Selective N-debenzylation of amides with p-TsOH

Article abstract of DOI:10.1016/S0040-4039(02)02738-7

N-Benzylamides were debenzylated efficiently with 4 equiv. of p-TsOH in refluxing toluene. Good to quantitative yields of the desired primary amides were obtained within 2-4 h from a wide variety of N-2,4-dimethoxybenzylamides. N-4-Methoxylbenzyl amides and N-benzylamides were also debenzylated cleanly. In the case of N-2,4-dimethoxylbenzylamides, selective N-debenzylation was possible in the presence of N-Fmoc, N-t-BOC or N-trityl-protection. Protected amino acid amides survived these conditions without any detectable epimerization.

Full text of DOI:10.1016/S0040-4039(02)02738-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 1039–1041
Selective N-debenzylation of amides with p-TsOH
Ching-Yuh Chern,a Yu-Ping Huangb and Wai Ming Kanb,*
aDepartment of Applied Chemistry, Chao Yang University of Technology, Taichung, Taiwan 413, ROC
bDepartment of Pharmacology, National Cheng Kung University, Tainan, Taiwan 701, ROC
Received 18 September 2002; revised 21 November 2002; accepted 29 November 2002
Abstract—N-Benzylamides were debenzylated efficiently with 4 equiv. of p-TsOH in refluxing toluene. Good to quantitative yields
of the desired primary amides were obtained within 2–4 h from a wide variety of N-2,4-dimethoxybenzylamides. N-4-Methoxyl-
benzyl amides and N-benzylamides were also debenzylated cleanly. In the case of N-2,4-dimethoxylbenzylamides, selective
N-debenzylation was possible in the presence of N-Fmoc, N-t-BOC or N-trityl-protection. Protected amino acid amides survived
these conditions without any detectable epimerization. © 2003 Elsevier Science Ltd. All rights reserved.
Table 1. Examples of N-debenzylation with p-TsOH in
refluxing toluene
N-Benzyl groups are used in amine and amide protec-
tion or as handles for amine introduction in organic
synthesis. However, they enjoy limited success because
their removal is often difficult. The most commonly
employed methods include acid hydrolysis in refluxing
TFA1 and hydrogenolysis with catalytic Pd/C.2 Even
under these conditions, their removal is not guaranteed.
Strongly reducing conditions such as Na/NH3(l)3 or Li
with catalytic naphthalene4 have been used or t-BuLi/
THF; O2 can be employed in extreme conditions.5
These harsh conditions can prevent the use of N-benzyl
protecting groups for labile compounds.
Many attempts have been made to develop milder
N-debenzylation conditions. 4-Methoxylbenzyl (MB),
2,4-dimethoxylbenzyl (DMB) and 2,4,6-trimethylbenzyl
groups with increased acid sensitivity have been intro-
duced. Nevertheless, the resultant benzyl cations are
highly electrophilic. Alkylation of the starting material
or the product are not uncommon. Addition of nucle-
ophilic scavengers such as thioanisole6 or anisole7 is
often necessary to suppress side reactions. N-MB and
N-DMB but not N-benzyl groups can also be removed
by oxidative N-debenzylation using CAN8,10 or
DDQ.9,10 Obviously, oxidation sensitive starting materi-
als are not compatible with these reagents.
Herein, we report the use of p-TsOH as a convenient
and effective N-debenzylating agent. The application of
this reagent is summarized in Table 1. The N-2,4-
Keywords: N-debenzylation; N-2,4-dimethoxybenzylamide; p-TsOH.
* Corresponding author. Tel.: +886-6-2353535, ext. 5469; fax: +886-
6-2749296; e-mail: vincent@mail.ncku.edu.tw
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)02738-7

