Chemistry Letters Vol.35, No.5 (2006)
479
Table 2. CsOH-induced asymmetric [1,2] Stevens rearrange-
ment
References and Notes
1
For reviews see: a) J. A. Vanecko, H. Wan, F. G. West,
´
sive Organic Synthesis, ed. by B. M. Trost, I. Fleming, Perga-
mon, Oxford, 1991, Vol. 3, Chap. 3.10.
CO2But
CsOH (5 equiv.)
CO2But
N
N
1,2-dichloroethane
Ar
(R)-5
Ar
I
−10 °C, 24 h
2
(1S)-4
Entry
Ar
Product
Yield/%a
ee/%b
3
4
For preparations of ꢀ-substituted proline derivatives, see:
references therein.
1
2
3
4c
4-Me-Ph
4-MeO-Ph
4-F-Ph
b
c
d
77
56
69
42d
84
86
90
>99
4-tBuOCOPh
e
aDetermined by 1H NMR assay using mesitylene or diphenyl-
methane as an internal standard. bDetermined by chiral
HPLC analysis after reduction of 5 with LiAlH4. cPerformed
at 0 ꢁC. dt-Butyl p-toluate was isolated in 55% yield.
5
6
Prepared from Cbz-L-proline in three steps: (i) isobutene, cat.
H2SO4, CH2Cl2, rt. (ii) H2 (1 atm), 10% Pd–C, EtOAc, rt. (iii)
PhCH2Cl, NaHCO3, CH3CN, reflux. Other substrates (3b–
3e) were prepared by the same procedure using the corre-
sponding benzylic chloride or bromide in step (iii).
Spectroscopic data; (1S, 2S)-N-benzyl-N-methylproline am-
Me
CO2But
CO2But
N
25
monium salt (1S)-4a; mp 155–156 ꢁC; ½ꢀꢂD ¼ ꢃ20:2ꢁ
(R)-5
N
Me
on (R)-face
(c ¼ 1:00, MeOH); 1H NMR (270 MHz, CDCl3) ꢁ 7.69–
7.65 (m, 2H, Ph), 7.53–7.42 (m, 3H, Ph), 5.41 (d, 1H,
J ¼ 12:7 Hz, CH2Ph), 5.22 (dd, 1H, J ¼ 10:5, 9.2 Hz,
CHCO2But), 5.12 (d, 1H, J ¼ 12:7 Hz, CH2Ph), 4.76 (ddd,
1H, J ¼ 10:5, 10.5, 10.5 Hz, 5-H), 3.46 (ddd, 1H, J ¼ 10:5,
8.4, 1.9 Hz, 5-H), 3.08 (s, 3H, CH3), 2.81–2.68 (m, 1H, 3-
or 4-H), 2.39–2.19 (m, 2H, 3- or 4-H), 2.09–1.95 (m, 1H,
.
CH2Ar
"solvent cage"
Ar
[outside the solvent cage]
(S)+(R)-5
Scheme 3. Pathways of the chirality transmission.
ammonium salts 4b–4e which were prepared in stereo-pure form
in the same way as described for 4a.9 As shown in Table 2, these
rearrangements afforded the corresponding ꢀ-(arylmethyl)-
proline t-butyl esters (5b–5d) with 84–90% ee in reasonable
yields except for the case of 4e where 5e was obtained as an
almost single enantiomer in only 42% yield, together with a
considerable amount of t-butyl p-toluate (Entry 4).
