Catalytic filtration: efficient C-C cross-coupling using Pd(II)-salen complex-embedded cellulose filter paper as a portable catalyst

A new approach has been developed for environmentally friendly C-C cross-coupling reactions using bi-functional Pd(ii)-salen complex-embedded cellulose filter paper (FP@Si-PdII-Salen-[IM]OH). A Pd(ii)-salen complex bearing imidazolium [OH]−moieties was covalently embedded into a plain filter paper, then used as an efficient portable catalyst for the Heck, Suzuki, and Sonogashira cross-coupling reactions under environmentally friendly conditions via the filtration method. The catalytic filter paper properties were studied by EDX, XPS, TGA, ATR, XRD, and FESEM analyses. The reactions were catalyzed during reactants' filtration over the catalytic filter paper. The modified filter paper was set up over a funnel and the reactants were passed through the catalytic filter paper several times. The effect of reaction parameters including loading of Pd(ii)-salen complex, temperature, solvent, and contact time were carefully studied and also the optimal model of conditions was presented by the design expert software. High to excellent yields were obtained for all C–C coupling types with 5 to 8 filtration times. Under optimal conditions, all coupling reactions showed high selectivity and efficiency. Another advantage of the modified filter paper was its stability and reusability for several times with preservation of catalytic activity and swellability.


Introduction
One of the main concerns of researchers is to develop and replace methods for the preparation of organic compounds, which mostly include the non-production of waste materials and consequently the preservation of the environment. From this point of view, heterogeneous and recoverable catalysts have a special place. 1 The use of lter paper is a smart strategy in organic synthesis for solid phase synthesis which benets from clean work-up. 2 Because of the high ability to modify (functionalize) and manipulate a cellulose paper, it has attracted the attention of many scientists in different elds of sciences to achieve various purposes. 3 In organic chemistry, chemical modications of cellulose can convert a cellulose paper to a catalyst and could be utilized as a suitable substrate.
Previously, Koga et al. reported the application of a methacryloxy-modied porous microstructure-cellulose paper as a support for lipase immobilization. The obtained paper showed signicant catalytic activity to the non-aqueous trans-esterication. 4 Filter paper containing silver nanoparticles was synthesized and utilized by Mourya et al. as a catalyst for methyl orange decomposition, cascade reaction, and reduction of nitroarenes. This silver nanoparticle-embedded lter paper was synthesized by immersing FP in hydrophobic AgNO 3 monodispersed solution. 5 Cellulose paper is also a good substrate for the NP preparation in situ, which was developed for Co 6 or Ni 7,8 NPs.
In another work, cellulose lter paper modied with 3mercapto-propanoic acid was used as an adsorbent to efficiently remove of arsenate from drinking water. 9 Wei and colleagues introduced a Fe-tannin-framework ink coating for the preparation of cellulose-based catalysts by cellulose lter paper. The obtained Fe 3 C/Fe-N-C catalysts were utilized to reduce oxygen under alkaline conditions. 10 In addition, the lter paper loaded with gold nanoparticles shows a high level of Raman efficiency (SERS) that supply real-time monitoring for chemical reactions. 11 Ag-doped cellulose FP as an antibacterial wound dressing, 12 synthesis of cellulose paper with superhydrophobic property for water drop energy harvesting, 3 paper-based electrodes, 11 and self-cleaning superhydrophobic cellulose paper, 13 were some of the recent advances in the eld of cellulosic lter paper manipulation, which reects its high potential for application in various elds of science.
The results well show that the surface modication of a cellulose paper with nanoparticles as well as organic compounds is possible through covalent bonding or electrostatic interactions. Despite some advantages associated with the heterogeneous catalysts such as recoverability and ease of operation (work-up), their catalytic activity is decrease during consecutive recycles due to metal leaching, lack of stability in the mixture, and poisoning of the active sites. 1 Contamination with the reactants also requires tedious and frequent washing and defective recycling from the reaction mixture were another disadvantages of heterogeneous catalytic systems which reduces its activity. The design of a catalytic system based on a portable catalytic lter paper with a ltration protocol eliminates the above-mentioned drawbacks and directs the reactions to green conditions. Although various studies have been performed on modied cellulose lter paper for its catalytic applications, their activity has not been studied from the point of view of catalyzing reactions through the sequential ltration of raw materials as a portable catalyst, and in all cases, the modied lter paper, has been used as an immersion (like a traditional heterogeneous reaction).
