CuI nanoparticles on modified poly(styrene-co-maleic anhydride) as an effective catalyst in regioselective synthesis of 1,2,3-triazoles via click reaction: a joint experimental and computational study

Abstract In situ immobilization of CuI nanoparticles (NPs) on modified poly(styrene-co-maleic anhydride) [SMA] was achieved. Proper immobilization of CuI on the prepared support was confirmed by scanning electron microscopy (SEM), energy dispersive X-ray analysis (EDAX) and inductively coupled plasma (ICP). In order to present a quantitative description for experimental features of CuI/SMI nanocatalyst, the coordination behavior and the nature of interactions between Cu(I) ions and modified SMI ligand in gas and solution phases was computationally assessed. In this line, the mathematical properties of electron density functions were calculated and analyzed topologically via density functional theory (DFT) and quantum theory of atoms in molecules (QTAIM) approaches for different coordination modes. Encouraged by our computational results, this new catalyst was employed in a one-pot, three-component reaction involving terminal alkynes, alkyl halides and sodium azide in water resulting in highly regioselective synthesis of 1,4-disubstituted 1,2,3-triazoles in high to excellent yields. The catalyst was reused without pre-activation and recycled for at least five runs without significant decrease in its activity being observed.


Introduction
The 1,2,3-triazole derivatives exhibit several diverse biological activities. They have also been widely employed as precursors in the synthesis of biologically active systems such as anticancer, antimicrobial, antifungal and other cytotoxic active agents [1]. These compounds have also shown high chemical stability towards severe hydrolytic, oxidizing and reducing conditions, even at high temperatures [2].
1,2,3-Triazoles were synthesized via Huisgen 1,3-dipolar cycloaddition [3]. A developed regioselective Huisgen 1,3-dipolarcycloaddition being performed under mild reaction conditions which has a great impact on biological, industrial and synthetic applications was in much demand. In 2002, Sharpless et al. [4] and Meldal and co-workers [5] independently reported that the use of catalytic amounts of Cu(I) resulted in the rapid and regioselective azide-alkyne cycloadditions at ambient temperature in organic medium, so-called click reaction [6][7][8][9]. The role of Cu(I) as a coordinating agent in the mechanistical pathway of click reactions has been well established via the coordination of Cu(I) with terminal alkynes and the activation of azide species to produce a copper-azide-acetylide complex [10,11].
It is known that the ligand protects the Cu ion from interactions leading to degradation and generation of side products and more importantly prevents the oxidation of the Cu(I) species to the Cu(II) [12,13]. Moreover, immobilization of Cu(I) nanoparticles on different kinds of supports is a useful method to make them heterogeneous, thus easily recoverable and reusable [14][15][16]. Applications of various forms of polymer-supported CuI as recoverable catalyst for the click synthesis of 1,2,3-triazoles between azides and alkynes have been reported [17].
Recently, the coordination of metal NPs on the polymer surfaces has attracted much attention by rational design of organic ligands on the polymer surface as coordinating agent [18,19]. In another words, to synthesize the desired polymers with predictable structure and coordination properties, it is substantial to apply appropriate multidentate ligands with strong metal-ligand interactions.
To achieve higher stability, better dispersion of the particles, as well as sufficient recyclability of the catalyst, CuI NPs have been immobilized on high surface area supports [20][21][22][23]. For the purpose of effective immobilization of CuI NPs, the support surface is modified in the multiple-step synthetic route so that it can be coordinatively bound to CuI NPs with a considerable metal-ligand interaction energy. Significantly metal polymer-supported catalysts maintain some important benefits of homogenous catalysis such as giving short reaction times and obtaining satisfactory yields in aqueous media [24]. SMA is an inexpensive commercially available copolymer of styrene and N-aryl succinimide. It contains reactive anhydride groups, which can be modified with different nucleophilic reagents [25,26], providing useful support for the design of numerous reagents and also catalysts for effective coordination with metal NPs [25,27]. However, since then, the modification and application of SMA have been largely overlooked.
