.5 mol-%)
were stirred in ethanol (0.75 mL) under Ar at room temperature.
After a stirring period of 3 to 5 hours, the aldehyde (1 equiv.), nitro-
methane (10 equiv.) and triethylamine were added and the reaction
mixture was stirred for 1 to 5 days. [b] Additional amount of trieth-
ylamine added after the complexation step. [c] Conversion deter-
[a] Reactions were performed on a 0.5 mmol scale: Cu(OTf)2
(5 mol-%), ligand 2b (5.5 mol-%) and triethylamine (5.5 mol-%)
were stirred in ethanol (0.75 mL) under Ar at room temperature.
After a stirring period of 3 to 5 hours, cyclohexanecarboxaldehyde
(1 equiv.), nitropropane (10 equiv.) and triethylamine were added
and the reaction mixture was stirred for 5 days. [b] Additional
amount of triethylamine added after the complexation step. [c] De-
termined by 1H NMR spectroscopy of the crude product. [d] Deter-
mined by chiral HPLC. [e] The absolute configuration of syn-6n
was determined by comparison with published HPLC retention
times.[5] [f] The absolute configuration of anti-6n was determined
by analogy with the absolute configuration of compound anti-7a.[6]
1
mined by H NMR spectroscopy of the crude product. [d] Deter-
mined by chiral HPLC. The absolute configuration of nitroalcohol
products were assigned by comparison with that reported in the
literature.[4a]
Table 4. Henry reaction with nitromethane catalyzed by 2b/Cu-
(OTf)2.[a]
Applying the optimal reaction conditions described in
Table 1 (5 mol-% of additional triethylamine), the reaction
between cyclohexanecarboxaldehyde and nitropropane af-
forded the corresponding nitro alcohol 6n with encouraging
diastereoselectivity and enantioselectivity [90:10 syn/anti,
84% ee (syn), Table 2, Entry 2]. With more base (5.5 mol-
% of additional triethylamine) the conversion increased to
64% but the stereoselectivity decreased [86:14 syn/anti, 82%
ee (syn), Entry 1]. When the amount of additional triethyl-
amine was reduced to 4.5 mol-%, the diastereoselectivity
and enantioselectivity as well as conversion increased [92:08
syn/anti, 91% ee (syn), Entry 3]. Essentially the same en-
antio- and diastereoselectivity were obtained with 3 and
1.5 mol-% of additional triethylamine (Entries 4 and 5).
When no additional base was added, the diasteroselectivity
remained high but the enantioselectivity dropped to 75% ee
(Entry 6).
Entry
R
3
Time
[days]
Yield
[%]
ee
[%][c]
1
2
3
4
5
6
7
8
Ph
3a
3b
3c
3d
3e
3f
3g
3h
3i
3j
3k
3l
3m
3n
3o
3p
1
1
1
1
1
1
1
1
1
1
5
5
5
5
5
5
94
95
86
93
90
70
80
63
95
70 (S)
10 (S)
2 (S)
2-NO2C6H4
4-NO2C6H4
2-ClC6H4
4-ClC6H4
3-MeC6H4
2-MeOC6H4
4-MeOC6H4
1-naphthyl
4-PhC6H4
iBu
nBu
tBu
iPr
Et
27 (S)
38 (S)
56 (S)
68 (S)
52 (S)
46 (S)
53 (S)
57 (S)
60 (S)
79 (S)
56 (S)
55 (S)
81 (S)
9
10
11
12
13
14
15
16
89
92[b]
80[b]
89[b]
83[b]
90[b]
93[b]
Cy
[a] Reactions were performed on a 0.5 mmol scale: Cu(OTf)2
(5 mol-%), ligand 2b (5.5 mol-%) and triethylamine (5.5 mol-%)
were stirred in ethanol (0.75 mL) under Ar at room temperature.
