Catalyst-Free [3+2] Cycloaddition of Electron-Deficient Alkynes and o-Hydroxyaryl Azomethine Ylides in Water

A catalyst-free [3 + 2] cycloaddition reaction of electron-deficient alkynes and o-hydroxyaryl azomethine ylides in water was developed, affording pyrroline derivatives in moderate to high yields (up to 90%).


■ INTRODUCTION
Nitrogen-containing heterocycles are widely present in various drugs, natural products, organic catalysts, and synthetic intermediates. 1 Among these structural skeletons, 3-pyrroline has a unique application in that they not only can be further reduced to saturated pyrrolidines but also can be oxidized to pyrrolidones. 2 As shown in Figure 1, 8-oxoerymelanthine (I) is a type of erythrina alkaloids and it has hypnotic activity and pharmacological effects such as sedation and antihypertensive effects. 3 MK-0731(II), as a type of kinesin spindle protein inhibitors, can be used as a good cytotoxic agent for the treatment of refractory solid tumors. 4 Spirodactylone (III), an alkaloid spirolactone isolated from marine sponge, showed activity in inhibitors of the carcinogenic PAX3-FOXO1 transcription factor. 5 Considering the huge potential in drug discovery of the pyrrolidine, it is highly desirable and a great challenge to develop facile and efficient methodologies to access structural diversity and complexity of pyrrolidines.
To date, because of its significant biological activity, a number of methods have been reported to obtain 3-pyrroline derivatives. Krause's group discovered in 2004 that when an amino group or a substituted amino group was present in the α-position of allene, under the catalysis of gold, intramolecular self-cyclization can occur to form an N-substituted pyrroline derivative (Scheme 1, eq 1). 6 Subsequently, the Fan group, Soriano group, and Harmata group successively discovered that the substrate can also undergo intramolecular cyclization under the catalysis of different transition metals such as copper, platinum, and silver. 7 In addition to metal catalytic systems, phosphine can be used to catalyze a series of unsaturated hydrocarbons to obtain pyrroline derivatives. Kwon's group discovered that α-aminoalkylallenic esters can undergo intramolecular γ-umpolung addition under the catalysis of phosphine to obtain 3-carbethoxy-2-alkyl-3-pyrrolines. 2 Compared to intramolecular addition, cycloaddition of two molecules was also a strategy for the synthesis of pyrrolines. In 2008, Lu's group found that the modified allylic compounds and N-tosylimines can undergo [3 + 2] cycloaddition under the catalysis of phosphine to obtain N-protected pyrroline derivatives (eq 2). 8 Subsequently, Marinetti's group found that conjugated dienes which were activated by electron-withdrawing groups on both ends can undergo [3 + 2] cycloaddition with imines under phosphine catalysis to obtain functionalized 3-pyrrolines. 9 Recently, Tong's group found that δ-acetoxy allenoates can undergo [3 + 2] cycloaddition with 2sulfonamidomalonate under the catalysis of phosphine to obtain highly substituted 3-pyrrolines. 10 Besides, Zhang's group found that under the catalysis of Lewis acid, alkynes and N-tosylaziridines can undergo [3 + 2] cycloaddition via a C−C bond cleavage to obtain highly substituted 3-pyrrolines (eq 3). 11 In addition to the methods described above, the use of azomethine ylides and electron-deficient alkenes was also an effective approach to obtain 3-pyrroline derivatives since Grigg first reported it in 1978. They used a very active dimethyl acetylenedicarboxylate and N-phenylmaleimide refluxing in toluene to obtain the corresponding pyrroline derivative. 12 It is significant to emphasize that Deng's group successfully catalyzed the cyclization of azomethine ylides with electrondeficient alkynes using silver and copper (eq 4) and achieved asymmetric synthesis of pyrroline derivatives. 13,14 However, under their catalytic system, only the terminal electrondeficient alkynes can react with azomethine ylides. When the end of alkyne was attached with substituents, the desired product cannot be obtained. In addition, our group also previously studied the synthesize of pyrroline derivatives by azomethine ylides and allenoates under phosphine catalysis or triethylamine. 15,16 Nevertheless, these strategies have two key disadvantages. First, the use of transition-metal catalysis may result in the presence of residual metal ions which may lead to heavy metal contamination, while the organic phosphine is also toxic and environmentally unfriendly. In addition, organic solvents are also harmful to the environment. Herein, it is of great importance to develop a facile, environmentally friendly approach for preparing pyrrole derivatives without toxic catalysts and avoiding the use of organic solvents.
Compared to traditional transition-metal catalysis 17 or phosphine-catalyzed reaction of azomethine ylide with electron-deficient alkynes, the highlight of our work is that no catalyst was used and no organic solvents were used. Using water as the solvent is completely nontoxic, environmentally friendly, and inexpensive. In addition, the reaction had a relatively high yield and no side reactions occurred during the process, and therefore, the atomic utilization rate can be recognized high. Last but not the least, the raw materials we used are easy to prepare.

