Reduction and reductive coupling of imines by Sm(II)-based reagents

Article abstract of DOI:10.1016/j.tet.2003.06.004

The reductive coupling of aldimines and ketimines by a series of Sm(II)-based reagents (SmI₂, SmI₂-HMPA, SmBr₂, Sm{N[Si(CH₃)₃]₂}₂, and SmI ₂/triethylamine/water) were examined. In general, aldimines and ketimines were efficiently reduced or coupled using reductants that are more powerful than SmI₂, and the use of Sm{N[Si(CH₃) ₃]₂}₂ led to higher diastereoselectivities in reductive coupling reactions. Surprisingly, only the combination of SmI ₂/triethylamine/water was capable of reducing and coupling para-substituted benzaldimines and coupling ketimines.

Full text of DOI:10.1016/j.tet.2003.06.004

Tetrahedron 59 (2003) 10397–10402
Reduction and reductive coupling of imines
by Sm(II)-based reagents
a
Myeongseob Kim, Brian W. Knettle, Anders Dahlen, Goran Hilmersson
a
b
b
,
*
´
¨
and Robert A. Flowers, IIa
,
*
aDepartment of Chemistry and Biochemistry, Texas Tech University, Box 41061, Lubbock, TX 79409-1061, USA
¨ ¨
Department of Chemistry, Goteborg University, SE-412 96 Goteborg, Sweden
b
Received 21 April 2003; accepted 11 June 2003
Abstract—The reductive coupling of aldimines and ketimines by a series of Sm(II)-based reagents (SmI2, SmI2–HMPA, SmBr2,
Sm{N[Si(CH3)3]2}2, and SmI2/triethylamine/water) were examined. In general, aldimines and ketimines were efficiently reduced or coupled
using reductants that are more powerful than SmI2, and the use of Sm{N[Si(CH3)3]2}2 led to higher diastereoselectivities in reductive
coupling reactions. Surprisingly, only the combination of SmI2/triethylamine/water was capable of reducing and coupling para-substituted
benzaldimines and coupling ketimines.
q 2003 Elsevier Ltd. All rights reserved.
1. Introduction
some of the more forcing conditions required for imine
reduction by SmI2, nitrones (iminium ion equivalents) have
been used to carry out cross coupling reactions between
aldehydes or ketones13 and a,b-unsaturated esters and
amides.14 In both reports, nitrones were efficiently reduced
under mild conditions with SmI2.
Reductants based on samarium have widespread applica-
bility in reductions and reductive couplings of numerous
functional groups including alkyl and aryl halides, alde-
hydes, epoxides, and a,b-unsaturated ketones, esters, and
amides.1–4 In many SmI2 mediated reactions, HMPA or
other additives are required for successful reduction
of alkyl bromides, chlorides and ketones.5 While SmI2 or
SmI2-additive combinations have been enormously useful
for the reduction and reductive coupling of several
functional groups, relatively little work has been carried
out using this reagent in the reduction of imines.
Recently, we have been examining two alternative
approaches to facilitate the reaction of functional groups
typically resistant to reduction by single electron
transfer from Sm(II). The first approach utilizes
Sm{N[Si(CH3)3]2}2. This reductant was first prepared by
Evans15 and has been found to reduce ketones at a much
faster rate than SmI2 or SmI2–HMPA.16 The second
approach utilizes the combination of SmI2 with amines
and H2O to drive precipitation of Sm(III) and I2. The
combination of SmI2/amines/H2O also reduces conjugated
double bonds17 and ketones.18 Initial results indicate that
both systems should be efficient reductants for imines.19
Herein, we communicate the results for reduction of an
iminium salt, aldimines, and ketimines under mild, room
temperature conditions. Comparisons of these reagents with
SmI2 (in the case of the iminium salt) and SmBr2 (in the case
of aldimines and ketimines) are presented.
