α-Hydroxy carboxylic acids as ligands for enantioselective addition reactions of organoaluminum reagents to aromatic and aliphatic aldehydes

Article abstract of DOI:10.1016/j.tetasy.2004.11.084

The first examples of the enantioselective titanium-mediated trialkylaluminum additions to aromatic and aliphatic aldehydes catalyzed by optically active α-hydroxy acids are presented. The reactions proceed with very good yields and in good asymmetric induction. Enantioselectivities up to 92% are obtained depending on the ligand and aldehyde used. A stereochemical model for the reaction is proposed.

Full text of DOI:10.1016/j.tetasy.2004.11.084

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 16 (2005) 851–855
a-Hydroxy carboxylic acids as ligands for enantioselective
addition reactions of organoaluminum reagents to aromatic
and aliphatic aldehydes
Tomasz Bauer* and Joanna Gajewiak
Department of Chemistry, Warsaw University, Pasteura 1, PL-02-093 Warsaw, Poland
Received 28 October 2004; accepted 23 November 2004
Available online 30 January 2005
Abstract—The first examples of the enantioselective titanium-mediated trialkylaluminum additions to aromatic and aliphatic alde-
hydes catalyzed by optically active a-hydroxy acids are presented. The reactions proceed with very good yields and in good asym-
metric induction. Enantioselectivities up to 92% are obtained depending on the ligand and aldehyde used. A stereochemical model
for the reaction is proposed.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
ther purified by crystallization. For industrial purposes,
such a solution could be more interesting than the use of
a ligand, which gives product with very high ee, but its
synthesis is tedious and expensive (Scheme 1).
Enantioselective organometallic addition to aldehydes is
a valuable method for the synthesis of optically active
secondary alcohols and several efficient systems have
been developed, asymmetric additions of organometallic
reagents to aldehydes remains a challenge. Chiral sec-
ondary alcohols are often found in natural products,
and are of great interest to the pharmaceutical industry.
From the economic point of view, catalytic systems
based on inexpensive chiral ligands and organometallic
reagents available in bulk quantities can be very attrac-
tive provided that chiral product is obtained with good
enantiomeric excess (better than 80%) and can be fur-
Recently, we have introduced new class of chiral ligands
for enantioselective organometallic additions. a-Hydr-
oxy carboxylic acids are excellent ligands for the enantio-
selective addition of diethylzinc to aromatic and
aliphatic aldehydes.1 Taking into account the above dis-
cussed literature reports2–4 we decided to investigate the
additions of triethyl- and trimethylaluminum to alde-
hydes, catalyzed by a-hydroxy carboxylic acids 1a–h
(Scheme 2).
R
O
HO
OH
1
OH
OH
1, Ti(OiPr)4
CHO
+
+
+
OH
Et3Al
(S)-3
(R)-3
4
2
Scheme 1.
*
Corresponding author. Tel.: +48 22 822 02 11x273; fax: +48 22 822 59 96; e-mail: tbauer@chem.uw.edu.pl
0957-4166/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2004.11.084

852
T. Bauer, J. Gajewiak / Tetrahedron: Asymmetry 16 (2005) 851–855
O
O
O
HO
1a
OH
HO
OH
HO
O
OH
1b
1c
O
O
O
HO
OR
HO
OH
HO
OH
HO
OH
OH
1g R=Me
1h R=iPr
1d
1e
1f
Scheme 2.
2. Results and discussion
2.1. Enantioselective triethylaluminum addition
To examine the effect of the solvent the reaction was car-
ried out in five typical solvents. The best enantiomeric
excess and yield were obtained for tetrahydrofuran
(Table 1, entry 1), but the reaction was much slower
than in other solvents. The fastest reaction was observed
in toluene, but the product was obtained with low ee.
We decided to test, whether the mixture of these two
solvents (toluene/THF 1:1) can give better results, but
the reaction appeared to be as slow as it was in THF
and gave the product with moderate ee (Table 1, entry
6). For further experiments THF was chosen as the
solvent.
2.1.1. Enantioselective triethylaluminum addition to alde-
hydes in the presence of (S)-mandelic acid. The first
reaction was performed under conditions proposed by
Chan et al.: 0.2 equiv of (S)-mandelic acid 1a as the
ligand, 1.4 equiv of Ti(OiPr)4, and 3 equiv of Et3Al in
THF at room temperature.2 The reaction was stirred
overnight, and after TLC showed complete conversion,
quenched with aqueous hydrochloric acid. After work-
up and chromatography 1-phenyl-1-propanol 3 was iso-
lated in 85% yield. The enantiomeric excess [74% ee in
favor of the (1S) product] was determined by HPLC
using a Chiracel OD column.