■ RESULTS AND DISCUSSION
According to our knowledge, the ynone 2a can undergo nucleophilic addition with Ph 3 P to generate zwitterions, 18 and azomethine ylide can generate carbon anions under the action of zwitterions. We initiated our investigation with the cyclization of o-hydroxyaryl azomethine ylide 1a and 4phenyl-3-butyn-2-one 2a under the phosphine-catalyzed condition. First, we used Ph 3 P as the catalyst and replaced different solvents such as methanol, toluene, acetonitrile, tetrahydrofuran (THF), diethyl ether, dimethylformamide (DMF), acetone, dichloromethane, and so on. Surprisingly,

ACS Omega
http://pubs.acs.org/journal/acsodf Article bases such as Cs 2 CO 3 , Na 2 CO 3 , and NaHCO 3 . We found that using NaHCO 3 as the base can further increase the yield (up to 78%), while the reaction time reached 18 h. Based on the experimental results, we speculated that the reaction can be catalyzed by the weak base. Then, we made a new bold attempt: whether the reaction can continue to occur without adding any catalyst. In the beginning, we used ethanol as the solvent; when no catalyst was used, we found that the reaction could not occur at room temperature. Then, we tried to increase the reaction temperature and to our excitement, the reaction could be completed in 3 h when ethanol was refluxed and the yield did not decrease. Then, we had a new attempt to use water as the solvent. To our excitement, the reaction can be successfully carried out by using water as the solvent and no catalyst was added while the yield would drop slightly (70%) ( Table 1). Considering that water is a solvent which is completely harmless to the environment and is cheaper than ethanol, we determined to use water as the solvent.
With the optimal reaction conditions secured, we tried to establish the scope of the synthesis of pyrrolines. In general, the reaction proceeded well to form the desired products 3 in good yields. When azomethine ylides reacted with alkynyl ketones (Scheme 2), the desired product can be obtained whether the alkynyl ketone was a terminal alkyne or a substituent. In addition, both ends of the alkynyl ketone can either be an alkane compound or an aromatic compound (3ai−3al). However, when both ends of the alkyne were alkane compounds, the yield of the resulting product (3ai) decreased (40%). It is worth noting that the substituent can only occur at the 5-position of the o-hydroxyaryl azomethine ylides (3aa−3ah).
In order to study the reaction mechanism, we synthesized diethyl (E)-2-(benzylideneamino)malonate 7 to react with alkynyl ketone; however, the expected target product cannot be obtained under our standard conditions (eq 1) (Scheme 4). In addition, we also synthesized ethyl (E)-2-((2hydroxybenzylidene)amino)acetate 8; it still cannot react with alkynyl ketone to yield the target product (eq 2). At the same time, we also guessed whether it was because of the self-catalysis of phenolic hydroxyl groups. Therefore, we still used compound 7 and alkynyl ketone, under the same conditions; stoichiometric and catalytic amounts of phenol were added to the reaction system (eq 3), and it was found that no corresponding product was produced. Therefore, we believe that this reaction is not because of the self-catalysis of the phenolic hydroxyl group. In addition, according to previous reports of Grigg, some evidence for a tautomeric equilibrium was provided by heating the imines. 12 We first dissolved compound 1a in a mixed solution of CDCl 3 and CD 3 OD, heated it to 70°C in a sealed tube, and observed whether there are 1,3-dipolar isomers 9. According to our previous work and related knowledge, the hydroxyl group at the 2-position of the azomethine ylide can form intramolecular hydrogen bonds with nitrogen to stabilize ylide. 19 When the reaction is carried out using the substrate 8 from which an ester group is removed, transition-metal catalysis is required because of its insufficient activity. 20 Based on the results of our control experiments, we speculated the possible mechanism (Scheme 5): o-hydroxyaryl azomethine ylide in the polar protic solvent, heating will produce intramolecular proton transfer isomer 9, 12 the carbon anion on the α-position of the two ethyl ester groups attacks one side of the alkynyl ketone, forming a double bond while negative charge transfer to form an intermediate 10. Then, intermediate 10 undergoes intramolecular Morita−Baylis− Hillman reaction to generate product 3aa.
To highlight the synthetic potential of the current method, the transformations of 3 were investigated (Scheme 6). The pyrrolidine derivatives can be readily transformed into other interesting compounds because of the presence of a free hydroxyl group. Treatment of 3aa with a formaldehyde

■ CONCLUSIONS
In conclusion, we have developed a catalyst-free, o-hydroxyassisted [3 + 2] cycloaddition of azomethine ylides with electron-deficient alkyne. This reaction allows a wide range of o-hydroxyaryl azomethine ylides and alkynyl ketones or alkyne esters to provide useful and densely functionalized pyrroline derivatives. Notably, our discovery is not only the first example of the catalyst-free 1,3-dipolar cycloaddition reaction of azomethine to electron-deficient alkynes in the water phase but also greatly complementary to Deng's metal catalyst system. However, it is regrettable that our scheme cannot achieve asymmetric synthesis of pyrroline; therefore, this is the direction our laboratory is working on.

■ EXPERIMENTAL SECTION
General Methods. All reactions were refluxed in water unless otherwise stated. Reactions were monitored through thin layer chromatography (TLC) on 0.30 mm SiliCycle silica gel plates and visualized under UV light. NMR spectra of the new products were recorded using Bruker AC-300 and Bruker AC-400 instruments, calibrated to CDCl 3 as the internal reference (7.26 and 77.0 ppm for 1 H and 13 C NMR spectra, respectively). Chemical shifts (δ) and coupling constants (J) were expressed in ppm and Hz, respectively. The following abbreviations indicate the multiplicities: s, singlet; d, doublet; t, triplet; q, quartet; and m, multiplet. High-resolution mass spectrometry (HRMS) was obtained on a Thermo Scientific LTQ Orbitrap XL Instrument equipped with an ESI source.