Instantaneous SmI2/H2O/amine-mediated reductions in THF

Article abstract of DOI:10.1002/chem.200390129

The SmI₂-mediated reductions of ketones, imines, and α,β-unsaturated esters have been shown to be instantaneous in the presence of H₂O and an amine in THF. The SmI₂-mediated reductions are not only shown to be fast and quantitative by the addition of H₂O and an amine, but the workup procedures are also simplified. Competing experiments with SmI₂/H₂O/amine confirmed that α,β-unsaturated esters could be selectively reduced in the presence of ketones or imines. Comparison of analogue ligands showed that nitrogen and phosphorus ligands are superior to oxygen and sulfur ligands in these reductions. The trialkylphosphine 1,2-bis(dimethylphosphino)ethane (DMPE) provided a primary kinetic isotope effect, yielding a k_H/k_D of 4.5.

Full text of DOI:10.1002/chem.200390129

FULL PAPER
Instantaneous SmI2/H2O/Amine-Mediated Reductions in THF
Anders Dahle¬n and Gˆran Hilmersson*[a]
Abstract: The SmI2-mediated reductions of ketones, imines, and a,b-unsaturated
esters have been shown to be instantaneous in the presence of H2O and an amine in
THF. The SmI2-mediated reductions are not only shown to be fast and quantitative by
the addition of H2O and an amine, but the workup procedures are also simplified.
Competing experiments with SmI2/H2O/amine confirmed that a,b-unsaturated esters
could be selectively reduced in the presence of ketones or imines. Comparison of
analogue ligands showed that nitrogen and phosphorus ligands are superior to
oxygen and sulfur ligands in these reductions. The trialkylphosphine 1,2-bis(dimeth-
ylphosphino)ethane (DMPE) provided a primary kinetic isotope effect, yielding a
kH/kD of 4.5.
Keywords: functional groups
N ligands ¥ reduction ¥ selectivity ¥
samarium
¥
diamine (TMEDA) and N,N,N',N'',N''-pentamethyldiethyl-
enetriamine (PMDTA), reduced dialkyl ketones to alcohols
instantaneously at 20.08C (Scheme 1).[26] It was shown that
Introduction
Ever since Kagan developed the versatile coupling and
reducing reagent samarium diiodide in THF in the late
1970×s,[1, 2] there has been a remarkable activity in discovering
new reactions in which this reagent can be utilized.[3 10] It has
been found to be especially useful in coupling reactions
between alkyl halides and ketones, also known as the Barbier
reaction,[11, 12] between aromatic carbonyls (pinacol cou-
pling),[13, 14] and between aldehydes and a,b-unsaturated
esters (ketyl olefin coupling).[15, 16] The reducing power of
samarium diiodide in THF is sensitive to the presence
OH
O
SmI2 in THF
Amine/H2O
< 10 seconds
R1
R2
R1
R2
> 99 %
Scheme 1. Instantaneous reduction of ketones induced by amine/H2O.
of
co-solvents,
for
example,
hexamethylphosphor-
two equivalents of amine and three equivalents of water had
to be used for each oxidized samarium diiodide in order to
achieve the instantaneous and complete reduction. This
powerful combination was also shown to be superior to other
methods, including the widely used HMPA/alcohol mix-
tures.[5] The increase in rates of the SmI2-mediated reduction
exceed 100000 compared to that of water or amine alone.
Previously Cabri et al. reported that the addition of triethyl-
amine accelerates the SmI2/H2O-mediated cyclization of aryl
radicals and olefins.[27]
Herein we wish to report the use of the reagent mixture
SmI2/H2O/amine in various reactions. It has been shown that
this method can be applied to different types of functional
groups and that reduction takes place both quantitative and in
a matter of seconds (Figure 1).
