Reduction of amides with NaBH4 in diglyme at 162°C

Article abstract of DOI:10.1039/b208464c

High temperature (162°C) reductions of aromatic amides were studied to extend the useful range of functional group transformations by NaBH₄. Primary aromatic amides were reduced to the amines with NaBH₄-diglyme at 162°C. Reduction proceeds via fast initial loss of hydrogen, followed by formation of the corresponding nitrile, which is then more slowly reduced to the amine. N-Methylbenzamide is not reduced under these conditions, but it is reduced to benzylmethylamine when LiCl is added to NaBH₄-diglyme at 162°C. LiCl addition raised the rate of primary aromatic amide and aromatic nitrile conversions to both the nitrile, first, and the amine. An intermediate was isolated from the reaction of N-benzylformamide with NaBH₄-LiCl in diglyme at 162°C. It was examined by ¹H NMR, atomic absorption, IR and thermal decomposition. Possible structures are proposed. A mechanism for the reduction of primary aromatic amides is proposed based on the initial evolution of one mole equivalent of hydrogen and formation of the nitrile prior to further reduction to amine.

Full text of DOI:10.1039/b208464c

View Article Online / Journal Homepage / Table of Contents for this issue
Reduction of amides with NaBH4 in diglyme at 162 ꢀC
Hua-Jie Zhu, Kai-Tao Lu, Guang-Ri Sun, Jin-Bao He, Hai-Qing Li and Charles U. Pittman, Jr.
Department of Chemistry, Mississippi State University, Box 9573, Mississippi State,
MS 39762, USA
Received (in New Haven, CT, USA) 30th August 2002, Accepted 9th October 2002
First published as an Advance Article on the web 2nd January 2003
High temperature (162 ꢀC) reductions of aromatic amides were studied to extend the useful range of
functional group transformations by NaBH4 . Primary aromatic amides were reduced to the amines with
NaBH4–diglyme at 162 ꢀC. Reduction proceeds via fast initial loss of hydrogen, followed by formation of the
corresponding nitrile, which is then more slowly reduced to the amine. N-Methylbenzamide is not reduced
under these conditions, but it is reduced to benzylmethylamine when LiCl is added to NaBH4–diglyme at
162 ꢀC. LiCl addition raised the rate of primary aromatic amide and aromatic nitrile conversions to both the
nitrile, first, and the amine. An intermediate was isolated from the reaction of N-benzylformamide with
1
NaBH4–LiCl in diglyme at 162 ꢀC. It was examined by H NMR, atomic absorption, IR and thermal
decomposition. Possible structures are proposed. A mechanism for the reduction of primary aromatic amides
is proposed based on the initial evolution of one mole equivalent of hydrogen and formation of the nitrile prior
to further reduction to amine.
The reduction of amides by hydride reagents using a variety of
different metal cations and solvents is well known.1–4 For
can be reduced to amines in moderate yields in NaBH4–
diglyme at 162 ꢀC in the presence of LiCl. Furthermore, we
show that LiCl addition enhances the rate of primary amide
reduction by speeding the reduction of the intermediate nitrile.
Ellzey et al.12 obtained 59% and 71% yields, respectively, of
benzonitrile when attempting to reduce benzamide to benzyl-
amine with 1.1 equiv. or a substantial excess of NaBH4 in
refluxing diglyme. No benzylamine was obtained. Newman
and Fukunga13 reported that ‘‘appreciable’’ yields of benzoni-
trile were obtained from reductions of benzamide by LiAlH4 in
THF. A mechanism, proceeding through initial deprotonation
of the amide nitrogen followed by conversion to the dianion,
PhC(Oꢁ Li+) ¼ Nꢁ Li+, was postulated based on the evolu-
tion of 2 equiv. of H2 . Both NaBH4 and LiAlH4 are good
hydride donors but LiAlH4 is far more active. In addition,
the relative roles of the Li+ versus Na+ counterions are differ-
ent. Thus, simple parallels between these two reagents do not
always exist. No mechanistic information is now available in
the literature on amide reductions by NaBH4 in refluxing
diglyme.
