PVA/Fe(NO₃)₃ nanofiber mats: an efficient, heterogeneous and recyclable catalyst for the synthesis of quinolines via Friedländer annulations

Abstract

PVA/Fe(NO₃)₃ nanofiber mats are found to be a heterogeneous recyclable catalyst for the efficient synthesis of quinoline derivatives in the Friedländer condensation of 2-aminoarylketones with carbonyl compounds and β-keto esters. The novel catalytic system could be recovered and reused at least 5 times without any loss of catalytic activity.

---

Quinoline derivatives are an important class of heterocycles, structural unit in alkaloids, therapeutics and synthetic analogues.¹ These compounds have interesting biological activities, such as antimycobacterial,² antimalarial,³ anti-inflammatory⁴ and anticancer.⁵ Quinolines are also valuable reagents for the synthesis of nano- and mesostructures with enhanced electronic and photonic properties.⁶ Because of their importance in a wide range of synthetic and natural products, significant effort continues to be directed towards the development of novel procedures and catalysts for the preparation of quinoline derivatives including the Skraup,⁷ Pfitzinger,⁸ Friedländer⁹ and Combes reactions.¹⁰ Among them, the Friedländer reaction is one of the most straightforward and excellent methods. This procedure usually involves the acid or base catalyzed condensation reaction between a 2-aminoarylketone or aldehyde and a carbonyl compound possessing a reactive α-methylene group.¹¹ Lewis acid catalysts such as Zr(NO₃)₄,¹² Y(OTf)₃,¹³ AgOTf,¹⁴ ZnCl₂ ¹⁵ and PdCl₂ ¹⁶ have been used in the Friedländer reaction. However, many of these procedures have some limitations, such as low yield, low selectivity, long reaction time, harsh reaction conditions and the need of an excess amount of catalyst. Recently, FeCl₃,¹⁷ FeCl₂ ¹⁸ and Fe(NO₃)₃ ¹⁹ have been reported for the effective synthesis of quinolones as iron based Lewis acid catalysts. Despite their high yield and short reaction time, the homogeneity of the catalyst, tedious work-up, non-reusability, deactivation (due to aggregation of particles) and product contamination by residual transition metal are the main problems arising in the use of these catalysts. A general scope in catalysis is the conversion of a homogeneous catalyst into a heterogeneous catalytic system.¹⁹ The immobilization of an homogeneous catalyst onto a solid support opens up a new route for the design and engineering of novel catalysts. These structures are easily separable from the reaction mixture, allowing the recovery of the solid and its eventual reuse. However, the heterogenization of a homogeneous catalyst decreases its catalytic activity because it reduces the accessible catalytic centers and makes the diffusion of reactants to the catalytic sites difficult.²⁰ To enhance the availability of the active sites, materials with porous structures and different smaller size structures, such as nanoparticles and nanofibers, have been utilized.²¹

Poly(vinyl alcohol) (PVA) is a water soluble and biodegradable synthetic polymer. The polymerization of vinyl acetate, followed by hydrolysis of the product, leads to the production of PVA. It is white (colourless), odourless and granular in its commercial form. The polyhydroxy structure of this polymer is shown below.

PVA is used in various industries, such as membrane, coatings, textile sizing, adhesive and medical devices, because of its biocompatibility, good chemical resistance and physical properties.²²

This inexpensive and non-toxic polymer is a good candidate as a supporting matrix to immobilize Lewis acid catalysts. Transition metal species can be chelated with the numerous surface hydroxyl functional groups in PVA to prepare a heterogeneous solid support, which is insoluble in most organic solvents and can also be functionalized. However, the semi-crystalline structure, high melting point and low active sites of bulk PVA makes it difficult to fabricate it into a suitable solid matrix for catalytic applications.²³ Electrospinning is a versatile method for the preparation of ultra-long, one-dimensional PVA nanofibers. This method uses an electromagnetic field with a voltage sufficient to overcome the surface tension of a polymer solution and to draw it into an ultrafine nanofiber. The electrospun nanofibers exhibit a number of unique features and properties such as 1D morphology, extraordinary length, high surface area, and hierarchically porous structure.²⁴ When compared to nanoparticles, electrospun nanofibers can be easily prepared as membranes (supported or free-standing, and non-woven or woven) that can be conveniently handled and manipulated during an application.²⁵ Few studies have been reported on the catalytic activity of PVA nanofibers. L. Shao used Pd/PVA nanofiber mats as a catalyst in a coupling reaction.²⁶ However, to the best of our knowledge, no study has been conducted on the use of electrospun PVA nanofiber mats as a heterogeneous iron salt catalyst. We herein report the use of a PVA/Fe(NO₃)₃ nanofiber mat as an effective heterogeneous catalyst for quinoline synthesis via a Friedländer annulation. This catalyst showed good catalytic activity and high selectivity under mild conditions. Moreover, it can be easily separated using forceps or simple filtration and reused at least 5 times without any loss of catalytic activity.

