W. Wu et al. / Chinese Chemical Letters 21 (2010) 18–22
19
Scheme 1. (1) Reaction conditions. 1 mmol propargyl ether, 1 mmol phenylacetylene, in 4 mL H2O/THF (1:1), at room temperature, 1 mol%
catalyst, 1 h; (2) yields were determined by GC–MS using acetophenone as standard.
Scheme 2. a 1%[Rh(cod)Cl]2, H2O/THF (1:1), r.t., 1 h.
To start our research, cross-coupling between dipropargyl ether and phenylacetylene was used as a model reaction
(Scheme 1) and several commercially available catalysts were examined for the cyclization. Because all the catalysts
screened are insoluble and always results in a solid ball in water, pure water is not a good choice as the reaction media.
Otherwise the reaction is difficult to be carried out completely. Therefore, a water/THF (1:1) mixed solvent was
employed as the reaction media for a balanced solubility and reactivity. The use of [Ir(cod)Cl]2 or [Ir(coe)2Cl]2 led to
either low yield or no desired reaction product, respectively, while as it was reported that the reaction can occur in the
presence of specific ligand (like DPPE) [7]. The use of rhodium-based catalyst such as Rh(acac)(C2H4)2 and
[Rh(cod)Cl]2 provide the cyclization product smoothly in 65% and 72% isolated yield respectively. Other metal
complexes such as RuCHPhCl2[P(C6H12)3]2, Ru(CO)(H2)(PPh3)3, (CH2 CHCH2PdCl)2, Pd2(dba)3, PdCl2(PPh3)2, as
well as Re(CO)5Cl were virtually inactive under the same reaction conditions.
Subsequently, various diynes and alkynes including both terminal alkynes and internal alkynes were employed
under our standard condition (Scheme 2). The results with [Rh(cod)Cl]2 catalyst are listed in Table 1. In each case,
when N,N-dipropargyl p-toluenesulfonamide was used as the diyne (Table 1, entries 1–7) essentially no by-product
(such as due to oligomerizations) was observed when reacting with a terminal alkyne to yield the corresponding
cyclization product in high yields. However, the use of internal alkynes decreased the reactivity, possibly due to steric
reasons.
Recently, Yus and co-workers reported synthesis of 4-phenylphthalan via cyclization of dipropargyl ether with
phenyl acetylene, by use of Wilkinson’s catalyst in toluene [8]. The reaction gave only a moderate yield (50%). In our
cases, good yields can still be observed for most reactions when dipropargyl ether was used as the diyne (Table 1,
entries 8–16). In addition, when an internal alkyne with electron-withdrawal groups was employed as the monoalkyne,
the reaction gave a rather low yield (entry 17, 34%). It is worthy to note that the yields (71–84%) for dipropargyl ether
reactions were roughly lower than those for N,N-dipropargyl p-toluenesulfonamide reactions. Obviously the center
atom (N or O) played a crucial role for this cyclization. To further test this center atom effect of the diynes, 1,6-
heptadiyne (Table 1, entry 18), 2,2-dipropargyl-1,3-cyclohexadione(Table 1, entry 19) and dipropargylthioether
(Table 1, entry 20), in which the center atom is either C or S, were used as diynes for the cyclization. Interestingly, very
low yields of the cyclization products were observed. These reactions showed a significant drop in reaction rate and
resulted in complicated mixture of products.
To the best of our knowledge, although a variety of diynes with either electron-donating or electron-withdrawing
groups were investigated as substrates for the cyclization [9], there is no comment on the role of center atom. It is more
reasonable that coordination between the metal and the center atom played an important role for certain kind of diyne
substrates, than something related to its electronic effects. To verify such a hypothesis (see Table 2), N,N-dipropargyl
p-toluenesulfonamide was reacted with 1,6-heptadiene under the same reaction condition, again virtually a single
cyclization product was obtained, with an excellent yield (98%). The reaction of dipropargyl ether with 1,6-heptadiyne
lead to the cross-cyclization product as major product (yield 67%), together with some oligomerization products. No
Wu, Wei