We continued investigation of factors influencing yield
and asymmetric induction in this reaction. The reaction
performed in the absence of Ti(OiPr)4 is slow and
Table 1. Et3Al addition in various solvents
Entry
Ligand
Solvent
Temp [ꢁC]
Time [h]
Yield [%]
Ee (config.) [%]
1
2
3
4
5
6
1a
1a
1a
1a
1a
1a
THF
n-Hexane
Ether
17
28
28
18
28
19
19
3
85
83
86
34
86
96
74 (S)
10 (S)
10 (S)
3 (R)
2.5
4
CH2Cl2
Toluene
Toluene/THF
2.5
19
11 (S)
62 (S)
Table 2. Influence of various parameters on the addition of Et3Al to benzaldehyde
Entry
Ligand
Amount [equiv]
Concn [mmol/mL]
Temp [ꢁC]
Time [h]
Yield [%]
Ee (config.) [%]
1a
2b
3c
4
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
0.2
0.2
0.2
0.1
0.2
0.4
0.4
0.4
0.6
0.6
0.8
1.0
0.04
0.04
0.04
0.03
0.04
0.05
0.1
27
18
27
19
43
19
19
19
4
88
90
76
77
85
85
91
75
95
90
74
23
14 (R)
44 (S)
4 (R)
18
18
38 (S)
74 (S)
82 (S)
81 (S)
53 (S)
81 (S)
78 (S)
73 (S)
66 (S)
5
18
18
6
7
18
8
0.1
ꢀ23
18
140
19
19
43
43
9
0.06
0.04
0.07
0.08
10
11
12
18
18
18
a Reaction in the absence of Ti(OiPr)4.
b Addition order was reversed from Ti(OiPr)4/Et3Al to Et3Al/Ti(OiPr)4.
c Reaction in the presence of Ti(OtBu)4 instead of Ti(OiPr)4.

T. Bauer, J. Gajewiak / Tetrahedron: Asymmetry 16 (2005) 851–855
853
Table 3. Addition reactions of Et3Al to benzaldehyde in the presence
of ligands 1a–h
proceeds with low asymmetric induction yielding (R)
product (Table 2, entry 1). Addition of aluminum and
titanium reagents in the reversed order influenced the
enantioselectivity greatly (Table 2, entries 2 and 3),
which suggests a transition state different from the one
proposed for the analogous diethylzinc addition.1 The
reaction is even more susceptible to bulkiness of the tita-
nium reagent. In the presence of Ti(OtBu)4 the R enan-
tiomer was obtained in 4% ee and accompanied by
27.6% of benzyl alcohol (Table 2, entry 4).5 The optimal
amount of ligand 1a was found to be 0.4 equiv at the
concentration 0.1 mmol/mL at room temperature (Table
2, entries 5–13).
Entry Ligand Amount Temp Time Yield Ee (config.)
[equiv]
[ꢁC]
[h]
[%]
[%]
1
2
3
4
5
6
8
9
1a
1b
1c
1d
1e
1f
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
18
18
18
18
18
19
19
19
4
4
91
96
95
96
88
95
95
95
81 (S)
90 (S)
87 (S)
88 (S)
80 (S)
81 (S)
84 (S)
82 (S)
4
4
19
19
4
1g
1h
4
ligand 1b the amount of benzyl alcohol increased only
slightly when compared to ligand 1a (Table 4, entry 4
2.1.2. Enantioselective triethylaluminum addition to alde-
hydes in the presence of various a-hydroxy acids. After
the best conditions had been established for the ligand
1a, we performed a series of reactions in the presence
of ligands 1b–h (Table 3). The best result, 90% ee, was
obtained for ligand 1b.