5
example, TiCl4–NaBH4 or NaBH4 in trifluoroacetic acid6,7
reduce amides to amines. When used alone, NaBH4 reduces
aldehydes, ketones, acid chlorides, and in some cases esters,
but not carboxylic acids, amides, nitriles, nitro compounds
or halogenated organic molecules. Since NaBH4 is safe to han-
dle and far cheaper than more active hydride agents, such as
LiAlH4 , NaBEt3H or NaB(OR)3H, we have sought to expand
the range of reductions that NaBH4 can achieve by investigat-
ing its use at higher temperatures in glyme solvents.8–10 Thus,
the reductions of carboxylic acids,8 esters,8 nitriles,8 benz-
amide,8 4-chlorobiphenyl9 and pentachlorophenol10 with
NaBH4 were successfully achieved at temperatures from 120–
290 ꢀC in glyme solvents. Despite the high thermal stability
of NaBH4 , very few studies of its use in reductions at high
temperatures have been carried out because it is usually used
in alcohol solvents. Alcohols or other protic solvents react with
NaBH4 to generate hydrogen as temperature increases. Addi-
tionally, NaBH4 has a very low solubility in most solvents at
elevated temperatures.
We now report that aromatic primary amides are converted
first to their corresponding nitriles with the evolution of one
equivalent of H2 upon treatment with NaBH4 in diglyme at
162 ꢀC. Subsequently, the nitriles are further reduced to amines
under these conditions in the presence of LiCl. The nitrile con-
version is very slow without LiCl present. Furthermore, evi-
dence is presented that the reduction mechanism does not
involve thermal dehydration to the nitrile nor dehydration to
the corresponding imine.
Results and discussion
Reductions with NaBH4 at 162 ꢀC
The reduction of 2-chlorobenzamide, 1, with NaBH4 in
diglyme at 162 ꢀC produced 2-chlorobenzonitrile 2 in 2 h [entry
1, Table 1; eqn. (1)].
NaBH4 reduction of amides in refluxing pyridine to modest
yields of amines was previously reported.11 Primary aromatic
amines dehydrated to nitriles in either very low or trace
yields.11 Thus, benzamide and p-methoxybenzamide produced
a ‘‘trace’’ of benzonitrile and 27% of p-methoxybenzonitrile,
respectively, using 3 equiv. of NaBH4 .11 Secondary amides
were inert under these conditions and tertiary amides gave
moderate yields of the corresponding tertiary amines.11
Herein, we now demonstrate that secondary aromatic amides
ð1Þ
Nearly one mole of H2 gas per mole of amide present was
formed quickly (within 15 min). However, at this time
only ꢂ16% of the substrate had been converted to nitrile.
DOI: 10.1039/b208464c
New J. Chem., 2003, 27, 409–413
409
This journal is # The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2003

View Article Online
Table 1 Conversion of amides to nitriles with hydride reductants
Hydride:substrate
mole ratio
% Substrate
consumeda
Entry Substrate (Hydride reagent)
Time/h T/ꢀC
% Yield nitrileb
% Yield amineb
1
2-Chlorobenzamide, 1 (NaBH4)
1:1
0.25
0.5
1.0
2.0
2.6
1.0
1.0
1.0
1.0
2.0
4.0
6.3
0.5
1.0
2.0
3.0
1.0
2.0
3.3
2.0
2.0
162
16
43
2
40
67
84
77
2
0
0
68
86
0
0
93
92
4
2
3
2-Chlorobenzamide, 1 (LiAlH4)
2-Chlorobenzamide, 1 (NaBH4)
1:1c
1:1d
0.6:1d
1.5:1
78
78
14
0
100
74e
37
42
36
33
61
63
46
8
78
162
1.5
0
77
96
0
3.4
9
100
11
4
5
4-Chlorobenzamide, 6 (NaBH4)
4-Fluorobenzamide, 7 (NaBH4)
1:1
1:1
162
162
0
65
86
60
73
68
68
74
64
55
–
0
0
99
83
0
0
93
0
100
100
0
0
6
7
2,2-Dimethylpropanamide, 8 (NaBH4)
N-Methylbenzamide, 9 (NaBH4)
1.5:1:2
1:1
162
162
44
–
a
b
The calculation was based on GC analysis using a DB-5 column. The yields of the nitriles and amines were obtained by GC analysis and are
c
based on the amount of amide substrate present initially. LiAlH4–THF solution was injected via syringe into a refluxing solution of 2-chloro-
d
benzamide in THF. LiAlH4–THF solution was added to the solution of 2-chlorobenzamide in THF at room temperature; the solution then
e
was heated to reflux. There are three unknown compounds (23%, 3% and 4% area percents in GC) also formed in the reaction.