The question arises as to why the solid–liquid biphasic
condition provides such a remarkably enhanced % ee. While
the exact reason cannot be advanced at present, it is safe to
say that under the biphasic conditions, the recombination of
the radical pair initially formed from the N-ylide occurs more
rapidly in a solvent cage and hence more preferentially in the
retentive fashion (on the bottom side) to give an enhanced %
ee (Scheme 3). In other words, the recombination outside the
solvent cage leading to a decrease in % ee would be suppressed
under the biphasic conditions. The stability of the benzylic radi-
cal involved might be another factor in dictating the % ee. The
more unstable benzylic radical involved is, the more rapidly
the recombination would occur inside the solvent cage,
thus leading to a higher % ee as actually observed in Entry 4
(Table 2), although the p-(t-butoxycarbonyl)benzyl radical is
so reactive (unstable) and hence abstracts a hydrogen leading
to the formation of t-butyl p-toluate as a by-product.
t
3- or 4-H), 1.54 (s, 9H, Bu); 13C NMR (68 MHz, CDCl3) ꢁ
164.8, 132.5, 130.7, 129.2, 127.5, 85.7, 71.9, 65.9, 63.8,
44.0, 28.0, 24.6, 18.4; IR (KBr) 1744, 1148 cmꢃ1; Anal. calcd
for C17H26INO2: C, 50.63; H, 6.50; N, 3.47%. Found: C,
50.79; H, 6.61; N, 3.51%.
7
8
Similar rearrangements at 0 and ꢃ10 ꢁC gave 5a with 80% ee
(71% yield) and 84% ee (91% yield), respectively.
Reaction procedure: To a solution of (1S)-4a (180 mg,
0.447 mmol) in 1,2-dichloroethane (5 mL) was added CsOH
(0.37 g, 2.5 mmol) in one portion at ꢃ10 ꢁC under a nitrogen
atmosphere. After stirring for 24 h at the same temperature,
the resulting mixture was diluted with H2O and extracted
with Et2O. The combined extracts were washed with brine,
dried over Na2SO4 and concentrated. The yield of (R)-5a
was determined by 1H NMR spectroscopy of the crude
product using mesitylene as an internal standard (73% yield).
The pure (R)-5a was obtained after chromatography on silica
gel (Hex/EtOAc = 20:1 as eluent) as a colorless oil (84.1
22
mg, 68% yield). ½ꢀꢂ ¼ 8:4ꢁ (c ¼ 1:00, EtOH); 1H NMR
(270 MHz, CDCl3) ꢁD7.27–7.15 (m, 5H, Ph), 3.27 (d, 1H,
J ¼ 13:2 Hz, CH2Ph), 3.06–2.99 (m, 1H, 5-H), 2.65 (d, 1H,
J ¼ 13:2 Hz, CH2Ph), 2.70–2.58 (m, 1H, 5-H), 2.46 (s, 3H,
CH3), 2.06–1.93 (m, 1H, 3- or 4-H), 1.81–1.53 (m, 3H, 3-
t
and 4-H), 1.45 (s, 9H, Bu); 13C NMR (68 MHz, CDCl3) ꢁ
171.9, 138.0, 130.3, 127.6, 126.0, 80.9, 71.4, 54.4, 40.5,
35.5, 33.8, 28.4, 21.6; IR (film) 1714, 1160 cmꢃ1; Anal. calcd
for C17H25NO2: C, 74.14; H, 9.15; N, 5.09%. Found: C,
74.23; H, 9.39; N, 5.11%. The ee was determined to be
92% ee by chiral HPLC analysis of the amino alcohol 6a
which prepared by reduction of (R)-5a with LiAlH4 in THF
[Daicel CHIRALPAK AD-H, Hex/EtOH = 85:15, 0.50
mL/min, tR ¼ 15:7 min for the (S)-isomer and 18.4 min for
the (R)-isomer].
In summary, we demonstrated that the [1,2] Stevens rear-
rangement of the (S)-N-benzylic proline-derived ammonium
salts, when carried out under the solid–liquid biphasic condi-
tions, exhibits a remarkably high level of the N-to-C chirality
transmission to afford the corresponding ꢀ-arylmethyl-substitut-
ed proline derivatives in high enantio-purities. Further works on
other asymmetric Stevens rearrangements are underway.
This work was financially supported by the Uchida Energy
Science Promotion Foundation.
9
Recrystallization solvents: Hex/CH2Cl2 for 4b and 4d,
Hex/THF for 4c, and Hex/CH2Cl2/PhH for 4e.