Chemoselective and one-pot synthesis of novel coumarin-based cyclopenta[c]pyrans via base-mediated reaction of α,β-unsaturated coumarins and β-ketodinitriles

In this paper, the base-mediated cascade reactions of 4-chloro-3-vinyl coumarins with β-ketodinitriles were demonstrated, allowing the efficient synthesis of coumarin-based cyclopenta[c]pyran-7-carbonitriles with interesting chemoselectivity. These transformations include the domino-style formation of C–C/C–C/C–O bonds through a base-mediated nucleophilic substitution, Michael addition, tautomerization, O-cyclization, elimination, and aromatization. The presented synthetic strategy has many advantages such as simple and readily available starting materials, green solvent, highly chemoselective route, synthetically useful yields, and easy purification of products by washing them with EtOH (96%), described as GAP (Group-Assistant-Purification) chemistry.


Introduction
It is highly of interest to efficiently prepare various heterocyclic structures based on privileged frameworks from the point of view of synthetic organic chemistry and drug discovery. Coumarins are important privileged heterocycles because their derivatives are valuable structures for the discovery of novel pharmaceutically active molecules and therapeutic agents. [1][2][3][4][5][6][7][8][9] For instance, hypocrolide A is a natural antibiotic, 10 derived from the fungus Hypocrea sp, and contains the coumarin core ( Fig. 1). Furthermore, some substituted coumarins are very favorable to use in perfumes, cosmetics, lasers, radiometric chemosensors, bio-sensors, and living cells imaging. [11][12][13][14] Accordingly, the synthesis of molecules containing the coumarin motif is highly desirable.
Based on the literature and our previous reports, 15-19 4chloro-3-vinyl coumarin has three potential electrophilic active sites, which can selectively be attacked by various nucleophiles. Considering these active sites, we decided to investigate the base-mediated reaction of 4-chloro-3-vinyl coumarin 2a and bketodinitrile 20 1a as a bisnucleophilic synthon. Indeed, we envisioned that this designed reaction offers an efficient pathway for the structural unity between coumarin and cyclopenta[c]pyran 21-23 moieties (Fig. 2).
It is noteworthy that the cyclopenta[c]pyran scaffold is a privileged heterocyclic system that serves as the structural core in various functionalized natural products. For example, iridoids [24][25][26][27][28] are an expansive family of natural monoterpenoids, which are characterized by their cyclopenta[c]pyran ring systems. The structures of some natural iridoids are shown in Fig. 3.
Members of this family have marine and terrestrial origins 29-31 and they have attracted broad attention owing to their various pharmaceutical activities. In this context,