We have also achieved the reaction of SMA with 4-aminopyridine and 2-aminothiazole and used it as modified co-polymer as support for immobilization of CuI and PdCl 2 NPs to prepare nanocatalysts and applied them successfully in click and coupling reactions. Moreover, we analyzed and interpreted information on the coordination of metal NPs, based on observed IR stretching frequency of coordinating residues, in conjunction with our obtained computational results [34,35].
Due to our interest in design and preparation of nanocatalysts, and also employing the computational methods for the coordination chemistry of metal complexes with modified SMA ligand in the nanocatalysts [32, 33,35,37], herein, we report in situ preparation and characterization Cu(I) NPs immobilized on modified SMA with 2-aminopyridine to obtain poly(styrene-co-maleimide) (SMI) copper catalyst. The composite was then characterized by different methods such as SEM, EDAX and ICP. More importantly, to find a satisfactory correlation between experiment and theory, we focused on theoretical interpretation of immobilization behavior of CuI on SMI support surface using density functional theory (DFT) [38] and quantum theory of atoms in molecules (QTAIM) [39] approaches. We have also considered the solvent effect through the polarized continuum model (PCM) computations [40].
It is worthwhile to state that molecular information such as structural and electronic properties of metal complexes give deeper insights into modulation of the coordination chemistry of metal complexes and so the rational design of new catalysts. The recent advances in computational chemistry approaches have led to present a theoretical description for catalytic reactions with detail and accuracy required for reliable comparison with experiments [41,42].
In the next step, the composite was used as an efficient heterogeneous catalyst in a onepot, three-component regioselective synthesis of 1,4-disubstituted 1,2,3-triazoles via click reaction in water as a green solvent. This heterogeneous catalyst exhibited an excellent catalytic performance for click synthesis of various 1,2,3-triazole derivatives and was recovered easily and reused several times with no considerable loss of activity.
It should be mentioned that in our previous research, we just reported the synthesis and characterization of CuI NPs immobilized on SMA modified by 4-aminopyridine and used it in the regioselective synthesis of 1,4-disubstituted 1,2,3-triazoles [34]. We actually overlooked an important issue which could have been helpful in interpretation of the experimental observations. That issue was the computational analysis of metal-ligand interactions in CuI/ SMI complex. As compensation, in this research, we focused on the employment of 2-aminopyridine instead of 4-aminopyridine, reacting it with SMA as a modifier. Since, in 2-aminopyridine-modified SMA, the pyridine nitrogen is nearer to carbonyl functional group of maleimide, we assumed that the former acts as a stronger coordinating agent towards metal ions in comparison with SMA modified by 4-aminopyridine. This structural feature is expected to influence the coordination behavior of CuI/SMI complex leading to more effective interaction between copper and modified SMA. This phenomenon has been observed and reported previously in the case of a similar system [43][44][45].
Thus we investigated these two different CuI/SMI complex models via quantum chemistry computations. Strictly speaking, we made a comparative assessment on the coordination behavior of CuI/SMI nanocatalyst modified with 2-and 4-aminopyridines to predict the immobilization behavior of copper NPs on SMI polymeric surface. In this line, we also compared our experimental results when SMA was modified with 2-aminopyridine and used as support for immobilization of Cu(I) NPs catalyst, with those obtained when 4-aminopyridine was used under similar conditions, reported previously [34].
It should be noticed that the lower cost of commercially available 2-aminopyridine and also easily accessible via Chichibabin reaction [46] relative to 4-aminopyridine is motivating to use former for the modification of SMA support being used as heterogeneous catalyst in the regioselective click synthesis of 1,4-disubstitued 1,2,3-triazoles.

Materials and methods
N,N-Dimethylformamide (DMF) and triethylamine (TEA) were distilled and stored over 4 Å molecular sieve before use. Other reagents were purchased from Aldrich and Merck and used as received. SMA was obtained from KARABOND. The general formula of SMA used in this study is [ Proton resonances are designated as singlet (s), doublet (d), triplet (t) and multiplet (m). FTIR spectra were recorded using KBr disks on a FTIR Bruker Tensor 27 instrument from 500 to 4000 cm −1 . Band intensities are assigned as weak (w), medium (m), and strong (s). Cu content was measured by ICP on a Varian Vista pro analyzer. Scanning electron micrographs were recorded using a Lecia Cambridge S 360 SEM instrument. All yields refer to isolated products.