The undeniable importance of coupling reactions such as Stille, Heck, Suzuki, Sonogashira, etc. in various elds from the preparation of important pharmaceutical compounds to agriculture and the chemical and petrochemical industries, has led researchers to seek new cost-effective methods, with high efficiency as well as easy work-up to use in the industry and larger scales. Pd-based catalytic systems are one of the most reliable methods for synthesizing these couplings, which have been used in various forms of (1) solid-supported metal ligand complexes, (2) discrete soluble palladium complexes, (3) soluble Pd NPs, (4) supported Pd nano-and macroparticles, (5) palladium-exchanged oxides, and (6) soluble ligand-free Pd so far. 14 On the other hand, research has shown that imidazolium tails bearing hydroxide counter ions in the heterogeneous catalytic systems have a high basicity towards coupling reactions (as a basic reagent), so that their presence in the catalyst structure in the coupling reactions causes the reaction to take place in the absence of any basic reagent. [15][16][17][18] In this way, in the present work, a modied cellulose lter paper was prepared by immobilization of a Pd(II) salen complex bearing imidazolium hydroxide tails on a silica treated cellulose lter paper, as a portable catalytic system (Scheme 1). Scheme 1c shows an original picture of the resulting cellulose lter paper aer surface modication. The application of this system was recently demonstrated by the transfer hydrogenation of nitroarenes using cellulose lter paper-supported Pd/C by ltration methods. 19 The catalytic activity of the modied cellulose paper was studied on Heck, Sonogashira, and Suzuki cross-coupling reactions in depth (Scheme 2), which are among the key reactions used in organic synthesis. 1 Different reactions could be the subject of this study, but due to the importance of such reactions and the need for a stable and available catalyst, these reactions were selected. On the other hand, due to the involvement of several parameters in this protocol, the coupling reactions were selected so that their product is soluble in the reaction medium and does not deposit on the lter paper.
In this method, one of the most challenging problems in the synthesis of organic compounds, namely "effective concentration", was overcome. In common methods for the synthesis of organic compounds, as the reaction progresses, the concentration of the raw materials decreases and the effective collisions of the raw materials with each other as well as with the catalyst surface (in the case of heterogeneous catalysts) are reduced and stop the reaction progresses. But with the design of a catalytic lter paper, the catalytic process is subject to the passage (ltration) of raw materials and forced contact with the catalyst surface. Thus, the reaction takes place only on the part of the material that is being ltered, causing the reaction to proceed to the completion.

Characterization of 2,3,4,5 compounds
Characterization studies were performed in two phases on compound 5 as well as on lter paper 7 step by step and aer conrming formation of the desired product in each step, the synthesis continued and the compound was used for the next step. Fig. S1,S3-S10 † shows the results of FTIR and NMR ( 1 H & 13 C) spectroscopy for 1-5. The presence of characteristic peaks at 674 cm À1 and 2856 cm À1 , respectively, related to the stretching vibrations of C-Cl and C-H bonds (aliphatic, belonging to the methylene group) conrms the structure of 5-chloromethyl salicylaldehyde (1) in full agreement with the previous reports ( Fig. S1a †). [20][21][22] The peak corresponding to the stretching vibration of C]O (aldehyde group) also appeared at 1660 cm À1 . Fig. S1b † shows 5-iodomethyl salicylaldehyde (2) FTIR spectrum, in which the peak appearing at 610 cm À1 (weaker stretching vibration than C-Cl) 23 was attributed to C-I stretching vibration. Stretching vibrations related to C-H (aliphatic) and C]O (aldehyde group) also appeared at 2841 cm À1 and 1663 cm À1 , respectively ( Fig. S1b †), in agreement with the literature. 24,25 The salen ligand (3) was characterized by the shi of the peak belonging to the aldehyde group towards weaker vibrations due to the formation of C]N bond at 1635 cm À1 . 25,26 Also, the peaks related to C-I and C-H (aliphatic) and C-H (aromatic) have appeared at 582 cm À1 , 3078 cm À1 and 2916 cm À1 , respectively (Fig. S1c †). 23,24,26 Substitution of methyl imidazolium groups has caused a series of stretching vibrations related to the C-H of the imidazole ring at 1500-2950 cm À1 in agreement with the literature (Fig. S1d †). 15,16,22 The peaks that appeared at 1625 cm À1 and 1543 cm À1 were also attributed to the stretching vibrations of the amidine (-N-C]N) and C]C groups in the imidazole ring. 15,16,22 The peak at 1650 cm À1 also shows the stretching vibration of C]N bond with strong intensity. Pd coordination to the salen ligand shied the C]N stretching vibration from 1650 cm À1 to the weaker 1620 cm À1 wavenumbers, which was strong evidence for the formation of the Pd II -salen complex (Fig. S1e †). 25,27 In addition, the two peaks at 462 cm À1 and 568 cm À1 , respectively, related to the stretching vibrations of Pd-O and Pd-N bonds, conrm the coordination of Pd to the salen ligand. 25,27 Also, the intensity of stretching vibration related to O-H (phenolic) was signicantly reduced due to this coordination. It is also worth mentioning that based on previously published articles, different synthetic pathways were evaluated to prepare compound 5 as well as lter paper 7, and the synthetic pathway shown in the Scheme 1 was the most optimal and reproducible pathway in order to prepare the catalytic lter paper.

