A highly stereoselective synthesis of optically active trisubstituted 1,2-ethylenediamines: The first example of Grignard addition to N-diphenylphosphinoyl ketimines derived from amino acids

Article abstract of DOI:10.1021/jo049405i

The efficient synthesis of optically active trisubstituted 1,2-ethylenediamines is described. Addition of aryl and/or alkyl Grignard reagents to α-amino N- diphenylphosphinoyl ketimines derived from α-amino acids was demonstrated to afford the desired tri

Full text of DOI:10.1021/jo049405i

A High ly Ster eoselective Syn th esis of Op tica lly Active
Tr isu bstitu ted 1,2-Eth ylen ed ia m in es: Th e F ir st Exa m p le of
Gr ign a r d Ad d ition to N-Dip h en ylp h osp h in oyl Ketim in es Der ived
fr om Am in o Acid s
Yoshinori Kohmura* and Toshiaki Mase
Process Research, Process R & D, Laboratories for Technology Development, Banyu Pharmaceutical Co.
Ltd., 9-1, Kamimutsuna 3-Chome, Okazaki, Aichi 444-0858, J apan
yoshinori_kohmura@merck.com
Received April 9, 2004
The efficient synthesis of optically active trisubstituted 1,2-ethylenediamines is described. Addition
of aryl and/or alkyl Grignard reagents to R-amino N-diphenylphosphinoyl ketimines derived from
R-amino acids was demonstrated to afford the desired trisubstituted 1,2-ethylenediamines in good
yields and with high diastereoselectivities. Subsequent removal of the diphenyphosphinoyl group
from the adduct was smoothly accomplished in reasonable yield without racemization under newly
developed reductive conditions.
In tr od u ction
of chiral R,R-dibranched amines containing a quaternary
carbon.6,7 Ellman and others have reported the diaste-
reoselective addition of organometallics to a ketimine
with chiral auxiliaries (i.e. tert-butanesulfinylimine).6 In
general this method provides high diastereoselectivity
and allows for the facile removal of the sulfinyl group
from the resulting sulfinamides.6c We expected that
diastereoselective nucleophilic addition to a ketimine
would be feasible without chiral auxiliaries on nitrogen
if the ketimine has a chiral center at the R-position of
the imine. We envisioned that chiral R-aminoketimines,
which are readily available from R-amino acids, can
undergo diastereoselective nucleophile addition to readily
provide the desired chiral trisubstituted 1,2-ethylenedi-
amines.
Optically active multisubstituted 1,2-ethylenediamines
are recognized as superb ligands/auxiliaries in asym-
metric synthesis.1 In addition, their derivatives are
valuable key intermediates in drug discovery efforts and
heavily used for the preparation of pharmacophoric
heterocycle substructures.2 Therefore, much effort has
been directed toward an efficient synthesis of these
compounds. Application of optically active monosubsti-
tuted, vicinal disubstituted, and trisubstituted 1,2-eth-
ylenediamines for asymmetric synthesis has been well
documented.1 Among them, trisubstituted 1,2-ethylene-
diamines offer an advantage. For example, it is reported
that (2S)-1,1-bis(4-methoxyphenyl)-3-methyl-1,2-butane-
diamine acts as a more efficient chiral ligand for a
ruthenium-catalyzed enantioselective hydrogenation than
(R,R)-trans-cyclohexane-1,2-diamine and (R,R)-1,2-diphen-
ylethylenediamine in terms of both turnover number of
the catalyst and enantioselectivity.3 Syntheses of optically
active monosubstituted and vicinal disubstituted 1,2-
ethylenediamines have been successfully reported.1 How-
ever, few examples of stereoselective synthesis of trisub-
stituted 1,2-ethylenediamines have been reported,4 due
to the difficulty in the construction of the quaternary
chiral center.5
(4) (a) For substitution of the hydroxyl group with azide under acidic
condition and consecutive hydrogenation, see: Wey, S.-J .; O’Connor,
K. J .; Burrows, C. J . Tetrahedron Lett. 1993, 34, 1905-1908. (b) For
[3+2] cycloadditions between nonracemic p-toluenesulfinimines and
iminoesters, see: Viso, A.; Ferna´ndez de la Pradilla, R.; Garc´ıa, A.;
Guerrero-Strachan, C.; Alonso, M.; Tortosa, M.; Flores, A.; Mart´ınez-
Ripoll, M.; Fonseca, I.; Andre´, I.; Rodr´ıguez, A. Chem. Eur. J . 2003, 9,
2867-2876 and references therein. (c) For stereoselective alkylation
of chiral 3-amino-4-styryl-â-lactams, see: Ojima, I.; Pei, Y. Tetrahedron
Lett. 1990, 31, 977-980.
(5) For a recent review of the construction of quaternary carbon
stereocenters, see: Corey, E. J .; Guzman-Perez, A. Angew. Chem., Int.
Ed. 1998, 37, 388-401 and references therein.
(6) (a) Hua, D. H.; Miao, S. W.; Chen, J . S.; Iguchi, S. J . Org. Chem.
1991, 56, 4-6. (b) Spero, D. M.; Kapadia, S. R. J . Org. Chem. 1997,
62, 5537-5541. (c) Cogan, D. A.; Ellman, J . A. J . Am. Chem. Soc. 1999,
121, 268-269. (d) Steinig, A. G.; Spero, D. M. J . Org. Chem. 1999, 64,
2406-2410. (e) Cimarelli, C.; Palmieri, G.; Volpini, E. Tetrahedron:
Asymmetry 2002, 13, 2011-2018. (f) Cimarelli, C.; Palmieri, G.;
Volpini, E. J . Org. Chem. 2003, 68, 1200-1206.
The addition of nucleophiles to ketimines is one of the
most direct and promising methods for the preparation
* To whom correspondence should be addressed. Fax: +81 564 51
7086.
(1) For a recent review of the chemistry of 1,2-ethylenediamines,
see: Lucet, D.; Gall, T. L.; Mioskowski, C. Angew. Chem., Int. Ed. 1998,
37, 2580-2627 and references therein.
