Some aspects of NaBH4 reduction in NMP

Article abstract of DOI:10.1016/S0968-0896(02)00110-4

In our solvent optimization study of NaBH₄ reduction, NMP was found to enhance the reactivity. A chemoselective debromination of the bromide and sulfonates can be attained in the new borohydride reagent system: NaBH₄-LiOTf-NMP. This mixed system worked as an alternative to NaBH₃CN and Bu₃SnH for the S_N2 type displacement of alkylbromide and sulfonate. Also mentioned is an expedient reduction of an azide group into amine by NaBH₄ in NMP without any additive, which offers a convenient protocol for the direct transformation of halides into amines via azide in one flask. Some examples of other reductions were also presented.

Full text of DOI:10.1016/S0968-0896(02)00110-4

Bioorganic & Medicinal Chemistry 10 (2002) 2583–2587
Some Aspects of NaBH4 Reduction in NMP
Yasuhiro Torisawa,* Takao Nishi and Jun-ichi Minamikawa
Process Research Laboratory, Second Tokushima Factory, Otsuka Pharmaceutical Co. Ltd.,
Kawauchi-cho, Tokushima, 771-0182 Japan
Received 24 January 2002; accepted 21March 2002
Abstract—In our solvent optimization study of NaBH4 reduction, NMP was found to enhance the reactivity. A chemoselective
debromination of the bromide and sulfonates can be attained in the new borohydride reagent system: NaBH4-LiOTf-NMP. This
mixed system worked as an alternative to NaBH3CN and Bu3SnH for the SN2 type displacement of alkylbromide and sulfonate.
Also mentioned is an expedient reduction of an azide group into amine by NaBH4 in NMP without any additive, which offers a
convenient protocol for the direct transformation of halides into amines via azide in one flask. Some examples of other reductions
were also presented. # 2002 Elsevier Science Ltd. All rights reserved.
Introduction
In the contemporary synthesis of important drug sub-
stances, selection of a suitable solvent or replacement of
a harmful reagent is obviously the urgent topic to be
concerned. Particularly, safety and ecologically sound
systems are demanded for all basic organic transforma-
tions from the bench-side synthesis to plant scale pro-
duction.1 It is particularly desirable to attain high
degree of efficiency by the use of a safe and conventional
reagent in the optimized solvent system or with new
additives.2 Recent progress for the improved and
chemoselective reduction by NaBH4 is a typical example
of this kind.3 We also focused, among many useful
reactions by NaBH4, on the debromination of alkyl-
bromides to alkanes, which is a straightforward hydride
substitution, but relatively a few options are available
by NaBH4 itself. Actually, we had to carry out a safe
conversion of the bromide (1) into the alkane (2) in our
process research on Aripiprazole.4
Scheme 1. Survey for debromination.
Model Study for Hydride Displacement in NMP
The reduction of alkyl halides and sulfonates by NaBH4
or NaBH3CN was often carried out in a polar aprotic
solvent such as DMSO, sulfolane and HMPA as shown
in Scheme 1.5 Reaction in glyme is also reported.3e We
were interested in the possibility that common amide
We also extended our attention to the reduction of
simple azides into amines. Our aim is to carry out a
convenient conversion of the bromides (or chlorides)
into amines without isolation of the azide intermediate
often claimed to be explosive. Other useful information
is also included with regard to the solvent selection in
NaBH4 reduction.
solvent might work as a convenient alternative.
Although uncontrolled reaction of NaBH4 in N,N-di-
methylformamide (DMF) was noted in the authentic
book,5a no report dealt with the utility of its cyclic con-
gener N-methylpyrrolidone (NMP, shown below) in
NaBH4 reduction. NMP is more stable and less harmful
than DMF.
