Organic Letters
Letter
unique triplicate roles of Ti(OiPr)4 are involved, which is
thoroughly different from the known modes. Furthermore, the
furan products display promising fluorescence, which repre-
sents an uncommon outcome in the field of material chemistry.
Herein, we report the results.
Our initial test of the annulation of 1a under a
stoichiometric amount of Ti(OiPr)4 in CH2Cl2 resulted in
no product (Table 1, entry 1). However, increasing the
Having identified the optimal conditions for the formation
of highly functionalized furan, we then focused on the survey
of the substrate scope. The variation of the R1 groups by
introducing electron-rich phenyl rings showed better effects on
the yields, releasing 2b−2d in ≤91% yield (Scheme 2, 2b−2d).
When R1 was a fluoro-substituted phenyl or naphthyl group,
the reaction proceeded well to give the corresponding
products, albeit in relatively lower yields (Scheme 2, 2e and
2f). In addition, heterocyclic furan and thiophene substituents
showed little impact on the reactions (Scheme 2, 2g and 2h),
and the aliphatic cyclopropyl group proved to be suitable for
the protocol (Scheme 2, 2i). Subsequently, we evaluated a
series of R2 groups on the substrates and found that an
electron-rich OMe group installed on a phenyl ring showed
little effect on the reaction under the optimal conditions
(Scheme 2, 2j), but a lower concentration of 0.5 M and rt have
to be used to suppress side reactions when the R2 groups were
electron-deficient phenyl rings (Scheme 2, 2k and 2l).
Additionally, when R2 was a thienyl group, a 63% yield of
2m was produced (Scheme 2, 2m), and the furan product with
a cyclopropyl substituent was also obtained, albeit in a poor
yield (Scheme 2, 2n). Furthermore, the simultaneous variation
of both R1 and R2 groups also proved to be possible, and
substrates having different substitution patterns such as
electron-rich and electron-poor (Scheme 2, 2o and 2p), both
electron-rich (Scheme 2, 2q and 2r), and electron-poor
(Scheme 2, 2s) could all produce the corresponding products.
Finally, a substrate with a camphor unit was tested and allowed
access to 2t in 67% yield (Scheme 2, 2t). The structure of the
products was confirmed unambiguously by the single-crystal X-
ray analysis of 2j, and other compounds were assigned by
analogy.
The scale-up reaction can be easily achieved using 1.17 g of
1a, which could afford 0.975 g of 2a in 83% yield (Scheme 3,
eq 1). Then further transformations on 2a were also
conducted, and we found that the alkyne moiety could be
readily transferred to ketone 3a in 91% yield under gold
catalysis (Scheme 3, eq 2). Then the hydrogenation of 2a
released 3b in good yield (Scheme 3, eq 3), and the 1,4-
diketone unit of 2a was found to undergo the condensation
with hydrazine to yield 3c in 76% yield (Scheme 3, eq 4).
To our delight, we found that the solid furan products
showed fluorescence emission under ultraviolet irradiation at
365 nm. As shown in Figure 1a, 2j, 2a, 2t, 2f, 2o, and 2d show
a gradient color change from green to yellow according to the
substitution patterns. Fluorescent materials have found wide
applications in organic OLEDs, semiconductor lasers, and
fluorescent sensors.8 However, troubles caused by aggregation-
caused quenching (ACQ) have greatly restricted the
corresponding applications because fluorescent materials are
usually used in the aggregated or solid state.9 Furthermore,
furan compounds have been conventionally considered inferior
in displaying photochemical properties.10 In this context, our
work provides an uncommon exception. As shown in panels b
and c of Figure 1, obvious red shifts can be observed from the
emission spectra of the six products, which range from 434 to
494 nm (see the Supporting Information for more details).
The origin of the unusual fluorescent emission could be
deduced from the molecular packing arrangement in the crystal
of 2j (Figure 1d). One can see that close π−π stacking exists
between the alkynyl 4-OMe-phenyl unit of one molecule of 2j
and the 4-OMe-phenyl moiety of another molecule of 2j,
which is essential for the regular arrangement and the
a
Table 1. Condition Optimizations
concn
(M)
yield
(%)
entry Ti reagent
solvent
temp
rt
rt
rt
rt
rt
rt
rt
rt
additive
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Ti(OiPr)4
Ti(OiPr)4
Ti(OiPr)4
Ti(OiPr)4
Ti(OiPr)4
Ti(OiPr)4
Ti(OiPr)4
Ti(OiPr)4
Ti(OiPr)4
Ti(OiPr)4
Ti(OiPr)4
Ti(OiPr)4
Ti(OiPr)4
Ti(OiPr)4
TiCl4
CH2Cl2
CH2Cl2
DMF
CH3CN
THF
MeOH
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
0.2
0.5
0.5
0.5
0.5
0.5
0.5
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
2.0
2.0
−
−
−
−
−
−
−
−
0
33
trace
0
15
0
41
46
53
53
79
77
75
13
0
b
rt
−
−
−
40 °C
40 °C
40 °C
40 °C
40 °C
40 °C
40 °C
40 °C
c
Zn(OTf)2
AlCl3
−
−
−
c
d
e
Ti(OtBu)4
Al(OiPr)3
0
0
e
−
a
Reaction conditions: 1a (1 mmol), under an argon atmosphere, 12 h.
b
All isolated yields were based on 1a. With 2 equiv of Ti(OiPr)4
c
d
added. With 20 mol % additive added. With 20 mol % Ti(OiPr)4
e
added. A complicated reaction mixture was obtained.
concentration to 0.5 M led to a 33% yield of 2a (Table 1, entry
2). The formation of the product undergoes a formal scission
of the CO−CO bond and the internal CO bond, and then
recombination with the second alkynyl diketone molecule. The
process is in accordance with the “cut and sew” strategy
proposed by Dong and co-workers.7 Other solvents such as
DMF, CH3CN, THF, and MeOH were also surveyed but
proved to be less suitable (Table 1, entries 3−6, respectively).
To our delight, toluene increased the yield to 41% (Table 1,
entry 7), and a concentration of 1.0 M could further improve
the yield (Table 1, entry 8). Then, 2.0 equiv of Ti(OiPr)4 or 40
°C can promote the yield to 53% (Table 1, entry 9 or 10,
respectively). A further modification of the reaction concen-
tration to 2.0 M resulted in 2a at an acceptable level (Table 1,
entry 11). The addition of a Lewis acid such as Zn(OTf)2 and
AlCl3 as the additives showed almost no influence on the yield
(Table 1, entries 12 and 13, respectively). Catalytic Ti(OiPr)4
liberated 2a in only 13% yield (Table 1, entry 14), and another
titanium reagent such as TiCl4 was also studied but produced
only a complicated reaction mixture (Table 1, entry 15). In
sharp contrast, the use of Ti(OtBu)4 led to no conversion of 1a
(Table 1, entry 16), and Al(OiPr)3 resulted in a messy mixture
despite the full consumption of the starting material (Table 1,
entry 17).
1505
Org. Lett. 2021, 23, 1504−1509