Therefore, the mechanism could not have proceeded via reduc-
tion of the carbonyl group to a hydroxyl function, followed by
dehydration and subsequent dehydrogenation (Scheme 1).
This dehydration route requires the generation of two moles
of H2 per mole of nitrile formed and needs an available proton
source, which was not present in the thoroughly dried diglyme
used. Simple thermal dehydration (Scheme 2) would generate
one mole of water, which would rapidly give one mole of
hydrogen per mole of amide consumed due to rapid reaction
in refluxing diglyme are consistent with the mechanism in
Scheme 3, where 1 gives the azenolate borohydride 4 via 3.
Elimination via transition state 5 generates 2.
NaBH4 reductions of 4-chlorobenzamide, 6, 4-fluorobenz-
amide, 7, and 2,2-dimethylpropanamide, 8, all produced the
corresponding nitriles (Table 1). In contrast, the secondary
amide, N-methylbenzamide, 9, was inert to NaBH4 in refluxing
diglyme.
Amide 1 also reacted with LiAlH4 (0.6 and 1.0 equiv.) in
THF at 78 ꢀC over 1 h to produce nitrile 2 (entry 2, Table
1). Formation of nitrile during the LiAlH4 reduction of amides
agrees with Newman’s previous observations.13 Only a small
amount (2%) of nitrile was observed at 92% consumption of
substrate 1 after 1 h using LiAlH4 (1 equiv.), which had been
injected as a THF solution directly into a refluxing THF solu-
tion of 1. In this case 14% amine was formed. Addition of
LiAlH4 to 1 in THF at room temperature followed by heating
to 78 ꢀC produced 42% and 36% nitrile in 1 h at 100% and 74%
substrate consumptions, respectively. Clearly, nitrile formation
occurs prior to the reduction to amine. Furthermore, substan-
tial amounts of an intermediate species are present at various
times. In agreement with Newman’s results,13 a dehydration
route is not operating. Replacing NaBH4 with LiAlH4 in the
dehydration mechanism (shown in Scheme 2) shows that only
0.25 equiv. of LiAlH4 is required to remove the 1.0 equiv. of
water that would be generated per mole of amide consumed.
However, about 26% of the amide substrate remained
unreacted when 0.6 equiv. of LiAlH4 was used (entry 2, Table
1). Taken together, these results for both NaBH4 and LiAlH4
ꢁ
of that water with BH4 at 162 ꢀC. However, this route
requires that the generation of this hydrogen must coincide,
stoichiometrically, with the consumption of amide. An equiva-
lent of hydrogen could not be generated before an equivalent
of the amide disappeared in the mechanism shown in Scheme
2. In contrast to Schemes 1 and 2, our experiments demon-
strated that the mole of H2 formed was produced long before
an equivalent of the amide was consumed. Furthermore, no
reaction was observed and amide 1 was recovered unchanged
when 1 was heated to 162 ꢀC in diglyme for 1 h in the absence
of NaBH4 . Thus, borohydride must be involved in the
mechanism. Finally, it is clear from entry 1 (Table 1) that sub-
stantial amounts of the substrate consumed are present as
some intermediate species that is increasingly converted into
products with time. All these results for NaBH4 reductions
Scheme 1
Scheme 2
410
New J. Chem., 2003, 27, 409–413

View Article Online
ð2Þ
ð3Þ
Scheme 3
ð4Þ
rule against the dehydration mechanism during conversion of 1
to 2. Furthermore, about two moles of H2 was evolved using
LiAlH4 versus only one mole when NaBH4 was used. This
major difference is consistant with the mechanism shown in
Scheme 3 for the NaBH4 reductions. A different route for
LiAlH4 reductions is indicated by this finding [such as forma-
tion of an intermediate dianion, o-ClC6H4C(Oꢁ Li+) ¼ Nꢁ
Li+, originally advanced by Newman13].
ð5Þ
One equivalent of LiCl per mole of amide increased the rate of
reduction of primary aryl amides, 1, 6 and 7, to the corre-
sponding amines 14, 11 and 16, respectively (entries 1–4, Table
2). However, the nitrile was still formed first when LiCl is pre-
sent (see entry 1, Table 2). Reduction of o-chlorobenzamide, 1,
with 1 equiv. of LiCl and NaBH4 produced 55% amine in 5.5 h
Reduction with NaBH4–LiCl at 162 ꢀC
Addition of LiCl to the NaBH4 amide reductions increased the
rate of amine formation [eqns. (2)–(5), Table 2].