As summarized in the Table 1, NaBH4 showed good
solubility in NMP, and debromination reaction of the
model bromide (3) or mesylate (4) did proceed at room
*Corresponding author. Tel.: +1-88-665-2126; fax: +1-88-637-1144;
e-mail: torisawa@fact.otsuka.co.jp
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00110-4

2584
Y. Torisawa et al. / Bioorg. Med. Chem. 10 (2002) 2583–2587
Table 1. Model study for the debromination in NMP
Substrate
Reagentsa
Conditions
5 (yields)b
1 3 (1.0 g)
2 3 (3.0 g)
3 3 (3.0 g)c
4 4 (3.0 g)
5 3 (1.0 g)d
NaBH4/DMSO
NaBH4/NMP
NaBH4/NMP
NaBH4/NMP (4 eq)
NaBH3CN/NMP
rt, 1h
rt, 3 h
rt, 3 h
rt, 3 h
ꢁ80 ꢀC, 1h
70%
78%
75%
70%
75%
Scheme 2. Preparation and debromination of 1.
aReagent (2 equiv) was added portionwise to a solution of 3 or 4.
bIsolated yields after silica gel short path are shown.
cCompound 3 was added to a solution of NaBH4 in NMP.
dCompound 3 was added to a solution of NaBH3CN in NMP.
As expected, the reaction of 1 with NaCNBH4 in NMP
gave 2 in high yield after heating at 125 ꢀC for 1h.
This reaction could be performed up to 100 g scales
without any difficulty in the work up stage. Our result
thus clearly demonstrated the practical superiority of
NMP than the other conventional protocols shown in
Scheme 2.
temperature with slight exothermic way (75–78% iso-
lated yields) within 3 h. Reaction of the mesylate (4) was
somewhat slower and larger amount of reagent was
required. In the case of DMSO as a solvent (run 1), a
foamy mixture was developed with unpleasant smell. It
should be noted the same reaction with LiBH4 in NMP
was very slow and took long time to complete the
debromination at room temperature. This indicates a
well-solvated LiBH4 was no more effective for the
reduction of the simple bromide. Our survey indicated
the superiority of NMP as a solvent of choice for the
hydride displacement reaction by NaBH4. As shown in
the last example of Table 1, NMP could supplant
hazardous HMPA in the reaction with NaBH3CN
under gentle heating. Numerous examples are known in
the literature for the reaction of NaBH4 and NaBH3CN
in polar solvents,2 but no example has ever been repor-
ted on the usefulness of NMP in such reductions.
Protocol with Li Salts in NMP
We further searched for the reagent system that is
alternative (or supplant) to the toxic reagent NaBH3CN
especially for the conversion of 1 to 2. Reaction of 1
with NaBH4 in NMP at room temperature was not
satisfactory to afford 2 in less than 20% yield after 3 h
stirring. We however found that the reduction of 1 with
added Li-salt was the most convenient and successful in
affording 2 after 10 h stirring at room temperature. In a
large-scale reaction with LiCl (0.5ꢁ1equiv), it was
found that insoluble NaCl formed a suspension and
reaction mixture was heated at 60 ꢀC for the completion.
In a search for more effective salt, we found that LiOTf
was a convenient alternative. Addition of LiOTf (0.5ꢁ1
equiv) resulted in a more homogeneous solution than
LiCl and reaction was almost complete at room tem-
perature. In the reactions with 20 g of bromide, it was
convenient to warm the mixture (room temperature to
60 ꢀC) to complete the reaction within 1h. Further
heating after the complete consumption of 1 slowly
caused a conversion of the initial product (2) into the
polar product. The tentative structure (6), in which
demethylation of one methoxy group took place, was
most probable. This unwanted side-reaction was com-
pletely suppressed by the addition of ethereal co-solvent
such as DEM (diethoxymethane, ethylal) as shown
below, which is detailed in Experimental.
Debromination of the Bromide (1)
Next, investigation focused on the debromination of 14
with the conditions mentioned above as well as conven-
tional conditions just for comparison. An efficient
method for the debromination of 1 into 2 was required
in our process research on Aripiprazole and its
congeners.4 Reaction sequence for the preparation of 1
was summarized in Scheme 2 and detailed in Experi-
mental.6
Reduction of 1 under various conditions is summarized
in the Scheme 2 and Table 2. The routine radical reac-
tion of 1 with 2 equivalents of Bu3SnH afforded the
product (2) in 45% yield along with recovered bromide
(1). Use of excess of tin reagent was effective but incon-
venient in a large-scale preparation of 2. Another pro-
tocol using excess NaBH4 in DMSO was satisfactory to
give 2 in 60% but again unpleasant work up (smell
badly) was an obvious nuisance.