Preparation of SMI
SMA (1) was modified in two steps in accord with the procedure previously reported by Lee et al. [47]. In a 100 mL glass reactor SMA (1.00 g), 2-aminopyridine (1.54 g, 16 mmol) and dry DMF (15 mL) were placed. Then gas N 2 was charged into the reactor and sealed. The reactor was located into an oil bath, and the reaction mixture was oscillated for 3 h at 35 °C. During this period of time, reagents were dissolved in DMF. Then, acetic anhydride (0.6 mL, 6 mmol), sodium acetate (0.33 g, 4 mmol) and triethylamine (0.3 mL, 2 mmol) were consecutively added into this mixture by syringe (figure 1). The reaction temperature was raised to 75 °C, while oscillation was sustained for 3 h. The mixture was then cooled to ambient temperature and poured into 300 mL of methanol under stirring. The fiber-like polymer was separated by filtration, washed with methanol thoroughly and dried under reduced pressure at 70 °C. The SMI polymer (3) was further purified by reprecipitation in DMF.
The amine content of SMI was estimated according to previously reported procedures [48,49]. Accordingly, 10 mL of HCl (0.2 N) was added to 0.05 g of the SMI, and the obtained mixture was stirred for 30 min at room temperature. The obtained SMI was washed with deionized water. Excess HCl was then titrated with NaOH (0.2 N) in the presence of phenolphthalein as an indicator. Amine site content of the synthesized SMI was determined to be 3.42 mmol/g.

Preparation of polymer-supported catalyst
CuI (0.247 mg, 0.001 mmol) was dissolved in acetonitrile (2 mL) under ultrasonic irradiation. A transparent pale yellow solution was formed. To this solution, DMF (20 mL) was added at ambient temperature, generating CuI NPs. To this mixture, dry SMI (1.00 g) was added. This mixture was refluxed under nitrogen atmosphere for 5 h. Then, the generated CuI/SMI (4) was washed, filtered off and washed with acetonitrile and dried under vacuum at 60 °C (figure 1).

Determination of the copper content in SMI-CuI
A sample of CuI/SMI (100 mg) was extracted with concentrated HCl and HNO 3 (1:1, 10 mL) for digestion of the metal complex using a screw-capped vessel. Then, the mixture was relocated into a volumetric flask (100 mL), diluted and made ready for ICP analysis. The Cu concentration was determined by conventional atomic emission (324.754 nm) by reference to a linear (R = 0.99) calibration curve of (1-4 ppm), similar to other CuI sample preparation. The Cu content was 6.92% w/w. The same procedure was employed to measure the leaching accounts of the Cu catalyst after five cycles.

Synthesis of 1,4-disubstituted 1,2,3-triazoles: General procedure
A mixture involving appropriate α-haloketones (1 mmol) or alkyl halide (1 mmol), a suitable alkyne (1 mmol), sodium azide (1.1 mmol) and CuI/SMI (0.03 g) was refluxed for the indicated time. Progress of the reaction was monitored by TLC using hexane and ethyl acetate as eluent. After completion of the reaction, the supported catalyst was filtered off and washed with hot ethanol. The recovered catalyst was then washed with acetone and dried under reduced pressure at 70 °C for 3 h. This recovered catalyst was stored for recycling. The filtrate was evaporated under reduced pressure and the residue was recrystallized from EtOH/H 2 O (3:1v/v) to give the pure crystalline products. All the products were known and their physical data were compared with those of authentic samples and found being identical [28,[50][51][52][53][54].

Results and discussion
In 2005, initially SMA polymer was reacted with di(2-pyridyl)methylamine to obtain the modified SMA [27]. Subsequently, palladium was immobilized on this modified SMA. This system was used as an efficient and recyclable heterogeneous catalyst in different cross coupling reactions including Heck, Suzuki and Sonogashira coupling reactions [55]. The high reactivity of cyclic anhydride is obvious and known [56]. The combination of this high reactivity with high concentration makes SMA a seamless and an ideal polymeric support for the immobilization of metal nanoparticles being used as heterogeneous catalysts in different organic transformation [25,27].
In this light, we recently used 4-aminopyridine as a simpler amine for the successful modification of SMA, prepared and characterized Cu(I) NPs immobilized on modified SMA [34]. In continuation of this work, we thought that it is worthwhile to study the use of a less expensive and more readily accessible isomer, 2-aminopyridine instead of 4-aminopyridine, for the modification of SMA and convert it to SMI to give our previous work a broader overview such as substrate scope.