After a stirring period of 3 to 5 hours, the aldehyde (1 equiv.), nitro-
methane (10 equiv.) and triethylamine (4.5 mol-%) were added and
the reaction mixture was stirred for 1 to 5 days. [b] Conversion
determined by 1H NMR spectroscopy of the crude product. [c]
Determined by chiral HPLC. The absolute configuration of nitro-
alcohol products 3 were assigned by comparison with that reported
To see if reactions with other substrates showed the same
base dependence, we carried out a series of Henry reactions
with nitromethane and several aldehydes (Table 3). As
found for the reaction of nitropropane with cyclohexanecar-
baldehyde, reducing the amount of additional base from 5.5
to 5.0 mol-% had a beneficial effect on enantioselectivity
and conversion for the aliphatic aldehydes tested (Table 3, in the literature.[4a]
Eur. J. Org. Chem. 2008, 4591–4597
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4593

A. Toussaint, A. Pfaltz
FULL PAPER
Reactions with Nitromethane
Reactions with Nitropropane
With the exception of 2- and 4-nitrobenzaldehyde, which
With an optimized procedure in hand, the scope of the
Henry reaction with nitromethane was further investigated again gave nearly racemic products, aromatic aldehydes re-
(Table 4). Aromatic aldehydes in general gave only low to acted with moderate to good enantioselectivity (Table 5,
moderate enantioselectivities. The highest enantiomeric ex- Entries 1–8). The diastereoselectivities on the other hand
cesses were obtained with benzaldehyde and derivatives were rather low. The highest ee values were recorded for 2-
with electron-rich aryl groups (46–70% ee; Entries 1 and chlorobenzaldehyde (85%) and 4-methoxybenzaldehyde
6–10). Reactions with electron-deficient aromatic aldehydes (82%). Interestingly, the corresponding 4-chloro and 2-
gave very poor results: 2- and 4-nitrobenzaldehyde led to methoxy derivatives afforded lower enantioselectivities,
nearly racemic products, while 2- and 4-chlorobenzaldehyde whereas in the reaction with nitromethane the opposite
reacted with 27 and 38% ee, respectively (Entries 2–5).
selectivity order was observed (4-Cl Ͼ 2-Cl; 2-MeO Ͼ 4-
Aliphatic aldehydes, such as propionaldehyde, isobutyral- MeO).
dehyde, valeraldehyde, and isobutyraldehyde were converted
Much higher syn/anti ratios were obtained in the ali-
into the corresponding nitro alcohols with moderate phatic series with isobutyraldehyde, propionaldehyde and
enantioselectivities of 55–60% ee (Entries 11, 12, 14, 15). cyclohexanecarboxaldehyde (Entries 12–18). The best re-
Much better results were obtained with the more sterically sults were achieved with the latter two substrates: 6m:
hindered pivaldehyde and cyclohexanecarboxaldehyde: the syn/anti 85:15, 87% ee (syn); 6n: 92:08, 91% ee. In order to
Henry products 3m and 3p were formed with 79 and 81% increase the rate of reaction, the reactions with propional-
ee and high conversion (Entries 13 and 16). Overall, the dehyde and cyclohexanecarboxaldehyde were conducted at
borabox ligands induced lower enantioselectivity in the 35 °C (Entries 15 and 18). In the latter case, the diastereo-
Henry reaction with nitromethane than bisoxazoline li- selectivity dropped slightly to 85:15, but the enantiomeric
gands.[4a] However, the promising results obtained with ali- excess of the major syn isomer remained at 91% while the
phatic aldehydes prompted us to further evaluate the bora- conversion increased to 73% (Entry 18). For propionalde-
box ligand 2b in the Henry reaction with nitropropane.