We are particularly interested in the use of DEM,
because of its emerging utility in the process research.7
The same effect was observed with other metal triflates
such as In(OTf)3 and Bi(OTf)3, but obviously LiOTf is
most cost effective (Scheme 3).
In order to expand the utility of this protocol, we briefly
compared the reaction of the triflate (T) and that of the
alcohol (A) as shown below. Reaction of the triflate was
fast without any assistance of metal triflate, while
The reaction of 1 with excess NaBH3CN in HMPA was
more reliable to afford 2 in 70% yields after purification.

Y. Torisawa et al. / Bioorg. Med. Chem. 10 (2002) 2583–2587
2585
Table 2.
1Methods
Conditions
2 (yields)
1 4.5 g
2 2.5 g
3 2.5 g
(a) Bu3SnH/toluene
(b) NaBH4/DMSO
(c) NaBH3CN/HMPA
reflux, 3 h
ꢁꢀ, 1h
45%
60%
70%
ꢁ100 ꢀC, 1h
4 21.0 g
5 2.0 g
6 2.0 g
7 10.0 g
NaBH3CN/NMP
NaBH4/LiCl/NMP
NaBH4/LiOTf/NMP
NaBH4/LiOTf/NMP
ꢁ125 ꢀC, 1h
89%
85%
85%
88%
rt, 30 h
rt, 30 h
rtꢁ60 ꢀC, 1h
Scheme 3.
Table 3. Azidation-Reduction Sequence in NMP
Scheme 4.
Substrate
Aa
Bb
9c
[NMR]d By-producte
1 7 (2.0 g) ꢁ70 ꢀC, 3 h 50ꢁ70 ꢀC, 3 h 0ꢁ5% [7%]
11, 13f
2 7 (2.0 g) ꢁ80 ꢀC, 2 h 50ꢁ70 ꢀC, 2 h 55% [60%] 11
3 8 (2.0 g) ꢁ70 ꢀC, 2 h 50ꢁ70 ꢀC, 2 h 75% [75%] 13f (minor)
4 9 (2.0 g) ꢁ75 ꢀC, 2 h 50ꢁ70 ꢀC, 2 h 65% [70%] 12 (minor)
almost no reaction was observed in the reaction of
alcohol even though triflate salt was present in the mix-
ture (Scheme 4).
aSubstrates (7, 8, 9)-NaN3 (1.2 equiv)-H2O were mixed at rt and
heated as indicated.
bNaBH4 (2 equiv) was delivered at rt and heated as indicated with
adequate stirring.
cAmine obtained after extractive isolation.
dYields based on NMR analysis of the crude extracts with a standard.
eCharaterized by-products are shown.
Reduction of Azides
After the initial progress shown above, we further
focused on some advantageous aspect of the reaction in
NMP. We observed that the reaction of NaBH4 was
faster than that of LiBH4 in NMP in the model experi-
ment with 3 as mentioned. These results prompted us to
carry out two successive substitution reactions by
sodium salt in NMP.
fIsolated by preparative TLC.
temperature around 50 ꢀC. Yields of the free amine after
extractive isolation were approximately 75% for bro-
mide (8) and 55% for the less reactive chloride (7) as
shown in the Table 3.
Nucleophilic substitution of halides by NaN3 affords
azide intermediate in NMP, whose reduction might fur-
nish the corresponding amine in one flask by the present
reduction. Although a convenient one-pot conversion
from halides into amines via azide is reported via Stau-
dinger reaction,8 no reports have attained such trans-
formation by the use of NaBH4 as the sole reagent.
Recent reports indicated some successful reduction of
azides to amine by NaBH4 with proper selection of the
additives such as CoCl2, ZrCl4, and tin-based reagent.9
Our improved protocol by the simplest reagent compo-
sition without additives are summarized below.