(2) (a) Anastassiadou, M.; Danoun, S.; Crane, L.; Baziard-Mouysset,
G.; Payard, M.; Caignard, D.-H.; Rettori, M.-C.; Renard, P. Bioorg. Med.
Chem. 2001, 9, 585-592. (b) Szabo, B. Pharmacol. Ther. 2002, 93,
1-35.
(3) Ohkuma, T.; Koizumi, M.; Doucet, H.; Pham, T.; Kozawa, M.;
Murata, K.; Katayama, E.; Yokozawa, T.; Ikariya, T.; Noyori, R. J . Am.
Chem. Soc. 1998, 120, 13529-13530.
(7) Marco and Carda have reported nucleophilic alkylation to
R-oxyketimines derived from erythrulose; see: (a) Marco, J . A.; Carda,
M.; Murga, J .; Gonza´lez, F.; Falomir, E. Tetrahedron Lett. 1997, 38,
1841-1844. (b) Marco, J . A.; Carda, M.; Murga, J .; Rodr´ıguez, S.;
Falomir, E.; Oliva, M. Tetrahedron: Asymmetry 1998, 9, 1679-1701.
(c) Portole´s, R.; Murga, J .; Falomir, E.; Carda, M.; Uriel, S.; Marco, J .
A. Synlett 2002, 711-714.
10.1021/jo049405i CCC: $27.50 © 2004 American Chemical Society
Published on Web 08/21/2004
J . Org. Chem. 2004, 69, 6329-6334
6329

Kohmura and Mase
SCHEME 1. Syn th esis of N-Dip h en ylp h osp h in oyl Ketim in es 4a -e
TABLE 1. Nu clep h ilic Ad d ition of P h en ylm eta la ted
Rea gen ts to 4a
Selection of the proper protecting group for the ketimine
is a key issue. Addition of alkyllithiums or Grignard
reagents to N-p-toluenesulfonyl ketimines8a and N-aryl
ketimines8b has been reported; however, deprotection of
these adducts, in general, requires harsh conditions.
Many protecting groups, which are cleaved under mild
condition, have been reported for the amino group.9
Among them, we expected that the N-diphenylphosphi-
noyl protecting group would be one of the most attractive
candidates because of its electron-withdrawing ability
and its ease of removal under acidic conditions.9a There-
fore, our research efforts focused on the nucleophilic
addition to N-diphenylphosphinoyl ketimines.10 Herein,
we wish to report an efficient and stereoselective syn-
thesis of trisubstituted 1,2-ethylenediamines via N-
diphenylphosphinoyl ketimines.
yield (%)a
entry
PhM
PhLi
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
temp (°C)
5a a
4a
6a
1
2
3
4
-78
-78
-20
0
20
40
18
0
36
94
18
0
0
0
37
3
61
77
73
31
60
25
22
18
27
27
5
6
7b
0
0
a
b
HPLC assay yield of crude product. The reaction was carried
out in THF.
Resu lts a n d Discu ssion
Preparation of optically active R-amino N-diphenylphos-
phinoyl ketimines 4a -e was accomplished from the
Weinreb amide of N-Boc-protected amino acid 1a and 1b
in three steps: (i) alkylation of 1, (ii) formation of oximes
from ketones 2,11 and (iii) rearrangement of O-phosphino-
oximes generated in situ from oximes 3 and diphenyl-
chlorophosphine (Scheme 1).12,13 The stereochemistry of
the CdN bond of 4a was confirmed by its X-ray crystal-
lography to be Z. The configurations of 4b, 4c, 4d , and
4e were similarly assigned.
With the desired substrates in hand, we examined the
key addition reaction to the ketimine. To simplify the
analysis of the products of reaction, we selected phenyl-
metalated reagents as a model nucleophile and 4a as a
model substrate since the expected products were not
diastereomeric (Table 1). Though it is reported that
strong bases such as alkyl Grignard reagents are sus-
ceptible to enolization of imines to enamines,6d,14 we
expected that the CdN double bond would be activated
enough by the electron-withdrawing diphenylphosphinoyl
group to promote 1,2-addition while avoiding enolization.
Furthermore, N-metalation on the R-amino group should
reduce the acidity of the R-proton and might offer a
positive chelating effect with the imine nitrogen. Thus,
metalation would prevent enolization. To explore this
possibility, we first examined the reaction using phenyl-
lithium as a nucleophile in noncoordinating toluene as
solvent at -78 °C. The desired reaction occurred but the
yield of adduct 5a a was only 18%, and 36% of ketimine
4a remained unreacted (entry 1). Upon further investiga-
tion of the reaction mixture, we found considerable
amounts (37%) of enamine 6a , which was formed via
R-proton abstraction of the imine. To address this issue,
we proposed the notion that the R-proton abstraction of
the imine could be suppressed by reducing the basicity
of nucleophile. We next examined the reaction with
phenylmagnesium bromide instead of phenyllithium
under the same condition; however, 5a a was not observed
with almost complete recovery of 4a (entry 2). Fortu-
(8) (a) Wolfe, J . P.; Ney, J . E. Org. Lett. 2003, 5, 4607-4610. (b)
Gilman, H.; Eisch, J . J . Am. Chem. Soc. 1957, 79, 2150-2153.
(9) Green, T. W.; Wuts, P. G. M. Protective Groups in Organic
Synthesis, 3rd ed.; Wiley: New York, 1999; p 598.
(10) For a contemporaneous study of addition to N-diphenylphos-
phinoyl R-ketimino ester, see: Wipf, P.; Stephenson, C. R. J . Org. Lett.
2003, 5, 2449-2452.
(11) Davis, F. A.; Haque, M. S.; Przeslawski, R. M. J . Org. Chem.
1989, 54, 2021-2024.
(12) Krzyzanowska, B.; Stec, W. J . Synthesis 1978, 521-524.
(13) It is presumed that slight loss of ee in 4d and 4e would take
place at the formation of oximes 3, which required the longer reaction
time under reflux condition due to steric bulkiness in 2d and 2e.
(14) Stork, G.; Dowd, S. R. J . Am. Chem. Soc. 1963, 85, 2178-2180.