amide (HMPA),[17, 18] N,N'-dimethyl-N,N'-propyleneurea
(DMPU),[19
and
1,8-diazabicyclo[5.4.0]undec-7-ene
22]
(DBU).[23] HMPA has been found to increase the oxidation
potential of SmI2 from À1.33 to À2.05 V, indicated by linear
sweep voltammetry,[24] or its standard potential from À0.98 to
À1.75 V vs SCE, indicated by cyclic voltammetry.[25] Addition
of these additives makes it possible to fine-tune the reactivity,
thereby altering the experimental conditions and the outcome
of the reaction. However, the very popular HMPA is
carcinogenic, and the discovery of a replacement for this
additive is therefore of great concern. The influence of other
solvents or additives on SmI2-mediated reactions has been
reported, although HMPA appears to be superior.[4]
Recently we reported that SmI2, water, and simple amines,
such as triethylamine (Et3N), N,N,N',N'-tetramethylethylene-
O
¬
[a] Dr. G. Hilmersson, A. Dahlen
R2
O
Organic Chemistry, Department of Chemistry
Gˆteborg University, 412 96 Gˆteborg (Sweden)
Fax : (‡46)31-772-3840
R1
OR3
R1
R1
R2
N
R2
R2
E-mail: hilmers@organic.gu.se
Figure 1. Functional groups investigated in this study.
0947-6539/03/0905-1123 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 5
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1123

¬
G. Hilmersson and A. Dahlen
FULL PAPER
Table 1. Relative rates in the SmI2/H2O/amine-mediated reduction of ketones
Results and Discussion
c]
compared to 2-methyl-heptan-3-one.[a
Entry
Substrate
Relative rate
(2 equiv Et3N)
Relative rate
(1 equiv
TMEDA)
Relative rate
(0.67 equiv
PMDTA)
Reduction of ketones by diols: The impact of chelating
alcohols for the SmI2-mediated reduction of heptan-3-one was
investigated in a previous study.[28] It was found that proton
donors that contain coordinating ether functions, for example,
mono-, di-, and triglycol methyl ethers, reduced heptan-3-one
more efficiently than the nonchelating proton-donor meth-
anol. Diols were found to be superior to monoalcohols in this
reduction. The existence of a primary kinetic isotope effect
proved that the proton transfer is part of the rate-determining
step (rds) (Scheme 2).
1
1
1.0
1.0
1.0
2
3
2
3
2.6
7.1
2.7
8.3
2.5
7.0
4
4
13
14
12
To verify the generality of this conclusion several other
dialkyl ketones were studied. The results initially found for
heptan-3-one turned out to be general for all dialkyl ketones
tested, and the diol di(ethyleneglycol) was the most efficient
proton source. Each ketone was quantitatively reduced, into
its respective alcohol, in excess SmI2 (7 equiv) and diglycol
(7 equiv), to ensure minimization of variations in concentra-
tion (pseudo-first-order rate conditions).[28] A direct compar-
ison of the initial rates of the diglycol-mediated reductions of
ketones revealed that the steric crowding around the carbonyl
carbon is reflected in the rate of these reductions. It also
became evident that the phenyl substituents significally affect
the rates (Figure 2).
5
7
5
6
23
31
30
39
21
33
6
7
59
77
63
[a] The amount of amine corresponds to two R3N per SmI2. [b] In a typical
experiment two or more substrates were introduced into a mixture of SmI2 and
amine. Water was added slowly under vigorous stirring until a white precipitate
and a clear organic layer was obtained. The respective yields of the reactants
were determined by GC. [c] The data were obtained from several separate
experiments and the relative error is less than 10%.
with TMEDA a slightly larger
increase in relative rates was ob-
served upon going from the more
sterically hindered to the less steri-
cally hindered ketones. This may
be an effect of the steric require-
Scheme 2. SmI2-mediated ketone reduction accelerated by alcohols.
ments of the amines. We can not
exclude a possible change in the
O
O
O
O
O
Substrate:
<
<
<
<
reducing power of SmI2 coordinat-
ed to amines. However, it is most
unlikely that a bulky amine like
Et3N can compete with the bulk
solvent THF as a ligand for SmI2.
1
7
50
130
140
Relative rate:
Figure 2. Comparison of the reactivity of different ketones by using SmI2/diglycol in THF.
Reduction of ketones by the SmI2/H2O/amine mixtures:
Applying the SmI2-mediated reduction promoted by amine/
H2O to these ketones resulted in complete reduction in less
than ten seconds for all dialkyl ketones investigated. Thus, the
very fast rate of reduction did not allow initial rate studies by
GC analysis as above. However, we performed competitive
experiments, by introducing mixtures of two or more sub-
strates into the same reaction mixture, and could in this
manner compare the very fast reactions. Large differences in
relative rates were achieved between the ketones (Table 1).