In the case of the chloride (7), crude mixture contained
starting chloride and reduced 3-phenylpropane (11).
This side product was obtained in the case of chloride
through the reduction with the unreacted chloride. In a
separate experiment, direct reduction of the chloride (7)
by NaBH4 (2 equiv) at 70ꢁ85 ꢀC for 2 h gave 11 in a
65% isolated yield. Reaction of the mesylate (9) was
also successful, but we observed a small amount of the
alochol (12) was always accompanied through the
hydrolysis of 9 during azide formation under heating.
In this way, we have succeeded in a one-pot conversion
of alkyl halide into amine, but caution must be taken to
detect side-products present in the mixture including the
amine borane complex such as 13. We are now investi-
gating the utility of this one-pot sequence in a co-solvent
system with more demanding secondary halides.
The commercially available phenylpropyl chloride (7)
and bromide (8) were selected as a model to test for our
one-pot protocol. Formation of the primary azide from
halides in NMP was a conventional pathway as in
DMF. We have optimized this step by the addition of
H2O (1ꢁ2 equiv) under careful temperature control
(<70 ꢀC). Thus, the reaction of the bromide (8) with
NaN3 was a straightforward reaction at 60ꢁ70 ꢀC,
while the reaction of the chloride (7) required higher
temperature than 70 ꢀC. Reduction of the resulting
azides was done in one-pot with 2 equiv NaBH4 at
In the related study, we have briefly surveyed on the
reduction of acid chlorides into aldehyde with 1equiv
NaBH4 in the presence of amine (TMEDA). Our initial
survey indicated a facile conversion of simple acid
chlorides (benzoyl chloride or toluoyl chloride) into the
corresponding aldehydes at temperature below 0 ꢀC in

2586
Y. Torisawa et al. / Bioorg. Med. Chem. 10 (2002) 2583–2587
ca.50ꢁ60% yield after one distillation. In these study,
we employed toluene, hexane, DME, TBME, and DEM
as a co-solvent. As mentioned before, DEM was a con-
venient solvent for the direct extractive isolation of the
crude product after quench by H2O.
and dilution with CH2Cl2. The insoluble reagent was
then recovered through filtration and washing with
CH2Cl2. The filtrate was directly concentrated to give
1
crude product as a brown oil (61g). H NMR (CDCl3):
d 1.28 (3H, t, J=4 Hz), 3.90 (6H, s), 4.25 (2H, q,
J=4 Hz), 5.70 (1H, s), 6.82 (2H, d, J=8.7 Hz), 6.94
(2H, d, J=8.7 Hz). 13C NMR (CDCl3): d 171.2; 155.4;
134.0; 128.6; 127.2; 122.3; 109.5; 61.5; 56.2; 52.0; 13.9.
MS (EI): m/z 436 (M+).
Conclusion
We have surveyed the reaction of borohydrides in NMP
or other solvents as a convenient alternative for the
reduction of some halides and azides. Although some
explosive reaction of NaBH4 in DMF was reported, we
observed no such undesirable reaction in NMP, which is
less harmful than DMF. Usually, the reaction mixture
formed a homogeneous solution, in which NaBH4 was
not easily decomposed. Simple work up (dilution with
H2O) gave nearly pure product as indicated in the
experimental procedures shown below. The addition of
LiOTf was particularly effective to attain chemoselective
debromination of the labile bromide. The present sys-
tem thus worked as an alternative to LiBH4 and
NaCNBH3, both of them obviously hazardous in large-
scale operation.
To a stirred solution of the above ester (61g, ca.
128.5 mM as ester) in DME (280 mL) was added a
powder of ZrCl4 (Merck, 30 g, 128.5 mM) and the sus-
pension was stirred for 30 min.