The first reports of SmI2 mediated reduction of aldimines
were carried out independently by Enholm6 and Imamoto7
in 1990. In both reports, reductions and reductive coupling
reactions required prolonged reaction times or elevated
temperatures. More recently, Namy has reported that the
use of catalytic amounts of NiI2 with SmI2 facilitates the
reductive coupling of aldimines.8 Although the diastereo-
selectivity is relatively low, the reactions proceed under
mild conditions and tolerate a wide variety of phenyl and
N-substitution. Another recent report utilized a combination
of SmI2 and Yb(OTf)3 to reductively couple aldimines to the
corresponding 1,2-diamines.9 Other Sm-based reductants
including [Sm(OTf)2(DME)2],10 SmBr2,11 and Sm metal
12
2. Results
containing catalytic amounts of I2 have been used to
reduce imines with some success as well. To circumvent
2.1. Reduction of iminium salt by Sm(II) reductants
Keywords: reductive coupling; aldimines; ketimines; samarium.
*
Reduction of the N-benzyliminium perchlorate of benz-
aldehyde (1) was carried out using SmI2, SmI2–HMPA,
Corresponding authors; e-mail: robert.flowers@ttu.edu; Tel.: þ46-31-
772-2904; fax: þ46-31-772-3840; e-mail: hilmers@organic.gu.se
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.06.004

10398
M. Kim et al. / Tetrahedron 59 (2003) 10397–10402
Table 1. Reactions of iminium perchlorate 1 with Sm(II) reductants
Entry
Substrate
Sm(II) reagent
Yield of 1a and 1b (1a:1b)
Reduced product 1c
1
2
3
4
1
1
1
1
SmI2 (1.5 equiv.)
SmI2 (1.5 equiv.) HMPA (15 equiv.)
Sm{N[Si(CH3)3]2}2 (3.0 equiv.)
SmI2 (1.5 equiv.)/Et3N (3.0 equiv.)/H2O (3.75 equiv.)
80a (70:30)b
80a (60:40)b
Trace (0:100)b
45b (65:35)b
–
–
–
55b
a
Isolated yield.
Determined by 1H NMR and GC.
b
Sm{N[Si(CH3)3]2}2, and SmI2/Et3N/H2O. All of these
reactions were carried out in 50% acetonitrile in THF
since 1 was insoluble in THF alone. In all cases the reactions
were determined to be quantitative by GC and NMR
examination of the crude product mixture. The results of
these experiments are contained in Table 1. Reaction of 1
with SmI2 was completed within minutes and examination
Sm(II) reductants was examined to determine the influence
of reagent and reductant structure on the outcome of the
reactions. A number of reactions of substrate 2 with SmI2
or SmI2 additive combinations have been reported.6–8
Furthermore, reduction of aldimines by SmI2 requires long
reaction times or elevated temperatures so we did not
examine this reagent any further. Kagan and co-workers
have reported that SmBr2 was useful in imine couplings, so
it was examined in place of SmI2.11 Table 2 contains the
results of the reduction of the N-benzyl and N-butyl imines
of benzaldehyde. All of the reactions were found to be
quantitative by GC and NMR analysis and good isolated
yields were obtained in some cases. Reduction of substrate 2
by SmBr2 and SmI2/Et3N/H2O produced diamines 2a and
2b in good yields. Unfortunately, there was no diastereo-
selectivity in the coupling reaction. The combination of
SmI2/Et3N/H2O also produced a significant amount (35%)
1
of the H NMR showed that only coupled product was
obtained. The isolated yield for this reaction was 80% and
the ratio of the meso and d/l coupled products was 70:30.
Next, the reaction was performed using SmI2–HMPA as the
reductant. The isolated yield was 80% with a slight change
in the observed diastereoselectivity. Since little change was
found in the diastereoselectivity upon addition of HMPA to
SmI2, Sm{N[Si(CH3)3]2}2 was examined as an alternative.
While 1b was the only observed coupling product, the
yield was low and products consistent with reduction of
acetonitrile were observed. Finally, the combination of
SmI2/Et3N/H2O was examined. This protocol was recently
of the reduced product 2c. Reduction of
2 by
Sm{N[Si(CH3)3]2}2 resulted in an isolated yield of 76%
and a much higher diastereoselectivity with a preference of
4:1 for the anti (d/l) coupled product.