Filter paper characterization
Catalytic lter paper 7 was characterized analytically and spectroscopically step by step. In rst, the papers were studied by ATR-IR analysis (ESI, Fig. S2 †). In comparison with the ATR-IR spectrum of the pristine cellulose paper (Fig. S2a †), ATR-IR spectrum of the silylated paper conrmed the silylation by the presence of the vibrations at 1160 cm À1 (strong) and 805 cm À1 (weak) attributed to the stretching vibrations of asymmetric and symmetric Si-O-C, respectively (Fig. S2b †). 28 Also, a peak appeared at 1518 cm À1 shows the stretching vibration related to Si-C. 29,30 The stretching vibration related to C-Cl bond at 636 cm À1 with strong intensity was another poof for the successful silylation of FP. 29,30 The characteristic peaks correspond to the imine and olen bonds at 1645 cm À1 and (1442-1643) cm À1 , respectively, represent the successful immobilization of complex 5 on the silylated cellulose paper (Fig. S2c †). Also, two peaks appeared at 534 cm À1 and 558 cm À1 show the stretching vibrations related to Pd-O and Pd-N bonds. EDX and ICP analyses conrmed the successful functionalization of the plain cellulose lter paper, wherein % wt Si was found to be 14.68, and 11.36 respectively. 19 This amount provides suitable substrate for the next functionalization on the silylated sites.
The loading contents of Si and Pd were measured on the prepared lter papers 6 and 7, respectively (Scheme 1) by ICP-MS analysis of the resulting ash from the lter paper. For this purpose, the modied lter paper was placed in a crucible (0.9 g) and then calcinated in an oven at 500 C under air atmosphere. At the same time, an unmodied (pristine) lter paper was calcinated as a control, under exactly the same conditions. Based on the results of ICP-MS analysis, the modied lter paper 7 contains 9.45 wt% Pd and 13.2 wt% Si. The elemental composition of lter papers 6 and 7 (catalytic lter paper) was also studied by EDX analysis. Fig. 1a shows the FTIR spectrum of the plain lter paper with the detection of only C, O elements that reects its purity. Two peaks at 2.6 and 2.8 eV binding energies in the EDX spectrum of SiCFP (Fig. 1b) were assigned to Cl Ka and Cl Kb respectively, demonstrating successful modication of the lter paper with CPTES.
The identication of Pd and N elements in the modied lter paper with Pd II -salen complex 5 at binding energies of (2.9 and 3.1 eV) and 0.3 eV, respectively, proves the successful immobilization of the complex on the FP. Complete removal of peaks related to Cl-related binding energies indicates that all silica groups were involved in the immobilization of the Pd II -salen complex and also the immobilization of the complex occurred through covalent bonding according to what is shown in Scheme 1 (Fig. 1c).
The study on the thermal behavior of the modied and unmodied papers clearly conrmed the silylation and functionalization of the lter paper with the Pd II -salen complex. Decomposition of the unmodied lter paper takes place in one step, starting at 280 C and ending with a steep slope at 410 C (Fig. 2a). 31 Silylation of the lter paper causes a signicant increase in the thermal stability of the lter paper. As shown in Fig. 2b, the thermal decomposition was performed with a gentle slope and at the end, the residual weight was about 35%, which by abstraction from the remaining weight of the lter paper in the previous step, the residual weight of 28% could be attributed to the remaining silica groups. This thermal behavior is similar to the thermal behavior of the previously reported crosslinked polymers. 32 The presence of silica groups on cellulose brous causes Si-O-Si bonds formation and consequently reduces the mobility of cellulose brous and provides high thermal stability of the lter paper.
Immobilization of the Pd II -salen complex due to the presence of C]N bonds has given more stability than the silylated lter paper (Fig. 2c). Schiff base compounds have a special place in heat-resistant polymers, 33 therefore, the immobilization of the Pd-salen complex (with coordinated Pd ions) has resulted in greater thermal stability in the catalytic lter paper. The presence of three peaks in the thermal decomposition of FP@Si-Pd II -Salen-[IM]OH was evidence of the presence of multiple immobilized phases on the surface of lter paper, which was in agreement with previous analyzes. As shown in Fig. 2c, two peaks appearing at 275 C (corresponding to 13% weight loss) and 366 C (corresponding to 3% weight loss) could be attributed to the decomposition of the imidazolium moieties (according to their weight loss percentages) and the Pd II -salen ligand decomposition, respectively. The nal peak at 420 C was also related to the residual decomposition of silicafunctionalized cellulose brous. Fig. 3a shows the XPS overall survey analysis (full range) of the unmodied lter paper with the presence of C and O elements, in agreement with the corresponding EDX analysis. Also, C 1s (Fig. 3b) and O 1s (Fig. 3c) deconvolution spectra showed the atomic center of the related functional groups for the cellulose lter paper. 34,35 The C 1s spectrum for the unmodied cellulose corresponds to three types of carbon atoms that can be attributed to (C-C, C-H), C-O and C-O-C bonds at 284.8 eV, 286.3 eV and 287 eV binding energies, respectively (Fig. 3b). 34 In full agreement with the previous reports, 34,36 oxygen-containing bonds such as C-O and C-O-C have higher intensities than C-C, due to the high content of these bonds in the cellulose structure. The O 1s spectrum was consistent with the presence of two types of oxygen atom as C-O-H and C-O-C bonds at 531 eV and 534 eV, respectively, in agreement with the structure of the brous cellulose (Fig. 3c). 