Table 2 Reduction of amides with NaBH4 at 162 ꢀC in diglyme in the presence of LiCl
NaBH4:substrate:LiCl
mole ratio
% Substrate
Time/h consumeda
% Isolated
aminec
Entry Substrate
% Yield nitrileb
% Yield amineb
1
2-Chlorobenzamide, 1
1:1:1
1.0
3.0
5.5
8.0
2.0
4.0
2.0
4.0
6.0
2.0
4.0
6
39
83
19
32
16
11
18
2
7
23
55
36
37
46
6
–
99
100
100
100
94
2
3
2-Chlorobenzamide, 1
4-Fluorobenzamide, 7
1.5:1:2
1.5:1:2
41
53
30
21
1
97
43
69
62
64
52
53
19
38
38
55
34
50
67
100
100
100
59
4
5
6
7
4-Chlorobenzamide, 6
N-Methylbenzamide, 9
N-Methylbenzamide, 9
N-Benzylformamide, 10
1.5:1:2
1.5:1:1
1.5:1:2
1.5:1:1
7
59
52
–
1
0
10
2
64
32
0
0
9
51
51
0
1.5
3.5
6.0
2.0
4.0
9.0
5.0
6.0
5.0
5.0
2.0
1.5
0
–
55
55
0
0
8
9
N-Benzylformamide, 10
N,N-Diethylacetamide, 23
1.5:1:2
62
74
0
74
52
0
83
0
0
83
d
1:1:0
1:1:1
0
d
d
d
5
19
0
1.5:1:1
1.5:1:2
0
87
0
10
11
N,N-Dimethylbenzylamide, 24 1.5:1:1
d-Valerolactam, 25 1.5:1:1
100
100
0
65
e
68
62
–
a
The calculation was based on the GC analysis using a DB-5 column. b The yields of the nitrile and amine were based on the consumed substrate
and calculated based on GC using internal standard quantitation techniques. c The yield was based on the consumed substrate. d The low boiling
triethylamine was not quantitated by GLC but it could be captured by a cold trap at ꢁ78 ꢀC. e Quantitation of piperidine by GLC was difficult due
to the presence of an overlapping solvent degradation product.
New J. Chem., 2003, 27, 409–413
411

View Article Online
8.3 ppm for the proton at the carbonyl carbon. About 30%
(see entry 1, Table 2) versus only 9% amine in 6.3 h without
LiCl (despite having 1.5 equiv. of NaBH4 present; see entry
3, Table 1). Similarly, at 2 and 4 h the amine yields obtained
from 1 were, respectively, 0 and 3% without LiCl versus 37
and 55% (isolated) when 2 equiv. of LiCl were present under
identical conditions (compare entry 2, Table 2 to entry 3, Table
1). Plots of the product distributions versus time for all the
primary amides indicated that the effect of LiCl is largely the
result of speeding up the conversion of the intermediate nitrile
to amine. p-Chlorobenzamide, 6, and p-fluorobenzamide, 7,
did not give any amine formation unless LiCl was added.
LiCl is required for practical conversion of aromatic amides
to their amines with NaBH4 in refluxing diglyme. The presence
of Li+ ion probably enhances the rate of nitrile reduction to
amine due to its Lewis acid-like coordination to the nitrile
nitrogen. Furthermore, the lithium cation, whicꢁh can be tightly
coordinated by diglyme, may enhance BH4 solubility in
diglyme at 162 ꢀC. The secondary amide, N-methylbenzamide,
9, remained unreacted in NaBH4–refluxing diglyme. However,
9 was converted slowly to benzylmethylamine when 1 equiv. of
LiCl was added (entry 5, Table 2). Adding a second equivalent
of LiCl was not further beneficial (entry 6, Table 2).
of the intermediate had also decomposed by the time the spec-
trum had been completed (calculated from the integration).
The oxygen atoms of nitrobenzene are known to strongly
coordinate with Lewis acids. Thus, the nitro group was
expected to coordinate with Li+, displacing diglyme from its
coordination sites in the complex. Indeed, nitrobenzene rapidly
replaced diglyme and only free diglyme was observed (3.00
ppm, s, –OCH3 ; 3.32 ppm, t and 3.45 ppm, t, –CH2CH2O–).
