The use of diazabicyclo[2.2.2]octane as a novel highly selective dechloroacetylation reagent

Article abstract of DOI:10.1021/ol005567z

(equation presented) In a study directed toward the use of the chloroacetyl protecting group in carbohydrate synthesis, the sterically hindered tertiary amine diazabicyclo[2.2.2]octane (DABCO) was found to give complete and selective cleavage of the chlor

Full text of DOI:10.1021/ol005567z

ORGANIC
LETTERS
2000
Vol. 2, No. 5
701-703
The Use of Diazabicyclo[2.2.2]octane as
a Novel Highly Selective
Dechloroacetylation Reagent
Dirk J. Lefeber, Johannis P. Kamerling, and Johannes F. G. Vliegenthart*
BijVoet Center, Department of Bio-Organic Chemistry, Utrecht UniVersity,
P.O. Box 80.075, NL-3508 TB Utrecht, The Netherlands
Vlieg@pobox.uu.nl
Received January 19, 2000
ABSTRACT
In a study directed toward the use of the chloroacetyl protecting group in carbohydrate synthesis, the sterically hindered tertiary amine
diazabicyclo[2.2.2]octane (DABCO) was found to give complete and selective cleavage of the chloroacetyl group in the presence of other ester
functions such as benzoyl and acetyl groups at primary and/or secondary positions.
The chloroacetyl group is used for the temporary protection
of hydroxyl groups, as it is a very base-sensitive group that
can be selectively cleaved in the presence of other acyl
functions by a number of methods. In addition to, for
example, aqueous ammonia1 or hydrazine acetate,2 the most
widely used deprotection reagent is thiourea.3 The latter
reaction is easy to perform but requires relatively long
reaction times. The nucleophilic sulfur of thiourea displaces
the chloride atom of the chloroacetyl group, and one of the
amine functions subsequently attacks the ester bond intramo-
lecularly.4 Deprotection using hydrazine dithiocarbonate
(HDTC), introduced by van Boeckel et al.,5 is based on the
same reaction mechanism. It gives rise to fast and clean
reactions but requires the preparation of the reagent just
before use, since it is only stable for 15-20 min.6 This paper
reports a novel dechloroacetylation reagent, diazabicyclo-
[2.2.2]octane (DABCO), which turned out to be highly
selective, fast, and easy to handle.
In the course of the synthesis of oligosaccharides, we were
confronted with problems in the deallylation of compound
1 to give 2 (Figure 1). The yield of the isomerization of the
(1) Reese, C. B.; Stewart, J. C. M.; van Boom, J. H.; de Leeuw, H. P.
M.; Nagel, L.; de Rooy, J. F. M. J. Chem. Soc., Perkin Trans. 1 1975,
934-942.
(2) Udodong, U. E.; Rao, C. S.; Fraser-Reid, B. Tetrahedron 1992, 48,
4713-4724.
(3) See, for example: Blatter, G.; Jacquinet, J. C. Carbohydr. Res. 1996,
288, 109-125. Naruto, M.; Ohno, K.; Naruse, N.; Takeuchi, H. Tetrahedron
Lett. 1979, 3, 251-254. Glaudemans, C. P. J.; Bertolini, M. J. Meth.
Carbohydr. Chem. 1980, VIII, 271-275.
Figure 1. Deallylation of 1 (f 2) and the formed side product 3.
(4) Masaki, M.; Kitahara, T.; Kurita, H.; Ohta, M. J. Am. Chem. Soc.
1968, 90, 4508-4509.
double bond using tristriphenylphosphine rhodium(I) chloride
and DABCO (3 equiv) in refluxing ethanol was low, due to
(5) van Boeckel, C. A. A.; Beetz, T. Tetrahedron Lett. 1983, 24, 3775-
3778.
10.1021/ol005567z CCC: $19.00 © 2000 American Chemical Society
Published on Web 02/15/2000