37 The presence of the expected elements in FP@Si-Pd II -Salen-[IM]OH lter paper was conrmed by the overall survey XPS analysis (full range) of the lter paper including C, N, O, Si, and Pd elements (Fig. 3d). Also, the deconvulated high resolution C 1s, N 1s, O 1s, Si 2p, and Pd 3d spectra gave useful information about functional groups and elemental composition for FP@Si- Pd II -Salen-[IM]OH lter paper (Fig. 3e-i). The C 1s spectrum for FP@Si-Pd II -Salen-[IM]OH corresponds to six different types of bonds for carbon, which in comparison with the C 1s spectrum for the unmodied lter paper (Fig. 3b), indicates the successful silylation and immobilization of the Pd-salen complex on the lter paper. 38 As shown in Fig. 3e, a peak at 284 eV, related to Si-C bond in C 1s spectrum of the lter paper, conrms the successful immobilization of CPTES groups on the lter paper. 39 Also, two peaks at 283.6 eV and 285 eV for C-N and C]C bonds, respectively, conrmed the covalent immobilization of Pd-salen complex on the lter paper. 39,40 The 284.6 eV, 285.5 eV, and 286.8 eV binding energies can also be attributed to the C-C/C-H, C-O, and C-O-C bonds, respectively. 39,40 The two characteristic peaks at 398.7 eV and 401.0 eV were attributed to Nsp 2 -C and Nsp 3 -C bonds in the N 1s deconvulated spectrum, respectively (Fig. 3f). 38 The Nsp 2 -C peak indicates the presence of C-N]C imidazole groups in the lter paper framework. 39 In addition, the Nsp 2 -C bond was also consistent with the ionic moiety of the imidazole rings. 41 As shown in O 1s high resolution region (deconvulated), ve different oxygen-related functional groups were detected in the spectrum (Fig. 3g). Two characteristic peaks at 530.0 eV and 532.2 eV shows the binding energies related to Si-O and O-Si-O bonds in agreement with the previously reported XPS O 1s results for the silylated cellulose bers (Fig. 3g). 30,42 Also, the coordinated Pd to O atom (C-O-Pd in the Pd-salen complex) was appeared with a low intensity at 533.4 eV (Fig. 3g). 40 Two other high intensity peaks were assigned to C-O and C-O-C bonds at 528.5 and 534.7 eV, respectively (Fig. 3g). 35 200) and (004) planes respectively, were in complete agreement with the monoclinic cellulose type 1 crystal structure (PDF les: 000561717, 000561718 and 000561719), 45,46 that conrmed the crystalline structure of the cellulose FP in agreement with the previous reports (Fig. 4a). 6,19 XRD pattern of the FP 7 conrmed the existence of two different phases, including silicate groups and the Pd complex. According to the lter paper X-ray diffraction pattern of 7, the amorphous peak appearing at 2q ¼ 4.3 corresponds to the amorphous structure of silicate groups on the cellulose brous. As shown in Fig. 4b, the immobilization of Pd complex on the cellulose causes the XRD pattern of the cellulose to be deviated from crystalline state in full accordance with the X-ray diffraction patterns reported from immobilized Pd complexes (Fig. 4b). 47 The surface area and porosity of the modied and plain FPs were evaluated by BET method. The specic surface area and the average pore size of the cellulose FP, FP@Si-Cl, and FP@Si-Pd II -Salen-[IM]OH was in agreement with other analyses and modication of the FP in each step. Based on the results, the plain FP has a pore size and a specic surface area of 8.66 microns and 1.52 m 2 g À1 respectively, which reaches to 6.46 microns and 1.66 m 2 g À1 aer silylation. 19 The specic surface area increased to 3.55 m 2 g À1 aer surface modication of the silylated FP with Pd-salen complex. Also, the average pore size was decreased to 1.60 microns upon this modication. Fig. 5 shows the SEM images taken from the plain and modied papers. FESEM images from the plain and silylated lter paper showed an increase in average diameter of 7 mm (from about 16 mm for the plain lter paper to about 23 mm for  the silylated lter paper) 19 with a decrease in porosity aer the silylation of the lter paper, which conrmed the success of the functionalization ( Fig. 5a and b). Different plate-shaped morphology for the functionalized lter papers can be attributed to the crosslinking between the silica groups. In addition, the brighter spots in the SEM image of FP@Si-Pd II -Salen- [IM] OH was related to the immobilized complex 5 on the silylated lter paper (Fig. 5c).

Screening of the reaction parameters
Four effective parameters including Pd loading amount, reaction temperature, paper surface area and N 2 inlet gas pressure were analyzed by Design Expert soware and the designs were studied on the reaction of Sonogashira model (coupling of phenylacetylene with iodobenzene). Tables S6 † to S8 show the results of some of these tests. Based on these analyzes, Sonogashira efficiency depends on all 4 parameters (ESI, Fig. S11 -S13 and Tables S1-S8 †). First, the effect of contact surface duration (ltration time) on the coupling reaction efficiency was studied. Increasing the contact time of the reaction mixture with the lter paper surface placed on the funnel was done by applying low N 2 pressure into the vacuum Erlenmeyer ask placed under the funnel. Table S1 † shows the results of this study in which, by applying 0.3 bar pressure during 5 consecutive ltrations for 70 min, the efficiency reaches 96%. As shown in Table S1, † as the pressure inside the Erlenmeyer decreases, the ltration time also decreases exponentially, resulting in reduced efficiency. In this case, the number of ltration times should be increased, which is not desirable (Table S1, † entries 1 and 2). On the other hand, increasing the pressure inside the Erlenmeyer ask to 0.7 bar did not have a signicant effect on reaction efficiency (Table S1, † entries 4 and 5).
