Enantioselective synthesis of optically active homoallylamines by allylboration of N-diisobutylaluminum imines

Article abstract of DOI:10.1016/S0022-328X(99)00048-0

Diisobutylaluminum hydride (DIBAL-H) reduces nitriles to give N-diisobutylaluminum imines, which were asymmetrically allylated with chirally modified allylboron reagents. The corresponding chiral primary homoallylamines were obtained with up to 87% ee.

Full text of DOI:10.1016/S0022-328X(99)00048-0

Journal of Organometallic Chemistry 581 (1999) 103–107
Enantioselective synthesis of optically active homoallylamines by
allylboration of N-diisobutylaluminum imines
Katsuhiro Watanabe, Shizue Kuroda, Ayako Yokoi, Koichi Ito, Shinichi Itsuno *
Department of Materials Science, Toyohashi Uni6ersity of Technology, 441-8580 Toyohashi, Japan
Received 6 May 1998; received in revised form 17 July 1998
Abstract
Diisobutylaluminum hydride (DIBAL-H) reduces nitriles to give N-diisobutylaluminum imines, which were asymmetrically
allylated with chirally modified allylboron reagents. The corresponding chiral primary homoallylamines were obtained with up to
87% ee. © 1999 Elsevier Science S.A. All rights reserved.
Keywords: Enantioselective synthesis; Optically active homoallylamine; N-Metalloimine; N-Diisobutylaluminum imine; Chirally modified allyl-
boron reagents
loimines such as N-boryl- [9] and N-aluminum imines
[10] could react with organolithium or Grignard
reagents to afford primary amines. In the presence of
chiral ligand, optically active primary amines were ob-
tained [9c, 10]. These N-metalloimines are easily pre-
pared from partial reduction of nitrile with aluminum-
or borohydride. Further study on the reaction of N-
metalloimines revealed that treatment of N-diisobutyla-
luminum imines 2 and N-trimethylsilyl imines with
triallylborane gave the desired homoallylamines in good
yield. These results encouraged us to apply to the
asymmetric version of allylboration reactions. We have
already communicated the enantioselective allylbora-
tion of the N-trimethylsilyl imines by chirally modified
allylboron reagents to afford primary homoallylamine
[11]. In this paper we wish to describe the first applica-
tion of N-aluminum imines to enantioselective allylbo-
ration. Since this appeared to be a potentially useful
method for preparing certain chiral homoallylamines,
we examined the scope and limitations of this proce-
dure. Reactivity and stereoselectivity in asymmetric al-
lylboration of N-aluminum imines are also discussed.
1. Introduction
The enantioselective synthesis of optically active
amines by nucleophilic addition of organometallic
reagents toward imines is a topic of current interest [1].
Asymmetric addition of allylmetal species to imines is
one of the useful approaches to synthesis of optically
active homoallylamines which are indeed important
compounds as starting or intermediate materials in the
synthesis of biologically active substances [2]. They are
also useful as resolving agents and chiral auxiliaries for
asymmetric reactions [3]. However, enantioselective
methods for homoallylamine synthesis are scarcely de-
veloped. Only an example of reaction of chiral allylzinc
reagents with aldimines using chiral bis(oxazoline) pro-
moter was described [4]. The major limitation of this
methodology is that only cyclic aldimines can be allyl-
ated with high enantioselectivity.
On the other hand, the ability of N-masked imine
derivatives of ammonia including N-metallo- [5], N-
thio- [6], and N-phosphinylimines [7] to undergo vari-
ous nucleophilic addition reactions is now well
established and has been utilized in the synthesis of
various kinds of primary amines and their derivatives
[8]. In previous papers, we have reported that N-metal-
2. Results and discussion
We have first investigated the allylboration of N-bo-
* Corresponding author. Tel.: +81-532-446813; fax: +81-532-
446813.