´
introduced by Hilmersson and Dahlen and appeared to have
great potential for the reduction of functional groups not
typically reduced by SmI2 and SmI2–HMPA.19 Reduction
of 1 by SmI2/Et3N/H2O was instantaneous. Analysis by GC
Substrate 3 was examined to determine the impact of
N-substitution on the product distributions and yields.
Reaction of 3 with SmBr2 and SmI2/Et3N/H2O provided
nearly identical results. Both reductants provided nearly
80% of the coupled product with the balance being reduced
product. Both reductants provided essentially no diastereo-
selectivity. In contrast, reaction of 3 with Sm{N[Si(CH3)3]2}2
provided predominantly the coupled diamine and the dia-
stereoselectivity was 5:1 with a preference for the d/l pair.
1
and H NMR showed only reduced (1c) and reductively
coupled products (1a and 1b). The reductively coupled
products showed a low diastereoselectivity similar to that
obtained with SmI2 and SmI2–HMPA.
2.2. Reduction of aldimines by Sm(II) reductants
Reduction of a series of aldimines by the previous series of
Table 2. Reactions of 2 and 3 with Sm(II) reductants
Substrate
R
Sm(II) reagent
Yield of a and b (a:b)
Reduced product c
2
2
2
3
3
3
Bn
Bn
Bn
n-C4H9
n-C4H9
n-C4H9
SmBr2 (3.0 equiv.)
Sm{N[Si(CH3)3]2}2 (3.0 equiv.)
SmI2 (1.5 equiv.)/Et3N (3.0 equiv.)/H2O (3.75 equiv.)
SmBr2 (3.0 equiv.)
Sm{N[Si(CH3)3]2}2 (3.0 equiv.)
70a (50:50)b
76a (20:80)b
65b (50:50)b
78b (55:45)b
95b (16:84)b
79b (54:46)b
–
–
35b
22b
5b
SmI2 (1.5 equiv.)/Et3N (3.0 equiv.)/H2O (3.75 equiv.)
21b
a
Isolated yield.
Determined by 1H NMR and GC.
b

M. Kim et al. / Tetrahedron 59 (2003) 10397–10402
Table 3. Reactions of substrates 4–6 with Sm(II) reductants
10399
Substrate
R
Sm(II) reagent
Yield of a and b (a:b)
Reduced product c
4
4
4
5
5
5
6
6
6
p-CH3C6H5
p-CH3C6H5
p-CH3C6H5
p-CH3OC6H5
p-CH3OC6H5
p-CH3OC6H5
p-CF3C6H5
p-CF3C6H5
p-CF3C6H5
SmBr2 (3.0 equiv.)
Sm{N[Si(CH3)3]2}2 (3.0 equiv.)
SmI2 (1.5 equiv.)/Et3N (3.0 equiv.)/H2O (3.75 equiv.)
SmBr2 (3.0 equiv.)
Sm{N[Si(CH3)3]2}2 (3.0 equiv.)
SmI2 (1.5 equiv.)/Et3N (3.0 equiv.)/H2O (3.75 equiv.)
SmBr2 (3.0 equiv.)
Sm{N[Si(CH3)3]2}2 (3.0 equiv.)
–
–
Trace
–
65a (45:55)
35a
Trace
–
–
–
70a (30:70)
30a
Trace
–
–
–
SmI2 (1.5 equiv.)/Et3N (3.0 equiv.)/H2O (3.75 equiv.)
65a (45:55)
35a
a
Determined by 1H NMR and GC.
Next substrates 4 and 5 with electron donating para-
substituents, and substrate 6 with an electron withdrawing
para-substituent were examined to determine the electronic
effects on reductions and reductive coupling reactions of
imines. The results of these experiments are contained
in Table 3. Reactions of substrates 4–6 with SmBr2 and
Sm{N[Si(CH3)3]2}2 were unsuccessful, and resulted in
predominantly unreduced starting material.
4–6, there is a significant amount of reduction using SmI2/
Et3N/H2O.