6330 J . Org. Chem., Vol. 69, No. 19, 2004

Grignard Addition to N-Diphenylphosphinoyl Ketimines
TABLE 2. Gr ign a r d Ad d ition to N-Dip h en ylp h osp h in oyl Ketim in es 4a -e
entry substrate ee (%)
R3
temp (°C) time (h) adduct yield (%)a de (%)b ee (%)c Bz-imine yield (%)a ee (%)c
1
2
3
4
5
6
7
8
9
4a
4a
4a
4a
4b
4b
4c
4d
4d
4e
99
99
99
99
99
99
99
96
96
92
Ph
0
0
0
0
0
0
0
25
25
25
3
3
3
3
3
3
3
7
7
5
5a a
5a b
5a c
5a d
5ba
5bb
5ca
5d a
5d b
5ea
66 (77)
75
(84)
70
49
74
98
99
99
99
98
99
99
97
96
92
4-F-C6H4
4-MeO-C6H4
Me
Ph
4-F-C6H4
Ph
Ph
4-F-C6H4
Ph
94
80
>96
83
72
96
(60)
(62)
(58)
7d a
7d b
7ea
(40)
(38)
(39)
96
96
92
>96d
>96d
10
a
b
Isolated (HPLC assay) yield. Yields were calculated based on the amount of imine 4. Determined by 1H NMR analysis (500 MHz).
c Determined by chiral HPLC analysis. Diastereomer was not detected.
d
nately, the reaction proceeded smoothly when the reac-
tion temperature was warmed to -20 °C to provide the
desired product 5a a as a major product (entry 3). The
best result of 77% yield was attained when the reaction
was performed at 0 °C (entry 4). Elevated temperatures
beyond 0 °C resulted in diminished product yield (entries
5 and 6). The reaction in toluene gave much better results
than that in coordinating solvents such as THF (entry
7). Addition of several Lewis acids did not improve the
reaction.15
F IGURE 1. Confirmation of absolute configurations of 8a b
and 8ba by NOE.
Absolute configuration of the newly formed stereogenic
center in 5a b and 5ba was confirmed by NOE studies of
the corresponding imidazolidinones 8a b and 8ba as
shown in Figure 1.18 The enantiomeric excesses of 8a b
and 8ba were the same as those of 5a b and 5ba ,
therefore, no epimerization occurred at C-5. The absolute
configuration at C-5 of both imidazolidinones was 5S
because of its origin from (S)-alanine (natural amino
acid). For 8a b, a characteristic NOE was observed
between H-5 (imidazolidinone) and the ortho proton of
the 4-fluorophenyl group at C-4 (7.6%), and between the
methyl proton at C-5 and the ortho proton of the phenyl
group at C-4 (2.1%). For 8ba , a characteristic NOE was
observed between H-5 (imidazolidinone) and the ortho
proton of the phenyl group at C-4 (7.5%), and between
the methyl proton at C-5 and the ortho proton of the
4-fluorophenyl group at C-4 (2.0%). These results sup-
ported the absolute configurations of 8a b and 8ba as
4S,5S and 4R,5S, respectively. The absolute configura-
With the optimal conditions in hand, the addition of
various Grignard reagents to 4 was examined to expand
the scope of the reaction. The results are summarized
in Table 2. The addition to the imines derived from
L-alanine (4a , 4b, and 4c: R1 ) Me) proceeded smoothly
in toluene at 0 °C and the adducts were obtained in
moderate to good yields with good to excellent diastereo-
selectivities (entries 1-7).16 The reaction with alkyl
Grignard reagents such as methylmagnesium bromide
afforded the desired product 5a d in reasonable yield
(70%) with >96% de (entry 4). In efforts to improve
the diastereoselectivity while suppressing the R-proton
abstraction from the imine, we next examined the addi-
tion to bulkier imines derived from L-valine (4d and 4e:
R1 ) i-Pr). Higher temperature and prolonged reaction
times were required to complete the reaction. As ex-
pected, the reactions proceeded with excellent diastereo-
selectivity without formation of enamine 6, although the
yields were moderate due to formation of benzoyl imines
(Bz-imine: 7d a , 7d b, and 7ea ) caused by nucleophilic
attack of Grignard reagents on the Boc group (entries
8-10).17
(17) The enantiomeric excesses of the Bz-imines 7d a , 7d b, and 7ea
were the same as those of the substrate imines 4d and 4e. From these
results, no enolization occurred because of the bulkiness of the
isopropyl group.
(15) CeCl3, Ti(Oi-Pr)4, Me3Al, BF3‚Et2O, ZnCl2, MgBr2, and LiBr
were used.
(16) The main byproduct was corresponding enamide 6.
(18) 8a b and 8ba were prepared by removal of Boc and diphe-
nylphosphinoyl groups from 5a b and 5ba consecutive treatment with
1,1′-carbonyldiimidazole, respectively.
J . Org. Chem, Vol. 69, No. 19, 2004 6331

Kohmura and Mase
TABLE 3. Con ver sion of th e Ad d u cts 5 to Dia m in es 10
entry substrate ee (%) R1
R2
R3
product time (h) yield (%)a product time (h) yield (%)a de (%)b ee (%)c
1
2
3
4
5a a
5a b
5ba
5bb
98
99
98
99
Me Ph
Me Ph
Me 4-F-C6H4 Ph
Ph
4-F-C6H4
9a a
9a b
9ba
9bb
4
97
98
95
97
10a a
10a b
10ba
10bb
2
4
4
2
64 (89)
72
64
99
98
98
99
1.5
1.5
1.5
95
86
Me 4-F-C6H4 4-F-C6H4
73
a
b
Isolated (HPLC assay) yield. Determined by reversed-phase HPLC analysis. c Determined by chiral HPLC analysis.