hyde, both the diastereoselectivity and enantioselectivity de-
Table 5. Henry reaction with nitropropane catalyzed by 2b/Cu(OTf)2.[a]
Entry
R
6
Time
[days]
Yield
[%][b]
syn/anti[c]
ee (syn)
ee (anti)
[%][d,e]
[%][d,f]
1
2
3
4
5
6
7
8
Ph
6a
6b
6c
6d
6e
6f
6g
6h
6i
6j
6k
6l
6m
6m
6m
6n
6n
6n
1
1
1
1
1
1
2
1
2
5
4
5
4
4
5
5
5
5
74
90
80
59
96
83
58
70
62:38
68:32
73:27
53:47
65:35
61:39
75:25
63:37
63:37
66:34
64:36
87:13
85:15
82:18
61:39
92:08
93:07
85:15
65 (1S,2S)
3 (1S,2S)
2 (1S,2S)
14 (1S,2R)
3 (1S,2R)
8 (1S,2R)
38 (1S,2R)
12 (1S,2R)
10 (1S,2R)
48 (1S,2R)
2 (1S,2R)
11 (1S,2R)
21 (1S,2R)
8 (1S,2R)
19 (1S,2R)
49 (1S,2R)
58 (1S,2R)
8 (1S,2R)
33 (1S,2R)
74 (1S,2R)
27 (1S,2R)
2-NO2C6H4
4-NO2C6H4
2-ClC6H4
4-ClC6H4
2-MeOC6H4
4-MeOC6H4
2-naphthyl
PhCH2CH2
iBu
nBu
iPr
Et
Et
85 (1S,2S)
48 (1S,2S)
50 (1S,2S)
82 (1S,2S)
54 (1S,2S)
39 (1S,2S)
63 (1S,2S)
48 (1S,2S)
75 (1S,2S)
87 (1S,2S)
89 (1S,2S)
30 (1S,2S)
91 (1S,2S)
94 (1S,2S)
91 (1S,2S)
9
45[g]
53[g]
64[g]
55[g]
62[g]
55[h]
97[g][i]
60[g]
53[h]
73[g][i]
10
11
12
13
14
15
16
17
18
Et
Cy
Cy
Cy
[a] Reactions were performed on a 0.5 mmol scale: Cu(OTf)2 (5 mol-%), ligand 2b (5.5 mol-%) and triethylamine (5.5 mol-%) were stirred
in ethanol (0.75 mL) under Ar at room temperature. After a stirring period of 3 to 5 hours, the aldehyde (1 equiv.), nitropropane (10 equiv.)
and triethylamine (4.5 mol-%) were added and the reaction mixture was stirred for 1 to 5 days. [b] Combined yields of syn and anti
1
isomers. [c] Determined by H NMR spectroscopy of the crude product. [d] Determined by chiral HPLC. [e] The absolute configuration
of nitroalcohol products syn-6 were assigned by analogy with compound syn-6n. [f] The absolute configuration of nitroalcohol products
1
anti-6 were assigned by analogy with compound anti-7a. [g] Conversion determined by H NMR spectroscopy of the crude product. [h]
Reactions were performed on a 2.5 mmol scale. [i] Reactions were carried out at 35 °C.
4594
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 4591–4597

Asymmetric Henry Reactions
terioriated significantly at elevated temperature (Entry 15).
Reactions with these two aldehydes were also carried out
on a larger scale. Using 2.5 mmol of aldehyde, the enantio-
selectivities were improved in both cases (89 and 94% ee for
the syn isomers 6m and 6n, Entries 14 and 17).
Conclusions
The results of this study show that boraboxcopper(II)
complexes are active and stereoselective catalysts for asym-
metric Henry reactions of nitroalkanes with aldehydes. The
highest enantio- and diastereoselectivities were obtained in
reactions between cyclohexanecarboxaldehyde and nitro-
ethane or nitropropane. For reactions of this type, which
build up two new sterogenic centers, borabox-Cu complexes
rival the most selective catalysts known to date.[3] With ee
values of up to 94% and syn/anti ratios of Ն 90:10, the
borabox ligand 2b clearly outperformed the corresponding
box derivative 1a and other privileged ligands. On the other
hand, for reactions with nitromethane the box ligand 1a
proved to be superior to borabox analogue 2b. Overall,
these findings are in line with previous studies, which have
shown that the performance of borabox and box ligands
can differ substantially despite their similarity.[2] Thus, re-
placement of box by borabox ligands may serve as a useful
strategy for reactions that give unsatisfactory results with
box complexes.
Reactions with Nitroethane
In addition, nitroethane was also briefly investigated as
reactant (Table 6). Again the reaction with cyclohexanecar-
boxaldehyde was highly stereoselective [90:10 syn/anti, 90%
ee (syn); Entry 2]. However, the Henry adducts 7a and 7c
derived from benzaldehyde and propionaldehyde were
formed with low diastereo- and enantioselectivity.