The substrates with the least steric requirements and the
substrates with electron-withdrawing groups, for example,
phenyls, were again proven to be reduced in preference of the
aliphatic, bulky substrates. Three different amines were used
in these experiments: Et3N, TMEDA, and PMDTA. The
relative rates between the more reactive ketone cyclohex-
anone and the less reactive ketone 2-methyl-heptan-3-one is
approximately 77:1 in the presence of TMEDA. Interestingly
the results with Et3N or PMDTA are almost identical, but
SmI2 has been known to mediate pinacol couplings ever
since the early days of the samarium diiodide era, especially
with aryl ketones, under mild conditions.[13] Acetophenone is
reported to undergo reductive coupling in 30 seconds yielding
the pinacol product (95%).[13] However, reduction of aceto-
phenone is also reported in the presence of MeOH.[1]
Applying the SmI2/H2O/amine mixtures to the substrates
described in Table 1 led to quantitative reduction of the
ketones to their corresponding alcohols, independently of the
addition order of the additives. However, when aryl-substi-
tuted ketones was studied, that is, acetophenone, mainly
pinacol coupling products were obtained if water was added
after the addition of the aryl ketone. (All products described
1
in this article were isolated and analysed by H-NMR and/or
GC/MS.) However, mixing SmI2, amine, and water prior to
dropwise addition of acetophenone yielded mostly phenyl-
ethanol (Scheme 3). Thus, it is apparent that the order of
additions of the additives is crucial for the SmI2/H2O/amine
reduction of aryl-substituted ketones.
1124
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0905-1124 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 5

SmI2-Mediated Reductions
1123 1128
Figure 4. Primary kinetic isotope effect for the reduction of 2-methyl-
heptan-3-one using excess SmI2, DMPE, and H2O/D2O.
Scheme 3. Different pathways for acetophenone in SmI2-mediated reac-
tions.
Rate comparison of nitrogen, phosphorus, oxygen, and sulfur
analogues in the reduction of ketones: The discovery of the
highly efficient SmI2/H2O/amine-mediated reductions of
ketones opened our eyes for analogue Lewis bases with
oxygen, phosphorus, and sulfur.
the amine analogue could not be achieved, since the reactivity
of this mixture is far too high for initial rate measurement in
the reduction of ketones with our current method. However,
we believe that the rate-determining steps of the N and P
ligands are similar.
Reduction of 4-phenyl-2-butanone (1 equiv) with seven
equivalents of 1,2-dimethoxyethane (DME) or 1,2-bis(meth-
ylsulfanyl)ethane (MSE) gave only 40% conversion after
24 hours in the presence of excess water (17.5 equiv). How-
ever, the phosphorus-containing analogue, 1,2-bis(dimethyl-
phosphino)ethane (DMPE), resulted in complete reduction
of 4-phenyl-2-butanone in approximately 30 minutes with
excess water. As mentioned earlier 4-phenyl-2-butanone is
completely reduced in less than ten seconds by using SmI2/
H2O/TMEDA (Figure 3).
Why are the phosphorus and nitrogen ligands so incredibly
effective relative to their oxygen and sulfur analogues? Both
trialkylphosphorus ligands and amines have the ability to
form insoluble hydrohalide salts. Oxygen and sulfur lack this
ability; this makes them considerably less effective in these
reactions. Thus, the formation of amine and phosphorus salts
during the progress of the reaction increases the rate of
reduction, because of the increase in the equilibrium constant
(Scheme 4).
Scheme 4. Balanced equation for reductions induced by nitrogen or
phosphorus ligands.
<<
<<
P
P
N
N
<
–
O
O
S
S
Figure 3. Comparison of analogous ligands in the reduction of 4-phenyl-2-
butanone.
Based on this result we propose that the main reason for the
enormous rate enhancements observed for nitrogen and
phosphorus ligands is the rapid precipitation of byproducts;
this drives the reaction forward according to Le Chateliers
Principle. Amines are more effective than trialkylphosphorus
ligands in the reduction of dialkyl ketones because of more
effective formation and precipitation of the salts in the
reaction.