Results and discussion

Preparation and characterization of the PVA/Fe(NO₃)₃ nanofiber mats

Herein a PVA/Fe(NO₃)₃ solution was electrospun to produce nanofiber mats that are insoluble in most of the solvents unlike the Fe(NO₃)₃ starting material. This freestanding membrane-like structure was used as a novel reusable iron catalyst in the Friedlander reaction.

The uniform and bead-free structure of the nanofiber mats could be achieved when the PVA weight concentration was at 6.0–8.0% in water.²⁷ The PVA weight (8.0 wt%) was chosen for the preparation of the PVA solution and was mixed with ferric nitrate.²⁸ Then, the electrospinning of the solution produced PVA/Fe(NO₃)₃ nanofiber mats (Scheme 1). The SEM images shown in Fig. 1a and b show the morphologies of the PVA/Fe(NO₃)₃ nanofibers at different magnifications. As this figure shows the nanofibers appear uniform in size, with smooth surfaces and no significant beading in the samples. The resultant nanofiber mats offer unique properties such as a high surface area to volume ratio, porous membrane structure, good interconnectivity of pores and potential to functionalize on a nanoscale. The obtained histograms (Fig. 1c) confirm the narrow diameter distribution for 50 observed nanofibers. The diameter of the PVA/Fe(NO₃)₃ nanofibers is between 80–200 nm. Fig. 2 provides the FTIR spectra of PVA, Fe(NO₃)₃·9H₂O and PVA/Fe(NO₃)₃ nanofibers. The FTIR spectra of the PVA/Fe(NO₃)₃ nanofiber mats are similar to that of PVA with only a slight increase in the transmittance. Moreover, the peaks corresponding to the Fe(NO₃)₃ stretch were not observed indicating that the Fe(NO₃)₃ was well retained in the PVA matrix. X-ray diffraction analysis (Fig. 3) indicated that the nanofibers have an approximately amorphous structure and the crystallites of the PVA powder diminished during the solution preparation followed by electrospinning. In addition, there is no characteristic diffraction peak for the ferric nitrate, suggesting that the metal salt dispersed on the PVA nanofiber mats has an amorphous structure and was successfully chelated to PVA.²⁹ Previously, groups of researchers have studied the binding energy of O 1s and Fe 2p during the formation of the PVA/Fe(III) complex in nanofibers. These results indicated the bonding of the hydroxyl groups in PVA to the metal ion through a Fe–O–C bond.³⁰

Scheme 1 Preparation of PVA/Fe(NO₃)₃ nanofiber mats.

Fig. 1 (a and b) SEM image of electrospun PVA/Fe(NO₃)₃ fiber at different magnifications, (c) fiber diameter distribution of the electrospun PVA/Fe(NO₃)₃ fibers.

Fig. 2 FT-IR spectra of: (a) PVA, (b) Fe(NO₃)₃·9H₂O, and (c) PVA/Fe(NO₃)₃.

Fig. 3 XRD pattern of: (a) PVA, (b) Fe(NO₃)₃·9H₂O, and (c) PVA/Fe(NO₃)₃ nanofibers.

Synthesis of quinolines catalyzed by the PVA/Fe(NO₃)₃ nanofiber mats

After catalyst preparation, we investigated the reaction of 2-aminobenzophenone with ethyl acetoacetate in the presence of PVA/Fe(NO₃)₃ nanofibers as a model reaction to check the catalytic activity in the Friedländer reaction (Scheme 2).

Scheme 2 Synthesis of ethyl 2-methyl-4-phenylquinoline-3-carboxylate.