Table 6. Henry reaction with nitroethane catalyzed by 2b/Cu-
(OTf)2.[a]
Entry
R
7
Time
[days]
Yield syn/anti[c] ee (syn)
ee (anti)
[%][b]
[%][d,e]
[%][d,f]
Experimental Section
1
2
3
Ph 7a
Cy 7b
Et 7c
1
5
5
80
62:38
90:10
64:36
21 (1S,2S)
90 (1S,2S)
51 (1S,2S)
15 (1S,2R)
47 (1S,2R)
23 (1S,2R)
General Procedure for the Copper-Catalyzed Nitroaldol Reaction:
Ligand 2b (11.5 mg, 0.027 mmol), Cu(OTf)2 (9.00 mg, 0.025 mmol)
and Et3N (3.80 µL, 0.027 mmol) were added to a dry Young tube
containing a magnetic stirring bar. Ethanol (0.75 mL) was added
and the mixture was stirred at room temp. for 3 to 5 hours. To
the resulting dark green solution the nitroalkane (5.00 mmol), the
aldehyde (0.50 mmol) and Et3N (3.00 µL, 0.022 mmol) were added.
After stirring at room temperature for the indicated time (see
Tables 2, 3, 4, and 5), the volatile compounds were removed under
reduced pressure and the crude product was purified by column
chromatography on silica gel [Chemie Uetikon (C-560 D, 0.040–
0.063 mm) or Merck (silica gel 60, 0.040–0.063 mm)].
82[g]
80[g]
[a] Reactions were performed on a 0.5 mmol scale: Cu(OTf)2
(5 mol-%), ligand 2b (5.5 mol-%) and triethylamine (5.5 mol-%)
were stirred in ethanol (0.75 mL) under Ar at room temperature.
After a stirring period of 3 to 5 hours, the aldehyde (1 equiv.), nitro-
ethane (10 equiv.) and triethylamine (4.5 mol-%) were added and
the reaction mixture was stirred for 1 to 5 days. [b] Combined yield
1
of syn and anti isomers. [c] Determined by H NMR spectroscopy
of the crude product. [d] Determined by chiral HPLC. [e] The abso-
lute configuration of nitroalcohol products syn-7 were assigned by
analogy with compound syn-7a.[6] [f] The absolute configuration of
nitroalcohol products anti-7 were assigned by analogy with com-
pound anti-7a.[6] [g] Conversion determined by 1H NMR spec-
troscopy of the crude product.
Nitroalcohol (S)-3f: The crude product was purified by column
chromatography using 15% EtOAc/hexanes as eluent to give a col-
orless oil (63 mg, 70%, Rf = 0.33). [α]2D0 = +7.7 (c = 1.1, CHCl3,
1
3
56% ee). H NMR (400.1 MHz, CDCl3, 295 K): δ = 7.29 (t, JHH
= 7.6 Hz, 1 H, HAr), 7.22–7.20 (m, 1 H, HAr), 7.20–7.16 (m, 2 H,
3
2
HAr), 5.42 (br. d, JHH = 9.6 Hz, 1 H, CHOH), 4.60 (dd, JHH
=
3
2
3
13.4, JHH = 9.6 Hz, 1 H, CHHNO2), 4.50 (dd, JHH = 13.4, JHH
= 3.0 Hz, 1 H, CHHNO2), 2.85 (s, 1 H, OH), 2.37 (s, 3 H, CH3)
ppm. 13C{1H} NMR (100.6 MHz, CDCl3, 295 K): δ = 139.3 (CAr),
138.5 (CAr), 130.1 (CHAr), 129.3 (CHAr), 127.0 (CHAr), 123.4
(CHAr), 81.7 (CHOH), 71.4 (CH2NO2), 21.8 (CH3) ppm. IR
Comparison of Borabox 2b with Box and Other Ligands
In order to assess the efficiency of the borabox ligand 2b
in comparison to analogous bisoxazolines and other privi-
leged ligands,[7] box 1a, pybox 8, phox derivatives 9a and
9b, and binap 10 were tested in the Henry reaction between
cyclohexanecarboxaldehyde and nitropropane (Table 7). In
general, the CuII complexes of these ligands were more re-
active than the Cu-borabox catalyst. With ligands (R)- and
(S)-10 the reaction was complete after less than one day,
however, enantioselectivities were very low (Entries 6 and
7). Phox and pybox ligands also performed poorly. Better
results were obtained with the box derivative 1a, but both
the syn/anti ratio and the ee were significantly lower than
with the borabox analog 2b.