around the diphenylphosphinoyl group. To overcome this
difficulty, an alternative approach was required. In
efforts to decrease the bulkiness of the protecting group,
removal of the oxygen atom from the diphenylphosphi-
noyl group was investigated. Furthermore, it has been
reported that aminophosphines readily react with an
alcohol to afford an alkyl phosphinate under milder
reaction conditions than that for the corresponding
phosphorus(V) adduct.22 One can envision that reduction
to the aminophosphine would allow it to react smoothly
with water to liberate diphenylphosphinic acid. On the
basis of this idea, reduction conditions were extensively
studied with use of trichlorosilane,23a triethoxysilane,23b
and lithium aluminum hydride.23c We found that reduc-
tion with trichlorosilane in the presence of triethylamine
in toluene at 70 °C efficiently cleaves the diphenylphos-
phinoyl group even from a bulky R,R-dibranched amine
as shown in Table 3.24 Adduct 5 was typically trans-
formed to the desired diamine 10 via cleavage of the Boc
group as depicted in Table 3. Under these conditions, no
racemization was observed; 10a a , 10a b, 10ba , and 10bb
were obtained in 98-99% ee, and 10a b and 10ba were
obtained in 95% de and 86% de, respectively.
F IGURE 2. Proposed transition state model of the addition.
tions of 5a c, 5a d , 5ca , 5d b, and 5ea were similarly
assigned.
Although moderate yields were attained in the present
system, the addition of a strongly basic Grignard reagent
such as phenylmagnesium bromide or methylmagnesium
bromide to an enolizable imine has generally resulted in
lower yields because of their propensity to act as a base.
Nitrogen-magnesium bond formation on the R-amino
group would reduce the acidity of the R-proton of the
imines. Moreover, excellent diastereoselectivities were
also obtained. These results could be rationalized by
invoking a chelation mechanism through a five-mem-
bered transition structure by coordination of magnesium
ion on nitrogen atom to ketimine nitrogen as shown in
Figure 2.19 Magnesium ion activates unreactive ketimine
and Grignard reagent approaches the less-hindered Si
face of the CdN bond. This proposal is supported by the
results obtained from reactions performed in the coordi-
nating THF solvent: decreased adduct formation and
increased enamine formation (Table 1, entry 7).
It is well-known that the diphenylphosphinoyl group
can be removed under acidic conditions.9 Accordingly, we
expected that diphenylphosphinoyl and Boc groups in the
adducts would be simultaneously removed under acidic
conditions.20 However, the diphenylphosphinoyl group
was not removed by the usual acidic conditions, although
the Boc group was easily removed by treatment with HCl
in ethyl acetate to give monoamide 9 quantitatively.
Regardless of further trials under many acidic conditions,
removal of the diphenylphosphinoyl group from 9 to the
desired diamine 10 did not occur.21 We thought this
difficulty of removal could be due to steric hindrance
Con clu sion
We have demonstrated the first example of a diaste-
reoselective addition of aryl and alkyl Grignard reagents
to N-diphenylphosphinoyl ketimines derived from amino
acids. The reaction proceeds in moderate to good yield
(up to 75%) and good to excellent diastereoselectivity (up
to >96% de). In addition, cleavage of the resulting
phosphinoyl amide to the desired amine was efficiently
achieved via a trichlorosilane/triethylamine protocol. This
methodology provides a novel process for the preparation
of trisubstituted 1,2-ethylenediamines. Asymmetric reac-
tions with these chiral diamines are currently under
investigation.
Exp er im en ta l Section
Gen er a l P r oced u r e for Syn th esis of N-Dip h en ylp h os-
p h in oyl Ketim in es (4). A typical experimental procedure was
exemplified by the synthesis of 4a (Scheme 1): Phenylmagne-
sium bromide (1.03 M, THF solution, 52 mL, 53.8 mmol) was
dropwise added to a stirred suspension of 1a (5.00 g, 21.5
mmol) in THF (50 mL) at 0 °C, and the mixture was stirred
for 13.5 h at 25 °C. After the mixture was cooled to 0 °C,
aqueous KHSO4 solution (8.79 g, 64.6 mmol in 75 mL of water)
was added. The reaction mixture was stirred at ambient
temperature for 30 min and extracted with tert-butyl methyl
(19) We consider that Z-ketimines converted to E-isomers in the
reaction system and E-ketimines constructed a five-membered transi-
tion structure by coordination of magnesium ion on nitrogen atom to
ketimine nitrogen. It is reported that N-diphenylphosphinoyl ketimines
exist in
a very fast equilibrium between E- and Z-isomers, see:
Masumoto, S.; Usuda, H.; Suzuki, M.; Kanai, M.; Shibasaki, M. J . Am.
Chem. Soc. 2003, 125, 5634-5635.
6332 J . Org. Chem., Vol. 69, No. 19, 2004

Grignard Addition to N-Diphenylphosphinoyl Ketimines
ether (50 mL × 2). The organic layer was washed with water
(50 mL) and brine (25 mL), dried over Na2SO4, and concen-
trated in vacuo. Crude ketone 2a (5.85 g) was obtained as a
colorless solid. Crude 2a (5.85 g, 21.5 mmol) was dissolved in
ethanol (107 mL). To it was added hydroxylamine hydrochlo-
ride (3.74 g, 53.8 mmol), sodium acetate (4.41 g, 53.8 mmol),
and water (21 mL). The mixture was stirred for 1.5 h under
reflux, cooled to room temperature, neutralized with saturated
aqueous NaHCO3 (27 mL), and concentrated under reduced
pressure to remove ethanol. The residual solution was ex-
tracted with ethyl acetate (27 mL × 3). The organic layer was
dried over Na2SO4, and concentrated in vacuo. The residue was
passed through a pad of silica gel (heptane/ethyl acetate )
20/1 to 4/1). Crude oxime 3a (5.82 g) was obtained as a colorless
solid. Crude 3a (5.82 g, 21.5 mmol) was dissolved in THF (57
mL), followed by the addition of triethylamine (3.6 mL, 25.8
mmol), and the mixture was cooled to -30 °C. Chlorodiphe-
nylphosphine (4.1 mL, 22.6 mmol) was added dropwise at -30
°C, and the mixture was stirred for 3 h at the same temper-
ature. It was then warmed to room temperature over 30 min.
Water (57 mL) was added and the organic layer was separated.