The conclusion of these results is that neither oxygen nor
sulfur analogues accelerate the reduction of 4-phenyl-2-
butanone in the presence of excess water, since similar yields
are reached with water alone. However, DMPE/H2O gave a
considerable rate enhancement, relative to the rates of DME/
H2O-, MSE/H2O-, H2O-, or DMPE-accelerated reductions.
The results also confirmed that the TMEDA/H2O mixture was
superior to its Lewis base analogues, with O, S, or P, reducing
any dialkyl ketone tested in less than ten seconds in excess
water.
The DMPE/H2O-accelerated reduction is sufficiently slow
for detailed kinetic studies; therefore, several separate experi-
ments were conducted to determine the rate-determining step
for the SmI2/H2O/DMPE-promoted reduction of ketones.
Two substrates, 4-phenyl-2-butanone and 2-methyl-heptan-3-
one, were tested separately for the existence of a primary
kinetic isotope effect (PKIE) in the SmI2-mediated (7 equiv)
reduction in the presence of DMPE (7 equiv) and H2O or
D2O (17.5 equiv). A PKIE was attained for the reduction of
both substrates, yielding a kH/kD of 4.5 Æ 0.3 for 2-methyl-
heptan-3-one, indicating a rate-determining proton transfer
(Figure 4).
Reductions of a,b-unsaturated esters: Reductions and com-
peting reactions involving a,b-unsaturated esters were also
performed by using the novel method with SmI2/H2O/amine.
Reduction of a,b-unsaturated esters has previously been
studied in detail in the presence of HMPA, indicating that the
substituents affect the reactivity.[29 31] A possible side-reaction
in these reductions is reductive coupling between two double
bonds as reported by Inanaga et al.[32] However, no side-
reaction could be detected in the amine/H2O-induced reduc-
tion. The a,b-unsaturated esters were quantitatively and
instantaneously reduced by the powerful mixture of SmI2/
H2O/amine yielding the corresponding esters as the only
products (Scheme 5).
The expected reactivity order, also verified in a recent
report containing rate constants in SmI2-mediated HMPA/
alcohol reductions,[30] between the a,b-unsaturated esters
investigated confirmed that substrates containing bulky,
The reduction of 4-phenyl-2-butanone was very fast and the
evident PKIE could not be exactly determined. A PKIE for
Chem. Eur. J. 2003, 9, No. 5
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0905-1125 $ 20.00+.50/0
1125

¬
G. Hilmersson and A. Dahlen
FULL PAPER
R2
R2
O
O
SmI2
SmI2
R1
R1
R1
OR3
R1
OR3
N
R2
N
H
R2
Amine/H2O
<10s
Amine/H2O
< 10s
R2
R2
Scheme 6. Instantaneous SmI2/H2O-mediated reduction of imines induced
by amines.
Scheme 5. Instantaneous SmI2-H2O-mediated reduction of a,b-unsaturat-
ed esters induced by amines.
electron-donating groups react slower than substrates that
contain electron-withdrawing groups (Table 2).
above. The iminebenzylidenephenylamine does not react with
SmI2 in the absence of additives at room temperature. During
reflux in THF the imino pinacol coupling is completed in
30 minutes (Scheme 7);[33, 34] the addition of MeOH at room
temperature does not give reduction of the imine to the
corresponding amine, as would be expected, instead smaller
amounts of imino pinacol product were obtained. However,
The experimental data shows that the relative reactivity
between a,b-unsaturated esters and ketones is very high.
Competition between 2-methyl-heptan-3-one and any cinna-
mate gave exclusive reduction of the a,b-unsaturated esters,
leaving the ketone unreacted.
Reductions of imines: Knettle
and Flowers recently reported
on HMPA-induced SmBr2-
mediated reduction of ket-
imines.[33] Reaction times of
15 minutes yield 99% of the
corresponding amine. The
SmI2-mediated reduction pro-
moted by amine and water
quantitatively reduced the
imines below in less than ten
seconds (Scheme 6).
