Mechanistic considerations for the consecutive cyclization of 2,3-dibromopropylamine hydrobromide giving a strained molecule, 1-azabicyclo[1.1.0]butane

Article abstract of DOI:10.1248/cpb.52.89

The effective formation of 1-azabicyclo[1.1.0]butane (2) by treatment of 2,3-dibromopropylamine hydrobromide (1) with n-BuLi could be understood considering a rational reaction pathway via both transition states 10 and 19 based on the intramolecular Br...

Full text of DOI:10.1248/cpb.52.89

January 2004
Chem. Pharm. Bull. 52(1) 89—94 (2004)
89
Mechanistic Considerations for the Consecutive Cyclization of
2,3-Dibromopropylamine Hydrobromide Giving a Strained Molecule,
1-Azabicyclo[1.1.0]butane
a
a
a
b
,a
*
Kazuhiko HAYASHI, Yoshifumi IKEE, Satoru GOTO, Motoo SHIRO, and Yoshimitsu NAGAO
a Faculity of Pharmaceutical Sciences, The University of Tokushima; Sho-machi, Tokushima 770–8505, Japan: and
b Rigaku Corporation; 3–9–12 Matsubara-cho, Akishima, Tokyo 196–8666, Japan.
Received September 6, 2003; accepted October 8, 2003; published online October 9, 2003
The effective formation of 1-azabicyclo[1.1.0]butane (2) by treatment of 2,3-dibromopropylamine hydrobro-
mide (1) with n-BuLi could be understood considering a rational reaction pathway via both transition states 10
؉
…
and 19 based on the intramolecular Br Li coordination. A similar cyclization pathway starting from N-ben-
zyl-3-bromopropylamine hydrochloride (17) to afford N-benzylazetidine (18) could also be postulated on the
basis of a transition state 20 involving the intramolecular Br Li؉ coordination.
…
Key words azabicyclo[1.1.0]butane; lithium–bromine coordination; azetidene synthesis
Recently, we have developed an expeditious and good- in Table 1). Thus a lithium cation of the lithium amide gener-
yielding synthesis of 1-azabicyclo[1.1.0]butane (ABB) (2) ated by treatment of 1 with the lithium species should play an
bearing the high-strained bicyclic structure and established a essential role (vide infra) in the consecutive cyclization.
facile synthetic method for various 3-substituted azetidine
Subsequently, we investigated the first cyclization step
derivatives (3) from 2, as represented in Chart 1.1—3) The (i.e., path A toward aziridine 6 and/or path B toward azeti-
1
structure of 2 was comfirmed by H- and 13C-NMR analyses dine 7) in the consecutive cyclization reactions of a free
and its conversion to N-tosyl-3-chloroazetidine (4)4) (Fig. 1, amine 5,6) as shown in Chart 2. First, we inspected the X-ray
Table 1).1,3) The optimization of the structure was carried out crystallographic structure (Fig. 2) of 1, in which the relation-
by DFT molecular orbital computation, as shown in Fig. 1.1) ship between C2–Br and C3–Br adopts an approximate
Azetidine derivatives like 3 should be useful for syntheses of antiperiplanar arrangement (torsion angle: Br–C2–C3–Brϭ
new pendant molecules of b-lactam and quinolone antibi- 166.1°) due to the dipole–dipole and steric repulsion as
otics.1,5) As shown in Table 1, the desirable consecutive cy- anticipated.7) This Br–Br anti-conformation even in the mol-
clization of 2,3-dibromopropylamine hydrobromide (1) giv- ecule of the free amine 5 (Chart 2) seemed to be preferable,
1
ing ABB 2 proceeded only with the use of organolithium which was supported by the H-NMR analysis of a solution
compounds and lithium amides (entries 6—9 in Table 1).1) of 5 in THF-d8 at 23 °C based on the coupling constants
When other bases such as potassium, sodium, and magne- (Jϭ10.0, 5.4 Hz) between C2–H and C3–H2 (Ha and Hb in
sium species were employed, the reaction of 1 resulted in al- 5). Then, on the basis of the conformational analysis of 5 de-
most no yield of the product 2 (entries 1—5 and 10 in Table
Table 1. Conversion of 1 to 4 via 1-Azabicyclo[1.1.0]butane (2)
1). The same reaction with n-BuLi or LiNH2 in the presence
of a crown ether, 12-crown-4, trapping a lithium cation af-
forded no significant product, respectively (entries 11 and 12
Additive
(mol eq)
Yield (%)a)
Entry
Base
KOH
DBU
NaOMe
NaH
Temp.
Time
of 4
1
2
3
4
5
6
7
8
9
None
None
None
None
None
None
None
None
None
Reflux
Reflux
r.t.
r.t.
r.t.
r.t.
r.t.
Ϫ78 °C
Ϫ78 °C
Ϫ78 °C
Ϫ78 °C
1 h
1 h
2
NDc)
2
Chart 1
24 h
18 h
18 h
18 h
18 h
1 h
1 h
1 h
1 h
NDc)
NDc)
34
NaNH2
LiNH2
LDA
39
82
87
1
n-BuLib)
PhLib)
10
11
PhMgBrb)
n-BuLib)
None
12-Crown-4
(3.0)
12-Crown-4
(3.0)
1
12
LiNH2
Ϫ78 °C
1 h
NDc)
a) Determined by HPLC analysis. b) Quenched with 50% aq. KOH. c) Not de-
tected.