The unexpected results obtained with 4–6 led us to examine
reactions of the meta derivative, 7 with the Sm(II)
reductants. The results are contained in Table 4. These
reactions proceeded to completion and examination of the
products clearly showed the presence of both reduced and
coupled products. Surprisingly, SmBr2 provided predomi-
nantly reduced product 7c and 20% of 7a and 7b. Reaction
of 7 with Sm{N[Si(CH3)3]2}2 and SmI2/Et3N/H2O provided
70 and 75% yields, respectively of the coupled product
and only trace amounts of reduced 7c. Reductive coupling
of 7 by SmBr2 or SmI2/Et3N/H2O showed very little stereo-
selectivity while coupling by Sm{N[Si(CH3)3]2}2 showed a
5:1 preference for 7b.
In each of the reactions between substrates 4–6 and
Sm{N[Si(CH3)3]2}2, only starting material was recovered
after workup. In the case of SmBr2, a trace amount of the
reduced products 4c–6c were obtained, but the balance of
the reaction was unreduced starting material.
Reaction of substrates 4–6 with SmI2/Et3N/H2O were also
examined. Typically, this reductant reacts with substrates
within minutes, but reaction with the para-substituted
imines took about 30 min to proceed to completion.
Examination of the reactions after completion clearly
showed the presence of both reduced and coupled products
exclusively. No starting materials or other side products
were apparent from examination of the 1H NMR spectra or
GC chromatographic traces. While the reactions proceeded,
the yield of coupled to reduced products was approximately
2:1. The reductive coupling of substrates 4 and 6 provided
no diastereoselectivity. However, the reductive coupling of
substrate 5 showed a preference for the d/l pair over the
meso product in an approximate 2:1 ratio. With substrates
2.3. Reduction of ketimines by Sm(II) reductants
Reduction of 8 by Sm(II) reductants was examined to
determine the ability of the more powerful reagents to carry
out reduction or reductive coupling of a more sterically
congested ketimine. Not only are ketimines more sterically
demanding substrates, but they are also typically more
difficult to reduce through single electron transfer than
aldimines. Table 5 contains the results of reduction of
8 by SmBr2, Sm{N[Si(CH3)3]2}2, and SmI2/Et3N/H2O.
Reduction of 8 by SmBr2 and Sm{N[Si(CH3)3]2}2 provided
only reduced product 8c and no trace of coupling product
Table 4. Reaction of substrate 7 with Sm(II) reductants
Substrate
R
Sm(II) reagent
Yield of a and b (a:b)
Reduced product c
7
7
7
m-CH3C6H5
m-CH3C6H5
m-CH3C6H5
SmBr2 (3.0 equiv.)
Sm{N[Si(CH3)3]2}2 (3.0 equiv.)
SmI2 (1.5 equiv.)/Et3N (3.0 equiv.)/H2O (3.75 equiv.)
20a (44:56)
70b (15:85)
75b (46:54)
70a
Trace
Trace
a
Determined by 1H NMR and GC.
Isolated yield.
b

10400
M. Kim et al. / Tetrahedron 59 (2003) 10397–10402
Table 5. Reactions of substrate 8 with Sm(II) reductants
Sm(II) reagent
Yield of 8a and 8b (8a:8b)
Reduced product 8c
SmBr2 (3.0 equiv.)
Sm{N[Si(CH3)3]2}2 (3.0 equiv.)
SmI2 (1.5 equiv.)/Et3N (3.0 equiv.)/H2O (3.75 equiv.)
–
–
Quantitativea
Quantitativea
20a
80a (50:50)
a
Determined by 1H NMR and GC.
was observed. The reaction was quantitative in both cases.
Reduction of 8 by SmI2/Et3N/H2O was quantitative and led
to 80% of the diamines 8a and 8b with the balance being 8c.
While the reaction of 8 with SmI2/Et3N/H2O provides high
yield of coupled products, there is no diastereoselectivity in
this reaction.
yields of the coupled product, but SmBr2 showed no
diastereoselectivity whereas Sm{N[Si(CH3)3]2}2 showed a
reasonable preference for the anti-coupled product. The
reduction of substrates 2 and 3 with SmI2/Et3N/H2O
provided reasonable yields of the coupled products, but a
considerable amount of the reduced side product was also
detected. Unfortunately, this system also showed no
diastereoselectivity. Altogether, these results demonstrate
that N-substitution has little effect on the yield or
diastereoselectivity of the reductive coupling or reduc-
tion of the imine. Of the three Sm(II) reagents
studied, Sm{N[Si(CH3)3]2}2 provided the highest
diastereoselectivity.