Preparation and characterization of the modified SMA
In order to modify SMA (1) initially, it was reacted with 2-aminopyridine (2), following the procedure previously reported by Lee et al. for the reaction of SMA with substituted anilines [47]. FTIR spectroscopy confirmed the conversion of the maleic anhydride moiety in SMA to the corresponding maleimide. The absence of N-H peaks of the amino group reconfirms the presence of amido moieties in the polymer side-chain. In addition, the free amine site content of the resulted SMI was found 3.23 mmol/g, as calculated by back titration (for details see the Experimental section) [48,49].

Preparation and characterization of CuI/SMI
As discussed earlier, a general applicable strategy that prevents the presence of metal NPs of catalyst in the reaction products is the immobilization of metal NPs onto a support. In this line, in the previous step, we modified SMA support surface for the purpose of effective immobilization of CuI onto the polymeric surface. So, in this step, CuI NPs were coordinatively attached to the modified SMI polymer surface. After proper immobilization of CuI NPs, structure, coordination properties, spectroscopic data, amount of coordinatively bonded CuI NPs and surface morphology of the prepared CuI/SMI were discussed via SEM, EDAX and ICP analysis.
In order to avoid the agglomeration of the NPs, CuI NPs were prepared in situ and were refluxed with the obtained SMI under N 2 atmosphere in DMF as solvent (the molar ratio of amine to Cu was 2.4:1.0) [57] to obtain the Cu(I) NPs immobilized on SMI. To find the optimal reaction conditions, DMF and ethanol were used as solvent. Although the immobilization also occurred in ethanol, better immobilization was observed for the product obtained in DMF. The size of the immobilized copper particles on the surface of the catalysts determined by SEM images of the two obtained catalysts from two different solvents, were documented. As shown in figure 4, the superior immobilization and smaller size of CuI NPs on SMI is vividly observed for the product obtained in DMF. Due to insolubility of the CuI/SMI catalyst in most conventional organic solvents, we have to limit our structural elucidation to SEM, EDAX, IR and ICP analysis. The scanning electron micrographs of the CuI/SMI obtained from DMF (figure 4) vividly illustrate that the CuI NPs were well homogeneously immobilized on the SMI surface with average size of 70 nm. The other obtained data such as EDAX for the CuI NPs immobilized on SMI is illustrated in figure 5, which clearly approves the attachment of    CuI onto the surface of the polymer matrix. Concise comparative analysis of IR spectrum of SMI with CuI/SMI shows a slight negative shift in stretching vibration frequency of C=N of pyridine ring and C=O of maleimide for free ligand and its corresponding complex which can be attributed to the coordination of copper ion with pyridine nitrogen and oxygen of carbonyl group and so coordination is quantitatively distinguished (figure 2). In the following computational section, this behavior will be discussed and confirmed by the computational analysis.
Moreover, to estimate the Cu content, CuI NPs immobilized on SMI were initially treated with concentrated HCl for digestion of the Cu species for Cu content estimation. The digested prepared extract was immediately analyzed by inductively coupled plasma (ICP) analysis. In this way, 6.92% w/w Cu content was determined.
Comparative study of ICP results between 2-aminopyridine and 4-aminopyridine modified SMI catalyst shows slightly higher value for the copper content of 2-aminopyridine modified SMI which is about 0.3% more than 4-aminopyridine modified polymeric support [34]. This increase in copper content and the lower cost of 2-aminopyridine (about half ) in comparison with 4-aminopyridine make it more economical and affordable and more useful compound for the modification of SMA.
The immobilization technique on SMI as a polymeric support showed several merits. The observed advantages are facile procedure and non-required pre-activation. Bearing in mind that the Cu(I) catalyzed 1,3-dipolar Huisgen cycloaddition reaction via click reaction, we decided to examine the catalytic potency of this in situ CuI NPs immobilized on SMI which was fully characterized in the aforementioned reaction. The reaction of phenylacetylene as an alkyne and sodium azide was chosen as model reaction. The model reaction proceeded smoothly in the presence of 0.03 g of catalyst in water as the greenest and most abundant solvent at reflux temperature. The progress of reaction was monitored by TLC using (n-hexane: ethyl acetate; 3 : 1) as eluent. Upon completion of the reaction (indicated by TLC), the mixture was filtered off and the filtrate was washed with hot ethanol to separate the products. The filtrated catalyst was washed several times by acetone and then was dried and stored for use in subsequent reactions. The residue was then crystallized in ethanol to give the corresponding pure crystalline 1,4-disubstituted 1,2,3-triazole as sole product.
To prove the heterogeneous nature of catalyst, the recovered solid was washed with acetone dried and weighted. No appreciable loss in weight was observed. The recovered catalyst was used for further five runs in the same reaction conditions to give the products. The yield for each run is shown in figure 6. To establish the generality of method, various alkynes reacted with a wide range of organic azides under optimized conditions which provided the corresponding 1,2,3-triazoles in excellent regioselectivity and yields (table 3).