The resulting intermediate slowly decomposed at 25–29 ꢀC into
N-benzylformamide [eqn. (6)]. The diglyme:amide mole ratio
determined by 1H NMR integration was 0.48 and 0.50 in
two separately prepared and purified samples.
ð6Þ
Reduction of N-benzylformamide with NaBH4–LiCl and LiBH4
The quantity of borohydride hydrogens in the intermediate
was determined by treating fresh powder samples with 15%
by wt. aqueous H2SO4 and collecting the escaping hydrogen.
The hydrogen:N-benzylformamide mole ratio was almost
1.9:1.0. The lithium content (3.0% by wt) and boron content
is (4.6% by wt) were determined by atomic absorption. The
calculated molecular weight of the intermediate is 462 or 470
based on the Li or B contents, respectively, assuming that
two atoms each of Li and B are present.
The reaction of N-benzylformamide, 10, with LiCl—NaBH4 at
162 ꢀC in diglyme initially produced a white solid intermediate
that precipitates from the diglyme solution at 162 ꢀC. As the
reaction proceeds, benzylmethylamine, 15, was formed slowly
over 40 min. when 1 equiv. of LiCl per amide was present.
The conversion of amide to the corresponding amine increased
when the reaction time was extended and when the LiCl:amide
mole ratio increased to 2 (entries 7 and 8, Table 2). The white
solid, which initially precipitated, was isolated and thoroughly
washed with fresh ether (which had been previously dried with
LiAlH4 and distilled directly into glassware containing the
solid). Ether readily dissolves diglyme, N-benzylformamide,
10, and benzylmethylamine, 15. Thus, washing with ether
was continued until no further diglyme or amide was found
in the washings. Then this solid was dried at room temperature
under reduced pressure ( < 1 mmHg) overnight.
Further confirmation that diglyme takes part in metal chela-
tion in the intermediate comes from ether/diglyme exchange.
The isolated intermediate was treated with dried diethyl ether
and stirred strongly under dried N2 for over 30 h and then
dried in vacuo ( < 1 mmHg) for 3 h. After > 30 h exposure
of the intermediate to ether, a portion of the bound diglyme
was replaced by ether. The 1H NMR spectrum of this partially
exchanged sample was obtained in nitrobenzene-d5 where both
ether and diglyme were displaced. The proton signals from
both ethyl ether (3.07 ppm, q, J 6.99; 0.83 ppm, t, J 6.99 Hz,
area ratio 2:3) and diglyme were observed. The mole ratio of
ether to diglyme was ꢂ1.1 by NMR integration.
This isolated complex is very sensitive to moisture, decom-
posing when exposed to the laboratory atmosphere. The pure
intermediate cannot be recrystallized from many polar and
non-polar solvents, such as 1,2-dichlorobenzene, nitrobenzene,
THF, DMSO benzene, hexane and so on. Other attempts were
made to identify its structure. A portion of the solid was trea-
ted with 15% by wt. aqueous H2SO4 and the resulting solu-
tion’s pH was adjusted to 10 with 15% by wt. aqueous
KOH. Upon extraction with diethyl ether, diglyme and N-ben-
zylformamide, 10, in a 1:2 mole ratio were the only products
detected. Amine, 15, was not found in these products. The
diglyme liberated from the solid upon treatment with H2SO4
must have been tightly bound in the intermediate since this
solid had previously been washed many times with dried
diethyl ether, a good diglyme solvent, without loss of diglyme.
The white precipitate was not formed when LiCl was absent
during NaBH4 reductions of 10 in diglyme at 162 ꢀC. This
strongly suggests the Li+ ion is involved in the intermediate
or promotes formation of the intermediate. Furthermore, sam-
ples of this same complex were prepared using LiBH4 in place
of NaBH4–LiCl.
Thermal decomposition of the complex was investigated. It
decomposed into four products when heated in air to between
200 and 265 ꢀC [eqn. (6)]. Surprisingly, benzylamine, 18, (49%)
was the major product instead of benzylmethylamine, 15,
which had been produced in the NaBH4–LiCl reductions
1
(entries 7 and 8, Table 2). Benzylamine was identified by H
NMR and mass spectrometry. Five aromatic protons at 7.2–
7.4 ppm, two methylene protons at 3.9 ppm and two amino
protons at 1.6 ppm constituted its 1H NMR spectrum. The
GC-MS (EI) spectrum exhibited a molecular ion at m/z ¼
108 in accord with an M + 1 ion of 18 (MW 107). Its GC reten-
tion time was identical to that of authentic 18. The second pro-
duct (19%), N-benzylformamide, 10, exhibited a molecular ion
at m/z ¼ 135 in the GC-MS (EI) and its GC retention time
was the same as that of authentic 10. Benzonitrile, 19, (12%)
was identified as the third product. The final product N,N-
dibenzylhydroxyamine, 20, (19%) was identified by GC-MS
(EI) by comparing its fragmentation pattern and molecular
ion (m/z ¼ 194) with those of a standard N,N-dibenzyl-
hydroxyamine spectrum.