Table 4. Amount of benzyl alcohol in reaction mixture
Entry Ligand Ligand Concn
Time Ratio Amount
4:3a
amount [mmol/mL] [h]
[equiv]
of 4 [%]
1
2
3
4
1a
1a
1a
1b
0.2
0.4
0.4
0.4
0.04
0.05
0.1
19
19
4
1:14
1:57
7.5
1.2
As already mentioned, the main products were accom-
panied by benzyl alcohol 4. The amount of 4 was depen-
dent on the reaction conditions, ligand type, and
amount. Fortunately, in the presence of 0.4 equiv of
1:125 0.8
1:72 1.4
0.1
4
Table 5. Addition reactions of Et3Al to benzaldehyde in the presence of ligand 1b
Entry
Aldehyde
Temp [ꢁC]
Time [h]
Product
Yield [%]
92
Ee [%]
Cl
Cl OH
CHO
1
20
4
85
79
5
(S)-6
(S)-8
OH
Cl
Cl
CHO
Cl
Cl
2
3
4
20
21
21
6
4
4
95
85
85
7
CHO
9
OH
89
89
(S)-10
(S)-12
OCH3
CHO
H3CO
OH
11
OH
OH
H3CO
CHO
H3CO
5
6
7
8
9
20
20
21
21
21
4
6
4
4
4
95
90
80
70
82
92
66
63
77
83
13
(S)-14
CHO
H3CO
H3CO
15
(S)-16
OH
CHO
17
(S)-18
OH
CHO
19
(S)-20
OH
CHO
21
(S)-22

854
T. Bauer, J. Gajewiak / Tetrahedron: Asymmetry 16 (2005) 851–855
vs 1–3), while the enantiomeric excess was higher than
for other ligands.
hydrogen of the coordinated aldehyde. Such bonding
was already postulated by Corey and Lee,7 and was
proved to be a very important factor influencing asym-
metric induction in enantioselective reactions, also in
the reactions catalyzed by a-hydroxy acids.1 Its presence
substantially rigidifies transition states leading to good
enantiomeric excesses. The proposed model also ex-
plains the influence of the bulky substituents on tita-
nium tetraalkoxide on the enantioselectivity. The
increased steric clash of alkyl groups connected to the
oxygen bridge results in a weaker hydrogen bond and,
probably a seriously reorganized transition state.
The enantioselective addition of triethylaluminum to
some representative aromatic and aliphatic aldehydes
was performed in the presence of 0.4 equiv of ligand
1b. The resulting ee was highly dependent on the struc-
ture of the aldehyde. It is worth stressing that reason-
ably good enantioselectivities were obtained for
aliphatic and cycloaliphatic aldehydes (Table 5).
2.2. Enantioselective trimethylaluminum addition
Finally, we performed the addition of another organo-
aluminum reagent—trimethylaluminum. We chose
ligands 1b and 1d, which had given high ee for
trimethylaluminum additions to benzaldehyde. The
reactions proceeded with very good yield, but the
enantioselectivity was lower than that obtained for
Et3Al (Table 6).
3. Conclusions
We have shown that the addition reactions to the car-
bonyl group of aldehydes in the presence of a-hydroxy
carboxylic acids proceed with very good yields and good
enantiomeric excesses up to 92%. a-Hydroxy acids can
catalyze the addition of trialkylaluminum to both ali-
phatic and aromatic aldehydes, which are relatively
inexpensive and readily available. Further applications
of a-hydroxy acids as ligands for enantioselective orga-
nometallic additions are under investigation and will be
reported in due course.
2.3. Mechanistic considerations
The stereochemical outcome of the reaction performed
in the presence of a-hydroxy carboxylic acids can be
rationalized as follows: The reaction is highly dependent
on the order of addition of titanium and aluminum re-
agents. This suggests two different transition states,
and also, that transition state in the case of organoalu-
minum reagents differs from that observed for organo-
zinc compounds. We suppose that in this case reaction
proceeds through a bimetallic transition state similar
to that already proposed by Seebach et al. (Fig. 1).6
4. Experimental
4.1. General
Melting points were determined using a Kofler hot stage
apparatus and are uncorrected. Specific rotations were
recorded using a Perkin–Elmer PE-241 polarimeter with
a thermally jacketed 10-cm cell. Reactions were carried
under argon using Schlenk technique when necessary.
Flash column chromatography was made on silica gel
(Kieselgel-60, Merck, 230–400 mesh). High performance
liquid chromatography was conducted on Merck
Hitachi D-7000 with diode array detector L-7455 using
chiral column Diacel Chiracel OD. Gas chromatogra-
phy was conducted on Hewlett Packard 5890 series II
with FID detector using chiral column b-Dex,
30 m · 0.25 mm I.D. (Supelco, Bellefonte, USA). Reten-
tion time is given in minutes. a-Hydroxy carboxylic
acids were obtained according to published procedures.