rylimines prepared from nitrile reduction, since we have
0022-328X/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(99)00048-0

104
K. Watanabe et al. / Journal of Organometallic Chemistry 581 (1999) 103–107
shows that the N-aluminum imines could yield the
corresponding homoallylamines except for pyridyl and
benzyl derivatives (Runs 4, 5, and 10). In the case of 2d
and 2e, under the conditions used for the allylboration,
both imino functionality and pyridine ring seem to be
allylated with triallylborane to give a complex mixture
[16]. The DIBAL-H reduction of phenylacetonitrile 1j
involved undesired side reactions. For other nitriles this
reaction sequence would be efficient for the synthesis of
primary homoallylamines, since the N–Al bond could
be cleaved readily during the aqueous work-up proce-
dure. Another advantage in the use of such imines is
that N-aluminum imines derived from DIBAL-H re-
duction of aliphatic nitriles could be also allylated with
triallylborane to yield the homoallylic amines (Runs
6–9). In contrast to N-aluminum imines, the use of
N-trimethylsilyl imines is generally restricted to
nonenolizable imines since the enolizable ones give
poor yields of the expected amines in the reactions with
organometallic reagents such as Grignard reagents or
alkyllithiums [17]. A lower isolated yield for 2f derived
from propionitrile may be caused by its high volatility.
Bulkiness of naphthyl group in the imine 2c and its side
reaction resulted in low yield of the amine 4c. However,
the relatively higher reactivity of the other imines to-
ward triallylborane prompted us to investigate the
enantioselective synthesis of homoallylamines by nucle-
ophilic addition of chirally modified allylboron
reagents.
Various chirally modified allylboron reagents have
been proposed for the asymmetric allylboration of alde-
hydes in the literature [18]. Some of them are highly
effective for stereoselective allylboration of aldehydes.
For the enantioselective allylboration of N-aluminum
imines, we thus chose Brown’s reagents 5 [19] and 6 [20]
and Roush’s reagent 7 [20,21], which are a family of
readily accessible and synthetically convenient allyl-
boron reagents that exhibit excellent selectivities in the
allylboration of aldehydes. Another promising chiral
allylboron reagent is 8 exploited by ourselves, which
showed excellent enantioselectivity in the allylboration
of N-silyl imines [11]. Structures of the chiral allylboron
reagents used in this study are illustrated in Scheme 2.
Table 2 shows the enantioselective allylboration of N-
(diisobutylaluminum)benzaldehyde imine 2a with chi-
rally modified allylborating agents 5–8 at various
temperature. Although these chiral allylboron reagents
have quite bulky chiral substituents, the N-aluminum
imine was allylated with this reagent even at low tem-
perature to give homoallylic amine 4a. All chiral allylb-
oration reagents used successfully reacted with 2a to
afford the corresponding primary homoallylic amine
4a. The isolated yield of the amine was sometimes
moderate, since some of the amine product was lost
during isolation process mainly owing to the difficulty
of separation of the aluminum by-product.
Scheme 1.
proposed that the reaction of N-borylimines with alkyl-
lithiums at low temperature is one of the useful meth-
ods to obtain primary amines [9]. However,
unfortunately, the N-borylimine derived from benzoni-
trile and borane or diethylborane as reducing agent
reacted with triallylborane to yield only a complexed
mixture of products.
Next, we focused on the allylboration of N-alu-
minum imines 2, which are readily available by partial
reduction of nitriles with diisobutylaluminum hydride
(DIBAL-H) (Scheme 1). The imines 2 have been suc-
cessfully utilized for various organic transformations
[12,13]. For example, Panunzio and co-workers re-
ported the reaction of organometallic reagents includ-
ing allyl Grignard reagents with 2 to give primary
amines [14,15]. We thus examined the reaction sequence
of DIBAL-H reduction of nitrile followed by allylation
with triallylborane as shown in Scheme 1. We eventu-
ally found that 2a reacted with triallylborane at room
temperature to afford the desired primary homoallylic
amine 4a in 70% yield, although the imine possesses a
bulky N-substituent. Other various N-aluminum imines
derived from nitriles 1 were subjected to reactions with
triallylborane in THF at room temperature. Table 1
Table 1
Allylboration of various N-aluminum imines 2 with triallylboranea
Run number
N-Aluminum
Time (h)
Yield of homo-
allylamines (%)b
imines
1
2
3
4
5
6
7
8
9
2a
2b
2c
2d
2e
2f
2g
2h
2i
5
70
66
11
24
20
20
20
20
5
5
5
5
c
–
c
–
18
54
54
54
d
10
2j
–
a Reaction conditions: 2, 5 mmol (prepared from nitrile, 5 mmol;
DIBAL-H, 5 mmol; 0°C, 1 h); triallylborane, 6 mmol; THF 15 ml.
b Isolated yields based on the starting nitrile.
c Complexed mixture including ring allylated products were ob-
tained.
d Complexed mixture.