Fig. 1. Optimized Structure of 2 and Its 1H- and 13C-NMR Spectral Data
To whom correspondence should be addressed. e-mail: ynagao@ph2.tokushima-u.ac.jp
© 2004 Pharmaceutical Society of Japan

90
Vol. 52, No. 1
on the basis of the general aspect for the kinetic cyclization.8)
Thus, to clarify the first cyclization step, path A and/or
path B, a competitive cyclization reaction of N-benzyl-2,3-di-
bromopropylamine hydrochloride (12) giving only two types
of monocyclic products was examined. Treatment of 12 with
n-BuLi under the same reaction conditions (entry 8 in Table
1) as in the case of 1 furnished aziridine 139) in 14% yield, a
scarce amount of azetidine 14, and allylamine 1510) in 47%
yield together with recovery (33%) of 12, as shown in Chart
3 (HPLC analysis). The allylamine 15 was not derived from
13 or 14, respectively, under the same reaction conditions.
Compound 15 would be generated by the attack of n-BuLi on
a bromine atom of N-benzyl-2,3-dibromopropylamine in
competition with fairly difficult H-abstraction of the sec-
ondary NH group, followed by simultaneous exchange of Br
for Li and then elimination of lithium bromide. The aziridine
13 proved not to be obtained from the azetidine 14 (Chart 3).
Based on these results, it was strongly suggested that the first
cyclization step in the consecutive cyclization of 5 is almost
certainly path A. This outcome is consistent with the earlier
kinetic studies of the cyclization of w-(bromoalkyl)phenyl-
scribed above, we considered the first cyclization step in
terms of both “acyclic transition state” and “cyclic transition
state,” as shown in Fig. 3. In the former, two kinds of New-
man projection states 8 (toward aziridine 6) and 9 (toward
azetidine 7) may be plausible. In the latter, the five-mem-
bered cyclic state 10 (toward 6) and six-membered cyclic
state 11 (toward 7) consisting of the Br Li coordination
would also be plausible. Because of the greater instability of
state 9 bearing two types of severe gauche repulsion than
state 8 bearing one severe gauche repulsion (Fig. 3), path A
toward aziridine 6 is predominant over path B toward azeti-
dine 7 in the first cyclization step of 1. In the assumed cyclic
transition state, the five-membered cyclic state 10 must be ki-
netically predominant over the six-membered cyclic state 11
ϩ
…
Chart 2
Fig. 2. Computer-Generated Drawing Derived from the X-Ray Coordi-
nates of Compound 1
Fig. 3. Hypothetic Transition States in Path A and Path B
Chart 3

January 2004
91
amines11) or w-(bromoalkyl)dimethylamines12) under totally
different reaction conditions from our case. Namely, it was
demonstrated that the cyclization rate toward aziridine is ca.
200 or 70 times that toward azetidine.11,12)
class of amines one of the most difficult to prepare and the
cyclization of 3-halogenopropylamines can be complicated
by accompanying elimination, fragmentation, or dimeriza-
tion.14—16,17) The previous literature methods generally re-
quired drastic reaction conditions.16,18,19) With the back-
ground described above, the cyclization procedure using
n-BuLi should be more effective than the previous meth-
ods.14—16,20,21) Interestingly, we again observed that the cy-
clization of 17 with n-BuLi did not proceed in the presence
of 12-crown-4, but instead compound 17 was quantitatively
recovered (entry 4 in Table 2).
We are now interested in the cyclization of N-benzyl-3-
bromopropylamine hydrochloride (17) readily obtained from
3-bromopropylamine hydrochloride (16) to exploit our
method for exclusive azetidine formation (Chart 4). Treat-
ment of 17 with n-BuLi in THF at Ϫ78 °C for 1 h to give N-
benzylazetidine (18)13) in 73% yield (HPLC analysis) or in
64% isolation yield together with 22% or 14% recovery each
of 17 (entries 1 and 2 in Table 2). Surprisingly, this cycliza-
tion reaction proceeded very rapidly (7 min) to give 18 in
67% yield (HPLC analysis) with 26% recovery of 1714)
(entry 3 in Table 2). The ring strain of azetidine makes this
The fact that no n-BuLi-promoted cyclization reaction of 1
and 17 occurred in the presence of 12-crown-4 suggests that
the attack of the naked amide anionic species involving a
transition state 8 to the Br-attached carbon atom should be
ruled out. Hence the five-membered cyclic transition state 10
ϩ
…
consisting of the Br Li coordination must be the most
possible in the first cyclization step of the facile consecutive
cyclization of 5 with n-BuLi, giving the strained molecule
ϩ
…
ABB 2. The significance of such a halogen Li coordina-
tion had been suggested in the stereoselective alkylation of
the lithium enolate with alkyl halides by Meyers et al.22)
A
similar suggestion has been made in an asymmetric alkyla-
tion reaction of chiral enamines by Ando et al.23) Further, 1H-
ϩ
…
NMR analysis of the intermolecular Br Li coordination
between n-dodecylbromide and lithium diisopropylamide
1
(LDA) was examined, as shown in Fig. 4. On the H-NMR
spectrum chart of a solution of n-dodecylbromide alone in
THF-d8 at 23 °C, one set of triplet signals due to –CH2–Br
was observed at 3.40 ppm (A in Fig. 4), whereas on that of a
solution of n-dodecylbromide and 1 mol eq of LDA under the
same conditions, two sets of triplet signals probably due to
–CH2–Br and –CH2–Br Li N(i-Pr)2 could be observed in
the presence of a free n-dodecylbromide and a weakly n-do-
decylbromide-coordinated LDA at equilibrium (B in Fig.