Regioselective synthesis of 1,4-disubstituted 1,2,3-triazoles
To evaluate the potency of the copper-supported catalyst, it was applied in the click reaction. The Cu(I)-catalyzed reaction of benzyl chloride (1 mmol), sodium azide (1.1 mmol), and phenylacetylene (1 mmol) was selected as a model reaction and was performed in different solvents (table 1). As shown in table 1, the best results were obtained for entries 4 and 5 (solvent: EtOH/H 2 O and water). Naturally, clean water was selected as the solvent of choice (table 1, entry 5). To optimize the amount of catalyst, the model reaction was conducted in water using different amounts of the catalyst (table 2). The best result was obtained when the model reaction was performed in the presence of 0.03 g of CuI/SMI in refluxing water (table 2, entry 3). Increasing the catalyst concentration did not show any significant effect regarding the yield and reaction time. To establish the generality of the method, the reaction   of various α-haloketones or alkyl halides (5a-j), different alkynes (6), and sodium azide in water was performed in the presence of the optimized amount of CuI/SMI catalyst. The progress of the reaction reactions were monitored by TLC. After completion of the reactions and cooling the reaction mixtures to room temperature, all products were isolated by simple filtration, and recrystallized for further purification from ethanol. In this way a wide range of 1,4-disubstituted 1,2,3-triazoles were synthesized in excellent yields, with short reaction times and high regioselectivity. The physical and spectral data of all products were compared with those of authentic samples, which were identical (table 3). We have listed 1 H NMR and IR spectroscopic data for some of our synthesized 1,4-disubstituted 1,2,3-triazoles (Supplementary Material).
From the mechanistical viewpoint, in the first step of Cu(I) catalyzed azide-alkyne cycloaddition reaction, a Π-complex is formed between Cu(I) species with the triple bond of terminal alkyne. Then, Cu(I) coordinates to alkyne by deprotonation of the terminal hydrogen of alkyne to give Cu(I)-acetylide intermediate, in the presence of a base. So, in the consequent transition state structure, one Cu atom is coordinated to acetylide species while the other Cu atom activates the azide via the coordination with electrons on the nitrogen atom. Then, a weakly coordinating ligand is displaced on azide to produce a copper-azide-acetylide complex. Finally, cyclization occurs, is followed by protonation and the product is generated by dissociation. So, Cu(I)-ligand complex is reproduced as catalyst for further reaction cycles [10,11].
Comparative survey on the catalytic activity of 2-aminopyridine modified SMA copper catalyst with previously reported modification of SMA polymer with 4-aminopyridine [26] shows lower required reaction times and higher yields in the presence of CuI/SMI catalyst modified with 2-aminopyridine in the regioselective synthesis of 1,4-disubstituted-1,2,3triazoles via click reaction which in turn proves more effectiveness of this new catalyst. Furthermore, obtaining better yields for regioselective synthesis of 1,4-disubstituted-1,2,3triazoles somehow shows more regioselective potential of the new modified polymer catalyst. These results can be mainly attributed to the higher copper NPs content immobilized on the modified SMI support.
It is noteworthy to mention that the catalyst can be reused several times without loss of its catalytic activity. The recyclability of the catalyst was studied by performing the reaction of 2-bromoacetophenone, phenylacetylene, and sodium azide as model reaction ( figure 7). Consistent activity of the catalyst was observed over five consecutive cycles, without any need to reload and activate the catalyst ( figure 6). Moreover, we studied the morphology of CuI NPs for fresh and reused catalyst, after five runs via the comparative analysis of SEM images (presented in figure 8). It is worthwhile to note that we did not observe any considerable morphological changes of CuI NPs and the average diameter of CuI particles remain at the nano size even after five consecutive cycles.
The heterogeneous nature of the catalyst and possible leaching of CuI from the SMI was tested for by using ICP-AES on the reaction mixture. The difference in the Cu content for the fresh and reused catalyst, after the fifth run, was only 6%, approving a low rate of leaching.