Competitive experiments be-
tween ketones and imines were
also performed. (For imines the
entire reaction mixture had to
be quenched to obtain accurate
results, due to precipitation of
the products. n-Decan was used
as an internal standard.) Exclu-
sive reduction of the faster re-
acting ketone cyclohexanone 7
was accomplished versus the
slower reacting imine 11; the
relative rate is approximately
700:1 using TMEDA. Further
comparison of the relative rates
shows that the cinnamates are
reduced exclusively to both
imines and ketones. TMEDA
gave larger differences in rela-
tive rates, while almost identi-
cal relative rates were obtained
using Et3N and PMDTA (Ta-
ble 3).
Table 2. Relative rates in the SmI2/H2O/amine-mediated reduction of a,b-unsaturated esters compared to
d]
2-methyl-heptan-3-one.[a
Entry
Substrate
Relative rate
Relative rate
Relative rate
(2 equiv Et3N)
(1 equiv TMEDA)
(0.67 equiv PMDTA)
1
3
1
1.0
1.0
1.0
8
52
67
55
4
5
9
7.6 Â 102
15 Â 103
10 Â 102
24 Â 103
7.6 Â 102
18 Â 103
10
[a] The amount of amine corresponds to two R3N per SmI2. [b] In a typical experiment two or more substrates
were introduced into a mixture of SmI2 and amine. Water was added slowly under vigorous stirring until a white
precipitate and a clear organic layer was obtained. The respective yield was determined by GC. [c] The relative
rates were determined by competition with 6, and then arbitrarily relative to 2-methyl-heptan-3-one. [d] The data
were obtained from several separate experiments and the relative error is less than 10%.
Table 3. Relative rates in the SmI2/H2O/amine-mediated reduction of imines compared to 2-methyl-heptan-3-
c]
one.[a
Entry
Substrate
Relative rate
Relative rate
Relative rate
(2 equiv Et3N)
(1 equiv TMEDA)
(0.67 equiv PMDTA)
1
2
11
0.09
0.37
0.11
0.74
0.08
0.39
12
3
4
13
0.66
1.6
0.56
1
1.0
1.0
1.0
During the investigation of
the reduction of imines it was
also noticed that aryl-substitu-
tion on the imine carbon atom
led to imino pinacol coupling
as the main product, in close
analogy with the results of the
arylketone reduction described
5
14
3.4
8.3
3.0
[a] The amount of amine corresponds to two R3N per SmI2. [b] In a typical experiment two or more substrates
were introduced into a mixture of SmI2 and amine. Water was added slowly under vigorous stirring until a white
precipitate and a clear organic layer was obtained. The respective yields of the reactants were determined by GC.
[c] The data were obtained from several separate experiments and the relative error is less than 10%.
1126
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0905-1126 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 5

SmI2-Mediated Reductions
1123 1128
R2
R1
R2
R2
R2
R2
R2
N
HN
NH HN
NH HN
SmI2 (1.25 equiv.)
SmI2 (1.25 equiv.)
R1
R3N (2.5 equiv.)
H2O (3.1 equiv.)
O
+
reflux 65
0.5 h
C
R1 R1
R1 R1
O
20 C, < 10 s
> 95%
> 90%
> 10%
SmI2 (2.5 equiv.)
MeOH (5 equiv.)
R1 = H
R
2 = Ph
R2
R2
NH HN
R1 R1
< 20%
Scheme 7. Different pathways for benzylidenephenylamine in SmI2-mediated reactions.
the SmI2/H2O/TMEDA mixture resulted in primarily imino
pinacol product (>90%) and smaller amounts of the reduced
imine, when water was slowly added to SmI2 subsequent to
Effective workup procedures on larger scale
Ketones and a,b-unsaturated esters: In an ordinary reaction
two slightly different workup procedures may be employed; in
the first alternative the reaction mixture, which includes
inorganic samarium salts and precipitated amines, is dissolved
in HCl (0.1m), extracted with diethyl ether, and finally washed
with Na2S2O3 and brine. The organic layer is then dried over
Na2SO4 and after evaporation the product is sufficiently pure
for further use. This workup is the most common for
samarium diiodide reactions reported earlier in the litera-
ture.[13, 35]
In the second alternative, the precipitated materials are
removed initially by centrifugation or filtration. This is
particularly advantageous, since the product remains in the
solution. The supernatant is then diluted with diethyl ether,
washed with Na2S2O3 and brine, and finally dried over Na2SO4
and evaporated. The yield will not be affected by this slightly
faster workup.