To this was added very carefully NaBH4 (5 gÂ2,
264.5 mM) portionwise at room temperature. Resulting
suspension was stirred overnight before it was gradually
heated to 60 ꢀC. Resulting white mixture was cooled to
room temperature and kept in an ice-bath, to which was
added AcOEt (10 mLÂ4) followed by H2O (1 0 mLÂ3)
very carefully (gas evolution). Then the mixture was
stirred for a while before CH2Cl2 (800 mL) was added.
The insoluble salts were removed by filtration. The clear
filtrate was transferred to a separatory funnel to obtain
a clear two-phase solution. Lower organic layer was
separated, which was washed with further amount of
H2O. Resulting organic layer was dried over Na2SO4
and concentrated to afford the corresponding alcohol
colorless oil (50.5 g), which was solidified on standing.
1H NMR (CDCl3): d 1.65 (1H, OH), 3.89 (6H, s), 4.08
(2H, d, J=6 Hz), 4.97 (1H, t like), 6.82 (2H, d,
J=8.7 Hz), 7.12 (2H, d, J=8.7 Hz). 13C NMR (CDCl3)
d 154.5; 134.2; 131.3; 126.7; 122.3; 109.5; 63.3; 56.2;
47.0. MS (EI): m/z 394 (M+). MS (EI): m/z 394 (M+);
HRMS: calcd for C16H14Cl4O3: 393.9699; found:
393.9740.
Although NMP was frequently used in the reaction of
many transformations as a beneficial solvent, little is
known about the behavior of NMP during the reac-
tion. The most characteristic nature of the reduction in
NMP was that NaBH4 (or related species) has a long
lifetime to reduce halide and azides. Further applica-
tion to other functionality will give us more insight
into the nature of the reduction, as well as compar-
ison with other solvent system, especially with such
as ionic liquids, which are now in progress in this
laboratory.10
To a stirred solution of the above alcohol (50 g, ca.
120 mM) in CH2Cl2 (300 mL) was added Ph3P (40 g,
152.5 mM) and CBr4 (50 g, 150.7 mM) under ice-cooling.
(an exothermic reaction). Resulting faint yellow mixture
was stirred without cooling for 20 h, and further
amount of reagents were delivered. Resulting yellow
mixture was then concentrated to give a crude residue,
which was diluted with toluene (500 mL) and the whole
was stirred to precipitate the polar wastes (phosphine
oxide). After stirring for 1h at room temperature and
0.5 h under ice-cooling, the mixture was filtrated to
remove polar material. Evaporation of the resulting fil-
trate then gave a crude product (ca. 35.5 g), which was
first triturated with AcOEt/n-hexane=1/4 (ca. 200 mL)
to afford crude crystalline crop. Dilution of this solid
with AcOEt/n-hexane=1/10 (ca. 200 mL) then gave a
white precipitate, which was isolated by filtration fol-
lowed by washing with n-hexane and drying to afford
Experimental
Material and instrumentation
Chemicals were purchased from the commercial firms
indicated and used without further purification. TLC
analysis was carried out using Merck silica gel 60 F254
plate (Art 5715). H and 13C NMR spectra were mea-
sured at 300 and 75 MHz, respectively, with tetra-
methylsilane as the internal standard. Mass spectra
(LR-MS) were recorded with a Shimadzu GCMS-
QP1000 spectrometer at 70 eV.
1
Preparation of the bromide (1)
As in the same procedure described before,6 ethyl
glyoxylate polymer (TCI, 50% in toluene, 40 mL) was
concentrated under reduced pressure to obtain neat
glyoxylate (ca. 20 g). To this was added solid dichloro-
anisole (Aldrich, 45.5ꢀg, 257 mmol) and the whole was
heated at around 80 C to form a clear homogeneous
solution. To this was then added Bi(OTf)3 powder
(1.0 g)6 and the resulting mixture was gradually heated
to 110 ꢀC (ca. 1h). The mixture was further heated at
ꢁ135 ꢀC for 1h, before cooling to room temperature
1
nearly pure bromide (1, 21g). H NMR (CDCl3): d 3.76
(2H, d, J=7.5 Hz), 3.90 (6H, s), 5.17 (1H, d, J=7.5 Hz),
6.83 (2H, d, J=8.2 Hz), 6.69 (2H, d, J=8.2 Hz). 13C
NMR (CDCl3): d 155.2; 134.2; 131.3; 126.3; 126.3;
109.4; 56.2; 46.9; 31.6. MS (EI): m/z 456, 458 (M+). MS
(EI): m/z 456, 458 (M+); HRMS: calcd for
C16H13BrCl4O2: 455.8855; found: 445.8864.