3. Discussion
The goal of this study was to examine and compare the
ability of various Sm(II) reductants to reduce or couple
substrates containing a C–N double bond. Iminium ions are
relatively easy to reduce and a recent report describing the
SmI2-mediated coupling of nitrones with carbonyls suggests
that the CvN bond of the nitrone is reduced preferentially
to the CvO bond.13 The perchlorate salt of 1 was used as a
prototypical substrate to examine the Sm(II) mediated
reduction and reductive coupling of iminium salts. Many of
the perchlorate iminium salts we have synthesized in our lab
were found to be insoluble in THF and 1 was no exception to
this observed trend. To circumvent this, 1 was dissolved in
acetonitrile and the Sm(II) substrates in THF were added to
the acetonitrile solution. This combination of solvents,
reagent and substrates produced a homogeneous solution
and reactions were complete within 1 min. The reductive
coupling of 1 by SmI2 and SmI2–HMPA was clean and
quantitative. In both cases good yields of 80% were
obtained after column chromatography. Unfortunately the
addition of HMPA to SmI2 had little effect on the
diastereoselectivity of the coupling as expected. Exami-
nation of Sm{N[Si(CH3)3]2}2 in the reaction with 1
provided trace amounts of the d/l coupled product of 1b,
but unfortunately, it also reduced the acetonitrile co-solvent.
Examination of the recent protocol developed by
Hilmersson and Dahlen provided both the reduced and
coupled product. Interestingly, this more powerful reductant
did not reduce the acetonitrile.
Substrates 4–6 were examined to determine the importance
of electronic factors in aldimine reductions. Surprisingly,
Sm{N[Si(CH3)3]2}2 was ineffective at reducing or coupling
these substrates. Use of SmBr2 provided only trace amounts
of reduced product even after 12 h. Use of SmI2/Et3N/H2O
provided between 65–70% of the coupled diamine products
with the balance once again being the reduced amine.
Nearly every reaction carried out with SmI2/Et3N/H2O
is complete within a few minutes, but the reactions of
substrates 4–6 took approximately 30–60 min to proceed to
completion. Substrates 4 and 5 can be expected to be
somewhat more difficult to reduce than the unsubstituted 2,
but 6 should be easier to reduce through single electron
transfer from Sm(II).
To further examine the impact of phenyl substitution on
aldimines, meta substituted substrate 7 was examined.
Reactions of 7 proceeded to completion with all of the
Sm(II) reagents examined. All of the reactions between 7
and the various Sm(II)-based reagents proceeded in a
reasonable amount of time in a manner similar to reactions
involving substrate 2. These results indicate that para
substitution on the aldimines examined in this study inhibit
reduction by Sm(II)-based reagents.
Namy and Machrouhi recently reported that aldimines
similar to 5 are efficiently coupled using a combination
of SmI2/NiI2.8 Although the role of NiI2 in this system
is unclear, it is quite surprising that SmBr2 and
Sm{N[Si(CH3)3]2}2 are incapable of carrying out these
reactions at room temperature. Both are known to be strong
single electron reductants and their thermodynamic redox
potentials are within 100 mV of each other.16,20 Further-
more, the fact that SmI2/Et3N/H2O reduces these substrates
slowly (in comparison to other substrates) suggests that
the presence of para-substituents influence the rate of
Although these results are somewhat disappointing, they do
show that SmI2 alone is a suitable reductant for the coupling
of iminium salts. Recent elegant work by Skrydstrup and
Riber show that the iminyl radical formed upon reduction of
a nitrone (iminium equivalent) can efficiently couple with
a,b-unsaturated esters and amides as well.14
The next series of experiments focused on the reaction of
Sm(II) reagents with aldimines derived from benzaldehyde
and benzyl amine 2 or n-butyl amine 3. Reactions of either
substrate with SmBr2 or Sm{N[Si(CH3)3]2}2 provided good

M. Kim et al. / Tetrahedron 59 (2003) 10397–10402
10401
reduction. It is interesting that para substituted aromatic
substrates with both electron donating and electron
withdrawing substituents are more difficult to reduce.