However, the fact is that the only workup needed for these
reactions is simple filtration of the precipitated materials
followed by evaporation of THF, on condition that correct
proportions of amine and water are added to the samarium
diiodide reagent. Only the byproducts are precipitated, while
the product remains dissolved. This extremely short workup is
particularly convenient for water-soluble products, which
would dissolve into the water layer during workup.
1
amine and imine (determined by H and 13C NMR spectros-
copy). Mainly imino pinacol (>60%) was observed even
when the imine was added after the amine and water. No
reaction could be detected in the absence of H2O.
This is very interesting, since it opens up a new, milder
pathway for the somewhat more difficult imino pinacol
coupling. Refluxing conditions may be destructive for
functional groups and/or protective groups; this can
be avoided with this new, improved method, which is
instantaneous at room temperature. The generality of the
imino pinacol coupling reaction is currently under inves-
tigation.
IR and 1H NMR analysis and of precipitated salts: The white
salts formed in the reactions were filtrated, washed with
diethyl ether, and finally dried yielding fine, white powders.
The powders were analyzed by FTIR and 1H NMR spectros-
copy. The IR analysis gave strong signals in the region 2500
3000 cmÀ1, which reveal the existence of amine hydrohalides.
These signals disappeared after basic workup with sodium
hydroxide and diethyl ether, since the amines are released and
removed. The very strong signals around 3445 cmÀ1 point also
towards the existence of hydrogen bonds (OH stretching),
which may belong to Sm(OH)3. This signal did not disappear
after the basic workup.
The 1H NMR analysis of the precipitate, dissolved in
CDCl3, also supported the occurrence of hydrohalide salts,
since signals with shifts downfield of the reference ligands
were obtained, in addition to weak signals between 10
Imines: Since the method for SmI2/H2O/amine-mediated
reductions is based on precipitating all byproducts, that is,
Sm(OH)3 and HI salts of amines, the workup described above
can not be employed, since the product in this case will also
form HI and HCl salts. Instead excess NaOH (1m) is added to
the reaction mixture, which is then extracted by, for instance,
diethyl ether. The samarium salts, that is, Sm(OH)3, remain in
the water layer, while the amines are extracted into the
organic layer. The organic layer is then washed with Na2S2O3
and brine. Finally the organic layer is dried over Na2SO4 and
after evaporation the product is satisfactorily pure for further
‡
11 ppm, which belong to NH . Mixtures containing PMDTA
gave several signals, which may be an effect of unequal
amounts of HI per amine function. However, after basic
1
workup with sodium hydroxide only the H signals from the
pure amines were detected. (The IR and NMR data for the
Et3N, TMEDA, and PMDTA salts can be found in the
Experimental Section.)
Chem. Eur. J. 2003, 9, No. 5
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0905-1127 $ 20.00+.50/0
1127

¬
G. Hilmersson and A. Dahlen
FULL PAPER
IR and NMR data for TMEDA salts: IR (KBr): nÄ ˆ 3446 (vs), 2889 (s), 2650
(vs), 2448 (s), 1617 (m), 1480 (s), 1398 (s), 1150 (w), 994 (s), 972 cmÀ1 (s);
1H NMR (400 MHz, 258C, CDCl3): d ˆ 1.92 (brs, OH), 2.56 (s, 12H; CH3),
2.93 ppm (s, 4H; CH2).
use. The use of triethylamine is recommended as additive in
the SmI2/H2O/amine induced reduction of imines, since it can
be easily removed by evaporation.
IR and NMR data for PMDTA salts: IR (KBr): nÄ ˆ 3450 (vs), 2933 (s), 2720
(s) 2448 (w), 1636 (m), 1472 (s), 1394 (s), 997 (m), 962 cmÀ1 (m); 1H NMR
(400 MHz, 258C, CDCl3): d ˆ 1.80 (brs, OH), 2.38 (s, 3H; CH3), 2.74 (s, 9H;
CH3), 2.86 (s, 4H; CH2), 3.02 (s, 4H; CH2), 3.74 ppm (s, 3H; CH3).