the formation of the dechloroacetylated monosaccharide
derivative 3, as shown by 1H NMR (Figure 1).7 This
observation stimulated further investigations toward the use
of DABCO as a general dechloroacetylation reagent.
Complete dechloroacetylation of 1 to give 4 (Figure 2)
was achieved by incubating 0.02 M 1 in ethanol in the
reaction time was obtained (entry 6). By increasing the
concentration of 1 from 0.02 to 0.06 M (Table 1, entries 6
and 7), the reaction time was further decreased from 30 min
to less than 10 min. In all cases a quantitative conversion
1
was detected from TLC and H NMR, after neutralization
of the reaction mixtures with Dowex H+, filtration, and
concentration. No migration of the benzoyl group from O-2
to O-3 was observed. Only after 3 days of stirring, in the
case of entries 1 and 2, could about 5% migration (4 f 5,
Figure 2) be seen.10 This migration also started to occur after
more than 4 h of stirring for entries 4-6. Only in the case
of entry 7 was some migration already found after 30 min.
Therefore, in view of reaction time and practical ease, we
preferred the reaction conditions described for entry 6.
To compare the effectiveness of DABCO with thiourea,
the following experiments were performed. According to the
commonly used dechloroacetylation with thiourea, 3 equiv
of reagent, either thiourea or DABCO, was reacted with 0.03
M 1 in 5:1 ethanol-pyridine at 70 °C. Although in both
cases excellent isolated yields were obtained, the use of
DABCO resulted in much faster dechloroacetylation (Table
2, entries 1 and 2). Also following the established procedure
Figure 2. Dechloroacetylation of 1 afforded 4 and in some cases
the migrated product 5.
absence of Wilkinson’s catalyst with 5 equivs of DABCO
at room temperature for 6 h.8 Isomerization of the double
bond without affecting the chloroacetyl group was achieved
by reaction of 5 mM 1 with Wilkinson’s catalyst without
DABCO in refluxing ethanol or by using only a catalytic
amount of DABCO in either refluxing ethanol or refluxing
1:1 ethanol-toluene, the latter solvent system slowing down
the dechloroacetylation. After reaction of the intermediate
with N-iodosuccinimide and water, 2 was obtained in 77%
yield.9
Table 2. Comparison of DABCO and Thiourea in the
Dechloroacetylation of 1
entry
reagent
solvent
time (min)
yield (%)
To gain insight into the factors influencing the reaction
time needed for complete dechloroacetylation of 1 in ethanol,
several parameters were varied. By increasing the amount
of DABCO, the reaction time was shortened from 6 to 2 h
(Table 1, entries 1 and 2). Furthermore, the temperature of
1
2
3
4
thiourea
DABCO
thiourea
DABCO
ethanol/pyridine
ethanol/pyridine
ethanol
90
<10
270
30
93
95
97
94
ethanol
(Table 1, entry 6), DABCO was found to give a much faster
dechloroacetylation (Table 2, entries 3 and 4).
Table 1. Effect of the Amount of DABCO, Temperature, and
Concentration on the Reaction Time of the Dechloroacetylation
of 1 As Carried out in Ethanol
In additional studies, two other tertiary amines, triethyl-
amine (TEA) and N,N-diisopropylethylamine (DIPEA), were
examined for their dechloroacetylating properties. Incubation
of 1 with 15 equiv of TEA or DIPEA in ethanol at room
temperature resulted in partial dechloroacetylation. Before
the dechloroacetylation was complete, about 10% migration
of the benzoyl group had occurred from O-2 to O-3, resulting
in the formation of 5 (Table 3, entries 1 and 3). In
entry
DABCO (equiv) temp (°C) time (min) concn (M)
1
2
3
4
5
6
7
5
20
10
10
10
15
15
20
20
20
55
70
55
55
360
120
240
40
15
30
0.02
0.02
0.02
0.02
0.02
0.02
0.06
<10
Table 3. Dechloroacetylation of 1 Using the Tertiary Amines
TEA or DIPEA
the reaction mixture had a marked influence on the reaction
rate (entries 3, 4, and 5), and thereby on the reaction time.
Using 15 equiv of DABCO at 55 °C, a sufficiently short
entry base
solvent temp (°C) time (h) yield (%)a of 1:4:5
1
2
3
4
TEA
TEA
DIPEA ethanol
DIPEA Cl2CH2
ethanol
Cl2CH2
20
20
20
20
16
1.5
16
1.5
25:65:10
no reaction
25:65:10
(6) Smith, A. B., III; Hale, K. J.; Vaccaro, H. A.; Rivero, R. A. J. Am.
Chem. Soc. 1991, 113, 2112-2122.
no reaction
(7) 1H NMR analysis showed the removal of the chloroacetyl group (δ
4.00, 3.94: 2 d, each 1 H), a shift of H-3 from 5.87 to 4.40 ppm, and the
presence of a 1-propenyl group (δ 6.11: dq, 1 H, OCHdCHCH3. 5.19:
dq, 1 H, OCHdCHCH3. 1.52: dd, 3 H, OCHdCHCH3).
a Estimated from TLC.
(8) Characteristic 1H NMR data of 4: δ 5.87-5.74 (m, 1 H, OCH2CHd
CH2), 5.57 (s, 1 H, PhCH), 5.21 (d, 1 H, H-1), 5.30-5.22, 5.15-5.10 (2
m, each 1 H, OCH2CHdCH2), 5.06 (dd, 1 H, H-2), 4.40 (t, 1 H, H-3).
dichloromethane, no reaction was observed within 1.5 h at
room temperature (Table 3, entries 2 and 4). These results
702
Org. Lett., Vol. 2, No. 5, 2000