4).24) Tentative use of 3 mol eq of LDA showed one set of
triplet signals, presumably due to an n-dodecylbromide-coor-
dinated LDA alone, and two sets of triplet signals, in which
each height of the triplet signals at the higher magnetic field
is considerably higher than that of the corresponding triplet
Chart 4
Table 2. Cyclization of N-Benzyl-3-bromopropylamine (17) with n-BuLi
ϩ
Ϫ
…
Additive
(mol eq)
Yield (%)a) of 18
Entry
Time
[Recovery (%)a) of 17b)]
1
2
3
4
None
None
None
1 h
1 h
7 min
1 h
73 [22]
64c) [14]c)
67 [26]
12-Crown-4 (2.0)
0 [100]
a) Determined by HPLC analysis. b) Compound 5 was recovered as free amine.
c) Isolation yield.
Fig. 4. Selected –CH2–Br Peaks on the 1H-NMR Chart (400 MHz) of n-Dodecylbromide in the Presence or Absence of LDA or n-BuLi in THF-d8
Chart 5

92
Vol. 52, No. 1
Chart 6
hexane–AcOEt, 3/1) to give 4 (3.38 g, 83%) as colorless needles, mp 110 °C
(lit.,4) 106 °C). 1H-NMR (300 MHz, CDCl3) d: 2.47 (3H, s), 3.84 (2H, dd,
Jϭ5.6, 8.3 Hz), 4.23 (2H, t, Jϭ8.3 Hz), 4.3—4.5 (1H, m), 7.39 (2H, d,
signals at the lower magnetic field, were observed in the pres-
ence of 0.5 mol eq of LDA. The same phenomenon was ob-
served in the presence of n-dodesylbromide and 1 mol eq of
Jϭ8.2 Hz), 7.73 (2H, d, Jϭ8.2 Hz). IR (neat) 1596, 1343, 1162, 681 cmϪ1
.
ϩ
…
EI-MS calcd for C10H12NO2ClS MW245.0277, found m/z 245.0266 [Mϩ].
n-BuLi (C in Fig. 4). Intramolecular Br Li coordination in
state 10 seems to be stronger than that in the intermolecular Anal. Calcd for C10H12NO2ClS: C, 48.97; H, 4.94; N, 5.71. Found: C, 48.46;
ϩ
…
H, 4.86; N, 5.68.
Br Li coordination between n-dodecylbromide and LDA.
Optimization of the Structure of 1-Azabicyclo[1.1.0]butane (2) DFT
molecular orbital computation was performed with the GAUSSIAN 98 pro-
gram obtained from Gaussian Inc. (Pittsburgh).25) We used the B3LYP ap-
proach based on Becke’s three-parameter nonlocal-exchange function after
modification by using the nonlocal correlation of Lee, Yang, and Parr
(B3LYP).25) The structure was optimized using the 6-31G* basis set.
1H- and 13C-NMR Analysis of 1-Azabicyclo[1.1.0]butane (2) To a
suspension of 2,3-dibromopropylamine hydrobromide (1) (1.00 g,
3.36 mmol) in THF-d8 (10 ml) was added LiNH2 (0.23 g, 10.1 mmol) at
room temperature under an N2 atmosphere, and the reaction mixture was
stirred at room temperature for 18 h. Then the whole reaction mixture was
distilled at atmospheric pressure and the resulting distillate was analyzed by
Based on all the related experiments and considerations
described above, we propose a rational reaction pathway for
the consecutive cyclization of 1 with n-BuLi involving the
five-membered cyclic transition states 10 and 19 based on the
intramolecular Br Li coordination, as shown in Chart 5. A
similar reaction pathway for the cyclization of 17 via the six-
membered cyclic transition state 20 can be postulated as
shown in Chart 6.
ϩ
…
In conclusion, we have established an efficient and facile
method for the synthesis of small strained molecules such as
1-azabicyclo[1.1.0]butane and N-benzylazetidine using n-
1
1H- and 13C-NMR at 23 °C. H-NMR (400 MHz, THF-d8) d: 0.85 (2H, m),
2.07 (2H, m), 2.29 (1H, m). 13C-NMR (100 MHz, THF-d8) d: 17.2, 51.2.