K. Watanabe et al. / Journal of Organometallic Chemistry 581 (1999) 103–107
105
the decrease of the stereoselectivity. At a given temper-
ature the use of different reagents (varying chiral lig-
and) resulted in different selectivity. Although Brown et
al. reported that the allylboration of aldehydes by
chiral reagent 6 exhibited quite high enantioselectivity
(up to \99% ee) [19], the same reagent gave an
enantioselectivity of 52% in the allylboration of the
N-aluminum imine (Run 6). Roush’s reagent 7 led to
an R configuration of the product, while other reagents
preferred S amine as a major enantiomer. In our previ-
ous communication, we found that chiral B-allyloxaz-
aborolidine prepared from (−)-norephedrine exhibited
excellent selectivity in the allylboration of N-
trimethylsilyl imines [11a] and aldehydes [22]. However,
the same reaction on the N-aluminum imine afforded
only poor enantioselectivities (Runs 13–16, 15–33%
ee).
Encouraged by the good enantioselectivities achieved
with 5 at 25°C (Table 2), we extended the study to the
allylborations of representative N-diisobutylaluminum
imines derived from other nitriles. Table 3 shows the
results of enantioselective allylboration of N-aluminum
imines 2 with the chiral reagent 5. In contrast with
benzaldimine 2a, higher enantioselectivities were ob-
tained at lower temperatures in the allylboration of
these imines. N-Aluminum imine 2b derived from p-tol-
unitrile gave the highest enantioselectivities (up to 87%
ee) in the imines used for the enantioselective allylbora-
tion. In all cases, reaction proceeded in a homogeneous
system and significant change of enantioselectivity was
not observed until the reaction temperature was raised
to room temperature. It should be also emphasized that
Scheme 2.
In a number of asymmetric reactions, the level of
enantioselectivity increased by decreasing the reaction
temperature. Table 2 shows an unusual temperature
dependence of the asymmetric allylboration of 2a on
the enantioselectivity. We have observed that the selec-
tivity increased with increasing the temperature until an
optimal range was reached (0–20°C) where the selectiv-
ity then began to decrease slowly. When 5 was used as
the allylborating agent, no significant decrease of ee
was observed even at the reflux temperature of THF.
Whereas the THF solution of the reaction mixture is
homogeneous above 0°C during allylboration reactions,
a heterogeneous system was observed below −30°C.
Somewhat different reactive species might participate in
the reaction at lower temperatures, which would cause
Table 2
Enantioselective synthesis of homoallylamine by nucleophilic addition of chirally modified allylboron reagent to N-diisobutylaluminum imine 2a
Run number
Chirally modified allylboron reagent
Temperature (°C)
1-Phenyl-3-butenamine (4a)
Yield%a
Ee%b
Configurationc
1
2
3
4
5
6
7
8
5
5
5
5
6
6
6
6
7
7
7
7
8
8
8
8
–78
–45
25
66
–78
0
25
66
–78
0
25
66
–78
0
25
66
59
61
67
42
78
61
65
71
5
37
45
59
19
35
30
64
33
47
69
66
20
52
47
28
8
59
62
32
15
33
23
24
S
S
S
S
S
S
S
S
R
R
R
R
S
S
S
S
9
10
11
12
13
14
15
16
a Isolated yield.
b Determined by HPLC analysis with a chiral stationary phase column. See Experimental.
c The absolute configuration of the product was determined by the comparison of the specific rotation with the reported value 26.

106
K. Watanabe et al. / Journal of Organometallic Chemistry 581 (1999) 103–107
Table 3
were purchased from Aldrich, Inc. Reactions were
monitored by thin layer chromatography (TLC) using
Merck precoated silica gel plates (Merck 5554, 60F254).
Flash column chromatography was performed over
Wako silica gel (Wakogel C-200, 100–200 mesh).