The reaction is normally carried out in toluene, which is the normal solvent for the Friedlander reaction.³¹ It is well-known, that Fe(NO₃)₃ in its homogenous form is effective for the synthesis of quinoline (Table 1, entry 1). This reaction proceeded in a lower time and good yield because of the accessibility of the catalytic Fe³⁺center in solution, but the salt was soluble and homogenous in the solvent resulting in contamination of the desired product by salt and required a difficult work up without catalyst recycling. In the presence of the heterogeneous PVA/Fe(NO₃)₃ nanofibers, the reaction cleanly and simply proceeded, and the condensation product precipitated in excellent isolated yield (Table 1, entry 2). Our result showed that this catalyst was insoluble in toluene after 1 week at reflux. To optimize the reaction conditions and improve the yield, our attention was focused on the role of temperature. It was noticed that increasing the temperature to 80 °C considerably enhances the yield. By elevating the temperature up to 80 °C, the yield remained unchanged (Table 1, entries 3 and 4). In the next step, different amounts of catalyst were used for this transformation (Table 1, entry 5 and 6). Increasing the amount of catalyst increased the yield of quinoline. In the absence of the catalyst, the reaction takes more than 24 h for completion (Table 1, entry 7). The PVA nanofibers with no iron catalyst react to give a very low yield over a long time (Table 1, entry 8). Using the PVA nanofibers catalyst containing 3 mol% of Fe³⁺ is most effective in terms of both time and yield. Therefore, the optimum conditions for this reaction are as follows: PVA/Fe(NO₃)₃ nanofiber catalyst containing 3 mol% of Fe³⁺ at 80 °C.

Table 1 Optimization of the reaction conditions using the model system^a

Entry	Catalyst^b (mol%)	Temp. (°C)	Time (h)	Yield^c (%)
a Reaction conditions: 2-aminobenzophenone (2 mmol), ethyl acetoacetate (2 mmol), toluene (5 mL).b mol% of Fe^3+ in catalyst.c Isolated yield.d Reaction completed after 18 h.
1	Fe(NO3)3·9H2O (1)	60	2	84
2	PVA/Fe(NO3)3 nanofiber (1)	60	10	70
3	PVA/Fe(NO3)3 nanofiber (1)	80	10	78
4	PVA/Fe(NO3)3 nanofiber (1)	100	5	80
5	PVA/Fe(NO3)3 nanofiber (2)	80	4	85
6	PVA/Fe(NO3)3 nanofiber (3)	80	4	96
7	No catalyst	80	24	0
8^d	PVA nanofiber (0)	80	12	Trace

---

We extended these reaction conditions to a variety of 2-aminoarylketones with carbonyl compounds. As shown in Table 2, this method is equally effective for substituted 2-aminoarylketones such as 2-aminobenzophenone or 2-amino-5-chlorobenzophenone and 2-aminoacetophenone. Various 1,3-diketones and β-ketoesters reacted with 2-aminoarylketones to obtain the corresponding quinolines. Under the same conditions cyclohexane as a simple cyclic ketone afforded the tricyclic quinoline; however, the reaction needed an excess amount of ketone and proceeded slower than the other substrates (Table 2, entry i).

Table 2 Preparation of quinoline derivatives in the presence of PVA/Fe(NO₃)₃ nanofibers^a

Entry	Compound 1	Compound 2	Product 3	Melting point (°C)	Time (h)	Yield^b (%)	TON^c	TOF^d (h^−1)
a Reaction conditions: compound 1 (2 mmol), compound 2 (2 mmol), PVA/Fe(NO3)3 nanofibers (3 mol% Fe^3+), toluene (5 mL), 80 °C.b Isolated yield.c Turnover number (total number of product moles per mole of the catalyst).d Turnover frequency (turnover number per hour).e 2 eq. of compound 2 was used.
a				107	4	95	32	8
b				100	4	96	32	8
c				108	4	90	30	7.5
d				108	6	80	27	4
e				131	4	96	32	8
f				100	4	98	33	8
g				135	4	90	30	8
h				156	6	75	25	4
i^e				163	8	70	23	3
j				215	4	78	26	6.5
k				100	5	85	28	6
l				Oil	5	85	28	6
m				Oil	5	78	26	5
n				105	6	70	23	4

---

The reusability of the nanofibers was also examined. After completing the first run of model reactions, the catalyst was washed three times with CH₂Cl₂, and then was dried in a vacuum and reused in the subsequent reaction. The large structure of the nanofiber mat catalyst leads to its easy separation process using forceps. No significant loss of catalyst activity was observed even after 5 times of reuse (Fig. 4). A negligible decrease in the catalyst activity, rules out the deactivation of catalyst in the Friedländer reaction.

Fig. 4 The reusability of the PVA/Fe(NO₃)₃ nanofibers in the reaction between 2-amino benzophenone and ethyl acetoacetate.