The reported data4,26): 1H-NMR (100 MHz, CCl4) d: 0.93 (2H, m), 2.13 (2H,
m), 2.33 (1H, m). 13C-NMR (CDCl3) d: 18.1, 51.1.
BuLi and postulated a rational mechanistic pathway for their
ϩ
…
cyclization reactions involving the intramolecular Br Li
X-Ray Crystallographic Analysis of 2,3-Dibromopropylamine Hydro-
bromide (1) The crystalline compound 1 was prepared by bromination of
allylamine, as reported previously,2) and then recrystallized from MeOH for
coordination.
Experimental
All melting points were determined on a Yanaco micro melting point ap- X-ray crystallographic analysis. The measurement was done on a Rigaku
paratus and are uncorrected. IR spectra were obtained on a JASCO FT-IR RAXIS-RAPID Imaging Plate diffractometer with graphite monochromated
1
420 Fourier transform spectrometer. H-NMR (300, 400 MHz) spectra were MoKa radiation. The data were processed using the PROCESS-AUTO pro-
recorded on a JEOL JNM-AL300 or JEOL JNM-GSX400 spectrometer. 13C- gram package. The linear absorption coefficient, m, for MoKa radiation is
NMR (75, 100 MHz) spectra were recorded on a JEOL JNM-AL300 or 1.0 cmϪ1. A symmetry-related absorption correction using the program AB-
JEOL JNM-GSX400 spectrometer. Chemical shifts are given in d values SCOR was applied.27) The data were corrected for Lorentz and polarization
(ppm) using tetramethylsilane (TMS) as an internal standard. Mass spectra effects. The structure was solved by direct methods and expanded using
(MS) were recorded on a JEOL JMS SX-102A spectrometer. Elementary Fourier techniques.28,29) The nonhydrogen atoms were refined anisotropi-
combustion analyses were performed on a Yanaco CHN CORDER MT-5.
cally. Hydrogen atoms were included but not refined. Natural atom scatter-
HPLC analyses for the determination of yield were carried out on a SHI- ing factors were taken from Cromer and Waber.30) The values for the mass
MADZU LC-6A and UV detector of 254 nm. All reactions were monitored attenuation coefficients are those of Creagh and Hubbell.31) All calculations
by TLC employing 0.25-mm silica gel plates (Merck 5715; 60 F254). Prepar- were performed using the teXsan crystallographic software package.32)
ative TLC (PTLC) was performed on 0.5-mm silica gel plates (Merck 5744;
60 F254). Column chromatography was carried out on silica gel (Katayama crystal, monoclinic, space group P21/c (#14), aϭ7.973(3) Å, bϭ8.791(4) Å,
calcdϭ2.525 g/cm3,
Crystallographic Data for 1: C3H8NBr3, MWϭ297.81, colorless needle
Chemical K070; 70—300 mesh, Merck 9385; 230—400 mesh). All solvents cϭ11.225(5) Å, bϭ95.30°; Vϭ783.4(6) Å3; Zϭ4,
D
were distilled prior to use. All reagents were used as purchased.
Rϭ0.044, Rwϭ0.076; GOFϭ0.969.
General Procedure of HPLC Analysis (Table 1) The reactions were
1H-NMR Analysis of 2,3-Dibromopropylamine (5) The ether solution
performed by using a suspension of 2,3-dibromopropylamine hydrobromide of 5 was obtained from 2,3-dibromopropylamine hydrobromide (1) by the
(1) (5.00 g, 16.8 mmol) and each base (50.4 mmol) in THF (50 ml) under the method reported in the previous paper6) and evaporated in vacuo to afford
1
1
conditions shown in Table 1. After the reaction was completed, the solution the crude 5 which was analyzed by H-NMR at 23 °C. H-NMR (300 MHz,
was distilled at atmospheric pressure. In the cases of entries 7—10, after the THF-d8) d: 1.60 (2H, br s), 3.03 (1H, dd, Jϭ5.4, 14.2 Hz), 3.14 (1H, dd,
reaction was completed, the reaction mixture was quenched with 50% KOH Jϭ4.2, 14.2 Hz), 3.84 (1H, dd, Jϭ5.4, 10.0 Hz), 3.91 (1H, t, Jϭ10.0 Hz),
(10 ml), and the whole mixture was distilled at atmospheric pressure. To the 4.2—4.3 (1H, m).
resulting distillate was added p-toluenesulfonyl chloride (3.20 g, 16.8 mmol)
N-Benzyl-2,3-dibromopropylamine Hydrobromide (12) A suspension
at 0 °C, and the mixture was stirred at room temperature for 18 h to afford a of 2,3-dibromopropylamine hydrobromide (1) (4.47 g, 15 mmol), benzalde-
crude reaction solution, which was analyzed by HPLC under the following hyde (1.59 g, 15 mmol), Et3N (2.09 ml, 15 mmol), and MgSO4 (1.8 g) in
conditions. Mobil phase: phosphate buffer (pHϭ7.0, 0.05 mol/l)/MeCNϭ CH2Cl2 (60 ml) was refluxed for 1 h under an N2 atmosphere. The reaction
5/5, flow rate: 1.0 ml/min, column: TOSOH TSK-GEL 80Ts (4.6 mm f* 25
cm). N-Tosyl-3-chloroazetidine (4) (retention timeϭ9 min) was used as the was added to the residue and the precipitate was filtered off. The filtrate was
standard sample.