The aqueous layer was extracted with tert-butyl methyl ether
(57 mL × 2). The combined organic layer was washed with
brine (13 mL), dried over Na2SO4, and concentrated in vacuo.
The orange oily residue was chromatographed on silica gel
(heptane/tert-butyl methyl ether ) 4/1 to 2/1) to give 4a as a
colorless solid (7.21 g, 75%).
mixture was extracted with ethyl acetate (10 mL × 3). The
organic layer was washed with brine (5 mL), dried over Na2-
SO4, and concentrated in vacuo. The colorless solid residue was
chromatograped on silica gel (heptane/ethyl acetate ) 7/3) to
give 5a b as a colorless solid (454 mg, 75%).
ter t-Bu tyl N-{(1S,2S)-1-[(d ip h en ylp h osp h in oyl)a m in o]-
1-(4-flu or op h en yl)-1-p h en yl-2-p r op yl}ca r ba m a t e (5a b ):
26
A colorless solid. [R]
-66.0 (c 1.02, CHCl3, 99% ee, 94%
D
de). HPLC: DAICEL CHIRALPAK AD (hexane/2-propanol )
4/1, flow 0.5 mL/min) tR 18.4 min (minor) and 45.5 min (major).
1H NMR: δ 7.76 (m, 2H), 7.64 (m, 2H), 7.41-7.19 (m, 14H),
6.49 (dd, J ) 8.7, 8.7 Hz, 2H), 4.98 (dq, J ) 9.6, 6.8 Hz, 1H),
4.56 (br, 1H), 1.46 (s, 9H), 1.02 (d, J ) 6.6 Hz, 3H). 13C NMR:
δ 161.7 (d, J ) 247 Hz), 156.5, 142.0, 134.4 (d, J ) 127 Hz),
134.2 (d, J ) 131 Hz), 132.5 (d, J ) 6 Hz), 131.7 (d, J ) 10
Hz), 131.3, 131.2, 130.8, 128.3 (d, J ) 12 Hz), 127.9 (d, J ) 13
Hz), 127.7, 127.4, 127.2, 113.9 (d, J ) 21 Hz), 80.0, 68.5, 53.7,
28.4, 17.8. IR (KBr): ν 3431, 3239, 3057, 2978, 2931, 1701,
1605, 1511, 1439, 1366, 1161, 1121, 1053, 860, 754, 720, 698,
540 cm-1. HRFABMS m/z calcd for C32H34FN2O3PH (MH+)
545.2369, found 545.2364.
ter t-Bu tyl N-{(1S,2S)-1-[(Diph en ylph osph in oyl)am in o]-
1-(4-flu or op h en yl)-3-m et h yl-1-p h en yl-2-b u t yl}ca r b a m -
a te (5d b). Yields were determined by HPLC assay with the
following conditions: Column, YMC AQ-303 (0.1% aqueous
phosphoric acid/acetonitrile ) 50/50 to 5/95 in 30 min, flow
1.0 mL/min) tR 13.1 min (7d b), 19.7 min (4d ), and 21.5 min
(5d b). After workup, this product was obtained as a mixture
of 5d b and 7d b. Authentic samples were prepared as follows:
A mixture of 5d b and 7d b (551 mg) was treated with HCl (4
N, ethyl acetate solution, 2.8 mL) and stirred for 2 h at room
temperature. Aqueous NaOH (5 N, 2.3 mL) was added and
stirred for 30 min at room temperature. The mixture was
extracted with ethyl acetate (2.8 mL × 3). The organic layer
was dried over Na2SO4 and concentrated in vacuo. The pale
yellow oily residue was chromatographed on silica gel (hep-
tane/ethyl acetate ) 1/1 to CHCl3/methanol ) 9/1) to afford
des-Boc 5d b as a colorless solid (235 mg) and 7d b as a colorless
solid (204 mg). To a stirred solution of des-Boc 5d a (25.9 mg,
0.0548 mmol) in dioxane (0.52 mL) was added aqueous NaOH
(1 N, 0.055 mL, 0.0548 mmol) and di-tert-butyl dicarbonate
(14.3 mg, 0.0655 mmol). After being stirred for 9 h at 40 °C,
the mixture was directly subjected to column chromatography
(heptane/ethyl acetate ) 7/3) to afford 5d b as a colorless solid
ter t-Bu tyl (Z)-N-{(2S)-1-[(d ip h en ylp h osp h in oyl)im in o]-
1-p h en yl-2-p r op yl}ca r ba m a te (4a ): A colorless solid. Mp:
149-151 °C. [R] 26 -20.6 (c 1.02, CHCl3, 99% ee). HPLC:
D
DAICEL CHIRALPAK AD (hexane/2-propanol ) 9/1, flow 0.5
1
mL/min) tR 22.0 min (minor) and 27.9 min (major). H NMR:
δ 8.10-8.03 (m, 4H), 7.90 (m, 2H), 7.69 (d, J ) 9.5 Hz, 1H),
7.56 (dd, J ) 7.5, 7.3 Hz, 1H), 7.51-7.43 (m, 6H), 7.39 (m,
2H), 5.43 (dq, J ) 9.5, 7.2 Hz, 1H), 1.48 (d, J ) 7.2 Hz, 3H),
1.22 (s, 9H). 13C NMR: δ 186.9 (d, J ) 9 Hz), 159.7, 155.6,
138.2 (d, J ) 22 Hz), 134.2, (d, J ) 141 Hz), 133.9 (d, J ) 129
Hz), 132.5, 132.0 (d, J ) 9 Hz), 131.5 (d, J ) 3 Hz), 131.4,
128.7 (d, J ) 15 Hz), 128.5 (d, J ) 13 Hz), 128.4 (d, J ) 13
Hz), 79.0, 49.9 (d, J ) 11 Hz), 28.2, 19.4. IR (KBr): ν 3238,
3057, 2975, 1707, 1637, 1534, 1440, 1365, 1274, 1231, 1182,
1122, 1098, 1024, 856, 725, 698, 552, 525 cm-1. HRFABMS
m/z calcd for C26H29N2O3PH (MH+) 449.1994, found 449.1990.
ter t-Bu tyl (Z)-N-{(2S)-1-[(d ip h en ylp h osp h in oyl)im in o]-
(24.0 mg, 76%). [R] 26 -31.1 (c 1.00, CHCl3, 96% ee, >96%
3-m eth yl-1-p h en yl-2-bu tyl}ca r ba m a te (4d ): A colorless
D
27
solid. Mp: 49-54 °C. [R]
-22.6 (c 1.04, CHCl3, 96% ee).