Mechanistic studies are currently being planned to elucidate
the cause of this unexpected result.
filtered, and then washed with ether. The white solid was
dried under vacuum (1.32 g, 93%). 1H NMR (acetone-d6): d
5.38 (s, 2H), 7.44–8.27 (m, 10H), 9.51(s, 1H).
5.1.2. Reaction of (1) with Sm(II) reductants. The Sm(II)
reagent (1.5 equiv.) was added dropwise to an acetonitrile
solution of 1 (0.295 g, 1.0 mmol), under nitrogen atmos-
phere. The reaction mixture was stirred for additional three
hours then saturated ammonium chloride (aq.) was added
carefully. Then the solution was extracted with diethyl ether
(50 ml£3) and the organic layers were combined then
washed with saturated aqueous solution of Na2S2O3
(20 ml£2). The organic layer was dried under MgSO4
then the solvent was removed to give the mixture of two
diastereomers. 1H NMR (500 MHz, CDCl3): 1a; d 3.45
(center of AB system, 4H), 3.79 (s, 2H), 7.0–7.5 (m, 20H).
1b; d 3.61 (center of AB system, 4H), 3.74 (s, 2H), 7.0–7.5
(m, 20H). 1c; d 3.84 (4H, s), 7.0–7.5 (10H, m).
Substrate 8 was examined to determine the efficacy of
the Sm(II) reagents in ketimine reductions. While all of the
reagents examined were capable of reducing 8, only the
combination of SmI2/Et3N/H2O was capable of coupling to
provide diamine products 8a and 8b. Use of SmBr2 and
Sm{N[Si(CH3)3]2}2 provided only reduced 8c. Reduction of
ketimines by SmBr2-HMPA also provide amines exclu-
sively as products.21 It is unclear at this point whether the
reduction proceeds through H-atom abstraction from THF
or whether trace amounts of proton from water or solvent
impurities are responsible for exclusive amine formation.
Nonetheless, these findings suggest that SmI2/Et3N/H2O is
the best suited of the reagents examined for reductive
coupling of ketimines while SmBr2 and Sm{N[Si(CH3)3]2}2
are best suited for ketimine reduction.
When 1 was reacted with Sm{N[Si(CH3)3]2}2, NMR
showed only 1b as a product. Unfortunately 1b could
not be isolated since there were many other side
products resulting from reduction of acetonitrile by
Sm{N[Si(CH3)3]2}2.
4. Conclusions
For entry 4 in Table 1, SmI2 and Et3N, then H2O was added
to the acetonitrile solution of substrate. White precipitate
was immediately formed once the addition of H2O was
completed. Then the reaction mixture was worked up as
usual and evaporation of solvent gave the mixture of
coupled product and reduced product.
The experiments described herein show that SmI2 is capable
of reducing iminium perchlorate 1 to the corresponding
diamines 1a and 1b with a preference for the syn isomer.
Use of HMPA or other protocols show no advantage in
terms of product stereoselectivity or yield. In the reactions
of aldimines 2, 3 and 7 with Sm(II)-based reductants,
Sm{N[Si(CH3)3]2}2 showed the best diastereoselectivity in
the reductive coupling reaction. For more recalcitrant para-
substituted aldimines 4–6, the combination of SmI2/Et3N/
H2O was best suited for coupling although little or no
diastereoselectivity was observed in the reaction. Reductive
coupling of 8 to 8a and 8b is best achieved through the use
of SmI2/Et3N/H2O while reduction to 8c is accomplished
through the use of SmBr2 and Sm{N[Si(CH3)3]2}2. The
results of these experiments are currently being utilized to
develop coupling reactions of imines and other functional
groups.
5.2. General procedure for synthesis of imines 2–8
To a solution of aldehyde or ketone (50.0 mmol) in CH2Cl2
(50 ml) was added 5 g of 4A molecular sieves and benzyl
amine (50.0 mmol). The heterogeneous mixture was stirred
for an hour then the molecular sieves were filtered off.