(NaCl): ν = 3460 (m ), 2984 (m), 2927 (m), 1735 (s), 1556 (s), 1423
˜
br
(w), 1377 (s), 1247 (s), 1158 (w), 1046 (s), 918 (w), 889 (w), 848
(w), 790 (m), 707 (m) cm–1. C9H11NO3 (181.19 g/mol): calcd. C
59.66, H 6.12, N 7.73; found C 59.42, H 6.14, N 7.55. MS (EI,
70 eV): m/z (%) = 181 (4) [M+], 134 (45), 121 (23), 120 (75), 119
(90), 117 (11), 93 (19), 92 (18), 91 (100), 77 (10), 65 (19), 61 (19), 43
(47). HPLC: OD-H, n-heptane/iPrOH (85:15), 0.8 mL/min, 20 °C,
220 nm, tR = 10.7 min (R), 12.4 min (S).
The absolute configuration of 3f was assigned based on the sign
of optical rotation in comparison with the optical rotation of the
corresponding product 3a reported in the literature.[4a]
Eur. J. Org. Chem. 2008, 4591–4597
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4595

A. Toussaint, A. Pfaltz
FULL PAPER
Table 7. Comparison of borabox 2b with other ligands in the Henry reaction between cyclohexanecarboxaldehyde and nitropropane.[a]
[a] Reactions were performed on a 0.5 mmol scale: Cu(OTf)2 (5 mol-%), ligand (5.5 mol-%) and triethylamine (5.5 mol-%) were stirred
in ethanol (0.75 mL) under Ar at room temperature. After a stirring period of 3 to 5 hours, cyclohexanecarboxaldehyde (1 equiv.),
nitropropane (10 equiv.) and triethylamine (4.5 mol-%) were added and the reaction mixture was stirred for 1 to 5 days. [b] Determined
1
by H NMR spectroscopy of the crude product. [c] Determined by chiral HPLC.
Nitroalcohols syn-6n and anti-6n: The crude product was purified
by column chromatography using 15% EtOAc/hexanes as eluent to
HCy), 1.67–1.63 (m, 2 H, HCy), 1.39–1.33 (m, 1 H, HCy), 1.26–1.10
3
(m, 5 H, HCy), 0.98 (t, JHH = 7.4 Hz, 3 H, CH3) ppm. 13C{1H}
give syn-6n (50 mg, 50%, Rf = 0.29) and anti-6n (6 mg, 3%, Rf = NMR (125.0 MHz, CDCl3, 295 K): δ = 91.9 (CHNO2), 75.9
0.19) as colorless oils. syn-6n: [α]2D0 = +2.0 (c = 0.73, CHCl3, 94%
(CHOH), 40.3 (CHCy), 29.8 (CH2Cy), 26.8 (CH2Cy), 26.1 (CH2Cy),
26.0 (CH2Cy), 25.7 (CH2Cy), 24.0 (CH2), 10.2 (CH3) ppm. IR
(NaCl): ν = 3441 (s ), 2929 (s), 2855 (s), 1550 (s), 1451 (m), 1378
˜
br
1
3
ee). H NMR (500.1 MHz, CDCl3, 295 K): δ = 4.54 (ddd, JHH
10.7, JHH = 6.4, JHH = 4.4 Hz, 1 H, CHNO2), 3.63–3.58 (m, 1
=
3
3
3
H, CHOH), 2.20 (d, JHH = 6.9 Hz, 1 H, OH), 2.08–2.01 (m, 1 H, (m), 1306 (w), 1109 (w), 1028 (w), 892 (w) cm–1. C10H19NO3
CHHCH3), 1.88–1.82 (m, 1 H, CHHCH3), 1.80–1.72 (m, 3 H, (201.26 g/mol): calcd. C 59.68, H 9.51, N 6.96; found C 59.88, H
4596
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 4591–4597

Asymmetric Henry Reactions
9.35, N 6.79, HPLC: AD-H, n-heptane/iPrOH (97:03), 0.8 mL/min,
20 °C, 215 nm, tR = 18.9 min (1S,2S), 29.4 min (1R,2R).