Computational section
In the last decades computational chemistry and parallel synthesis are significant, chronically and recurrently applied tools for prime identification and prediction of reactivity, selectivity and optimization of reaction conditions before running the actual experiments, as well as the development of therapeutically important small molecules using ligand docking, ligandbased pharmacophores, molecular descriptors and quantitative structure-activity relationships [59].
In this content, due to our overgrowing theoretical interest in Cu(I) catalyzed Huisgen 1,3-dipolar cycloaddition via click reaction [32, 61,62], we expanded our undertakings to computational methods especially on its outstanding power in predicting regioselectivity which led to publication of a comprehensive review [29].
Going deep into the subject, in this work, we focused on theoretical interpretation of immobilization of CuI NPs on modified SMI, by performing DFT [38] and QTAIM [39] calculations. It should be emphasized that our computational modeling on the coordination behavior of CuI NPs with a confined size model of SMI polymeric system will make accurate simulation feasible, allowing detailed results to be produced about the immobilization of CuI NPs on polymeric SMI, therefore, can be used to modify the properties of existing polymer-supported catalyst and design the analogs of existing catalyst or propose designing similar new catalysts.
At the first step, we designed effective computational models for SMI and CuI/SMI compounds ( figure 9) considering that this model incorporates accuracy and time saving efficiency of computational procedure. As illustrated in figure 9, we have modeled three coordinating modes for description of metal-ligand interactions denoted as N-CuI/SMI (Cu(I) coordinates to pyridine nitrogen), O 1 -CuI/SMI and O 2 -CuI/SMI (Cu(I) coordinates to carbonyl oxygen), respectively. Thus, we determined initially the optimized structures of SMI, N-CuI/ SMI, O 1 -CuI/SMI and O 2 -CuI/SMI complexes via DFT calculations at M06/6-31G * level of theory. It is noteworthy to state that M06 functional has been introduced recently as a top performer within modern functional with a high performance and accuracy and was recommended for application in organometallic thermochemistry, kinetic studies and noncovalent interactions [62].
In the case of Cu(I) and iodine, the effective core potentials (ECP), LANL2DZ, were used together with the accompanying basis set to describe the valence electron density [63]. All DFT computations have been performed using GAMESS suite of programs [64]. In figure 10, the energy minimized structures of SMI and N-CuI/SMI, O 1 -CuI/SMI and O 2 -CuI/SMI complexes calculated at M06/6-31G* level of theory are displayed.
In order to assess the variation of bond orders via complexation process and interpret the interaction strength between copper ion and SMI, the bond order of some selected key bonds in SMI were compared with its corresponding bonds in N-CuI/SMI and O-CuI/SMI complexes. In table 4, the bond-order of selected bonds in SMI ligand and N-CuI/SMI, O 1 -CuI/ SMI and O 2 -CuI/SMI complexes calculated at M06/6-311+G** level of theory are listed. The reported results in table 4 indicate that the bond-order of carbonyl group and also C-N of pyridine ring decreases through complexation in N-CuI/SMI and O 2 -CuI/SMI complexes, and this does not happen in O 1 -CuI/SMI complex. This behavior can be mainly ascribed to the donation of shared electrons from these chemical bonds to copper ions and is in agreement with the reported FT-IR spectroscopic observations. Moreover, it has been demonstrated that Cu-N interactions in N-CuI/SMI complexes are stronger than Cu-O interactions in O 1 -CuI/SMI and O 2 -CuI/SMI complexes. On the basis of the obtained results, we can predict that copper NPs will bind more effectively to pyridine nitrogen than carbonyl oxygen. On the other hand, we investigated comparatively the interaction between copper ions and SMI modified with 2-aminopyridine and 4-aminopyridine (denoted as 2 N-CuI/SMI and 4 N-CuI/SMI complexes, respectively) via the analysis of calculated bond-order values. This comparison demonstrates that the interaction between pyridine nitrogen and copper ions in 4 N-CuI/SMI complex is slightly weaker than 2 N-CuI/SMI complex and so the copper NPs will coordinate more effectively to pyridinic nitrogen of 2 N-SMI in comparison to 4 N-SMI ligand.
Since QTAIM provides a great deal of information about the nature of bonding environment [65], in the next step we explored the topological properties of the electron charge density b and its Laplacian ∇ 2 b at various bond critical points (BCPs) in CuI/SMI complexes. In this respect, resulting 6-311+G** and 6-31G* wavefunction files for the optimized structures of SMI ligand, N-CuI/SMI, O 1 -CuI/SMI and O 2 -CuI/SMI complexes were employed as inputs to AIM2000 program package [66]. In table 5, the calculated values of electron density, b , its Laplacian, ∇ 2 b , electronic kinetic energy density, G b , electronic potential energy   Table 5. mathematical properties of BCPs associated to selected bonds in smi, n-Cui/smi and o-Cui/smi complexes. the properties, obtained via Qtaim analysis on m06/6-311+G** calculated wavefunctions of electron density. note that numbering of atoms is in accordance with Qtaim molecular graphs of smi ligand, n-Cui/smi, o 1 -Cui/smi and o 2 -Cui/smi complexes including all bond and ring critical points and their associated bond paths have been also displayed in figure 11.    and I-H BCPs that lead to some ring critical points and consequently stabilize complex electronically. It is noteworthy to mention that this electronic feature has not been observed in QTAIM topological analysis of wavefunctions in O 1 -CuI/SMI and O 2 -CuI/SMI complexes. In the next step, the bond ellipticity of selected bonds was considered. As reported in table 6, the calculated values of ellipticity of carbonyl BCP in SMI ligand are decreased during the coordination with CuI, indicating that the π-character of the carbonyl bonds is decreased while oxygen interacts with copper in O 1 -CuI/SMI and O 2 -CuI/SMI complexes.
Another QTAIM reliable indicator for quantifying covalent character of interatomic interactions is the electron delocalization index (DI), which is the average number of electrons delocalized (shared) between two bonded atoms. As reported in table 6, DI calculated values on C=O and C-N BCPs have decreased via complexation. Strictly speaking, the coordination between copper and oxygen or nitrogen in N-CuI/SMI, O 1 -CuI/SMI and O 2 -CuI/SMI complexes reduce the bond strength of C=O and C−N bonds that completely confirm the FT-IR spectroscopic observations.
In order to assess the solvent effect in the synthesis of CuI/SMI nanocatalyst, we focused on comparative evaluation of bond-orders of Cu-N (pyridine nitrogen), Cu-O 1 and Cu-O 2 (carbonyl oxygen) in the presence of ethanol and DMF as solvents. It is noteworthy to mention that better immobilization and smaller size of CuI NPs on SMI for the catalyst prepared in DMF solution in comparison with ethanol were practically observed.
Comparison of the calculated bond orders of Cu-N, Cu-O 1 and Cu-O 2 chemical bonds in DMF and ethanol solution phases confirmed that Cu-N bond in N-CuI/SMI complex is stronger than Cu-O bond in O 1 -CuI/SMI and O 2 -CuI/SMI complexes. Moreover, Cu-N bond order in N-CuI/SMI complex has higher calculated values in DMF solution in comparison to those calculated in ethanol and so corroborates with the observed practical preference of DMF as solvent rather than ethanol in the synthesis of catalyst.