1040
C.-Y. Chern et al. / Tetrahedron Letters 44 (2003) 1039–1041
dimethoxylbenzyl amides were synthesized by the reac-
tion of the corresponding carboxylic acids with 2,4-
dimethoxylbenzylamine in the presence of DCC at
ambient temperature in excellent yield. From entries
1–6, consisting of primary, secondary, tertiary, phenyl
and styryl carboxyamides, N-debenzylation occurred
smoothly with good to quantitative yields of the iso-
lated primary amides. Even with the 1,2-diamide 2d, a
65% yield of the desired product was isolated in spite of
steric hindrance.
starting materials. The results are summarized in Table
2. Neither of the established methods was useful for 3a,
no product formation was observed even after pro-
longed reaction (20 h). This came as no surprise since
there were literature reports that removal of N-benzyl
groups can be sluggish in TFA and Pd/C–H2. On the
other hand, p-TsOH afforded a 3% yield of 4 after 4 h,
while the reaction was complete after 11 h with a 93%
yield of the desired product. Some success was observed
for starting material 3b. After 4 h, 6 and 3% yields of
the desired amide 4 were isolated using hydrogenolysis
and acid hydrolysis, respectively, as the reactions were
incomplete. In comparison, with p-TsOH, a 46% yield
was obtained after 4 h with some starting material
remaining, while the reaction was complete after 6.5 h.
Similar results were obtained for 3c. After 4 h, the
reactions using the first two methods were incomplete
with 19 and 22% yields, respectively, while with p-
TsOH, a 98% yield of the desired amide was obtained.
Moreover, 6% of the reduced product (benzyl alcohol)
was isolated from the hydrogenolysis mixture. This
indicates that optimization of the hydrogenolysis condi-
tions by extending the reaction time may be futile since
it also increases the formation of the reduction product.
The optimum results were usually obtained when 0.3 M
of the starting material was allowed to react with 4
equiv. of p-TsOH in refluxing toluene. Lower reaction
temperature or the use of less p-TsOH led to inferior
results.
When we applied this reagent to N-Fmoc-, N-t-BOC-,
and N-trityl-protected amino acids, good to excellent
yields of the desired amides were isolated. Both entries
7 and 8 report good yields of primary amides without
any detectable racemization. The optical activities of 2g
and 2h were [h]1D9=−4.02 (c 1, MeOH) and [h]2D5=
−25.42 (c 1, MeOH), respectively [lit. [h]1D9=−3.53 (c 1,
MeOH)11 and [h]D25=−25.8 (c 1, MeOH),12 respectively].
Both acid sensitive t-BOC- and base sensitive N-Fmoc-
protections survived the reaction conditions. Moreover,
in the presence of N-trityl amide, selective N-debenzyl-
ation was possible as exemplified in 2i. For both of the
protected asparagine and glutamine, no epimerisation
occurred and N-Fmoc and N-trityl protection remained
intact.13
Using N-DMB-salicylamide 5 and N-DMB-5-phenyl-5-
hydroxypentamide 8 as model compounds, we com-
pared p-TsOH against CAN and DDQ. With
N-DMB-salicylamide 5 (Table 3), CAN failed to react
after 20 h with 96% of starting material being recovered
while DDQ afforded 85% of N-2,4-dimethoxylbenzoyl-
salicyamide 7. In the case of p-TsOH, a 92% yield of
salicylamide 6 was isolated cleanly after 2 h.
We then compared this method with two of the most
commonly used N-debenzylation methods, i.e. acid
With N-DMB-5-phenyl-5-hydroxypentamide 8 as sub-
strate (Table 4), CAN afforded 90% of the desired
amide 9 in 20 h (<5% in 4 h) whereas DDQ gave 28%
of 9 and 34% of the oxidized product 10 with 36% of
recovered starting material. Obviously, DDQ oxidized
14
hydrolysis in TFA/CH2Cl2 and hydrogenolysis in cat-
alytic Pd/C,15 using N-benzyl-3-phenacylpropanamide
3a, N-4-methoxybenzyl-3-phenacylpropanamide 3b and
N-2,4-dimethoxybenzyl-3-phenacylpropanamide 3c as
Table 2. Comparison of N-debenzylation methods
S.M.a
Reagents (deprotection)
Temp.
Time (h)
Product 4 (yield %)
3a
3b
3c
3a
3b
3c
3a
3b
3c
5% Pd/C, H2
5% Pd/C, H2
5% Pd/C, H2
CF3CO2H (1.2 equiv.)/CH2Cl2
CF3CO2H (1.2 equiv.)/CH2Cl2
CF3CO2H (1.2 equiv.)/CH2Cl2
p-TsOH (4.0 equiv.)/toluene
p-TsOH (4.0 equiv.)/toluene
p-TsOH (4.0 equiv.)/toluene
rt
rt
rt
9
9
9
9
9
9
11
6.5
4
11
6.5
4
11
6.5
4
0
7
19
0
4
22
93
92
98
a S.M., starting material.