Acknowledgments
Support of this work by the Swiss National Science Foundation is
gratefully acknowledged.
anti-6n: [α]2D0 = –1 (c = 0.12, CHCl3, 74% ee). 1H NMR
3
3
(500.1 MHz, CDCl3, 295 K): δ = 4.48 (ddd, JHH = 11.1, JHH
=
4.7, 3JHH = 3.0 Hz, 1 H, CHNO2), 3.76 (appt, 3JHH = 5.4 Hz, 1 H,
CHOH), 2.46 (br. s, 1 H, OH), 2.14–2.08 (m, 1 H, CHHCH3), 1.89–
1.86 (m, 1 H, CHHCH3), 1.79–1.71 (m, 3 H, HCy), 1.67–1.63 (m,
2 H, HCy), 1.53–1.48 (m, 1 H, HCy), 1.26–1.10 (m, 5 H, HCy), 0.96
[1] a) A. Pfaltz, Acc. Chem. Res. 1993, 26, 339–345; b) G. Desi-
moni, G. Faita, K. A. Jørgensen, Chem. Rev. 2006, 106, 3561–
3651; c) H. A. McManus, P. J. Guiry, Chem. Rev. 2004, 104,
4151–4202.
[2] a) C. Mazet, V. Köhler, A. Pfaltz, Angew. Chem. 2005, 117,
4966–4969; Angew. Chem. Int. Ed. 2005, 44, 4888–4891; b) C.
Mazet, S. Roseblade, V. Köhler, A. Pfaltz, Org. Lett. 2006, 8,
1879–1882; c) A. Toussaint, Dissertation, University of Basel
2008.
[3] For a recent review on catalytic asymmetric Henry reactions,
see: C. Palomo, M. Oiarbide, A. Laso, Eur. J. Org. Chem. 2007,
2561–2574.
[4] a) D. A. Evans, D. Seidel, M. Rueping, H. Wai Lam, J. T. Shaw,
C. W. Downey, J. Am. Chem. Soc. 2003, 125, 12692–12693; b)
C. Christensen, K. Juhl, R. G. Hazell, K. A. Jørgensen, J. Org.
Chem. 2002, 67, 4875–4881.
[5] Y. Sohtome, Y. Hashimoto, K. Nagasawa, Eur. J. Org. Chem.
2006, 2894–2897.
[6] T. Purkarthofer, K. Gruber, M. Gruber-Khadjawi, K. Waich,
W. Skranc, D. Mink, H. Griengl, Angew. Chem. Int. Ed. 2006,
45, 3454–3456.
[7] T. P. Yoon, E. N. Jacobsen, Science 2003, 299, 1691–1693.
[8] D. Seebach, A. K. Beck, T. Mukhopadhyay, E. Thomas, Helv.
Chim. Acta 1982, 65, 1101–1133.
3
(t, JHH = 7.4 Hz, 3 H, CH3) ppm. 13C{1H} NMR (125.8 MHz,
CDCl3, 295 K): δ = 91.7 (CHNO2), 76.4 (CHOH), 40.4 (CHCy),
29.4 (CH2Cy), 28.1 (CH2Cy), 26.1 (CH2Cy), 25.8 (CH2Cy), 25.5
(CH2Cy), 24.2 (CH ), 10.7 (CH ) ppm. IR (NaCl): ν = 3448 (m ),
˜
2
3
br
2929 (s), 2855 (s), 1550 (s), 1451 (m), 1379 (m), 1306 (w), 1108 (m),
1058 (w), 893 (w) cm–1. C10H19NO3 (201.26 g/mol): calcd. C 59.68,
H 9.51, N 6.96; found C 59.48, H 9.32, N 6.80, HPLC: AD-H, n-
heptane/EtOH (97:03), 0.5 mL/min, 20 °C, 215 nm, tR = 31.4 min
(1R,2S), 38.2 min (1S,2R).
The relative configuration was assigned by comparison of the 13C
NMR spectra of 6n with similar compounds reported in the litera-
ture.[8] The absolute configuration of syn-6n was determined by
comparison with published HPLC retention times[5] and the abso-
lute configuration of anti-6n was assigned by analogy with the ab-
solute configuration of nitroalcohol anti-7a.[6]
Supporting Information (see also the footnote on the first page of
this article): Characterization data including microanalytical analy-
ses for compounds 6c–6h, 6j–6m, 7a and 7c.
Received: June 12, 2008
Published Online: August 1, 2008
Eur. J. Org. Chem. 2008, 4591–4597
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4597