1042
N. Zhao et al. / Tetrahedron Letters 55 (2014) 1040–1044
With compounds 15 and 16 in hand, the direct condensation
Currently, there are two major methodologies in isoindigo syn-
thesis. One involves the condensation reaction between the corre-
sponding indoline-2,3-dione and indolin-2-one, in which reflux in
acetic acid is general needed. The other routes involve dimeriza-
tion of the dione in one step, using either unpleasant Lawesson’s
reagent31 or expensive tris(diethylamino)phosphine30 as the cou-
pling reagent. Compared to current isoindigo synthesis methodol-
ogies, the beauty of this reaction is that the alkylation, oxidation
and condensation were completed in one mild step.
The optical and electrochemical properties of C8-TII were stud-
ied and compared with that of C8-BII and the alkylated isoindigo
C8-II. Their UV–vis absorption spectra are shown in Figure 1a. All
three compounds display broad absorption bands from 230 to
600 nm with similar band edges (Fig. 1a), indicating that they have
similar HOMO–LUMO band gap, and fusing a thiophene ring or a
benzene ring to the isoindigo core does not lead to a lower band
gap. The difference lies in the intensity of the absorption at long
wave length. A much more intensed absorption band with the peak
around 496 and 478 nm is observed for C8-TII and C8-BII, respec-
tively, compared to the case of C8-II at the same concentration (the
insertion in Fig. 1a). This absorption is assigned to the HOMO to
LUMO transition of isoindigo derivatives.32 This band shows an
intramolecular charge-transfer character, since HOMO of the isoin-
digo is much more delocalized throughout the whole molecule,
while the LUMO is more localized on the central rings. A strength-
ened absorption of C8-TII and C8-BII implies a strengthened intra-
molecular charge-transfer process33 that occurs in them. This
might be due to the aromatic/heteroaromatic rings fused onto
the isoindigo core in C8-TII and C8-BII.
was then tested in acetic acid. Thiophene-fused isoindigo 17 was
obtained as a red solid in 80% yield, but it is only sparsely soluble.
Unfortunately, all the attempts to alkylate compound 17 for better
solubility failed, which ended up with either no reaction (K2CO3 or
KOH in DMF at room temperature or 100 °C) due to its poor solu-
bility in those solvents or decomposition (NaH in THF, reflux). This
was disappointing since excellent solubility of the building block is
needed to synthesize solution-processable conducting polymers
for OFETs. Then alkylation before condensation might be an alter-
native way to construct soluble thiophene-fused isoindigo. Surpris-
ingly and also delightfully, we found that when treated with excess
K2CO3 in the presence of 1-bromooctane in DMF at room temper-
ature, compound 16 was directly converted into alkylated thio-
phene-fused isoindigo (C8-TII) in 70% yield in one step. The
possible mechanism might involve the alkylation of compound
16, and the oxidation of the alkylated compound by air under basic
conditions to afford the alkylated dione 15a.26–29 The alkylated
dione then underwent condensation reaction with unoxidized
16a to give C8-TII, as shown in Scheme 5. The aerobic oxidation
is the rate-determine step so that compound 15a generated in situ
would be consumed in time to afford the final product.
We also noticed that the condition for aerobic oxidation of 16a
could also be applied to the oxidation of its isomer 6H-thieno[3,2-
e]indol-7(8H)-one to afford dione 10 in 80% yield (SI). But 10 could
not react with the starting material to give the isomer of 18 due to
the steric hindrance. However, indolin-2-one could not be con-
verted to indoline-2,3-dione under the same conditions. It seemed
that fusing one electron-rich aromatic/heteroaromatic ring to
indolin-2-one was crucial for the oxidation to occur.
The electrochemical properties of the three compounds were
compared using cyclic voltammetry (CV), as shown in Figure 1b.
The energy level of LUMO of the isoindigo derivatives was
calculated from the reduction potential onsets, using
LUMO ¼ 4:44 eV þ Eorendset equation (referring to ferrocene standard).
The optical (UV–vis) and electrochemical properties of C8-II, C8-TII
and C8-BII are summarized in Table 1. For each of these isoindigo
derivatives two pairs of reduction–oxidation waves were observed.
Similar methodology was also applied to synthesize the ben-
zene-fused isoindigo (C8-BII), but with a low yield in the key step
(ca. 4%), as shown in Scheme 6. It was presumably due to the insta-
bility of compound 23 under basic conditions, since obvious
decomposition was observed during the reaction. Other method
30
such as coupling of alkylated 22 using P(NEt2)3 as the coupling
agent also end up with low yield (15%).
Cl
Cl
Cl
C8H17
C8H17
H
K2CO3
K2CO3
N
N
N
O
O
O
C8H17 Br
S
S
S
16a
16
O2
C8H17
Cl
S
Cl
Cl
N
O
C8H17
C8H17
S
- OH-
N
N
O
O
S
S
H
N
O
Cl
O
O
15a
O
C8H17
18 C8-TII
Scheme 5. Plausible mechanism of the one-pot formation of C8-TII.
CCl3CH(OH)2, HCl,
NH2OH.HCl, Na2SO4
Cl
Cl
H
N
NH2
H2SO4
80oC, 1h
73%
NCS, THF
0oC, 1h
93%
NH2
Cl
OH
N
reflux, 2h
63%
O
19
Cl
20
21
C8H17
N
Cl
O
H
N
H
N
C8H17 Br, K2CO3
Zn/TiCl4
THF, rt, 5 min
O
O
DMF, air, rt
4%
74%
N
O
O
Cl
C8H17
22
23
24 C8-BII
Scheme 6. Synthetic route of benzene-fused isoindigo.