Y. Torisawa et al. / Bioorg. Med. Chem. 10 (2002) 2583–2587
2587
General procedure for the reduction in NMP (Table 2
entry 3)
Acknowledgements
We are grateful to Professors Bakthan Singaram
(UCSB) and Kozo Shishido (University of Tokushima)
for their informative commentary on the NaBH4
reduction and assistance for manuscript preparation.
We also thank Mr. I. Miura (NMR), Mr. K. Tada (MS)
and Mr. M. Nagasawa (MS) for their kind assistance in
the analysis.
To a stirred solution of NaBH4 (Wako, 1.0 g, 2 equiv) in
MNP (Wako, 30 mL) was added a solution of bromide
(3, 3.0 g)11 in NMP (2 mL) and resulting mixture was
stirred at room temperature for 3 h.
The mixture was first diluted with solvent and quenched
carefully with H2O in the cold. After vigorous gas evo-
lution subsided, the mixture was further diluted with
aqueous solvent. Repeated extraction gave products
after evaporation of the dried extracts. Through a SiO2
short column, a nearly pure material was obtained as a
colorless oil (5, 75%).
References and Notes
1. For example, see: Bolm, C.; Beckmannn, O.; Dabard, A. G.
Angew. Chem., Int. Ed. Engl. 1999, 907.
2. (a) As our own recent efforts on the topic, see: Torisawa,
Y.; Nishi, T.; Minamikawa, J. Bioorg. Med. Chem. Lett. 2000,
2493. (b) Torisawa, Y.; Nishi, T.; Minamikawa, J. Bioorg.
Med. Chem. Lett. 2000, 2489.
NaBH3CN reduction in NMP
To a stirred solution of NaBH3CN (Aldrich, 10 g) in
NMP (100 mL) was added carefully at 60 ꢀC the solid
bromide powder (1, 21g) and resulting mixture was
stirred at 100–125 ꢀC for 1h. The mixture was cooled in
ice bath and worked up as above. Repeated extraction
gave a nearly pure solid product (ca. 23 g) after eva-
poration of the dried extracts. By a simple trituration, a
nearly pure material was obtained as a white solid (2,
89%). 1H NMR (CDCl3): d 1.52 (2H, d, J=7.2 Hz) 3.88
(6H, s), 4.82 (1H, q, J=7.2 Hz), 6.83 (2H, d, J=8.2 Hz),
6.69 (2H, d, J=8.2 Hz). 13C NMR (CDCl3): d 154.5;
136.0; 133.5; 125.5; 122.0; 109.5; 56.2; 39.1; 19.9. MS
(EI): m/z 378 (M+). MS (EI): m/z 378 (M+); HRMS:
calcd for C16H14Cl4O2: 377.9750; found: 377.9750.
Anal. calcd for C16H14O2Cl4: C, 50.46; H, 3.72; found
C, 50.16; H, 3.75.
3. (a) As recent efforts for improved reduction protocols, see:
Gevorgyan, V.; Rubin, M.; Liu, J.-X.; Yamamoto, Y. J. Org.
Chem. 2001, 66, 1672 and references therein. (b) Narasimhan,
S.; Balakumar, R. Synth. Commun. 2000, 30, 4387 and
references therein. (c) Crich, D.; Neelamkavil, S.; Sartillo-Piscil,
F. Org. Lett. 2000, 2, 4029. (d) Chary, K. P.; Ram, S. R.;
Salahuddin, S.; Iyengar, D. S. Synth. Commun. 2000, 30, 3559.
(e) Desai, D. G.; Swami, S.; Hapase, S. B. Synth. Commun. 1999,
29, 1033. (f) Yang, C.; Pittman, C. U., Jr. Synth. Commun.