evaporated in vacuo to afford a crude imine product, whose structure was
Preparation of N-Tosyl-3-chloroazetidine (4)4) To a suspension of 2,3- confirmed by 1H-NMR analysis.1H-NMR (300 MHz, CDCl3) d: 3.8—4.0
dibromopropylamine hydrobromide (1) (5.00 g, 16.8 mmol) in THF (50 ml)
(2H, m), 4.0—4.2 (2H, m), 4.4—4.6 (1H, m), 7.3—7.5 (3H, m), 7.7—7.9
mixture was filtered, and the filtrate was concentrated in vacuo. Then Et2O
was added dropwise n-BuLi (1.61 mol/l, n-hexane solution, 31.3 ml, (2H, m), 8.30 (1H, s). The resultant crude imine was dissolved in MeOH
50.4 mmol) at Ϫ78 °C under an N2 atmosphere, and the reaction mixture (25 ml) and cooled to 0 °C. After addition of NaBH4 (567 mg, 15 mmol) at
was stirred at Ϫ78 °C for 1 h. Then the reaction mixture was quenched with 0 °C, the mixture was stirred at 0 °C for 0.5 h. Then the reaction mixture was
50% KOH (10 ml), and the whole mixture was distilled at atmospheric pres- quenched with 3.5% HCl at 0 °C followed by concentration in vacuo to re-
sure. The distillate was dried over K2CO3 and filtered. To the resulting fil- move MeOH. The concentrated solution was washed with Et2O, adjusted to
trate was added p-toluenesulfonyl chloride (3.20 g, 16.8 mmol) at 0 °C. Then pH 8—9 with NaHCO3, and extracted with Et2O. The Et2O layer was ex-
the mixture was stirred at room temperature for 18 h and evaporated in tracted with 3.5% HCl, and the water layer was evaporated in vacuo. The
vacuo. The residue was purified by column chromatography (eluent n- solid residue was recrystallized from MeOH–Et2O to afford compound 12

January 2004
93
(1.53 g, 30%) as colorless prisms, mp 168—169 °C. 1H-NMR (300 MHz, plates, mp 186—187 °C. 1H-NMR (400 MHz, CD3OD) d: 2.2—2.3 (2H, m),
CD3OD) d: 3.45 (1H, dd, Jϭ9.9, 13.9 Hz), 3.78 (1H, dd, Jϭ3.3, 13.9 Hz), 3.26 (2H, t, Jϭ7.8 Hz), 3.54 (2H, t, Jϭ6.4 Hz), 4.87 (2H, s), 7.4—7.6 (5H,
3.84 (1H, dd, Jϭ8.3, 11.2 Hz), 3.99 (1H, dd, Jϭ4.6, 11.2 Hz), 4.33 (2H, s), m). IR (KBr) 2941, 2799, 1445, 1245, 745, 697 cmϪ1. FAB-MS calcd for
4.5—4.7 (1H, m), 7.4—7.6 (5H, m). IR (KBr) 2691, 2581, 1569, 1459,
1413, 1215, 939, 758, 704, 581 cmϪ1. FAB-MS calcd for C10H14NBr2
MW305.9493, found m/z 305.9509 [(Mϩ1)ϩ]. Anal. Calcd for
C10H15NBrCl: C, 34.97; H, 4.11; N, 4.08. Found: C, 34.76; H, 3.99; N, 4.01.
C10H15NBr MW 228.0388, found m/z 228.0387 [(Mϩ1)ϩ]. Anal. Calcd for
C10H15NBrCl: C, 45.39; H, 5.71; N, 5.29. Found: C, 45.78; H, 5.71; N, 5.37.
Preparation of N-Benzylazetidine (18) (Table 2, Entry 2) To a sus-
pension of N-benzyl-3-bromopropylamine hydrochloride (17) (396 mg,
Preparation of N-Benzyl-2-(bromomethyl)aziridine (13) and N-Benzy- 1.50 mmol) in THF (4.5 ml) was added n-BuLi (1.56 mol/l, n-hexane solu-
lallylamine (15) To a suspension of N-benzyl-2,3-dibromopropylamine tion, 1.92 ml, 3.0 mmol) at Ϫ78 °C under N2 atmosphere. After being stirred
hydrobromide (12) (343 mg, 1.0 mmol) in THF (3.0 ml) was added n-BuLi for 1 h at Ϫ78 °C, the reaction mixture was quenched with a phosphate
(1.56 mol/l, n-hexane solution, 1.28 ml, 2.0 mmol) at Ϫ78 °C under an N2 at- buffer solution (pHϭ6.0, 0.05 mol/l, 10 ml) and then extracted with AcOEt.
mosphere. After being stirred for 1 h at Ϫ78 °C, the reaction mixture was The AcOEt extract was washed with brine, dried over MgSO4, and evapo-
quenched with a phosphate buffer solution (pHϭ6.0, 0.05 mol/l, 10 ml) and rated in vacuo. The residue was purified by column chromatography (eluent
extracted with AcOEt. The AcOEt extract was washed with brine, dried over CHCl3–MeOH, 9/1) to give 18 (142 mg, 64%), and the free amine of 17
MgSO4, and evaporated in vacuo. The residue was purified by PTLC (eluent (46 mg, 14%) as each colorless oil.