Conclusion
We prepared and characterized a polymer-supported CuI nanocatalyst and investigated its catalytic activity. The catalyst exhibited high activity and afforded a diverse range of products in good to excellent yields. Furthermore, the catalyst which is stable in the reaction conditions can be used with no pre-activation and can also be recycled for at least five consecutive cycles without an appreciable loss of activity.
We assessed also computationally the nature of interactions between copper ions with modified SMI ligand in several CuI/SMI complex models. In this respect, the mathematical properties of electron density functions have been calculated and analyzed via DFT and QTAIM in the gas and solution phases. Based on stringent analysis of calculated values for QTAIM electronic indicators, we have demonstrated that interaction of copper cation with pyridine nitrogen in N-CuI/SMI is stronger than with oxygen of carbonyl groups in O 1 -CuI/ SMI and O 2 -CuI/SMI models and subsequently predicted that CuI NPs immobilize mainly on nitrogen site of polymer-supported catalysts. Furthermore, our calculated results are in reliable agreement with IR spectroscopic data and solvent effect observations. Finally, our comparative computational and experimental analysis on CuI/SMI catalyst modified with 2-aminopyridine and 4-aminopyridine reveals the fact that 2-aminopyridine modified polymer support can be considered as an effective species with more catalytic activity and copper content value and making stronger metal-ligand interactions.