1998, 28, 2027.
4. (a) Torisawa, Y.; Nishi, T.; Minamikawa, J. Bioorg. Med.
Chem. Lett. 2001, 2787. (b) Oshiro, Y.; Sato, S.; Kurahashi,
N.; Tanaka, T.; Kikuchi, T.; Tottori, K.; Uwahodo, Y.; Nishi,
T. J. Med. Chem. 1998, 41, 658.
5. (a) For an overview, see: Banfi, L.; Narisano, E.; Riva, R.
In Encyclopedia of Reagents for Organic Synthesis; Paquette,
L. A., Ed.; John Wiley & Sons: New York, 1995; Vol. 7, p
4522. (b) Hutchins, R. O.; Hutchins, M. K. In Encyclopedia of
Reagents for Organic Synthesis; Paquette, L. A., Ed.; John
Wiley & Sons: New York, 1995; Vol. 7, p 4539.
Reduction by NaBH4-LiOTf in NMP
6. Torisawa, Y.; Nishi, T.; Minamikawa, J. Org. Process Res.
Dev. 2001, 84.
7. Boaz, N. W.; Venepalli, B. Org. Process Res. Dev. 2001, 5, 128.
8. Koziara, A.; Zwierzak, A. Synthesis 1992, 1063.
9. (a) Fringelli, F.; Pizzo, F.; Vaccaro, L. Synthesis 2000, 646.
(b) Bosch, I.; Costa, A. M.; Martin, M.; Urpi, F.; Vilarrasa, J.
Org. Lett 2000, 397.
To a stirred solution of NaBH4 (3 g) and LiOTf (Strem,
3 g) in NMP (50 mL) was added at room temperature
the crystalline bromide (1, 10 g) and resulting mixture
was stirred at room temperature for 0.5 h (a gentle exo-
thermic reaction) and further at 60 ꢀC for 0.5 h. The
mixture was carefully worked up as above. A solid pro-
duct (ca. 12.5 g) was obtained after evaporation of the
dried solvents, which was further purified by SiO2 col-
umn chromatography to give nearly pure material as a
white solid (2, 88%).
10. In our preliminary survey, aromatic nitro group was
reduced to aniline derivative on heating with excess of NaBH4
in NMP. In this reduction, a thermal runaway reaction was
observed in the reaction with >20 mM scale, which indicated
impractical aspect of NaBH4-NMP system for nitro reduction.
Nitrobenzene itself was not reduced to aniline even under
forced conditions. Some aromatic esters were also reduced to
the corresponding alcohols with excess NaBH4 after pro-
longed heating. The combination NaBH4-NMP-LiOTf was
particularly useful for the reductive amination reaction
between aldehydes and anilines. Imine-forming reactions
between anilines with aldehydes were successfully carried out
in DEM by the aid of LiOTf and MgSO4. Thus, one-pot
reductive amination was possible in LiOTf-DEM followed by
the treatment of NaBH4-NMP. Progress of this new protocol
will be reported in due course. Furthermore, addition of the
simplest ionic liquid such as EMIM(OTf) did improve the
course of these borohydride reduction.
NaBH4 reduction in NMP-DEM
To a stirred solution of NaBH4 (1g) in NMP (16 mL)
was added at room temperature LiOTf (1g) powder to
form a clear solution. A suspension of the bromide (1,
3 g) in DEM (TCI, 8 mL) was added in one portion and
resulting mixture was kept stirring at room temperature
for 0.5 h, followed by warming at around 40 ꢀC. After
exothermic process with gas evolution, the mixture was
further kept at 60 ꢀC for 1h. The resulting mixture was
cooled in a water-bath and carefully quenched by the
addition of H2O. Further dilution made the product
precipitated, which was filtered and washed with H2O,
furnishing nearly pure material as a white solid (2, 80%).
11. The bromide 3 was prepared from commercially available
5-phenyl-1-pentanol (Aldrich) by the treatment of CBr4 and
PPh3 in CH2Cl2 under cooling.