1H-NMR spectra were measured on a JEOL JNM-
GX270 spectrometer using Me4Si as an internal stan-
dard. Infrared spectra (IR) were recorded with a JEOL
JIR-7000 FT-IR spectrometer and are reported in re-
ciprocal centimeters (cm−1). Optical purity was deter-
mined by a Shimadzu Capillary Gas Chromatograph
14A with a chiral capillary column (Astec Chiraldex
G-TA, 30 m×0.25 mm). HPLC analyses were per-
formed with a TOSOH HLC-8020 equipped with a
chiral column (Chiralcel OD-H, Daicel) using hexane:2-
Enantioselective synthesis of homoallylamine 4 by nucleophilic addi-
tion of 5 to N-diisobutylaluminum imine 2
Run number Nitrile
Temperature
(°C)
Yield of 4%a Ee%b
1
2
3
4
5
6
7
8
1g
1g
1g
1g
1g
1h
1h
1h
1h
1h
1i
1i
1i
1i
1i
1b
1b
1b
1b
1b
–78
–30
0
60
54
63
47
39
32
22
21
41
37
32
49
53
53
51
46
73
70
66
27
73
71
67
63
45
79
85
66
60
36
80
67
78
70
43
87
83
82
78
53
25
Reflux
–78
–30
0
9
25
10
11
12
13
14
15
16
17
18
19
20
Reflux
–78
–30
0
propanol:diethylamine (90:10:0.1).
A UV detector
(TOSOH UV-8011) was used for the peak detection.
Optical rotations were taken on a JASCO DIP-140
digital polarimeter using 10 cm thermostated microcell.
25
Reflux
–78
–30
0
25
Reflux
3.2. Preparation of
N-(diisobutylaluminum)benzaldehyde imine (2a)
a Isolated yields based on the starting nitrile.
b Determined by GC analysis with a trifluoroacetylated g-cyclodex-
trin column (30 m×0.25 mm) (Astec, Chiraldex G-TA) for the
products 4g, 4h, and 4i and HPLC analysis using Daicel, Chiralcel
OD-H (250×4.6 mm) with hexane:2-propanol:diethylamine
(90:10:0.1) for the product 4b.
The preparation of 2a was based on that of Andreoli
et al. [13]. To a solution of benzonitrile (0.5 ml, 5
mmol) in THF (5 ml) was added dropwise a solution of
DIBAL-H (1.01 M, 5 mmol) at 0°C. The reaction
mixture was stirred at room temperature (r.t.) for 1 h.
Evaporation of the solvent in vacuo gave 2a as viscous
1
oil in quantitative yield. H-NMR (270 MHz, CDCl3) l
aliphatic imines derived from 1g, 1h, and 1i were asym-
metrically allylated to the corresponding homoally-
lamines in good to high enantioselectivities.
8.95 (s, 1H), 7.74–7.46 (m, 5H), 1.84 (m, 2H), 0.96–
0.74 (m, 12H); IR 1635 (CꢀN). For the following
allylboration reaction, a THF solution of 2a was used
without purification.
In conclusion, we have shown that optically active
homoallylamines, which are interesting intermediates in
the synthesis of biologically active natural products, are
readily prepared from enantioselective allylboration of
N-aluminum imines with chirally modified allylboron
reagents. This is a convenient and direct procedure for
the synthesis of such amines. A further advantage of
the use of N-aluminum imines is that primary homoal-
lylamines were easily obtained after usual aqueous
work-up.
3.3. Preparation of chirally modified allylboron
reagents
3.3.1. Chirally modified allylborane (5)
Chirally modified allylboranes were prepared from
the corresponding B-chlorodialkylboranes, such as
(−)-DIP-Cl or (−)-B-chlorodiisocaranylborane, with
allylmagnesium chloride in THF, according to Brown’s
procedure [23,24]. To a solution of (−)-DIP-Cl (1.92 g,
6 mmol) in THF (10 ml) was added dropwise allylmag-
nesium chloride in THF (1.1M, 6 mmol) at 0°C under
a nitrogen atmosphere. After the reaction mixture was
stirred for 1 h at r.t., the resulting solution of 5 was
used for the enantioselective allylboration reaction.