Conclusion
The generality of the instantaneous SmI2/H2O/amine medi-
ated reduction of ketones, a,b-unsaturated esters, and imines
has been shown. It has also been shown that high selectivity
could be obtained in the competition between a,b-unsatu-
rated esters and ketones or imines. The precipitation of
insoluble hydrohalide and samarium salts increases the rate of
reduction. The formation of hydrohalide salts is supported by
the fact that the trialkylphoshine DMPE as well as amines
strongly accelerate these reductions. This was also established
by IR and 1H NMR analysis of the precipitates of the nitrogen
ligands. As a result of the addition of H2O and amine to SmI2
the equilibrium of the reduction is pushed forward, thereby
driving the reaction to completion. The combination of SmI2
in THF, amine, and water certainly qualifies as an excellent
replacement for the toxic HMPA/alcohol mixtures, since it not
only has been shown to bring about quantitative and fast
reductions for several functional groups, but also to simplify
the workup procedures substantially.
Acknowledgement
We are grateful for financial support from the Swedish Research Council
and the Carl Trygger Foundation. We also would like to thank Docent äke
Nilsson for fruitful discussions.
[1] P. Girard, J. L. Namy, H. B. Kagan, J. Am. Chem. Soc. 1980, 102, 2693.
[2] H. B. Kagan, J. L. Namy, P. Girard, Tetrahedron, Suppl. 1981, 9, 175.
[3] D. P. Curran, T. L. Fevig, C. P. Jasperse, M. J. Totleben, Synlett 1992,
12, 943.
[4] H. B. Kagan, J.-L. Namy, Top. Organomet. Chem. 1999, 2, 155.
[5] a) G. A. Molander, Org. React. 1994, 46, 211; b) G. A. Molander,
Chem. Rev. 1992, 92, 29.
[6] P. G. Steel, J. Chem. Soc. Perkin Trans. 2001, 2727.
[7] J. Inanaga, Yuki Gosei Kagaku Kyokaishi 1989, 47, 200.
[8] H. B. Kagan, New J. Chem. 1990, 14, 453.
[9] G. A. Molander, C. R. Harris, Chem. Rev. 1996, 96, 307.
[10] T. Skrydstrup, Angew. Chem. 1997, 109, 355; Angew. Chem. Int. Ed.
Engl. 1997, 36, 345.
[11] G. A. Molander, J. B. Etter, Synth. Commun. 1987, 17, 901.
[12] Z. Huang, H. Jin, D. Duan, Synth. Commun. 2002, 32, 565.
[13] J. L. Namy, J. Souppe, H. B. Kagan, Tetrahedron Lett. 1983, 24, 765.
[14] N. Taniguchi, M. Uemura, J. Am. Chem. Soc. 2000, 122, 8301.
[15] M. Kito, T. Sakai, N. Haruta, H. Shirahama, F. Matsuda, Synlett 1996,
11, 1057.
[16] G. A. Molander, C. R. Harris, J. Am. Chem. Soc. 1996, 118, 4059.
[17] K. Otsubo, J. Inanaga, M. Yamaguchi, Tetrahedron Lett. 1986, 27, 5763.
[18] E. Prasad, R. A. Flowers, II, J. Am. Chem. Soc. 2002, 124, 6895.
[19] R. A. Batey, W. B. Motherwell, Tetrahedron Lett. 1991, 32, 6649.
[20] R. A. Flowers, II, A. Caracoti, J. R. Fuchs, J. Sealy, M. Shabangi,
Recent Res. Dev. Org. Chem. 1999, 3, 141.
[21] M. Nishiura, K. Katagiri, T. Imamoto, Bull. Chem. Soc. Jpn. 2001, 74,
1417.
[22] Z. Hou, Y. Wakatsuki, J. Chem. Soc. Chem. Commun. 1994, 10, 1205.
[23] a) M. Shabangi, J. M. Sealy, J. R. Fuchs, R. A. Flowers, II, Tetrahedron
Lett. 1998, 39, 4429.
