NATURAL PRODUCT RESEARCH
3
2. Results and discussion
Inspired by the intramolecular D–A approach developed by Stevenson and the
method of photochemical reaction in the preparation of naphthalenes reported by
Wessig et al. (2005). The photochemistry reaction of 3-(4-methoxyphenyl)prop-2-yn-1-
yl 3-(4-methoxyphenyl) propiolate (1a) was exploited. As expected, under a high-pres-
sure mercury lamp (175 W) irradiation for 12 h, most of the ester was converted to the
corresponding ANLs in 70% yield and the ratio of 2a (type I) to 3a (type II) is 2.3:1
(Scheme S1).
On the basis of this experiment, we attempted a continuous photoflow method for
the reaction. The reactor consists of a simple design in which fluorinated ethylene pro-
pylene (FEP) capillary tubing (ID 1.0 mm) was coiled around the UV lamp. A solution of
ester 1a in 1,2-DCE (0.005 M) was injected into the photoreactor used for the photo-
chemical reaction by a syringe pump. Encouragingly, after exposure for 6 min (the
average time that the reactants spend in the reactor), the reaction reached a modest
conversion rate of 50.3%. Moreover, the regioselectivity was also improved (ꢀ4.8:1)
(Entry 6, Table S1). In general, the selection of a proper solvent is a crucial step in con-
tinuous photoflow as the solvent needs to be able to solubilise the chemicals and
products to prevent reactor clogging and undesired light scattering. In the following
experiments, different solvents were selected for the reaction. Protonic solvents, such
as methanol, HFIP and acetic acid exhibited lower conversion rate of 19.5%, 22.8%
and 13.3%, respectively (Entries 10–12, Table S1). In contrast, most of the aprotic sol-
vents showed moderate conversions and excellent regioselectivity. The optimal choice
are haloalkanes, represented by DCM or 1,2-DCE (Entries 1, 6 and 13–17, Table S1).
The UV exposure time is also an important factor in the photoflow reactions. In Entries
1–8 (Table S1), experiments showed that the conversion of ester 1a to products 2a
and 3a was fast at the beginning several minutes, but the situation turned in after
10 min, which suggested that the prolonged irradiation may cause degradation of the
reactants or products. The initial concentration of 1a was also discussed. The conver-
sion rates were similar at the concentrations of 0.0025 and 0.005 M, however, with the
increase of concentrations (0.01 and 0.02 M), the conversion rates were decreased
(Entries 2, 6, 9 and 18–20, Table S1). In summary, the most efficient conditions of this
reaction seem to be: 1a in haloalkanes (0.005 M) was pumped through the FEP tubing
for a total exposure time (175 W Hg lamp) of 10 min.
The conditions of continuous photoflow method provide the possibilities for the
rapid preparation of ANLs and their analogs. For the preparation of those ANLs, the
corresponding 3-phenylprop-2-yn-1-yl 3-phenylpropiolates were prepared at first.
Herein, as shown in Scheme S3, we take the preparation of 1a as an example:
Anisaldehyde was first converted to dibromoalkene 4a, which was subjꢁected to the
Corey–Fuchs alkyne synthesis condition (2.5 equiv of n-BuLi in THF, À78 C) to gener-
ate alkyne anion, followed by addition of dry ice to give the desired 3-phenylpropiolic
acid 5a; while 6a was prepared by addition of paraformaldehyde to the alkyne anion
ꢁ
solution at À78 C. Coupling reaction of 5a and 6a was performed using DCC and
DMAP in DCM to afford the ester 1a.
With a series of 3-phenylprop-2-yn-1-yl 3-phenylpropiolates (1a–1e) in hands, rapid
pipelining preparation of naturally occurring ANLs were undertaken (Conditions:
Ge, Xiang