clearly show that the selectivity of DABCO in the dechloro-
acetylation of 1 does not result solely from being a tertiary
amine.
acetylated in 75 min using method A and in 2.5 h using
method B. The isolated yields were 95% and 97%, respec-
1
tively.13 No migration could be observed by H NMR.
To further explore the generality of the procedure,
compounds 6 and 8 (Figure 3) were subjected to the
Although method A is much faster than method B, method
B may be used when more basic conditions are expected to
cause problems.
Performing the reaction according to method B without
DABCO, complete dechloroacetylation of 6 and 8 required
6 and 7 h, respectively. This observation indicates that
pyridine alone is also able to affect the chloroacetyl group
but that DABCO greatly accelerates the reaction. Ethylene-
diamine, o-phenylenediamine, and 2-mercaptoethylamine14
have also been mentioned as dechloroacetylation reagents.
However, those reactions were performed with a large excess
of pyridine and TEA in methanol. Our findings with pyridine
and TEA show that these reagents themselves might not be
the actual dechloroacetylating agents.
Figure 3. Dechloroacetylation of 6 and 8.
The results clearly show that the use of DABCO as
dechloroacetylating agent provides a good method in addition
to the commonly used reagents. Benzoyl and acetyl groups
at primary and/or secondary positions are stable under the
reaction conditions employed. The procedure is fast, highly
selective, and easy to perform.
following reaction conditions:11 (A) 0.02 M 6 or 8, 15 equiv
of DABCO, 55 °C, ethanol; (B) 0.04 M 6 or 8, 1 equiv of
DABCO, 55 °C, 5:1 ethanol-pyridine.
Dechloroacetylation of 6 was found to be complete in 45
min using method A. Using method B, which is much milder,
2 h were needed for the conversion of 6 to 7. In both cases
the isolated yield was 95%.12 Compound 8 was dechloro-
OL005567Z
(11) Purification: after complete reaction, the mixture was diluted with
ethyl acetate, poured into 0.05 M HCl, and washed. The organic layer was
dried (MgSO4), filtered, and concentrated.
1
(9) Characteristic NMR data of 2: 1H, δ 5.54 (s, 1 H, PhCHR), 5.52 (s,
1 H, PhCHâ), 5.88 (t, 1 H, H-3), 5.57 (d, 1 H, H-1R), 5.19 (dd, 1 H, H-2â),
5.09 (dd, 1 H, H-2R), 4.92 (d, 1 H, H-1â); 13C, δ 101.5 (PhCH), 95.8 (C-
1R), 90.7 (C-1â), 40.4 (ClCH2COR), 40.3 (ClCH2COâ).
(12) Characteristic H NMR data: 6 δ 5.76 (t, 1 H, H-4), 4.86 (d, 1 H,
H-1), 3.95 and 3.89 (2 d, each 1 H, ClCH2CO). 7: δ 4.81 (d, 1 H, H-1),
3.93 (dt, 1 H, H-4), 3.58 (d, 1 H, HO-4).
1
(13) Characteristic H NMR data: 8 δ 5.72 (t, 1 H, H-4), 4.76 (d, 1 H,
1
(10) All signals in the H NMR spectrum were slightly shifted. Strong
H-1), 3.97 and 3.90 (2 d, each 1 H, ClCH2CO), 2.12 (s, 3 H, CH3CO). 9:
δ 4.73 (d, 1 H, H-1), 3.83 (t, 1 H, H-4), 2.12 (s, 3 H, CH3CO).
(14) Cook, A. F.; Maichuk, D. T. J. Org. Chem. 1970, 35, 1940-1943.
shifts were observed for H-3, from 4.40 to 5.61 ppm, and for H-2, from
5.06 ppm to the bulk region.
Org. Lett., Vol. 2, No. 5, 2000
703