CHCl3–MeOH, 9/1) to give the free amine of 12 (39 mg, 15%), 13 (72 mg,
39%), and 15 (20 mg, 19%).
18: Colorless oil, bp 70—75 °C/5 mmHg (lit.,13) bp 65—70 °C/4 mmHg).
1H-NMR (400 MHz, CDCl3) d: 2.06 (2H, quintet, Jϭ7.1 Hz), 3.20 (4H, t,
Free Amine of 12: Colorless oil. 1H-NMR (400 MHz, CDCl3) d: 1.82 Jϭ7.1 Hz), 3.55 (2H, s), 7.2—7.4 (5H, m). 13C-NMR (75 MHz, CDCl3) d:
(1H, br s), 3.02 (1H, dd, Jϭ6.8, 13.4 Hz), 3.20 (1H, dd, Jϭ3.4, 13.4 Hz), 17.7, 55.1, 63.9, 126.9, 128.3, 128.4, 138.3. IR (neat) 2957, 2818, 1493,
3.7—3.9 (4H, m), 4.2—4.4 (1H, m), 7.2—7.4 (5H, m). 13C-NMR (75 MHz, 1453, 1361, 1187, 733, 698 cmϪ1. EI-MS calcd for C10H14NBr MW
CDCl3) d: 34.2, 53.1, 53.3, 53.8, 127.7, 128.6, 129.0, 140.3. IR (neat) 3324,
3027, 2838, 1494, 1454, 1141, 738, 698, 570 cmϪ1. EI-MS calcd for
C10H13NBr2 MW304.9415, found m/z 304.9420 [Mϩ].
148.1126, found m/z 148.1108 [(Mϩ1)ϩ].
Free Amine of 17: Colorless oil.21) 1H-NMR (300 MHz, CDCl3) d: 2.07
(2H, m), 2.79 (2H, t, Jϭ6.9 Hz), 3.25 (1H, br s), 3.48 (2H t, Jϭ6.6 Hz) 3.81
(2H, s), 7.2—7.4 (5H, m).
13: Colorless oil.9) 1H-NMR (300 MHz, CDCl3) d: 1.54 (1H, d,
Jϭ5.9 Hz), 1.71 (1H, d, Jϭ3.3 Hz), 1.8—1.9 (1H, m), 3.20 (1H, dd, Jϭ6.3,
General Procedure for HPLC Analysis (Table 2, Entries 1, 3, 4)
Each reaction was carried out using a suspension of 17 (265 mg, 1.0 mmol)
10.2 Hz), 3.25 (1H, dd, Jϭ6.9, 10.2 Hz), 3.30 (1H, d, Jϭ13.2 Hz), 3.49 (1H,
1
d, Jϭ13.2 Hz), 7.1—7.3 (5H, m). The H-NMR data of 13 were identical to and n-BuLi (1.56 mol/l, n-hexane solution, 1.28 ml, 2.0 mmol) in THF (5 ml)
those of the known compound.9)
15: Colorless oil.10) 1H-NMR (300 MHz, CDCl3) d: 1.45 (1H, br s), 3.28 n-BuLi (1.56 mol/l, n-hexane solution, 1.28 ml, 2.0 mmol) was added to a
(2H, dt, Jϭ5.9, 1.3 Hz), 3.79 (2H, s), 5.1—5.2 (2H, m), 5.8—6.0 (1H, m), solution of 12-crown-4 (352 mg, 2.0 mmol) in THF (3 ml) at Ϫ78 °C, and
7.2—7.4 (5H, m). The 1H-NMR data of 15 were identical to those of the then the mixture was added to a suspension of 17 (265 mg, 1.0 mmol) in
known compound.10)
THF (2 ml). The whole mixture was stirred at Ϫ78 °C for 1 h. Each reaction
under the conditions shown in Table 2 (entries 1, 3). In the case of entry 4,
Preparation of N-Benzyl-3-bromoazetidine (14) To a solution of N- mixture was worked up in a similar manner to the case of preparative syn-
benzyl-3-hydroxyazetidine33) (163 mg, 1.0 mmol) and CBr4 (445 mg, thesis of 18. The corresponding resulting solution was analyzed by HPLC
1.34 mmol) in CH2Cl2 (1.0 ml) was added PPh3 (393 mg, 1.5 mmol) at 0 °C.
under the following conditions. Mobil phase: phosphate buffer (pHϭ6.0,
After being stirred for 0.5 h at 0 °C, the reaction mixture was evaporated in 0.05 mol/l)/MeCNϭ6/4, flow rate: 1.0 ml/min, column: Wakopack Fluofix
vacuo to give a residue. The residue was purified by column chromatogra- (4.6 mm f* 25 cm). The isolated compounds 18 (retention timeϭ10 min)
phy (eluent n-hexane–AcOEt, 2/8) to afford 14 (80 mg, 35%) as a colorless and 17 (retention timeϭ14 min) were used as the standard sample.