It is also important to note that the contact time could be increased by adding another raw lter paper. But more repetitive results were obtained using N 2 gas ow. On the other hand, lter paper contamination reduces efficiency. Control experiments (will be shown in the next section) showed that the presence of oxygen also plays an inhibitory role in contact with the catalytic lter paper and reduces efficiency.
The loading contents of Pd-salen complex 5 on 95 cm 2 surface area (on one lter paper with a diameter of 5.5 cm), water bath temperature, and solvent were studied as three effective and vital parameters in the Sonogashira model coupling reaction. It should be noted that due to the basic nature of the lter paper (which will also be shown in control experiments), no basic reagent was used in all reactions. The reaction of iodobenzene with phenylacetylene (Sonogashira coupling) was selected as the model reaction. In the rst step, loading contents of Pd-salen complex 5 on the lter paper with an area of 95 cm 2 was studied based on the Pd content. For this purpose, different concentrations of ethanolic Pd-salen complex solution including 0.3, 0.7, 1.2, 2.5 mmol Pd were prepared and immobilized on the FP based on the procedure described in the experimental section. Table S2 † shows the effect of different Pd loading amounts on the FP. The obtained results indicated that at a loading amount of 1.0 mmol of Pd, the highest possible efficiency for sonogashira coupling product, 1,2-diphenylethyne, was equal to 96% obtained during 5 consecutive ltrations (70 minutes) (Table S2, † entry 3). Increasing the loading amount to 2 mmol had no effect on the Sonogashira coupling product, but at 2 mmol of Pd loading, the reaction time increased signicantly and the efficiency decreased slightly. On the other hand, at loading amounts of 0.2 and 0.5 mmol of Pd, the efficiencies decreased to 72% and 86%, respectively. Considerably, a signicant increase in reaction time was consistent with an increase in Pd content in the lter paper. As the Pd loading on the complex increases, the porosity of the lter paper decreases and the reactants become more difficult to ltrate. Although this reduction in porosity increases the contact time between the lter paper surface and the reactants, this increase in contact time does not lead to a signicant increase in efficiency. On the other hand, polar-polar interaction between polar groups of silica (in addition to polar OH groups in cellulose) in the lter paper and the polar solvent EtOH : H 2 O, increases the reaction time, as observed by different solvents.
It seems that the reaction time (including successive ltrations) and consequently the reaction efficiency is affected by the two factors of viscosity and polarity of solvents. In order to use total surface of the lter paper (covering the whole of surface of the FP with the mixture) the amount of solvent was 5 mL for all tests. Table S3 † shows the effect of a wide range of solvents with different polarities and viscosities on the preparation of 1,2diphenylethyne. Water and ethanol gave 40% and 85% efficiencies for 60 and 50 minutes, respectively (a total of 5 consecutive ltrations) (Table S3, † entries 4, 5). But in a 1 : 2 mixture of H 2 O : EtOH, the efficiency reached 96% for 70 min. It seems that the addition of water to the reaction mixture causes better diffusion and interaction of the reactants in the lter paper layers and consequently their stronger interaction with the catalytic active sites. The reaction time of high viscosity solvents such as butyl acetate (2.98 cp) and DMSO (2 cp) than the H 2 O : EtOH mixture, increased to 94 and 70 min, respectively, but still provided lower efficiency than H 2 O : EtOH mixture. No observable products were found in the non-polar solvent of hexane (Table S3, † entry 9). Solvents such as CH 3 CN and DMF also did not give high efficiency (Table S3, †  entries 3,2).
To investigate the effect of temperature on the Sonogashira model coupling reaction, the whole set up was performed in a glove-box (air atmosphere). The reaction was placed on a water bath with adjustable temperature. According to the optimal performance of EtOH : H 2 O, the reaction was studied at three different temperatures of 27, 50 and 70 C. The results showed that the temperature was the least effective parameter (relative to the Pd content in the lter paper and solvent) in the Sonogashira coupling reaction. As shown in Table S4, † increasing the reaction temperature to 70 C only reduced the reaction time by 7 minutes. Therefore, the coupling reactions were performed at an optimum temperature of 50 C (on a water bath) and in EtOH : H 2 O solvent on FP@Si-Pd II -Salen-[IM]OH with a thickness and diameter of 0.2 mm and 5.5 cm, respectively, containing 1.0 mmol of Pd. Scheme 1 shows the steps of preparing the catalytic lter paper schematically. It should be noted that to prepare FP@Si-Pd II -Salen-[IM]OH, different synthetic pathways were examined and tested, and nally the synthetic pathway shown in Scheme 1 was used as a reliable and reproducible route. As shown in Scheme 1, in order to avoid the concerns related to metal leaching during the various stages of catalyst preparation, Pd coordination to the salen ligand takes place in the nal step.