3. Experimental
3.1. General
All experiments were carried out under an atmo-
sphere of dry nitrogen. Tetrahydrofuran (THF) was
dried over sodium-benzophenone ketyl, and was freshly
distilled just before use. All nitriles were distilled from
calcium hydride before use. DIBAL-H (diisobutyl-
aluminum hydride, 1.01 M solution in hexanes) and
(−)-B-chlorodiisopinocampheylborane [(−)-DIP-Cl™]
3.3.2. Chirally modified allylboronate (7)
The preparation of chirally modified allylboronates
was based on Roush’s method [25]. To a solution of
(+)-diisopropyl tartrate (1.41 g, 6 mmol) in THF (10
ml) was added dropwise a THF solution of triallylbo-
rane (6 mmol) at r.t. The reaction mixture was stirred

K. Watanabe et al. / Journal of Organometallic Chemistry 581 (1999) 103–107
107
for 2 h and then heated under reflux for another 1 h.
References
The solution was cooled, and all volatile components
were removed in vacuo. The resulting allylboronate 7
was dissolved in THF (10 ml) and then used for the
enantioselective allylboration reaction.
[1] (a) D. Enders, U. Reinhold, Tetrahedron: Asymmetry 8 (1997)
1895. (b) S.E. Denmark, O.J.-C. Nicaise, Chem. Commun.
(1996) 999.
[2] H. Moser, G. Rihs, H. Santer, Z. Naturforsch. 376 (1982) 451.
[3] J.K. Whitesell, Chem. Rev. 89 (1989) 1581.
[4] M. Nakamura, A. Hirai, E. Nakamura, J. Am. Chem. Soc. 118
(1996) 8489.
3.3.3. Chirally modified B-allyloxazaborolidine (8)
According to the procedure described before [11b], 8
was prepared from the N-sulfonylamino alcohol,
derived from (−)-norephedorine with triallylborane in
THF.
[5] G. Cainelli, M. Panunzio, P. Andreoli, G. Martelli, G. Spunta,
D. Giacomini, E. Bandini, Pure & Appl. Chem. 62 (1990) 605.
[6] (a) D.H. Hua, S.W. Miao, J.S. Chen, S. Iguchi, J. Org. Chem. 56
(1991) 4. (b) R.G. Lovey, A.B. Cooper, Synlett (1994) 167. (c)
T.-K. Yang, R.-Y. Chen, D.-S. Lee, W.-S. Peng, Y.-Z. Jiang,
A.-Q. Mi, T.-T. Jong, J. Org. Chem. 59 (1994) 914.
[7] (a) T. Suzuki, Y. Hirokawa, K. Ohtake, T. Shibata, K. Soai,
Tetrahedron: Asymmetry 8 (1997) 4033. (b) A.A. Cantrill, A.N.
Jarvis, H.M.I. Osborn, A. Ouadi, J.B. Sweeney, Synlett. (1996)
847.
3.4. Enantioselecti6e allylboration of
N-diisobutylaluminum imine 2a with chirally modified
allylboron reagent 5
[8] (a) D.J. Hart, K. Kanai, D.G. Thomas, T.-K. Yang, J. Org.
Chem. 48 (1983) 289. (b) S. Itsuno, H. Yanaka, C. Hachisuka,
K. Ito, J. Chem. Soc., Perkin Trans. 1 (1991) 1341.
[9] (a) S. Itsuno, C. Hachisuka, K. Ito, J. Chem. Soc., Perkin Trans.
1 (1991) 1767. (b) S. Itsuno, C. Hachisuka, K. Kitano, K. Ito,
Tetrahedron Lett. (1992) 627. (c) S. Istuno, C. Hachisuka, Y.
Ushijima, K. Ito, Synthetic Commun. 22 (1992) 3229.
[10] S. Itsuno, M. Sasaki, S. Kuroda, K. Ito, Tetrahedron: Asymme-
try 6 (1995) 1531.
[11] (a) S. Itsuno, K. Watanabe, K. Ito, A.A. El-Shehawy, A.A.
Sarhan, Angew. Chem. Int. Ed. Engl. 36 (1997) 109. (b) K.
