FULL PAPER
dihydrobenzofurans 4 embedding a disubstituted sul-
foxonium ylide moiety. The chemoselectivity of the
reaction depends on a combination of catalyst activa-
tion and substitution pattern on the salicylaldehyde
aromatic ring. In more detail, the use of a Brønsted
acid catalyst steers the reaction towards the formation
of 2H-chromene structures 3, while electron-withdraw-
ing groups on the aldehyde allow the reactions to
proceed without catalyst, leading to the 2,3-dihydro-
benzofuran counterparts 4. Two competing reaction
pathways accounting for the formation of these
structures were proposed. In these pathways, the
chemoselectivity derives from a competition between a
proton-transfer step, favoured in the absence of catalyst
and leading to 2,3-dihydrobenzofurans 4, vs a nucleo-
philic addition of a second ylide to a reaction
intermediate, which ultimately delivers 2H-chromenes
3. These hypotheses were validated by reversing the
chemoselectivity of the reaction, at least in some cases,
through the modulation of the catalyst loading and the
nucleophilicity of the sulfur ylide. In general terms, the
Figure 1. Catalyst- and ylide- controlled chemoselectivity
switch.
loading. On the contrary, the more acidic/reactive results herein reported introduce a new entry in the
salicylaldehyde 1k led to the exclusive formation of multifarious, and sometimes surprising, reactivity of
product 4ka, irrespective of the amount of catalyst sulfoxonium ylides with polyfunctional substrates.[1g,5]
employed (Figure 1, middle). Thus, to reverse the Besides, the capability of a catalyst to steer the reaction
selectivity of the reaction with 1k, we considered the towards the formation of 2H-chromenes 3, vs benzo-
use of an ylide species more nucleophilic than furans 4 obtained without catalyst, highlights the
sulfoxonium ylide 2a, such as the corresponding competency of catalytic species in outcompeting innate
sulfonium derivative 2’a. In fact, according to reaction pathways and reactivities.[16]
Scheme 3, a more nucleophilic sulfonium ylide may
favor a second nucleophilic attack to a catalyst-bound
intermediate I’. Furthermore, the less acidic nature of
Experimental Section
sulfonium vs sulfoxonium salts could hinder the General Procedure for the Synthesis of Products 3
proton-transfer step in intermediate V’. Both factors
should combine towards channeling the reaction
through the pathway leading to 2H-chromene 3ka.
In a small vial equipped with a magnetic stirring bar,
salicylaldehyde 1 (1.0 equiv., 0.25 mmol) sulfoxonium ylide 2
(2.5 equiv., 0.62 mmol), CH2Cl2 (500 μL) and catalyst
Indeed, performing the reaction between aldehyde 1k
and sulfonium ylide 2’a, we were delighted to observe
that only product 3ka was present in the reaction
mixture, with no traces of the corresponding 2,3-
dihydrobenzofuran derivative 4’ka (Figure 1, right).
The yield of 3ka could be even increased to a
moderate level, by using a larger amount of catalyst. In
contrast with sulfoxonium ylides, which do not form
products 3 in the absence of catalyst, sulfonium ylide
2’a could deliver small amounts of 3ka even without
catalyst being present.[14,15] On the other hand, the
reaction of ylide 2’a with the electron neutral neutral
salicylaldehyde 1a was found to give the product 3aa
in low yield (see SI), possibly due to the poor stability
of this ylide under the reaction conditions.
(PhO)2POOH (3.1 mg, 0.013 mmol, 5 mol% in two equal
portions, the first one immediately and the second one after 8 h)
were added. The resulting solution was stirred for 48 h at room
temperature and then directly purified by column chromatog-
raphy on silica gel, to afford the desired compound 3 as a solid.
Diethyl 2H-chromene-2,3-dicarboxylate (3aa)
Following the general procedure using salicylaldehyde 1a and
sulfoxonium ylide 2a, product 3aa was obtained as white solid
in 72% yield (49.7 mg) after column chromatography on silica
gel (n-hexane/Et2O=5:1). 1H NMR and 13C NMR analysis
reported below are in accordance with the literature.[17] 1H
NMR (400 MHz, CDCl3) δ=7.51 (d, J=0.7 Hz, 1H), 7.30–
7.24 (m, 1H), 7.17 (dd, J=7.5, 1.7 Hz, 1H), 6.99 (ddd, J=8.2,
1.3, 0.6 Hz, 1H), 6.94 (m, 1H), 5.78 (s, 1H), 4.36–4.24 (m, 2H),
4.20–4.05 (m, 2H), 1.34 (t, J=7.4, 3H), 1.18 (t, J=7.1, 3H).
13C NMR (101 MHz, CDCl3) δ=169.1, 164.4, 153.8, 133.4,
132.4, 129.1, 122.2, 121.4, 119.8, 116.5, 71.8, 61.6, 61.1, 14.2,
14.0. EI-MS (m/z, relative intensity): 276 (M+, 2%), 203 (M+
Conclusions
In conclusion, we discovered a chemodivergent reac-
tion between salicylaldehydes 1 and stabilized sulfoxo- À CO2Et, 100%).
nium ylides 2, leading to 2H-chromenes 3 or trans-2,3
Adv. Synth. Catal. 2021, 363, 1–8
5
© 2021 The Authors. Advanced Synthesis & Catalysis
published by Wiley-VCH GmbH
��
These are not the final page numbers!