The effect of surface area was studied by cutting lter paper with a diameter of 5.5 cm into square pieces with a length of 10 mm. For this purpose, the set up reaction was changed and the reaction was performed in the presence of cutted pieces of the lter paper and stirred by a magnet.
According to the results presented in Table S5, † the surface area (which is another reection of the loading amount of Pd complex) has a signicant effect on the reaction efficiency in full agreement with the loading effect of Pd (Table S5 †); because it reduces the surface area corresponds to a decrease in the amount of catalytically active sites. However, in this set up, a signicant difference was observed compared to the ltration set up.
As shown in Table S5, † the highest efficiency occurs in this set up at 95 cm 2 of the lter paper, which was equal to 80% for 70 minutes (50 C). One of the main reasons could be the sticking of the pieces of the lter paper to each other as well as the wall of the reaction balloon, and consequently the reduction of improper interaction between the reactants and the surface of the lter paper; In other words, homogeneous interaction is not achieved. The phenomenon of mass transfer in the ltration method also acts as a driving force that catalyzes the reaction while passing through the lter paper (ltration). Effective concentration was another important factor that probably affects the high efficiency observed for the ltration method. With the diffusion of the reactants to the lter paper, the effective concentration in this volume increases and leads to the achievement of maximum efficiency. This phenomenon has also been observed for cross-linked catalysts. [48][49][50][51] In other words, in the ltration method, only a part of the reaction mixture in contact with the catalyst surface intends to pass through the catalyst, and this issue increases the effective concentration at the catalyst surface and increases the efficiency. This experiment also showed that the ltration method, as a new and portable method, has an advantage over the heterogeneous method (cutted paper).
Functionalization of the lter paper with Pd II -salen complex reduces the polarity of the lter paper surface and consequently increases the interaction of the raw materials with the catalytically active surfaces in the modied lter paper. This, increases the contact time of the reactants with the lter paper during ltration and thus increases the efficiency.
The swelling amount of the lter papers was studied as an important parameter, because the high rate of swelling increases the contact time and interaction of the reactants with the lter paper (which contains catalytically active centers). The swelling rate for the prepared lter papers was given in Table 1.
As shown in Table 1, the swelling amount in the silylated FP was signicantly higher than in the plain lter paper, especially in protic solvents. The results were completely in agreement with the thermal behavior of the silylated lter paper with rigid and cross-linked structure that provides high swelling amount. The results of TGA also showed that the surface silylation, signicantly increases the thermal resistance of the lter paper; But with the immobilization of the Pd II -salen complex, the swelling rate for all solvents decreased. As shown in Table 1, FP@Si-Pd II -Salen-[IM]OH showed the highest swelling amount in aqueous solvent: ethanol equal to 14.9 mL g À1 .

Catalytic activity
The catalytic activity of FP@Si-Pd II -Salen-[IM]OH lter paper in the Suzuki, Sonogashira and Heck coupling reactions was studied. Tables 2-4 show the results of these couplings in the presence of a wide range of aryl halides. Aryl iodides produced good to excellent conversion for all three coupling reactions. For aryl iodides, the number of ltrations at 0.3 bar gas pressure was between 3-5 times, which increased to 8 times, due to the low efficiency for aryl chlorides and aryl bromides. The results showed that the coupling reactions was not very desirable for aryl chlorides and requires more time to react. However, in order to increase the contact time of the reaction mixture with the catalytic lter paper, the gas pressure inside the Erlenmeyer for aryl chloride derivatives (for all three coupling reactions) was increased to 0.5 bar. Under these conditions, the coupling reactions were also performed for aryl chlorides. It should be noted that at the inlet gas pressure of 0.3 bar, due to the high number of ltration times for aryl chlorides and also the lack of optimal efficiency, the inlet gas pressure increased to 0.5 bar, which caused the reaction to progress. However, as shown in Tables 2-4, there was no efficiency for aryl chlorides bearing electron donor groups (such as Me and O-Me,  Table 4, entries 2,3,9,10,16,17) even with inlet gas pressure of 0.5 bar and duration of 180 minutes (for 7 consecutive ltrations). For Suzuki coupling, % conversion was observed only with aryl chlorides bearing electron withdrawing groups (Table 4, entries 4,5,7,11,12,14,18,19).  The presence of cellulose brous as well as imidazolium moiety-OH counter ion terminals in the lter paper causes proper diffusion of the reactants in this solvent between the cellulose brous containing the catalytic active sites and consequently provides a suitable environment for the coupling reaction (Refers to the proposed mechanism). PdCl 2 salt also did not produce any detectable efficiency for the Sonogashira reaction (Table 5, entry 4). As expected, the lter paper alone did not produce any efficiency (Table 5, entry 5). Then, in order to prove the role of imidazolium-[OH] groups in the coupling reactions, a imidazolium-free lter paper was prepared in a similar way. The reaction under the exactly same conditions produced only 20% conversion. Despite the absence of any basic reagent (compared to Pd II -Salen complex 5, which did not produce any detectable efficiency), this amount of efficiency could be attributed to the appropriate environment created by the modied cellulose brous in the lter paper, which has the ability to give 20% conversion ( Table 5, entry 6). Preparation of a lter paper by immobilizing ligand 4 on the silylated lter paper (without the presence of any coordinated Pd) also did not produce any efficiency (Table 5, entries 7). The results of control experiments well showed that the components including Pd II -salen complex, the presence of imidazolium moieties along with the OH counter ions, immobilized on the   cellulose brous were highly correlated with each other, so that the removal of each of them led to the loss of catalytic activity. In the next step, the application of the catalytic FP was studied as small cut pieces or suspended in the reactions mixture for the preparation of 10a, 12a, and 14a. Table 6 shows the results of this study. As shown in Table 6, at similar time intervals, these two methods gave less efficiency than the a water bath), N 2 inlet (0.3 bar; 0.5 bar for aryl chlorides), catalytic lter paper (7, placed on a glass funnel, containing 1.0 mmol Pd/95 cm 2 ). b Total time spent in different cycles. Cycles refers to number of re-ltration of the residue (Scheme 2). c GC yield.   ltration method for all three model reactions. Also, set-up 1 (using cut pieces of the lter paper in the reaction mixture) created higher efficiency than the suspended mode. The results showed well that the highest efficiency of the catalytic lter paper for the coupling reactions occurs in the ltration mode of the reactants and has a unique performance.