Leaching of Fe(NO₃)₃ from the nanofibers to the solvent was analysed using ICP. We analysed the concentration of Fe³⁺ in the PVA/Fe(NO₃)₃ nanofibers catalyst before and after the reaction. The Fe³⁺ contents have been determined to be 0.055 and 0.052 mg for each mg of fresh and used catalyst, respectively. After the completion of the reaction (monitored by TLC), the catalyst and product were separated from the solvent. Having evaporated the solvent, water was added as the new solvent. The resultant aqueous solution was analysed using ICP. The content of Fe in the solution was 0.0002 mg for each mg of catalyst, indicating that the Fe leaching was less than 1%. These results show that the Fe leaching from the PVA/Fe(NO₃)₃ nanofiber mats is not significant during the reaction. SEM analysis of the nanofibers after the reaction (Fig. 5) showed that this catalyst is stable under the reaction conditions and was insoluble in toluene. Furthermore, the SEM image of the nanofibers after 5 runs showed no distinct change in morphology.

Fig. 5 SEM image of the electrospun PVA/Fe(NO₃)₃ fibers (a) after reaction (b) after 5 times of reuse.

Experimental

Instruments and characterization

All the chemicals were purchased from Merck chemicals Co. (Germany). Reagents were used without further purification and deionized water was used as solvent.

The electrospinning process was carried out using a Electroris (FNM Ltd., Iran, http://www.fnm.ir) as an electrospinner device having a high voltage and a syringe pump controllable in a range of 1–35 kV and 0.1–100 mL h⁻¹, respectively. This device can control the electrospinning parameters such as injection rate, drum rotation speed, working distance, needle scanning rate and temperature of the electrospinning media. Electroris has also been used for the industrial preparation of nanofibers. SEM images were observed using a SEM (Philips XL 30 and S-4160) equipped with gold coating. Powder XRD spectra were recorded at room temperature using a Philips X'pert 1710 diffractometer with Co Kα radiation (α = 1.54056 Å) in the Bragg–Brentano geometry (θ–2θ). FTIR spectra were obtained over the region of 400–4000 cm⁻¹ using a NICOLET IR100 FTIR and spectroscopic grade KBr. NMR spectra were recorded in CDCl₃ (Bruker, 250 MHz, Switzerland) and the proton chemical shifts reported in ppm relative to TMS as an internal reference.

Catalyst preparation

0.800 g of PVA powder was added to 10.0 mL of deionized water (8.0 wt%). Then, 0.400 g ferric nitrate (Fe(NO₃)₃·9H₂O) was added to the above PVA aqueous solution with stirring for 6 h. The electrospinning process was carried out using Electroris as an electrospinner. The obtained solution was loaded into a plastic syringe with 0.8 mm inner diameter of pinhead. In our experiment, a voltage of 20 kV was applied for electrospinning. The rate of flow was 1.0 mL h⁻¹ and aluminum foil served as the counter electrode. The distance between the capillary and the substrate electrode was 12 cm, while all the parameters were controlled by the device. The electrospun PVA/Fe(NO₃)₃ nanofibers were placed in a vacuum oven for 12 h at room temperature in order to remove the solvent residues, and then peeled off from the foil for use as a catalyst.

General procedure for the synthesis of quinolines

PVA/Fe(NO₃)₃ nanofiber mats (61 mg, 3 mol% Fe³⁺) was added to a mixture of 2-aminoaryl ketones (2 mmol) and β-ketoester/1,3-diketone/cyclic ketone (2 mmol) in toluene (5 mL) at 80 °C. The mixture was stirred until the completion of the reaction, as indicated by TLC. The nanofiber mats were separated from the reaction mixture using forceps. The precipitated product was separated by filtration and vacuum dried. Then, the product was recrystallized from CH₂Cl₂. All products were known compounds and were characterized by FTIR, Mass, and ¹H NMR spectroscopy and their respective melting points were compared with those reported in the literature.^21,32 Some selected spectral data are given in the ESI.†

Conclusions

In summary, PVA/Fe(NO₃)₃ nanofiber mats were prepared by electrospinning a solution of poly(vinylalcohol) (PVA) and Fe(NO₃)₃. The prepared nanofiber mats have been demonstrated as a novel, effective, heterogeneous and biodegradable catalyst used for the synthesis of quinoline derivatives in the Friedländer condensation of 2-aminoarylketones with carbonyl compounds and β-keto esters. The recovery of the catalyst was simple, allowing its reuse without any significant loss of its catalytic activity. The simple and fast catalyst preparation, mild reaction conditions, excellent yields, operational simplicity, product purity, and therefore cost efficiency are the advantages of this protocol. In addition, the environmentally benign membrane-like catalyst can be produced on a large scale using an industrial electrospinner.

Acknowledgements

We are grateful to the University of Tarbiat Modares for funding the project. We would also like to appreciate Dr Ali Ziyaeemehr for reading the paper and providing language assistance. This article is dedicated to memory of Majid Shahriyari.