Results and discussion
We started our investigation with the preparation of b-ketodinitrile 1a by the base-mediated condensation of phenacyl bromide, and malononitrile in absolute EtOH at room temperature. Aer 2 h the reaction was completed, and compound 1a was formed. Then, without further purication within a one-pot sequential process, 4-chloro-3-vinyl coumarin 2a, and two equiv. KOH were added to the reaction tube. The reaction was stirred magnetically at room temperature for 24 h, but no specic product was formed. Pleasingly, a moderate yield of the expected coumarin-based cyclopenta[c]pyran 3a (55% yield) was obtained when the reaction mixture was reuxed at 80 C for 15 hours (Scheme 1).
Encouraged by this very interesting result, the above reaction was chosen as the model reaction, and various bases were screened to optimize the reaction conditions. The desired product 3a was formed with moderate yields (49-60%) when inorganic bases such as NaOH, K 2 CO 3 , and Cs 2 CO 3 were used ( Table 1, entries 8-10). Then, piperidine (with relatively more nucleophilicity), DBU, and Et 3 N were tested for this process. The reaction promoted via piperidine didn't give 3a at all, probably owing to the nucleophilic substitution of piperidine with the chlorine atom of the substrate 2a (Table 1, entry 5). Moreover, product 3a was found with a negligible yield when DBU was used as base, and a mixture of overlapping spots were observed (Table 1, entry 6). Our examinations showed that Et 3 N is the best base for this transformation, affording the product 3a in a yield of 79% (Table 1, entry 1). In continue, different solvents were tested to obtain the optimal reaction solvent, and the best result was gained in absolute EtOH (Table 1, entries 1-4). It should be mentioned that the yield was remarkably decreased in non-anhydrous solvents. Furthermore, at temperatures below the reux temperature, the reaction was completed over longer periods of time. We established the optimal reaction conditions for the preparation of coumarin-based cyclopenta[c] pyran derivatives as follows: use of 2.0 equiv. Et 3 N as the base and absolute EtOH as the solvent to perform the reaction at 80 C (Table 1).
With the optimized reaction conditions for the synthesis of coumarin-based cyclopenta[c]pyrans in hand, we set out to investigate the generality of this cascade transformation using differently substituted a,b-unsaturated coumarins. Electrondonating andwithdrawing substitutes on aromatic rings of substrate 2 were all tolerated, affording the expected coumarinbased cyclopenta[c]pyrans (Scheme 2) in satisfactory yields (75-91%, 3b-3f). It is worth mentioning here that the nature of the substituent on a,b-unsaturated coumarins had a slight impact on the yields. Next, the inuence of different substituents (either electron-withdrawing or electron-donating) of b-ketodinitrile onto this process was investigated, and products 3g-3k were obtained in very good yields (72-83%). Notably, for substrate 1 with a NO 2 substituent attached on the benzene ring, no expected product was detected. With the aim of  exploring the synthetic utility of this novel domino process, a gram-scale experiment was performed with phenacyl bromide (3.5 mmol, 0.696 g), malononitrile (3.5 mmol, 0.233 g), and 4chloro-3-3-viyl coumarin (3.5 mmol, 1.08 g), yielding the desired  compound 3a in 71% yield without a remarkable loss of efficiency compared to small scale (79%). The structures of the synthesized products were characterized by Fourier transform infrared (FTIR), mass spectrometry, elemental analysis, and 1 H NMR. Notably, the solubility of the products was too low, and except for 3d which was more soluble than the others, we couldn't record 13 C{H} NMR spectra for them. The molecular structures of the synthesized compounds were undeniably conrmed by X-ray crystallographic analysis of product 3d (Fig. 4).
The mechanism of the rst step including b-ketodinitriles formation is known. 32

Conclusions
In summary, we presented a novel domino-type process for the synthesis of substituted coumarin-based cyclopenta[c]pyrans through a base-mediated nucleophilic substitution/Michael addition/tautomerization/O-cyclization/elimination/ aromatization reaction of 4-chloro-3-vinyl coumarins with bketodinitriles. Notably, two new carbon-carbon bonds and one carbon-oxygen bond were formed in these reactions, and highly stable polycyclic aromatic products were obtained. This unprecedented strategy was carried out under relatively green conditions, providing an efficient approach to structural unity between two valuable organic moieties.

Experimental
General Two melting points were measured on an Electrothermal 9100 apparatus. IR spectra were recorded as KBr pellets on a Nicolet FTIR 100 spectrophotometer. 1 H NMR (500 MHz, 300 MHz) and 13 C NMR (75 MHz) spectra were obtained using Bruker DRX-500 Avance and Bruker DRX-300 Avance spectrometers. All NMR spectra were recorded at r.t. in DMSO-d 6 and CDCl 3 . Chemical shis are reported in parts per million (d) downeld from an internal TMS reference. Coupling constants (J values) are reported in hertz (Hz), and standard abbreviations were used to indicate spin multiplicities. Elemental analyses for C, H, and N were performed using a Heraeus CHN-O-Rapid analyzer. Mass spectra were recorded on a Finnigan-MATT 8430 mass spectrometer operating at an ionization potential of 70 eV. All chemicals and solvents were purchased from Merck or Aldrich and were used without further purication. Starting materials were synthesized according to the procedures reported in the literature. 15-20 Single crystals of compounds 3d were formed in CH 2 Cl 2 .
General procedure for the preparation of compounds 3a-3k To a magnetically stirred solution of phenacyl bromide (1 mmol, 199 mg), and malononitrile (1 mmol, 66 mg), was added Et 3 N (1 mmol, 101 mg) in the absolute EtOH (5 ml). Aer 2 h, a,b-unsaturated coumarin 2 (1 mmol, 310 mg), and Et 3 N (2 mmol) were added to the reaction mixture. The reaction was carried out at 80 C, and