The filtrate was dried under MgSO4 then the solvent
was evaporated in vacuo to give colorless liquid as product
(73–97%).
Imine 8 was obtained from acetophenone (20.0 mmol) and
benzylamine (20.0 mmol) refluxed in benzene (120 ml)
fitted with a Dean–Stark trap overnight. After evaporation
of the solvent and drying in vacuo, imine 8 was collected as
a slightly yellow oil (70% yield).
5. Experimental
5.1. General
5.2.1. Reaction of imines 2–8 with Sm(II) reductants.
The Sm(II) reagent was added dropwise to a solution of
imine (0.6 mmol) in THF (10 ml), under nitrogen atmos-
phere. The reaction mixture was stirred for 12 hp then
saturated ammonium chloride (10 ml) was added. The
reaction mixture was extracted with diethyl ether (50 ml£3)
then the organic layers were combined and washed with
saturated Na2S2O3 solution (10 ml£2). The organic layer
was dried under MgSO4 and the solvent was removed in
vacuo to give product. pSmBr2 and Sm{N[Si(CH3)3]2}2
reductions. When SmI2/Et3N/H2O was used, the reaction
was completed within 30 min. 1H NMR (500 MHz, CDCl3):
2a; d 3.45 (center of AB system, 4H), 3.79 (s, 2H), 7.0–7.5
(m, 20H). 2b; d 3.61 (center of AB system, 4H), 3.74 (s,
2H), 7.0–7.5 (m, 20H). 2c; d 3.84 (4H, s), 7.0–7.5 (10H,
5.1.1. Preparation of (1). Approximately 16.5 g
(110 mmol) perchloric acid (70% in water) was added
dropwise to a 10.7 g solution of benzyl amine (100.0 mmol)
in ice bath. The reaction mixture was stirred overnight, then
benzyl amine was added slowly until the solution was
slightly basic. The solvent was removed in vacuo to give
white solid (17.8 g, 86%). 1H NMR (acetone-d6): d 4.51 (s,
2H), 7.42–7.62 (m, 5H). A mixture of 1.0 g (4.8 mmol) of
the benzylammonium perchlorate salt, 1.53 g, (14.4 mmol)
of benzaldehyde, five drops of diethyl amine, and benzene
(70 ml) were heated at reflux with a Dean–Stark trap. After
3 days of reflux, solvent was removed in vacuo and fresh
benzene (50 ml£2) was added and distilled off. The brown
oily residue was then crystallized in ether at room
temperature. The crystallized solid was ground in ether,

10402
M. Kim et al. / Tetrahedron 59 (2003) 10397–10402
m). 3a; d 0.73–0.76 (t, 6H), 1.03–1.12 (m, 4H), 1.21–1.43
(m, 4H), 2.20–2.30 (m, 4H), 3.72 (s, 2H), 7.00–7.32
(m, 10H). 3b; d 0.83–0.86 (t, 6H), 1.21–1.43 (m, 8H),
2.35–2.43 (m, 4H), 3.60 (s, 2H), 7.00–7.32 (m, 10H). 3c; d
0.89–0.92 (t, 3H), 1.21–1.51 (m, 4H), 2.62–2.65 (t, 2H),
3.79 (s, 2H), 7.00–7.32 (m, 5H). 4a; d 2.31 (s, 6H), 3.46
(center of AB system, 4H), 3.74 (s, 2H), 7.0–7.5 (m, 18H).
4b; d 2.41 (s, 6H), 3.60 (center of AB system, 4H), 3.73
(s, 2H), 7.0–7.5 (m, 18H). 4c; d 2.38 (s, 3H), 3.81 (s, 2H),
3.84 (s, 2H), 7.0–7.5 (m, 9H). 5a; d 3.43 (center of AB
system, 4H), 3.67 (s, 2H), 3.89 (s, 6H), 6.78–7.78 (m, 18H).
5b; d 3.58 (center of AB system, 4H), 3.65 (s, 2H), 3.84
(s, 6H), 6.78–7.78 (m, 18H). 5c; d 3.76 (s, 2H), 3.81 (s, 2H),
3.85 (s, 3H), 7.0–7.5 (m, 9H). 6a; d 3.47 (center of AB
system, 4H), 3.87 (s, 2H), 7.0–7.6 (m, 18H). 6b; d 3.57
(center of AB system, 4H), 3.88 (s, 2H), 7.0–7.6 (m, 18H).