D
de). HPLC: DAICEL CHIRALPAK AD (hexane/2-propanol )
9/1, flow 0.5 mL/min) tR 19.9 min (minor) and 33.4 min (major).
1H NMR: δ 7.74 (m, 2H), 7.64 (m, 2H), 7.37-7.26 (m, 13H),
6.51 (dd, J ) 8.6, 8.5 Hz, 2H), 5.28 (br, 1H), 4.79 (d, J ) 10.4
Hz, 1H), 4.49 (d, J ) 7.5 Hz, 1H), 1.86 (m, 1H), 1.48 (s, 9H),
0.95 (d, J ) 6.6 Hz, 3H), 0.26 (d, J ) 6.2 Hz, 3H). 13C NMR:
δ 161.8 (d, J ) 247 Hz), 157.3, 142.6, 134.9, 134.6 (d, J ) 128
Hz), 134.0 (d, J ) 131 Hz), 132.4 (d, J ) 8 Hz), 131.8 (d, J )
10 Hz), 131.2 (d, J ) 3 Hz), 131.1 (d, J ) 3 Hz), 130.8, 128.4,
128.2 (d, J ) 3 Hz), 127.9, 127.8, 127.3, 113.9 (d, J ) 21 Hz),
79.9, 68.6, 68.4, 28.4, 28.3, 23.7, 17.3. IR (KBr): ν 3441, 3222,
3058, 2964, 2930, 2873, 1703, 1605, 1511, 1439, 1391, 1366,
1232, 1166, 1121, 884, 833, 753, 720, 699, 539 cm-1. HR-
FABMS m/z calcd for C34H38FN2O3PH (MH+) 573.2682, found
573.2682.
HPLC: DAICEL CHIRALPAK AD (hexane/2-propanol ) 9/1,
flow 0.5 mL/min) tR 12.6 min (minor) and 19.2 min (major).
1H NMR: δ 8.10 (d, J ) 7.5 Hz, 2H), 8.04 (m, 2H), 7.90 (dd, J
) 12.0, 7.2 Hz, 2H), 7.61-7.37 (m, 10H), 5.03 (dd, J ) 9.5, 9.4
Hz, 1H), 2.34 (m, 1H), 1.27 (s, 9H), 1.02 (d, J ) 6.6 Hz, 3H),
0.67 (d, J ) 6.8 Hz, 3H). 13C NMR: δ 186.9, 156.1, 140.2 (d, J
) 22 Hz), 134.1 (d, J ) 133 Hz), 134.0 (d, J ) 138 Hz), 132.3,
131.8 (d, J ) 9 Hz), 131.6 (d, J ) 9 Hz), 131.6 (d, J ) 3 Hz),
131.5 (d, J ) 3 Hz), 128.8 (d, J ) 10 Hz), 128.6 (d, J ) 13 Hz),
128.4 (d, J ) 13 Hz), 78.9, 60.0 (d, J ) 10 Hz), 32.7, 28.2, 20.1,
19.8. IR (KBr): ν 3448, 3276, 3059, 2972, 2930, 2871, 1706,
1639, 1508, 1439, 1366, 1288, 1235, 1181, 1121, 1106, 1004,
726, 695, 550, 526 cm-1. HRFABMS m/z.calcd for C28H33N2O3-
PH (MH+) 477.2307, found 477.2325.
4b, 4c, and 4e were prepared analogously. Their analytical
and spectroscopic data are compiled in the Supporting Infor-
mation.
N-{(2S)-1-[(Dip h en ylp h osp h in oyl)im in o]-3-m et h yl-1-
ph en yl-2-bu tyl}-4-flu or oben zam ide (7db): A colorless solid.
[R] 27 -199.4 (c 0.76, CHCl3, 96% ee). HPLC: DAICEL
D
CHIRALPAK AD (hexane/2-propanol ) 4/1, flow 0.5 mL/min)
tR 16.0 min (major) and 41.3 min (minor). 1H NMR: δ 7.94
(dd, J ) 13.5, 7.1 Hz, 2H), 7.69 (m, 2H), 7.59 (dd, J ) 7.6, 6.4
Hz, 1H), 7.52 (ddd, J ) 7.6, 7.3, 3.4 Hz, 2H), 7.42 (m, 2H),
7.33 (dd, J ) 7.6, 7.3 Hz, 1H), 7.30-7.28 (m, 3H), 7.14 (ddd, J
) 7.8, 7.6, 3.7 Hz, 2H), 7.04 (m, 2H), 6.87 (dd, J ) 8.9, 8.5 Hz,
2H), 5.10 (s, 1H), 4.12 (d, J ) 1.8 Hz, 1H), 1.39 (m, 1H), 0.78
(d, J ) 7.1 Hz, 3H), 0.74 (d, J ) 6.4 Hz, 3H). 13C NMR: δ
162.3 (d, J ) 248 Hz), 160.9 (d, J ) 7 Hz), 142.0 (d, J ) 3 Hz),
Gen er a l P r oced u r e for Gr ign a r d Ad d ition to N-Dip h e-
n ylp h osp h in oyl Ketim in es (4). Typical experimental pro-
cedure was exemplified by the addition of 4-fluorophenylmag-
nesium bromide to 4a in toluene at 0 °C (entry 2 in Table 2):
To a stirred solution of 4a (500 mg, 1.11 mmol) in toluene (10
mL) at 0 °C was added 4-fluorophenylmagnesium bromide (1.0
M, THF solution, 3.3 mL, 3.33 mmol). After the mixture was
stirred at 0 °C for 3 h, to it was added aqueous KHSO4 (0.446
M, 10 mL, 4.46 mmol). After stirred for 30 min, the reaction
J . Org. Chem, Vol. 69, No. 19, 2004 6333

Kohmura and Mase
138.3, 133.1 (d, J ) 130 Hz), 132.3 (d, J ) 11 Hz), 132.0 (d, J
) 3 Hz), 131.9 (d, J ) 11 Hz), 131.8 (d, J ) 8 Hz), 131.5 (d, J
) 3 Hz), 131.4, 129.6, 128.3 (d, J ) 14 Hz), 127.3, 127.2, 126.9,
114.1 (d, J ) 21 Hz), 75.9, 69.8 (d, J ) 6 Hz), 30.0, 21.5, 15.1.
