ˇ
A. Krchnáková et al. / Tetrahedron Letters 54 (2013) 4749–4752
4750
O
R2
O
O
R2
R1
i, ii
N
R2
N
O
O
N
O
R2
N
O
Fmoc
X
N
H
L
N
O
O
R1
O
2a:
2b:
1a
R
R
2 = Et
: X = NH, L = Rink amide linker
1b: X = O, L = Wang linker
HX
3
2 = i-Pr
Scheme 2. Synthesis of triazolinones from polymer-supported Fmoc-amino acid, azodicarboxylate, and phosphine. Reagents and conditions: (i) PPh3, anhydrous THF, rt, 1 h;
(ii) 50% TFA in dichloromethane (DCM) (v/v), rt, 1 h.
Table 1
Synthesized triazolinones
Mitsunobu reaction with complete conversion of acid to an ester
in 1 h, but did not exhibit the formation of any significant amount
Entry
Product
X
R1
R2
Puritya (%)
Yieldb (%)
of triazolinone even after overnight reaction. Alternatively, the use
of four fold excess (0.4 M solution) quantitatively converted the
Fmoc-Ala-OMe to the triazolinone derivative in 1 h. This experi-
ment confirmed that the traditional Mitsunobu reaction can be
carried out independently and without competition from the for-
mation of triazolinone by avoiding a large excess of reagents.
LC/MS analysis of the reaction of Fmoc-Ala-OH, MeOH, PPh3,
and azodicarboxylates 2c and 2d in solution revealed the presence
of products 3c and 3d (XH@OCH3), thus confirming that formation
of 4 and 5 occurred in the cleavage cocktail used to release prod-
ucts from resin (LC/MS analysis 3c, rt 7.93 min, [M+H]+ 344,
[M+NH4]+ 361, [2M+NH4]+ 704, [2M+Na]+ 709; LC/MS analysis
3d, rt 8.53 min, [M+H]+ 412, [M+NH4]+ 429, [2M+NH4]+ 840,
[2M+Na]+ 845).
1
2
3
4
5
3a
3b
3e
4
O
O
NH
O
O
–CH3
–CH3
–CH3
–CH3
–CH3
–CH2CH3
–CH(CH3)2
–CH(CH3)2
–Bn
87
93
68
61
91
77
77
46
47
78
5
–C(CH3)3
a
Purity of the crude product before purification estimated from LC traces @
230 nm.
b
Total yield after purification of target compounds by semi-preparative reverse
phase HPLC.
Next, we carried out the reaction with dibenzyl (2c) and di-tert-
butyl (2d) azodicarboxylates. Whereas diethyl and di-iso-propyl
azodicarboxylates yielded triazolinone derivatives 3, dibenzyl azo-
dicarboxylate formed, after acid-mediated cleavage from the resin,
triazolinone derivative 4 (Scheme 3). The product prepared using
di-tert-butyl azodicarboxylate 2d decomposed in the cleavage
cocktail (50% TFA in DCM) and it was released from the Wang resin
by NaOH. The HRMS and 1H and 13C NMR spectra were consistent
with triazolinone derivative 5 in which the Boc groups had been
cleaved. To document that the transformations of triazolinone
derivatives prepared from azodicarboxylates 2c and 2d to com-
pounds 4 and 5 occurred during cleavage from the resin, we carried
out experiments in solution.
Solid-phase synthesis is traditionally carried out with several-
(typically three)-fold excess of reagents. Therefore, the triazolinone
synthesis was advantageously carried out on solid phase since the
excess of all components of the reaction mixture, except the resin-
bound product, was simply removed by washing the resin beads.
To address the effect of the use of excess Mitsunobu reagents
and competitive traditional Mitsunobu reaction course (formation
of an ester from acid and alcohol), we carried out experiments in
solution. Fmoc-Ala-OH and an equimolar quantity of MeOH in
anhydrous THF (both 0.1 M solution) were exposed to an excess
of diisopropyl azodicarboxylate (DIAD) and PPh3. Two molar excess
of DIAD and PPh3 (0.2 M solution) caused the expected standard
The high purity (61–91%) and high yield (47–78%) of the trans-
formation to triazolinone derivatives with excess Mitsunobu re-
agents prompted us to explore the potential of combining the
traditional course of Mitsunobu reaction with triazolinone forma-
tion on solid phase. As an example, we chose to use Miller’s Mits-
unobu-mediated b-lactam syntheses.8 Thus, hydroxylamine resin
69,10 was acylated with Fmoc-Ser-OH to yield resin 7 (Scheme 4).
Using an excess of azodicarboxylates 2a and 2b and triphenylphos-
phine on resin 7 led to the formation of b-lactams8 and at the same
time transformation of the Fmoc-protecting group to triazolinone
derivatives 8. It is worth mentioning that the synthesis of b-lac-
tams was previously reported on solid phase; however, the authors
did not mention the formation of triazolinone.11
Due to the inherent instability of N-hydroxy-b-lactams,12,13
especially upon exposure to strong acids and their potential rear-
rangement to five-membered ring isoxazolidin-5-ones
9
(Scheme 5), we acetylated product 8b with acetic anhydride. The
conversion to acetate 10b proceeded quantitatively in 30 min
(LC/MS analysis, rt 5.65 min, [M+H]+ 357, [M+NH4]+ 374,
[2M+NH4]+ 730, [2M+Na]+ 735). Then, we adjusted the pH with so-
dium bicarbonate to pH 8. After overnight exposure the acetyl
group was cleaved to allow clean regeneration of 8b (LC/MS
O
O
O
O
i, ii
O
N
Fmoc
O
N
O
L
N
H
N
O
O
HN
N
O
O
O
OH
4
O
O
2c
L = Wang linker
HN
O
N
i, iii
O
N
Fmoc
N
L
OH
N
H
N
O
O
O
O
5
2d
L = Wang linker
Scheme 3. Polymer-supported synthesis of triazolinone derivatives. Reagents and conditions: (i) PPh3, anhydrous THF, rt, 1 h; (ii) 50% TFA in DCM (v/v), rt, 1 h; (iii) 0.5 M
NaOH, THF/MeOH (1:1; v/v), rt, 30 min.