1
Attempts to determine the intermediate’s H NMR spectra
were complicated by its extreme sensitivity to moisture and
insolubility in most solvents. Nitrobenzene-d5 , dried by distil-
lation from NaH, was finally used because the solubility of the
intermediate was higher in this solvent than in solvents such as
1,2-dichlorobenzene. The 1H NMR spectra of the intermediate
at room temperature in nitrobenzene-d5 exhibited a new broad
singlet at 4.1 ppm for the benzylic protons and a singlet at
A hypothetical structure for this species that rationalizes all
of the experimental results would contain two moles of Li+,
two moles of N-benzylformamide, two moles of boron, and
one mole of diglyme per mole of this intermediate. Two possi-
ble structures, 21 and 22, are suggested for consideration. Both
412
New J. Chem., 2003, 27, 409–413

View Article Online
have the empirical formula C22H32N2O7Li2B2 (4.6% B, 2.9%
Li versus 4.6% B and 3.0% Li found by atomic absorption)
and a MW of 457.6. Each would liberate 2 moles of H2
per mole of N-benzylformamide present and each has a
diglyme:N-benzylformamide mole ratio of 0.5. The NMR che-
mical shifts observed appear reasonable for these structures.
Nevertheless, these are only postulated structures with idea-
lized coordination of the lithium ions and one can construct
related polymeric or oligomeric structures also. This species’
exact structure is unknown.
NMR spectra and selected MS data for other reduction pro-
ducts are given below.
4-Chlorobenzylamine, 11. 1H NMR (300 MHz, CDCl3)
d: 7.31–7.22 (4H, m), 3.83 (2H, s), 1.45 (2H, s).
N-Methylbenzylamine, 15. 1H (300 MHz, CDCl3) d: 7.37–
7.25 (5H, m), 3.75 (2H, m), 2.51 (3H, s).
4-Fluorobenzylamine, 16. 1H (300 MHz, CDCl3) d: 7.29–7.25
(2H, m), 7.10–6.98 (2H, m), 3.84 (2H, s), 1.60 (2H, s).
Piperidine. 1H (300 MHz, CDCl3) d: 2.80 (3H, br s), 2.10
(3H, br s), 1.53 (5H, br s).
4-Chlorobenzylnitrile. 1H (300 MHz, CDCl3) d: 7.63–7.58
(2H, m), 7.50–7.45 (2H, m).
4-Fluorobenzylnitrile. 1H (300 MHz, CDCl3) d: 7.71–7.66
(2H, m), 7.21–7.16 (2H, m).
The successful reduction of amides to amines by NaBH4 at
high temperature, when added to reports of NaBH4 reductions
of carboxylic acids,8 esters,8 nitriles,8 and chlorinated aromatic
hydrocarbons9,11 shows that this reagent can be used to reduce
a wide range of functional groups (nonselectively) at high
temperatures in glyme solvents.
1
2-Chlorobenzylnitrile, 2. H (300 MHz, CDCl3) d: 7.71–7.67
(1H, m), 7.59–7.55 (2H, m), 7.45–7.37 (1H, m).
Isolated intermediate. 1H NMR (300 MHz, C6D5NO2) d:
8.33 (1H, s), 7.19–7.04 (5H, m), 6.73 (1H, s), 4.41 (2H, d, J
6.10), 3.45 (2H, m), 3.43 (2H, m), 3.32 (3, s), 2.23 (1H, s). IR
(KBr, cmꢁ1): 3436, 2292, 2227, 1678, 1446, 1083, 618. Atomic
absorption analysis: B, 3.6%; Li, 3.0%.
Experimental
General
Benzonitrile, 19. Produced by decomposing the intermediate.
GC-MS (CI): m/z 103 (100%), 76 (52%), 50 (32%).