IR (KBr): ν 3432, 3221, 3061, 2962, 2933, 2875, 1707, 1602,
1509, 1438, 1385, 1315, 1226, 1121, 1105, 1061, 1016, 843, 806,
756, 719, 698, 628, 540 cm-1. HRFABMS m/z calcd for C27H26
-
FN2OPH (MH+) 445.1845, found 445.1841.
9a a , 9ba , and 9bb were prepared analogously. Their
analytical and spectroscopic data are compiled in the Sup-
porting Information.
Gen er a l P r oced u r e for P r ep a r a tion of Tr isu bstitu ted
Dia m in e 10. A typical experimental procedure was exempli-
fied by the preparation of 10a b (entry 2 in Table 3): To a
mixture of 9a b (100 mg, 0.225 mmol) and triethylamine (0.63
mL, 4.50 mmol) in toluene (3 mL) was added trichlorosilane
(0.23 mL, 2.25 mmol) at 0 °C. The reaction mixture was
warmed to 70 °C, stirred for 4 h, and cooled to 0 °C. Aqueous
NaOH (5 N, 3.0 mL) was carefully added followed by stirring
for 10 min at 70 °C and cooled to room temperature. After
phase cut, the aqueous layer was extracted with ethyl acetate
(3 mL × 3). The combined organic layers were washed with
water (3 mL × 2) and extracted with aqueous HCl (1 N, 3 mL
× 3). The aqueous layer was washed with ethyl acetate (3 mL
× 2), basified by aqueous NaOH (5 N, 1.5 mL), and extracted
with ethyl acetate (3 mL × 3). The organic layer was washed
with brine (3 mL), dried over Na2SO4, and concentrated in
vacuo. Crude 10a b was obtained as a colorless oil. Column
chromatography on NH-silica gel (heptane/ethyl acetate ) 4/1
to 7/3) afforded pure 10a b as a colorless solid (39.8 mg, 72%).
752, 727, 699, 591, 529 cm-1. HRFABMS m/z calcd for C30H28
-
FN2O2PH (MH+) 499.1951, found 499.1942.
5a a , 5a c, 5a d , 5ba , 5bb, 5ca , 5d a , 5ea , 7d a , and 7ea were
prepared analogously. Their analytical and spectroscopic data
are compiled in the Supporting Information.
(Z)-ter t-Bu tyl N-{R-[(Dip h en ylp h osp h in oyl)a m in o]-â-
m eth ylstyr yl} ca r ba m a te (6a ).25 An authentic sample was
prepared according to the following procedure: To a solution
of 4a (100 mg, 0.22 mmol) in THF (2 mL) was added lithium
diisopropylamide (2.0 M, 0.33 mL, 0.67 mmol) at -78 °C and
the reaction mixture was stirred for 2 h. To it was added water
(2 mL) and then the mixture was allowed to warm to room
temperature and extracted with ethyl acetate (2 mL × 3). The
organic layer was dried over Na2SO4 and concentrated in
vacuo. The pale yellow oily residue was chromatographed on
silica gel (heptane/ethyl acetate ) 3/2) to afford 6a as a
1
colorless solid (90 mg, 90%). H NMR: δ 7.73 (d, J ) 7.7 Hz,
2H), 7.72 (dd, J ) 12.4 1.1 Hz, 2H), 7.42 (m, 2H), 7.33-7.30
(m, 5H), 7.11-7.07 (m, 3H), 7.01 (m, 2H), 5.41 (d, J ) 6.4 Hz,
1H), 1.86 (d, J ) 1.8 Hz, 3H), 1.47 (s, 9H). 13C NMR: δ 153.8,
137.6, 132.2 (d, J ) 129 Hz), 131.9 (d, J ) 10 Hz), 131.7 (d, J
) 3 Hz), 130.2, 128.2 (d, J ) 13 Hz), 127.6, 127.2, 125.3, 123.4,
80.1, 28.4, 17.8. IR (KBr): ν 3239, 3058, 2977, 2928, 2871,
1717, 1484, 1440, 1366, 1318, 1247, 1169, 1123, 1069, 895, 780,
752, 729, 696, 540, 513 cm-1. HRFABMS m/z calcd for
Enantiomeric excess of 10a b was determined by analysis
of corresponding imidazolidinone. Diasteromeric excess of
10a b was determined by reversed-phase HPLC: YMC AQ-
303 (0.1% aqueous phosphoric acid/acetonitrile ) 95/5 to 90/
10 in 20 min, to 80/20 in 10 min, flow 1.0 mL/min) tR 17.5 min
(10a b) and 18.2 min (10ba ).
C
26H29N2O3PH (MH+) 449.1994, found 449.2014.
(1S,2S)-1-(4-F lu or op h en yl)-1-p h en yl-1,2-p r op a n ed i-
Gen er a l P r oced u r e for Dep r otection of th e N-ter t-
a m in e (10a b): A colorless solid. [R] 27 +17.6 (c 0.82, CHCl3,
D
Bu toxyca r bon yl Gr ou p fr om 5. A typical experimental
procedure was exemplified by the deprotection of the N-tert-
butoxycarbonyl group from 5a b (entry 2 in Table 3): 5a b (200
mg, 0.367 mmol) was treated with HCl (4 N, ethyl acetate
solution, 1.0 mL) and stirred for 1.5 h at room temperature.