1
؉
oil. 1H-NMR (400 MHz, CDCl3) d: 3.3—3.4 (2H, m), 3.62 (2H, s), 3.7—3.9
H-NMR Analysis of the Intermolecular Br Li Coordination (Fig.
(2H, m), 4.3—4.5 (1H, m), 7.2—7.3 (5H, m). 13C-NMR (75 MHz, CDCl3) 4) (A) To THF-d8 (1.0 ml) were added n-hexane (0.21 ml) and n-dodecyl-
d: 34.3, 63.3, 64.6, 126.9, 128.0, 128.1, 137.1. IR (neat) 3027, 2953, 2832, bromide (0.081 ml, 0.34 mmol) at 0 °C. The mixture was stirred at room
…
1495, 1453, 1362, 1234, 1178, 758, 712, 696 cmϪ1. EI-MS calcd for temperature for 10 min, and the resulting solution was analyzed by 1H-NMR
C10H12NBr MW225.0153, found m/z 225.0143 [Mϩ].
(400 Mz) at 23 °C.
General Procedure for HPLC Analysis (Chart 3) To a suspension of
(B) To a solution of diisopropylamine (0.047 ml, 0.34 mmol) in THF-d8
12 (172 mg, 0.5 mmol) in THF (1.5 ml) was added n-BuLi (1.56 mol/l, n- (1.0 ml) was added n-BuLi (1.58 mol/l, n-hexane solution, 0.213 ml,
hexane solution, 0.64 ml, 1.0 mmol) under an N2 atmosphere at Ϫ78 °C and 0.34 mmol) at 0 °C. Then n-dodecylbromide (0.081 ml, 0.34 mmol) was
the mixture was stirred at Ϫ78 °C for 1 h. In the reaction of 13 or 14 with n- added to the mixture at 0 °C and stirred at room temperature for 10 min. The
BuLi, a solution of 13 (20 mg, 0.089 mmol) or 14 (20 mg, 0.089 mmol) in
THF (0.3 ml), and n-BuLi (1.56 mol/l, n-hexane solution, 0.057 ml,
resulting solution was analyzed by 1H-NMR (400 Mz) at 23 °C.
(C) To a solution of n-dodecylbromide (0.081 ml, 0.34 mmol) in THF-d8
0.089 mmol) was employed. The resulting reaction mixture was worked up (1.0 ml) was added n-BuLi (1.58 mol/l, n-hexane solution, 0.213 ml,
in a similar manner to the preparative syntheses of 13 and 15 to furnish the 0.34 mmol) at 0 °C. Then the mixture was stirred at room temperature for
crude reaction solution, which was subjected to HPLC analysis. Mobil 10 min, and the resulting solution was analyzed by 1H-NMR (400 Mz) at
phase: phosphate buffer (pHϭ6.0, 0.05 mol/l)/MeCNϭ6/4, flow rate:
2.0 ml/min, column: Wakopack Fluofix (4.6 mm f* 25 cm). The isolated
compounds 12 (retention timeϭ13 min), 13 (retention timeϭ8.5 min), 14
23 °C.
Acknowledgments This work was partially supported by Grants-in-Aid
(retention timeϭ10.5 min), and 15 (retention timeϭ4.5 min) were used as for Scientific Research on Priority Areas (A), (2) (No. 14044075) from the
each corresponding standard sample. Ministry of Education, Culture, Sports, Science and Technology, Japan and
N-Benzyl-3-bromopropylamine Hydrobromide (17) A suspension of by a grant for Scientific Research (B) (2) (No. 14370723) from the Japan
3-bromopropylamine hydrochloride (16) (3.29 g, 15 mmol), benzaldehyde Society for the Promotion of Science.
(1.59 g, 15 mmol), Et3N (2.09 ml, 15 mmol), and MgSO4 (1.8 g) in CH2Cl2
(60 ml) was refluxed under an N2 atmosphere for 1.5 h. The reaction mixture References and Notes
was filtered, and the filtrate was concentrated in vacuo. Then Et2O was
added to the residue and the precipitate was filtered off. The filtrate was
evaporated in vacuo to afford a crude imine product, of which the structure
1) Hayashi K., Sato C., Hiki S., Kumagai T., Tamai S., Abe T., Nagao Y.,
Tetrahedron Lett., 40, 3761—3764 (1999).
2) Hayashi K., Kumagai T., Nagao Y., Heterocycles, 53, 447—452
(2000).
3) Hayashi K., Hiki S., Kumagai T., Nagao Y., Heterocycles, 56, 433—
442 (2002).
4) Funke W., Chem. Ber., 102, 3148—3150 (1969).
5) Tomioka H., Sato K., Kajitani H., Akaki T., Shishido S., Antimicrob.
Agents Chemother., 44, 283—286 (2000).