Mechanism
By switching the N 2 inlet gas to O 2 , the reaction efficiency of the Sonogashira model was drastically reduced. Due to the solubility of molecular oxygen in water, the catalytic lter paper dissolves the input molecular oxygen aer absorbing water (along with ethanol as the reaction solvent). Molecular oxygen also binds to Pd active sites and prevents its activity. The control Table 7 Reusability evaluation of the catalytic filter paper 7 using a sequential procedure consisting of three cross-coupling reactions of Sonogashira, Heck, and Suzuki towards the preparation of 10a, 12a, and 14a respectively using one filter paper a  [15][16][17]44 The observations obtained from the control experiments in this study (Table 5), were also in complete agreement with the previous reports. [15][16][17]44 As shown in Table 5 (entries 1,2), Pd IIsalen complex 5 and Pd II -salen-[IM]I, did not give any efficiency for the Sonogashira model reaction; While 60% efficiency was found in the presence of Pd II -salen-[IM]OH for 10a. Scheme 3 shows the proposed mechanism in accordance with the literature and observations in this work. Reagents are rst diffused into the lter paper by EtOH : H 2 O solvent. Due to the fact that very little concentration of the reactants enters the lter paper, the effective concentration inside the catalyst increases and causes a signicant increase in efficiency. This phenomenon has already been observed for the catalysts based on PVA, 48 sulfated zirconium oxide, 49,50 and saponin 51 with polar functional groups. Gravity causes this equilibrium to shi towards the lter paper. This phenomenon causes the reaction to proceed to completion and the effect of concentration does not affect the efficiency of the reaction too much.
Aer diffusion, the aryl halide (here iodobenzene) is attached to Pd via an oxidative-addition reaction (step I), that was in agreement with the previous reported mechanisms for the Pd-salen catalyzed coupling reactions. [52][53][54] Phenylacetylene is coordinated to the Pd sites via a p bond, and by removing the proton by the hydroxyl group (counter ion of the imidazolium moiety), it binds to Pd and takes the substituent of iodide (step II). Iodide ion also acts as a counter ion of imidazolium moiety, which are then converted to HI and OH by water, and the catalyst returns to its original state. The control experiments also showed that FP@Pd II -Salen (as a homologue of the FP 7 without the imidazolium hydroxide moiety), does not produce signicant efficiency for 10a, which conrms the function of this group as a basic agent (Table 5, entry 5). This step was conrmed by the recovery studies on the catalytic lter paper. As will be shown in the lter paper recovery studies in the next section, the wt% elemental composition (by EDX analysis) in the recycled FP@Si-Pd II -Salen-[IM]OH (aer four consecutive recoveries) was quite similar to the freshly prepared FP@Si-Pd II -Salen-[IM]OH, and no trace of iodide was observed. In addition, the decrease in the catalytic activity of FP@Si-Pd II -Salen-[IM]OH during several consecutive recoveries was negligible.
In the third step, by reductive-elimination, 52-54 the coupling product passes through the lter paper due to its solubility in ethanol (Step III). On the other hand, due to the higher efficiency of aryl halides bearing electron withdrawing group compared to the halides with electron donating substituent (Tables 2-4), it was suggested that the reaction proceed from the path including oxidative-addition and reductive-elimination steps in agreement with the literature. 18,47,55 As shown in Tables 2-4, the efficiency as well as the ltration time and cycles for aryl halides were as order of I > Br > Cl, and even for some aryl chlorides no efficiency was observed ( Table 2,   entry 14; Table 3, entries 13,14; and Table 4, entries [15][16][17]. These results were a strong evidence for the oxidative-addition and reductive-elimination steps. 18,47,[52][53][54] Recyclability and stability of the lter paper A signicant advantage of the modied lter paper bearing Pd IIsalen complex was its stability and reusability in the C-C coupling reactions. The catalytic activity of FP@Si-Pd II -Salen-[IM]OH in different cycles in the models coupling reactions of (1) coupling of iodobenzene with phenylacetylene, (2) coupling of iodobenzene with phenylboronic acid, and (3) coupling of iodobenzene with styrene, was studied. Catalyst recoverability was studied using one lter paper for all studies. Table 7 shows how these studies were performed. Using one lter paper, 10a, 12a, and 14a coupling products were prepared for four consecutive cycles. Aer each reaction, the lter paper was washed with water and ethanol and used immediately for the next reaction. As shown in Table 7, no signicant drop in catalytic activity was observed for the reactions up to the fourth cycle, and the lter paper appears to be reliable for several more cycles. The signicant advantage of FP@Si-Pd II -Salen-[IM]OH over heterogeneous catalysts was that there was no mass loss of catalyst, which means that no signicant drop in the FP performance was observed during successive cycles.