The a-hydroxy carboxylic acid makes an aggregate with
two titanium compounds, one of them being a coordina-
tion center for ligand and aldehyde, the second carries
an ethyl group transferred from Et3Al. The key factor
for good asymmetric induction is the presence of hydro-
gen bonding between the ligandÕs a-oxygen and the
Table 6. Addition reactions of Me3Al to benzaldehyde in the presence
of ligands 1b and 1d
Ligand Amount Temp [ꢁC] Time [h] Yield [%] Ee [%]
[equiv]
1b
1d
0.4
0.4
22
22
2.5
2.5
95
95
80 (S)
78 (S)
4.2. General procedure for the addition of triethyl-
and trimethylaluminum to aldehyde in the presence
(S)-mandelic acid
In the oven dried Schlenk tube filled with argon and
equipped with the stirring bar was placed (S)-mandelic
acid (30.2 mg, 0.2 mmol), followed by THF (2 mL),
Ti(OiPr)4 (0.42 mL, 1,4 mmol). After 1 h the solution
was cooled to 0 ꢁC and triethylaluminum (1.9 M in tolu-
ene, 0.8 mL, 1,5 mmol) was added. The stirring was
continued at this temperature for 30 min, aldehyde
(1 mmol) was added and after additional 30 min of stir-
ring at 0 ꢁC the reaction mixture was allowed to warm
up to room temperature. The progress of the reaction
was monitored with TLC using CH2Cl2/MeOH 70:2.
si
H
Et
O
O
O
O
O
O
Ti
Ti
O
O
O
Figure 1.

T. Bauer, J. Gajewiak / Tetrahedron: Asymmetry 16 (2005) 851–855
855
20
D
(c 1.73, CHCl3), ee = 59%.
20
D
After completion the reaction was quenched with slow
addition of 1 M HCl (10 mL) (CAUTION! Exothermic
reaction). The precipitate was filtered off using funnel
with cotton plug and the filtrate was extracted three
times with ethyl acetate (3 · 20 mL), combined extracts
were washed with brine (30 mL), dried over anhydrous
MgSO4, and evaporated in vacuo. The resulting oil
was purified by flash chromatography (CH2Cl2/MeOH
200:3 as eluent). All compounds have been previously
reported and were characterized by comparison with
their reported physical and spectroscopic data.
½aꢁ ¼ ꢀ4:9 (c 1, CHCl3 ), ee = 63%. lit.10 ½aꢁ ¼ ꢀ6:3
4.2.9. (S)-3-Octanol (S)-25. HPLC (as benzoate):
tS = 21.1, tR = 23.1 (hexane, flow 0.5 mL/min), ½aꢁ
20
D
¼
20
þ7:6 (c 1.03, CHCl3); ee = 77%, lit.8 ½aꢁ ¼ þ5:87 (c 1,
D
CHCl3), ee = 60%.
4.2.10. (S)-1-Cyclohexyl-1-propanol (S)-26. HPLC (as
benzoate): tS = 37.8, tR = 35.9 (hexane/iPrOH 99.9:0.1,
20
D
flow 0.2 mL/min); ½aꢁ ¼ ꢀ5:4 (c 1.14, CHCl3),
24
ee = 82.5%, lit.8 ½aꢁ ¼ ꢀ6:39 (c 1.05, CHCl3), ee = 97%.
D
4.2.1. (S)-1-Phenyl-1-propanol (S)-3. HPLC: tS = 12.1,
tR = 13.2 (hexane/iPrOH 97:3, flow 1 mL/min);
4.2.11. (S)-1-Phenyl-1-ethanol. HPLC: tS = 14.8, tR =
12.8
20
D
20
D
½aꢁ ¼ ꢀ42:0 (c 1.0, CHCl3); ee = 90.4%, lit.8 ½aꢁ
¼
(hexane/iPrOH
97:3,
flow
1 mL/min);
¼
20
D
20
D
ꢀ44:4 (c 1.01, CHCl3), ee = 95.5%.
½aꢁ ¼ ꢀ43:9 (c 1.38, CHCl3); ee = 79.5%, lit.11 ½aꢁ
ꢀ66:5 (c 0.6, CHCl3 ) ee = 98%.