All chemicals were purchased from Aldrich Company except
for diglyme, which was a gift from Ferro Corporation. 1H
NMR spectra were obtained on a General Electric QE-300
instrument. A Varian 3300 GC was used (DB-5, 30 meters).
GC/MS were obtained on a Varian Saturn 2000 instrument.
Decomposition temperatures were uncorrected.
Benzylamine, 18. Produced by thermal decomposition of the
1
intermediate. H NMR (300 MHz, CDCl3) d: 7.2–7.4 (5, m),
3.9 (2, s), 1.6 (2, br s). GC-MS (CI): m/z 103 (100%), 79
(50%). N-Benzylformamide produced by decomposing the
intermediate. GC-MS (CI): m/z 135 (100%), 106 (55%), 79
(77%).
Typical reduction procedure
2-Chlorobenzamide (155.5 mg, 1.0 mmol), LiCl (83 mg, 2.0
mmol) and the internal GC standard hexadecane (50.0 ml) were
added to diglyme (10 ml) at room temperature. NaBH4 (56 mg,
1.5 mmol) was added after the solution reached 162 ꢀC and the
reaction was continued for over 4 h [eqn. (1)]. The product dis-
tribution was followed versus time by GC analysis. Aliquots
(0.2 ml) were withdrawn and treated with H2SO4 (15% w/w)
to decompose the NaBH4 , followed by treatment with aqu-
eous KOH (15% w/w) to adjust the pH of the solution to
ꢂ10. These basic solutions were extracted with ether and the
products in the ether layer were analyzed by GC. After the
starting material was consumed completely and most of 2-
chlorobenzylnitrile was converted into 2-chlorobenzylamine,
the entire reaction solution was worked up by the method
described for the aliquots above. The ether solution was dried
for 2 h (anhydrous Na2SO4) and filtered to remove Na2SO4 .
Then HCl gas was bubbled into the solution to form the
hydrochloride salt of 2-chlorobenzylamine. After solvent
removal in vacuo, the amine hydrochloride was obtained. Then
it was dissolved in water and the solution pH value was
adjusted to 10 with aqueous KOH. After extraction with ether
(3 ꢃ 20 ml), the crude 2-chlorobenzylamine was purified by col-
umn chromatography over silica gel using ethyl acetate–hex-
ane (1:3) elution. Pure 2-chlorobenzylamine was obtained (78
N,N-Dibenzylhydroxylamine, 20. Produced by decomposing
the intermediate. GC-MS (CI): m/z 194 (42%), 165 (2%), 117
(11%), 91 (100%), 65 (27%).
Acknowledgements
The work was supported in part by the Morton Division of
Rohm and Haas and the Grant Division of Ferro Corp.
References
1
T. Satoh, S. Suzuki, Y. Miyaji and Z. Imai, Tetrahedron Lett.,
1969, 4555.
2
3
4
5
M. E. Kuehne and P. J. Shannon, J. Org. Chem., 1977, 42, 2082.
S. Raucher and P. Klein, Tetrahedron Lett., 1980, 21, 4061.
P. T. Thorsen and M. M. Kreevoy, J. Org. Chem., 1981, 46, 2579.
S. Kano, Y. Tanaka, E. Sugino and S. Hibino, Synthesis, 1980,
695.
6
7
8
9
N. Umino, T. Iwakuma and N. Itoh, Tetrahedron Lett., 1976, 763.
B. Malawska and M. Gorczyca, Pol. J. Chem., 1985, 59, 811.
C. Yang and C. U. Pittman, Jr., Synth. Commun., 1998, 28, 2027.
C. Yang and C. U. Pittman, Jr., Tetrahedron Lett., 1997, 38, 6561.
10 C. Yang and C. U. Pittman, Jr., Synth. Commun., 1998, 28, 517.
11 Y. Kikugawa, S. Ikegawi and S. Yamada, Chem. Pharm. Bull.,
1969, 17, 98.
12 S. E. Ellzey, Jr., C. H. Mack and W. J. Connick, Jr., J. Org.
Chem., 1967, 32, 846.
1
mg, 55% yield). H NMR (300 MHz, CDCl3) dH : 7.38–7.36
(2H, m), 7.35–7.33 (2H, m), 3.92 (2H, s), 1.59 (2H, br s).
The GC retention times of all products were compared with
those of authentic samples by spiking experiments. The 1H
13 M. S. Newman and T. Fukunaga, J. Am. Chem. Soc., 1960, 82,
693.
New J. Chem., 2003, 27, 409–413
413