C.-Y. Chern et al. / Tetrahedron Letters 44 (2003) 1039–1041
1041
Table 3.
variety of N-2,4-dimethoxybenzylamides leading to
reaction within 2–4 h. Under these reaction conditions,
no epimerization of the protected amino acid amides
was observed while N-Fmoc, N-t-BOC, and N-trityl
protection remained unchanged. Dehydration of benzyl
alcohol was also not observed. N-Benzyl- and N-4-
methoxybenzylamides may also be debenzylated in 11
and 6.5 h, respectively, while most methods failed. In
general, it was more efficient than most reported N-
debenzylation methods including acid hydrolysis with
TFA, hydrogenolysis with catalytic Pd/C and oxidation
methods with DDQ and CAN. In the cases of oxida-
tion or reduction sensitive starting materials, this new
method could be very valuable. Apparently, it might be
applicable to a wider variety of N-benzylamides.
Reagents
(deprotection)
Temp.
Time (h)
Product (yield %)
6
7
–
–
p-TsOH (4.0
9
rt
2
20
2
92
–
equiv.)/toluene
CAN (1.1 equiv.),
CH3CN/H2O (1:1)
DDQ (1.1
equiv.)/CH2Cl2
a
rt
–
85
Acknowledgements
a Recovered S.M. 96%.
We gratefully acknowledge generous support from the
National Science Council of the Republic of China.
Table 4.
References
1. (a) Schlessinger, R. H.; Bebernitz, P. L.; Poss, A. Y. J.
Am. Chem. Soc. 1985, 107, 1777; (b) Shimshock, S. S.;
Waltermire, R. E.; Deshong, P. J. Am. Chem. Soc. 1991,
113, 8791.
2. Gigg, R.; Conant, R. Carbohydr. Res. 1982, 100, C5.
3. (a) Ohgi, T.; Hecht, S. M. J. Org. Chem. 1981, 46, 1232;
(b) Webster, F. X.; Millar, J. G.; Silverstein, R. M.
Tetrahedron Lett. 1986, 27, 4941.
4. Alonso, E.; Ramon, D. J.; Yus, M. Tetrahedron 1997, 53,
14355.
5. Williams, R. M.; Kwast, E. Tetrahedron Lett. 1989, 30,
451.
6. Wood, J. L.; Stoltz, B. M.; Dietrich, H. J. J. Am. Chem.
Soc. 1995, 117, 10413.
Reagents
(deprotection)
Temp.
Time (h)
Product (yield %)
9
10
–
p-TsOH (4.0
9
rt
2
20
2
93
90
28
equiv.)/toluene
CAN (1.1 equiv.),
CH3CN/H2O (1:1)
DDQ (1.1
–
rt
34a
equiv.)/CH2Cl2
7. Wood, J. L.; Stoltz, B. M.; Goodman, S. N. J. Am.
Chem. Soc. 1996, 118, 10656.
a Recovered S.M. 36%.
8. Overman, L. E.; Osawa, T. J. Am. Chem. Soc. 1985, 107,
1698.
9. Mori, S.; Iwakura, H.; Takechi, S. Tetrahedron Lett.
the benzyl alcohol as fast as the benzylamide. On the
other hand, p-TsOH provided 93% of the desired amide
9 in just 2 h. No dehydration of the benzyl alcohol was
observed.
1988, 29, 5391.
10. Buenadicha, F. L.; Bartolame, M. T.; Aguirre, M. J.;
Avendano, C.; Sollhuber, M. Tetrahedron: Asymmetry
1998, 9, 483.
11. Gayathri, K.; Gopi, H. N.; Babu, V. V. S. Indian J.
From the results of Tables 3 and 4, we may conclude
that DDQ is more sensitive to the oxidizability of the
starting materials while CAN reacts more slowly. Obvi-
ously, these oxidants may not be applicable to N-ben-
zylamides in contrast to p-TsOH.
Chem., Sect. B 1998, 37, 151.
12. Eberle, A.; Fauchere, T. L.; Tesser, G. I.; Schwyzer, R.
Helv. Chim. Acta 1975, 58, 2106.
13. Pietta, P. G.; Cavallo, P.; Marshall, G. R. J. Org. Chem.
1971, 36, 3966.
14. Brooke, G. M.; Mohammed, S.; Whiting, M. C. Chem.
Commun. 1997, 1511.
15. Rigby, J. H.; Gupta, V. Synlett 1995, 547.
In summary, a versatile and high yielding N-debenzyla-
tion method is described. It is applicable to a wide