Notes and references

1. Z. Z. Ma, Y. Hano, T. Nomura and Y. J. Chen, Heterocycles, 1997, 46, 541; J. P. Michael, Nat. Prod. Rep., 2007, 24, 223.
2. Y. L. Zhao, Y. L. Chen, J. Y. Sheu, I. L. Chen, T. C. Wang and C. C. Tzeng, Bioorg. Med. Chem., 2005, 13, 3921.
3. A. P. Gorka, A. Dios and P. D. Roepe, J. Med. Chem., 2013, 56, 5231.
4. A. Baba, N. Kawamura, H. Makino, Y. Ohta, S. Taketomi and T. Sohda, J. Med. Chem., 1996, 39, 5176.
5. C. B. Sangani, J. A. Makawana, X. Zhang, S. B. Teraiya, L. Lin and H. L. Zhu, Eur. J. Med. Chem., 2014, 76, 549.
6. S. A. Jenekhe, L. Lu and M. M. Alam, Macromolecules, 2001, 34, 7315.
7. Z. H. Skraup, Monatsh. Chem., 1880, 1, 316.
8. P. K. Calaway and H. R. Henze, J. Am. Chem. Soc., 1939, 61, 1355.
9. P. Friedländer, Chem. Ber., 1882, 15, 2572.
10. E. Roberts and E. E. Turner, J. Chem. Soc., 1927, 1832.
11. R. C. Elderfield, in Heterocyclic Compounds, Wiley, New York, 1952, pp. 45–57; E. A. Fehnel, J. Heterocycl. Chem., 1966, 31, 2899; A. Kiss, A. Potor and Z. Hell, Catal. Lett., 2008, 125, 250.
12. M. A. Zolfigol, P. Salehi, A. Ghaderi and M. Shiri, Catal. Commun., 2007, 8, 1214.
13. S. Genovese, F. Epifano, M. C. Marcotullio, C. Pelucchini and M. Curini, Tetrahedron Lett., 2011, 52, 3474.
14. H. Li, C. H. Wang, H. Huang, X. Xu and Y. Li, Tetrahedron Lett., 2011, 52, 1108.
15. P. Gunasekaran, P. Prasanna, S. Perumal and A. I. Almansour, Tetrahedron Lett., 2013, 54, 3248.
16. N. Sharma, M. Asthana, R. Kumar, K. Mishra and R. M. Singh, Tetrahedron Lett., 2014, 55, 2348.
17. H. Li, X. Xu, J. Yang, X. Xie, H. Huang and Y. Li, Tetrahedron Lett., 2011, 52, 530.
18. J. Wu, L. Zhang and T. N. Diao, Synlett, 2005, 2653.
19. M. B. Gawande, P. S. Branco and R. S. Varma, Chem. Soc. Rev., 2013, 42, 3371.
20. A. J. Sandee, J. N. Reek, P. C. Kamer and P. W. Van Leeuwen, J. Am. Chem. Soc., 2001, 123, 8468.
21. M. Sheykhan, L. Ma'mani, A. Ebrahimi and A. Heydari, J. Mol. Catal. A: Chem., 2011, 335, 253.
22. M. L. Hallensleben, Polyvinyl Compounds, Others, Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH, Weinheim, 2000.
23. L. D. Rampino and F. F. Nord, J. Am. Chem. Soc., 1943, 65, 429.
24. W. E. Teo and S. Ramakrishna, Nanotechnology, 2006, 17, 89.
25. L. M. Bellan and H. G. Craighead, Polym. Adv. Technol., 2011, 22, 304.
26. L. Shao, W. Ji, P. Dong, M. Zeng, C. Qi and X. M. Zhang, Appl. Catal., A, 2012, 413–414, 267.
27. M. G. McKee, G. L. Wilkes, R. H. Colby and T. E. Long, Macromolecules, 2004, 37, 1760.
28. H. Ziyadi, A. Heydari and S. M. Rezayat, Ceram. Int., 2014, 40, 5913.
29. T. N. Glasnov, S. Findenig and C. O. Kappe, Chem.–Eur. J., 2009, 15, 1001.
30. (a) N. Mahanta and S. Valivaveettil, RSC Adv., 2013, 3, 2776; (b) T. Tao, Z. Wu, X. Wang, M. Li and J. Zhang, Acta Polym. Sin., 2006, 1, 387.
31. J. Marco-Contelles, E. Perez-mayoral, A. Samadi, M. Carreiras and E. Soriano, Chem. Rev., 2009, 109, 2652.
32. J. Akbari, A. Heydari, H. R. Kalhor and S. A. Kohan, J. Comb. Chem., 2010, 12, 137.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra07643e

---