6c; d 3.77 (s, 2H), 3.80 (s, 2H), 7.0–7.5 (m, 9H). 7a; d
2.40 (s, 6H), 3.45 (center of AB system, 4H), 3.82 (s, 2H),
6.93–7.40 (m, 18H). 7b; d 2.29 (s, 3H), 2.40 (s, 3H), 3.62
(center of AB system, 4H), 3.82 (s, 2H), 6.93–7.39
(m, 18H). 7c; d 2.39 (s, 3H), 3.82 (s, 2H) 3.85 (s, 2H),
6.93–7.40 (m, 9H). 8a; d 1.54 (s, 6H), 3.44 (center of AB
system, 4H), 7.0–7.35 (m, 20H). 8b; d 1.65 (s, 6H), 3.44
(center of AB system, 4H), 7.0–7.35 (m, 20H). 8c; 1.39
(d, 3H), 3.64 (center of AB system, 2H), 3.83 (q, 1H),
7.2–7.4 (m, 10H).
References
1. Steel, P. G. J. Chem. Soc. Perkin Trans. 1 2001, 2727.
2. Molander, G. A.; Harris, C. R. Chem. Rev. 1996, 96, 307.
3. Curran, D. P.; Fevig, T. L.; Jasperse, C. P.; Totleben, M. J.
Synlett 1992, 943.
4. Girard, P.; Namy, J. L.; Kagan, H. B. J. Am. Chem. Soc. 1980,
102, 2693.
5. Molander, G. A. Chem. Rev. 1992, 92, 29.
6. Enholm, E. J.; Forbes, D. C.; Holub, D. P. Synth. Commun.
1990, 20, 981.
7. Imamoto, T.; Nishimura, S. Chem. Lett. 1990, 1141.
8. Machrouhi, F.; Namy, J.-L. Tetrahedron Lett. 1999, 40, 1315.
9. Annunziata, R.; Benaglia, M.; Cinquini, M.; Cozzi, F.;
Raimondi, L. Tetrahedron Lett. 1998, 39, 3333.
10. Collin, J.; Giuseppone, N.; Machrouhi, F.; Namy, J.-L.; Nief,
F. Tetrahedron Lett. 1999, 40, 3161.
11. Lebrun, A.; Rantze, E.; Namy, J.-L.; Kagan, H. B. New
J. Chem. 1995, 19, 699.
12. Banik, B. K.; Zegrocka, O.; Banik, I.; Hackfeld, L.; Becker,
F. F. Tetrahedron Lett. 1999, 40, 6731.
13. Masson, G.; Py, S.; Vallee, Y. Angew. Chem. Int. Ed. 2002, 41,
1772.
14. Riber, D.; Skrydstrup, T. Org. Lett. 2003, 5, 229–231.
15. Evans, W. J.; Drummond, D. K.; Zhang, H.; Atwood, J. L.
Inorg. Chem. 1988, 27, 575.
16. Prasad, E.; Knettle, B. W.; Flowers, R. A., II. J. Am. Chem.
Soc. 2002, 124, 14663.
´
17. Dahlen, A.; Hilmersson, G. Tetrahedron Lett. 2003, 44, 2661.
Acknowledgements
´
18. Dahlen, A.; Hilmersson, G. Tetrahedron Lett. 2002, 43, 7197.
´
19. Dahlen, A.; Hilmersson, G. Chem. Eur. J. 2003, 9, 1123.
R. A. F. is grateful to the National Science Foundation
(CHE-0196163) for support of this work. Financial support
from the Carl Trygger foundation and the Swedish Natural
Research Council (VR) is also gratefully acknowledged
by G. H.
20. Miller, R. S.; Sealy, J. M.; Fuchs, J. R.; Shabangi, M.; Flowers,
R. A., II. J. Am. Chem. Soc. 2000, 122, 7718.
21. Knettle, B. W.; Flowers, R. A., II. Org. Lett. 2001, 3, 2321.