1
1
was confirmed by H-NMR. H-NMR (300 MHz, CDCl3) d: 2.25 (2H, m),
3.49 (2H, t, Jϭ6.3 Hz), 3.74 (2H, t, Jϭ6.6 Hz), 7.3—7.5 (3H, m), 7.5—7.8
(2H, m), 8.32 (1H, s). The resulting crude imine was dissolved in MeOH
(25 ml) and cooled to 0 °C. After the addition of NaBH4 (567 mg, 15 mmol)
at 0 °C, the mixture was stirred at 0 °C for 0.5 h. Then the reaction mixture
was quenched with 3.5% HCl at 0 °C and concentrated in vacuo to remove
MeOH. The concentrated solution was washed with Et2O, adjusted to pH
8—9 with NaHCO3, and extracted with Et2O. The Et2O layer was extracted
with 3.5% HCl, and the water layer was evaporated in vacuo. The solid
residue was recrystallized from EtOH to afford 17 (3.14 g, 79%) as colorless
6) Gensler W. J., J. Am. Chem. Soc., 69, 1966—1968 (1947).
7) Barili P. L., Bellucci G., Ingrosso G., Marioni F., Morelli I., Tetrahe-
dron, 28, 4583—4589 (1972).
8) Galli C., Illuminati G., Mandolini L., Tamborra P., J. Am. Chem. Soc.,

94
Vol. 52, No. 1
99, 2591—2597 (1978).
9) De Kimpe N., De Smaele D., Sakonyi Z., J. Org. Chem., 62, 2453—
23) Ando K., Takemasa Y., Tomioka K., Koga K., Tetrahedron, 49, 1579—
1588 (1993).
2457 (1997).
24) Feigel M., Kessler H., Chem. Ber., 111, 1659—1669 (1978).
10) Tehrani K. A., NguyenVan T., Karikomi M., Rottiers M., De Kimpe 25) Kohn W., Becke A. D., Parr R. G., J. Phys. Chem., 100, 12974—12980
N., Tetrahedron, 58, 7145—7152 (2002).
11) Bird R., Knipe A. C., Stirling C. J. M., J. Chem. Soc. Perkin Trans. 2,
1973, 1215—1220 (1973).
12) DeTar D. F., Brooks W., Jr., J. Org. Chem., 43, 2245—2248 (1978).
13) Lai G., Synth. Commun., 31, 565—568 (2001).
14) De Kimpe N., De Smaele D., Tetrahedron Lett., 43, 8023—8026
(1994).
(1996).
26) Bartnik R., Cal D., Synth. Commun., 28, 3949—3954 (1998).
27) ABSCOR: Higashi T., Program for Absorption Correction, Rigaku
Corporation, Tokyo Japan, 1995.
28) SIR97: Altomare A., Burla M. C., Camalli M., Cascarano G. L., Gia-
covazzo C., Guagliardi A., Moliterni A. G. G., Polidori G., Spagna R.,
J. Appl. Crystallogr., 32, 115—119 (1999).
15) Sammes P. G., Smith S. J., J. Chem. Soc. Perkin Trans. 1, 1984, 29) DIRDIF94: Beurskens P. T., Admiraal G., Beurkens G., Bosman W. P.,
2415—2419 (1984).
16) Vaughan W. R., Klonowski R. S., McElihinney R. S., Millward B. B.,
J. Org. Chem., 26, 138—144 (1961).
de Gelder R., Israel R., Smits J. M. M., The DIRDIF94 program sys-
tem, Technical Report of the Crystallography Laboratory, University
of Nijmegen, The Netherlands, 1994.
17) Freundlich H., Kroepelin H., Z. Physik, Chem., 122, 39—48 (1926).
18) Cromwell N. H., Phillps B., Chem. Rev., 79, 331—358 (1979).
30) Cromer D. T., Waber J. T., “International Tables for X-Ray Crystallog-
raphy,” Vol. IV. Kynoch Press, Birmingham, 1974, pp. 72—98.
19) Davies D. E., Storr R. C., “Comprehensive Heterocyclic Chemistry,” 31) Creagh D. C., Hubbell J. H., “International Table for X-Ray Crystal-
Vol. 7, ed. by Lwowski W., Pergamon Press, Oxford, 1984, pp. 237—
284.
lography,” Vol. C, ed. by Wilson A. J. C., Kluwer Academic Publish-
ers, Boston, 1992, pp. 200—206.
20) Moore J. A., Ayers R. S., “Small Ring Heterocycles, Part 2,” ed. by 32) teXsan: Crystal Structure Analysis Package, Molecular Structure Cor-
Hassner A., John Wiley & Sons, New York, 1983, pp. 1—217.
21) Wadsworth D. H., J. Org. Chem., 32, 1184—1187 (1967).
22) Meyers A. I., Knaus G., Kamata K., Ford M. E., J. Am. Chem. Soc.,
97, 567—576 (1975).
poration (1985, 1999).
33) Higgins R. H., Eaton Q. L., Worth L., Jr., Peterson M. V., J. Hetero-
cyclic Chem., 24, 255—259 (1987).