Watanabe, K. Ito, S. Itsuno, Tetrahedron: Asymmetry 6 (1995)
1531.
[12] P. Andreoli, G. Cainelli, M. Contento, D. Giacomini, G.
Martelli, M. Panunzio, J. Chem. Soc. Perkin Trans. 1 (1988) 945.
[13] J.R. Zietz Jr., G.C. Robinson, K.L. Lindsay, Compr.
Organomet. Chem. 7 (1982) 431.
[14] P. Andreoli, L. Billi, G. Cainelli, M. Panunzio, G. Martelli, G.
Spunta, J. Org. Chem. 55 (1990) 4199.
[15] G. Cainelli, E. Mezzina, M. Panunzio, Tetrahedron Lett. 31
(1990) 3481.
[16] Y.N. Bubnov, in G.W. Kabalka (Ed.), Current Topics in the
Chemistry of Boron, Royal Society of Chemistry, Cambridge,
1994, p. 44.
[17] D.J. Hart, K. Kanai, D.G. Thomas, T. -K. Yang, J. Org. Chem.
48 (1983) 289.
[18] H.C. Brown, P.V. Ramachandran, Pure & Appl. Chem. 63
(1991) 307 and references cited therein.
To a solution of 5 (6 mmol) prepared from (−)-DIP-
Cl (1.92g, 6 mmol) and allylmagnesium chloride (6
mmol), a THF solution of 2a prepared from benzoni-
trile (0.5 ml, 5 mmol) and DIBAL-H (4.95 ml, 5 mmol)
in THF (5 ml) was added. The reaction mixture was
then stirred for 5 h at r.t., and quenched with 2M HCl.
The aqueous layer was separated, washed with ether,
neutralized with NH4OH and extracted with ether. The
combined extracts were dried over MgSO4 and concen-
trated by rotary evaporator to yield a colorless oil
which is essentially pure 1-phenyl-3-butenamine (0.49 g,
67%). 1H-NMR (270 MHz, CDCl3) l 7.34–7.22 (m,
5H), 5.81–5.70 (m, 1H), 5.15–5.06 (m, 2H), 3.98 (dd,
J=7.81 and 5.37 Hz, 1H), 2.46–2.32 (m, 2H), 1.53 (br,
2H); IR 3300 (NH2), 1630 (CꢀC), 1300 (C–N). The
product was purified by flash-column chromatography
on silica gel (ether). The enantioselectivity 69% ee was
determined by HPLC analysis using a chiral stationary
phase column (Daicel, Chiralcel OD-H; hexane:2-
propanol:diethylamine = 90:10:0.1, flow rate; 0.5 ml
min−1): tR 16.3 min (R), tR 20.7 min (S). The absolute
configuration of the product was correlated to that
described in the literature [26].
[19] U.S. Racherla, H.C. Brown, J. Org. Chem. 56 (1991) 401.
[20] W.R. Roush, L.K. Hoong, M.A.J. Palmer, J.C. Park, J. Org.
Chem. 55 (1990) 4109.
[21] W.R. Roush, L.K. Hoong, M.A.J. Palmer, J.A. Straub, A.D.
Palkowitz, J. Org. Chem. 55 (1990) 4117.
Acknowledgements
[22] S. Itsuno, A.A. El-Shehawy, A.A. Sarhan, Reactive & Func-
tional Polym. 37 (1998) 283.
[23] H.C. Brown, P.K. Jadhav, J. Am. Chem. Soc. 105 (1983) 2092.
[24] U.S. Racherla, H.C. Brown, J. Org. Chem. 56 (1991) 401.
[25] W.R. Roush, A.E. Walts, L.K. Hoong, J. Am. Chem. Soc. 107
(1985) 8186.
This work was partially supported by a Grant-in-Aid
for Scientific Research from the Ministry of Education,
Science, Sports, and Culture, Japan and The Mitsubishi
Petrochemical (Mitsubishi Chemical) Foundation. The
Yamanouchi Award in Synthetic Organic Chemistry,
Japan (to S.I.) is also gratefully acknowledged.
[26] T. Basile, A. Bocoum, D. Savoia, A. Umani-Ronchi, J. Org.
Chem. 59 (1994) 7766.
.