Aqueous NaOH (5 N, 0.8 mL) was added with stirring over 30
min. The mixture was extracted with ethyl acetate (1.0 mL ×
3). The organic layer was dried over Na2SO4, and concentrated
in vacuo. 9a b was obtained as a colorless solid (159 mg, 98%).
98% ee, 95% de). HPLC: DAICEL CHIRALPAK AD (hexane/
2-propanol ) 4/1, flow 0.5 mL/min) tR 14.0 min (minor) and
1
20.6 min (major). H NMR: δ 7.51 (dd, J ) 8.7, 5.4 Hz, 2H),
7.45 (d, J ) 7.5 Hz, 2H), 7.29 (dd, J ) 7.9, 7.5 Hz, 2H), 7.19
(dd, J ) 7.1, 6.9 Hz, 1H), 6.97 (dd, J ) 8.7, 8.6 Hz, 2H), 4.03
(q, J ) 6.3 Hz, 1H), 1.54 (br, 4H), 1.00 (d, J ) 6.3 Hz, 3H). 13
C
NMR: δ 160.1 (d, J ) 246 Hz), 146.3, 142.8 (d, J ) 3 Hz),
128.3 (d, J ) 8 Hz), 128.2, 126.5, 126.4, 115.0 (d, J ) 21 Hz),
64.0, 52.1, 17.9. IR (KBr): ν 3371, 3288, 3067, 3037, 3000,
2980, 2938, 2875, 1600, 1504, 1446, 1377, 1223, 1154, 1107,
(1S,2S)-1-[(Diph en ylph osph in oyl)am in o]-1-(4-flu or oph e-
n yl)-1-p h en yl-2-p r op yla m in e (9a b): A colorless solid. [R] 26
D
1038, 932, 867, 829, 809, 765, 723, 705, 634, 579, 542 cm-1
.
-7.8 (c 1.05, CHCl3, 99% ee, 94% de). 1H NMR: δ 7.75 (m,
2H), 7.54 (m, 2H), 7.39-7.28 (m, 9H), 7.22-7.15 (m, 4H), 6.53
(dd, J ) 8.7, 8.6 Hz, 2H), 5.80 (br, 1H), 4.14 (q, J ) 6.4 Hz,
1H), 1.55 (br, 2H), 1.06 (d, J ) 6.5 Hz, 3H). 13C NMR: δ 161.8
(d, J ) 247 Hz), 141.2 (d, J ) 6 Hz), 135.0 (d, J ) 129 Hz),
134.9, 133.9 (d, J ) 131 Hz), 133.0 (d, J ) 8 Hz), 131.6 (d, J
) 9 Hz), 131.3 (d, J ) 9 Hz), 131.0 (d, J ) 3 Hz), 130.7 (d, J
) 3 Hz), 129.3, 128.2 (d, J ) 13 Hz), 127.8 (d, J ) 13 Hz),
127.1, 127.1, 113.6 (d, J ) 21 Hz), 68.5, 53.5 (d, J ) 3 Hz),
21.4. IR (KBr): ν 3390, 3315, 3210, 3057, 2971, 2932, 2875,
1605, 1511, 1439, 1377, 1236, 1200, 1119, 1071, 1032, 870, 831,
HRFABMS m/z calcd for C15H17FN2H (MH+) 245.1454, found
245.1459.
10a a , 10ba , and 10bb were prepared analogously. Their
analytical and spectroscopic data are compiled in the Sup-
porting Information.
Ack n ow led gm en t. We are grateful to Dr. Nobuy-
oshi Yasuda and Dr. Michael Palucki for extensive
editing and advice on design of the manuscript. We
thank Dr. Takayuki Nemoto and Mr. Moriaki Ishikawa
for NMR spectroscopic and mass spectrometric as-
sistance, Mr. Hisaki Kojima and Ms. Ikuko Nishimura
for X-ray structural determination, and Mr. Hiromu
Amano for support of data collection.
(20) The diphenylphosphinoyl group is slightly less stable to acid
than the Boc group, see ref 9a.
(21) We examined HCl, trifluoroacetic acid, methanesulfonic acid,
BF3‚Et2O, ZnCl2, trimethylsilyl chloride, AlCl3, and TiCl4.
(22) (a) Denis, P.; Mortreux, A.; Petit, F.; Buono, G.; Peiffer, G. J .
Org. Chem. 1984, 49, 5274-5276. (b) Kolodiazhnyi, O. I.; Gryshkun,
E. V.; Andrushko, N. V.; Freytag, M.; J ones, P. G.; Schmutzler, R.
Tetrahedron: Asymmetry 2003, 14, 181-183.
(23) (a) Takaya, H.; Akutagawa, S.; Noyori, R. Org. Synth. 1988,
67, 20-31. (b) Coumbe, T.; Lawrence, N. J .; Muhammad, F. Tetrahe-
dron Lett. 1994, 35, 625-628. (c) Wang, Y.; Guo, H.; Ding, K.
Tetrahedron: Asymmetry 2000, 11, 4153-4162.
(24) The assay yield of the desired product in the crude reaction
mixture by HPLC did not concur with the isolated yield since isolation
of diamines is difficult because of their polar and protic nature (Table
3, entry 1).
(25) Stereochemistry was assigned by consideration of the reaction
mechanism.
Su p p or tin g In for m a tion Ava ila ble: Spectroscopic and
analytical data for 4b, 4c, 4e, 5a a , 5a c, 5a d , 5ba , 5bb, 5ca ,
5d a , 5ea , 7d a , 7ea , 9a a , 9ba , 9bb, 10a a , 10ba , 10bb, 8a b,
1
and 8ba ; H and 13C NMR spectra of 4a -e, 5a a -5a d , 5ba ,
5bb, 5ca , 5d a , 5d b, 5ea , 6a , 7d a , 7d b, 7ea , 9a a , 9a b, 9ba ,
9bb, 10a a , 10a b, 10ba , 10bb, 8a b, and 8ba ; X-ray crystal-
lographic data for 4a , and CIF file of 4a . This material is
available free of charge via the Internet at http://pubs.acs.org.
J O049405I
6334 J . Org. Chem., Vol. 69, No. 19, 2004