G Model
CCLET 5114 No. of Pages 4
2
T. Lan et al. / Chinese Chemical Letters xxx (2019) xxx–xxx
Scheme 1. Nitrogen-containing benziodoxolone derivatives.
Scheme 2. Scope of carbazole-containing hypervalent iodine reagents. Reaction
conditions: KF (10 mol%), 1 (4.0 mmol), acetoxybenziodoxolone (4.0 mmol), in dry
Fig. 1. Single-crystal X-ray diffraction structure of 2a. The thermal ellipsoids are set
at 35% probability (CCDC 1909836).
CH
3
CN (10 mL) at room temperature. Isolated yields. a10 mmol scale.
reagents or Kiyokawa’s reagents, presumably due to the steric
hindrance.
Cu-catalyzed C ꢀꢀ N coupling reaction of carbazole and aromatic
heterocyclic substrates (Scheme 1B).
To highlight the potential utility of these novel reagents, we
attempted to investigate their intrinsic reactivity in organic
synthesis. We treated 3-methylbenzo[b]thiophene 3a and 1.0
equiv. of carbazole-containing hypervalent iodine reagent 2c as the
model substrates, and the reaction was performed in the presence
We first focused on the synthesis of carbazole-containing
hypervalent iodine reagents. We chose N-TMS-carbazoles as the
carbazole source to react with acetoxybenziodoxolone for synthe-
sizing the desired compounds. To our delight, with 10 mol% KF as
an additive, acetoxybenziodoxolone and 1a were mixed in dry
of 10 mol% CuCl as a catalyst, with dry acetonitrile as solvent at
CH
hypervalent iodine product 2a was obtained in 91% yield
Scheme 2). When non-dry CH CN was used as the solvent, the
3
CN and stirred for 24 h at room temperature, the desired
ꢁ
8
0 C for 24 h. The desired C ꢀꢀ N coupling product 4ac could be
obtained in 18% yield (determined by NMR analysis), meanwhile,
N ꢀꢀ N coupling bicarbazole by-product 5 could be obtained in 52%
yield (Table 1, entry 1), presumably via the dimerization of
carbazole radical A. Subsequently, the effects of other copper
catalysts were examined for the reaction. CuBr or CuTc was found
to give bicarbazole product 5 beyond 65% yield and no product 4ac
(
3
yield of product 2a decreased obviously, to only 32%. Furthermore,
hypervalent iodine reagents containing 3,6-dichloro-, dibromo-,
and diiodine-substituted carbazole groups all went smoothly to
afford the desired products in good yield. In addition, the synthesis
of hypervalent iodine product 2a and 2c could be easily scaled up to
the 10 mmol scale without yield decreased. All the carbazole-
containing hypervalent iodine reagents are stable under ambient
was detected (Table 1, entries 2 and 3). Surprisingly, using Cu(OTf)
or (CuOTf) as the catalyst, 4ac was obtained in good yields
with trace amount of byproduct 5 (Table 1, entries 4 and 5). And
when Cu(CH CN) PF was used (Table 1, entry 6), product 4ac was
formed in a higher yield of 63% (isolated in 58%). Then, a solvent
screening showed that switching the solvent from CH CN to DCM,
2
2 6 6
C H
condition, and soluble in DCM, CHCl
insoluble in CH CN at room temperature. When N-TMS-carbazoles
were replaced by carbazoles (N ꢀꢀ H) or metal carbazolide
metal = Li, Na or K) to react with acetoxybenziodoxolone in a
3
and DMSO, but almost
3
4
6
3
3
(
DCE, THF or DMSO resulted in lower yield of product 4ac (Table 1,
entries 7–10). Additionally, the change of the reaction temperature
could not improve the yield of 4ac (Table 1, entries 11 and 12). In
particular, the reaction only afforded byproduct 5 in absence of the
copper catalyst (Table 1, entry 13).
With the optimal reaction conditions in hand, the scope of this
C ꢀꢀ N coupling reaction was explored. As shown in Table 2, diverse
carbazole-containing hypervalent iodine reagents were first
variety of reaction conditions, no desired products 2 were
obtained.
The structure of 2a was further confirmed by XRD. As shown in
ꢁ
Fig. 1, the N1-I1-O1 angle of 170.74(15) and the N1-I1-C1 angle of
ꢁ
9
2.9(2) show the molecule has a distorted T-shape, in accord with
the typical structure of benziodoxolone derivatives. The I1-N1
bond length of 2.069(4) Å is shorter than the one in Zhdankin’s