.
Angewandte
Communications
Figure 2. Mechanistic rationale for selective formation of anthraqui-
nones I–IV.
alkoxy semiquinone aminals (VI and VII, Figure 2). In order
to develop this strategy, we undertook studies to determine
the feasibility of the expected carbon–carbon bond forming
and breaking reactions and explored the optimization of the
required base, conditions, and substrates. Upon extensive
experimentation with substrates 1 and 6 (Figure 3) (see
Supporting Information (SI) for details), it was found that
pathway a [V+ VI!I, Figure 2] could be achieved smoothly
by the use of LiHMDS in THF (or DME) at À78!258C to
afford p-aminophenolic anthraquinone 10 as demonstrated in
Figure 3 (e.g. 1 + 6!10, 88% yield). This process is presumed
to proceed through intermediate G (Figure 2), which prefer-
entially collapses via departure of the methoxy group rather
than the amino group to afford the observed product.
Exploration of the generality and scope of the reaction
using a variety of cyanophthalides V (1–5, Figure 3) and p-
methoxy semiquinone aminals VI (6–9, Figure 3) employing
the above optimized conditions led to a series of novel p-
aminophenolic anthraquinones I (10–17) in good to excellent
yields as shown in Figure 3. Noteworthy is the applicability of
this process to the construction of substituted p-aminophe-
nolic anthraquinones with additional fused rings in their
structures (i.e. compounds 12, 13, and 16, Figure 3).
Figure 3. Selective formation of p-aminophenolic anthraquinones I
(10–17). Abbreviations: Alloc=allyloxycarbonyl; MOM=methoxy-
methyl; Phth=phthalimide. Reactions were carried out on 0.09–
0.2 mmol scale. For preparation of substrates and further details, see
the Supporting Information.
Attempts to implement a similar strategy for the synthesis
of o-aminophenolic anthraquinones II (pathway c, Figure 2)
employing cyanophthalides V and o-methoxy semiquinone
aminals (VII) under the same optimized conditions at À60!
258C (LiHMDS, THF) led to the targeted o-aminophenolic
anthraquinones II (Figure 2) as only the minor products.
Their o-methoxyphenolic counterparts (III, Figure 2) were
the major products in these reactions formed via pathway d
(Figure 2) as demonstrated in Figure 4a with substrates
cyanophthalide 1 and semiquinone 18 (1 + 18!22, 20%
yield; 21, 70% yield). o-Methoxyphenolic anthraquinone 21
is presumably formed through intermediate H (Figure 4a),
Figures 4b and 5, respectively. Thus, employing substrates
1 and 18, and using LiOtBu, instead of LiHMDS, at lower
temperature (i.e. À78!258C) resulted in the formation of the
o-methoxyphenolic anthraquinone 21 (85% yield) as
depicted in Figure 4b. This reaction proved of general
applicability and scope, accommodating cyanophthalides V
(e.g. 1–3) and semiquinone aminals VII (R2 = Me) (e.g. 18 and
24) as substrates, furnishing a variety of o-methoxyphenolic
anthraquinones III (e.g. 21, 25–27) in good yields as summar-
ized in Figure 4b. It should be noted that the more obvious o-
dimethoxy semiquinone 29[21] (Figure 4c) is a fleeting inter-
mediate, undergoing rapid and quantitative self [4+2]-cyclo-
addition to form dimer 30 upon generation from o-methoxy-
phenol (28) at 08C through the action of PhI(OAc)2, and
therefore, cannot be conveniently used as a precursor to this
type of anthraquinones.
whose formation from 18 and collapse to 21 (expulsion of
À
= =
MeN C O and CN ) are shown in Figure 4a. Unexpected
and not ideal, this observation prompted us to optimize the
process further in order to deliver either product (i.e. 21 or 22)
in high yield. Our results from this study are summarized in
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 12687 –12691