Experimental Section
General: In a standard procedure, SmI2 in THF (2.5 equiv, 0.1m, Aldrich)
was added to a dry Schlenk tube, containing a magnetic stirrer bar and
fitted with a septum, inside a glove box under nitrogen atmosphere. The
ligand (5 equiv R3N) and the proton donor, that is, H2O (6.25 equiv), were
added under stirring. The substrate (1 equiv), at 20.08C was then added to
this mixture. The reaction was finished in a few seconds, as soon the
reaction mixture was mixed thoroughly. A white precipitate and a colorless,
clear organic layer indicated the completion of the reaction.
In experiments involving initial rate measurements small portions of the
mixture (100 mL) was removed by syringe and quenched with I2 in n-hexane
(0.1m, 0.1 mL) including n-decan as internal standard when needed.
Diethyl ether (1 mL) and HCl (0.1 mL, 0.12m), to dissolve the inorganic
salts, to the quenched solution was added and finally saturated Na2S2O3
(5 dr.) to remove excess iodine. The clear organic layer was transferred to a
vial and analyzed by GC. This procedure was repeated at least two times to
ensure accurate results.
[24] M. Shabangi, R. A. Flowers, II, Tetrahedron Lett. 1997, 38, 1137.
[25] M. Shabangi, M. L. Kuhlman, R. A. Flowers, II, Org. Lett. 1999, 13,
2133.
[26] A. Dahlen, G. Hilmersson, Tetrahedron Lett. 2002, 43, 7197.
[27] W. Cabri, I. Candiani, M. Colombo, L. Franzoi, A. Bedeschi,
The ketones/alcohols and imines/amines were separated using a CPWAX
52CB column (È ˆ 25 mm, length ˆ 25 m), with hydrogen as carrier gas at a
¬
flow rate of 2 mLminÀ1
. The standard method includes an injector
temperature of 2258C, a column temperature at initially 1008C for 8 min,
then heated to 1658C (108CminÀ1) for 20 min, and finally to 2008C
(108CminÀ1) for 10 min. The detector temperature was 2508C (FID).
Tetrahedron Lett. 1995, 36, 949.
[28] A. Dahlen, G. Hilmersson, Tetrahedron Lett. 2001, 42, 5565.
¬
[29] J. Inanaga, S. Sakai, Y. Handa, M. Yamaguchi, Y. Yokoyama, Chem.
Lett. 1991, 12, 2117.
[30] T.-Y. Lin, M.-R. Fuh, I. S. Chan, J. Chin. Chem. Soc. 2001, 48, 843.
[31] A. Cabrera, H. Alper, Tetrahedron Lett. 1992, 33, 5007.
[32] J. Inanaga, Y. Handa, T. Tabuchi, K. Otsubo, Tetrahedron Lett. 1991,
32, 6557.
[33] B. W. Knettle, R. A. Flowers, II, Org. Lett. 2001, 3, 2321.
[34] F. Machrouhi, J. L. Namy, Tetrahedron Lett. 1999, 40, 1315.
[35] M.-R. S. Fuh, T.-Y. Lin, S.-C. Chang, Talanta 1998, 46, 861.
The a,b-unsaturated esters/saturated esters were separated by using a CP-
Sil 8CB low-bleed column (È ˆ 25 mm, length ˆ 30 m), with hydrogen as
carrier gas at a flow rate of 1 mLminÀ1. The standard method includes an
injector temperature of 2258C, and a column temperature at initially 708C
for 4 min, followed by heating to 2508C (108CminÀ1) for 10 min. The
detector temperature was 2508C (FID).
IR and NMR data for Et3N salts: IR (KBr): nÄ ˆ 3442 (vs), 2950 (vs), 2768
(s) 2680 (s), 2478 (m), 1616 (m), 1464 (s), 1420 (s), 1399 (s), 1166 (m),
1035 cmÀ1 (s); 1H NMR (400 MHz, 258C, CDCl3): d ˆ 1.47 (t, 9H; CH3),
‡
3.17 (q, 6H; CH2), 10.7 ppm (brs, NH ).
Received: July 26, 2002 [F4292]
1128
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0905-1128 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 5