4.2.2. (S)-1-(2-Chlorophenyl)-1-propanol (S)-18. GC:
tR = 51.0, tS = 54.2 (Tcolumn = 130 ꢁC, P = 100 kPa),
Acknowledgements
20
½aꢁ ¼ ꢀ48:6 (c 1.25, benzene), ee = 85%.
D
Financial support from the State Committee for Scien-
tific Research (Grant 3 T09A 039 16) is gratefully
acknowledged.
4.2.3.
(S)-1-(3-Chlorophenyl)-1-propanol
(S)-19.
HPLC: tS = 17.1, tR = 18.6 (hexane/iPrOH 96:4,
20
D
flow 0.5 mL/min); ½aꢁ ¼ ꢀ23:3 (c 1.21, benzene), ee =
20
78.8%, lit.9 ½aꢁ ¼ þ26:6 (c 2.36, benzene) for the (R)-
D
enantiomer, ee = 97%.
References
1. (a) Bauer, T.; Tarasiuk, J. Tetrahedron Lett. 2002, 43, 687–
689; (b) Bauer, T.; Gajewiak, J. Tetrahedron 2004, 60,
9163–9170.
2. Chan, A.; Zhang, F.; Yip, C. J. Am. Chem. Soc. 1997, 119,
4080–4081.
4.2.4.
(S)-1-(4-Chlorophenyl)-1-propanol
(S)-20.
HPLC: tS = 22.4, tR = 24.3 (hexane/iPrOH 97:3,
20
D
flow 0.5 mL/min); ½aꢁ ¼ ꢀ23:6 (c 1.77, benzene), ee =
20
88.5%, lit.8 ½aꢁ ¼ ꢀ23:6 (c 1.73, benzene), ee = 93%.
D
3. Lu, J.; You, J.; Gau, H. Tetrahedron: Asymmetry 2000, 11,
2531–2535.
4.2.5.
(S)-1-(2-Metoxyphenyl)-1-propanol
(S)-21.
4. You, J.; Hsieh, S.; Gau, H. Chem. Commun. 2001, 1546–
1547.
HPLC: tS = 27.5, tR = 29.6 (hexane/iPrOH 97:3,
20
D
5. For positive effect of Ti(OtBu)4 on enantioselectivity see:
Novotny, S.; Vettel, S.; Knochel, P. Tetrahedron Lett.
1994, 35, 4539–4540.
flow 0.5 mL/min); ½aꢁ ¼ ꢀ50:8 (c 1.22, toluene), ee =
22
89.1%, lit.8 ½aꢁ ¼ ꢀ52:9 (c 1.02, toluene), ee = 91%.
D
4.2.6.
(S)-1-(3-Metoxyphenyl)-1-propanol
(S)-22.
6. Seebach, D.; Plattner, D.; Beck, A.; Wang, Y.; Hunziker,
D.; Petter, W. Helv. Chim. Acta 1994, 77, 2111–
2116.
7. Corey, E. J.; Lee, T. Chem. Commun. 2001, 1321–
1329.
HPLC: tS = 80.3, tR = 76.6 (hexane/iPrOH 97:3,
20
D
flow 0.3 mL/min); ½aꢁ ¼ ꢀ26:8 (c 0.8, toluene),
ee = 92%.
8. Asami, M.; Watanabe, H.; Inoue, S. Tetrahedron: Asym-
metry 1998, 9, 4165–4173.
9. Dai, W.; Zhu, H.; Hao, X. Tetrahedron: Asymmetry 2000,
11, 2315–2337.
10. Hwang, C.-D.; Uang, B.-J. Tetrahedron: Asymmetry 1998,
9, 3979–3984.
4.2.7.
(S)-1-(4-Metoxyphenyl)-1-propanol
(S)-23.
HPLC: tS = 38.1, tR = 33.6 (hexane/iPrOH 97:3, flow
20
D
0.5 mL/min); ½aꢁ ¼ ꢀ28:7 (c 1, benzene), ee =
20
83%, lit.8 ½aꢁ ¼ ꢀ25:8 (c 1.1, benzene), ee = 74%.
D
4.2.8. (S)-1-Phenyl-1-penten-3-ol (S)-24. HPLC: tS =
13.0, tR = 9.0 (hexane/iPrOH 90:10, flow 1.0 mL/min);
11. Kamal, A.; Sandbhor, M.; Ramana, V. Tetrahedron:
Asymmetry 2002, 13, 815–820.