In addition, as will be discussed below, the lter paper owes this high stability due to the presence of silica groups as well as Pd II -salen complex on the FP surface, which minimizes the shrinkage, and consequently preserves swelling due to the high stability of the lter paper.
Another advantage was the lack of leaching of any metal (especially Pd) as a result of residual solution analysis. In each reaction (aer completion of the reaction), the remaining solution was investigated by ICP-MS to determine the leaching of Si and Pd. The analyzes did not show any trace of Si and Pd metals until the end of the fourth cycle, which reects the absence of any metal leaching in the solution. In fact, the catalyst was designed in such a way that it does not create any soluble part due to the contact and successive passage (ltration) of solvents, and this issue, along with the stable Scheme 4 Metal leaching determination of FP 7 using the hot filtration assay.
coordination of Pd to the ligand framework, has caused no metal leaching from the FP. Therefore, the slight decrease in efficiency can be related to the decrease in FP quality in successive wetting-drying cycles. As will be shown below, during successive wetting-drying cycles, the plain paper undergoes some shrinkage and reduced swellability.
In addition, the hot ltration test was also studied on the catalytic lter paper. For this purpose, in the model Sonogashira reaction, aer the third ltration (80% conversion, 30 minutes), the ltration was stopped and the lter paper was removed (Scheme 4). Then, a magnet was added to the collected mixture and stirred at 50 C. GC analysis of the reaction mixture aer 1 hour also recorded an efficiency of 80%. The results showed well that despite successive ltrations, no Pd leaching occurs due to the coordination to a stable salen ligand structure. The results of EDX and leaching analyses were also completely consistent with these results and conrm the stability, no contamination, and reproducibility of the modied lter paper.
In addition, the FP was characterized by FESEM and EDX analyses aer the 4th cycle in the reaction of Sonogashira model (coupling of phenylacetylene with iodobenzene). This analysis was performed to ensure the adsorption of any impurities (including raw materials and products) on the lter paper aer repeated use. As shown in Fig. 6a, the % wt of C, O, N, Pd elements has not changed signicantly, which reects the ltration and subsequent complete washing of the lter paper aer ltration. In addition, this analysis showed that the absence of any impurities in the lter paper aer the reaction allows its repeated use with condence. In addition, the FESEM image of the recovered lter paper (Fig. 6b) showed that its morphology did not change compared to the freshly prepared lter paper. The results demonstrating the high stability of the immobilized palladium complex on the FP through strong covalent bonding.
To investigate the stability of the immobilized groups on the lter paper, various reagents were studied for this purpose, passed through the lter paper and the ltrate as well as the lter paper were analyzed. Table 8 shows the effect of acidic, alkaline and oxidative reagents on the elemental composition of the lter paper. Also, to investigate the possible leaching of Si and Pd metals on the lter paper, the ltrate was examined using ICP-MS.
The lter paper stability studies provided useful information on behavior of the lter paper in various media. Table 8 shows the results of this study. Nitric acid causes signicant leaching of Pd into the solution. In addition, due to the reduction in the percentage of nitrogen in the resulting lter paper, it appears that the Schiff base ligand was also hydrolyzed in an acidic medium and enters the solution phase (Table 8, entries 2 and 7). The lter paper has complete stability in basic environment (NaOH 0.1 N) which shows its applicability in high pH solutions. No traces of Si and Pd in the remaining solution were detected by ICP-MS (Table 8, entries 4 and 9). In the presence of HCl 0.1 N, the lter paper has a relative stability, so that the leaching rate was negligible for both Pd and Si metals (Table 8, entries 3 and 8). Lack of effect of oxidants such as NaOCl and H 2 O 2 on the composition of the lter paper reects the high stability of the lter paper structure against oxidants and therefore makes it possible for use in oxidation and epoxidation reactions (Table 8, entries 1, 5, 6, and 10).
In order to investigate the possibility of using the FP at different pHs, the amount of Pd leaching from the lter paper at pHs 1-14 (using HCl and NaOH solutions) was studied. Fig. S14 † shows the amount of Pd leaching at each pH aer 5  consecutive ltrations of each solution by ICP analysis. As shown in Fig. S14, † the highest Pd leaching rate occurs at pH 1 equal to 4.89%. At this pH, almost all the Pd loaded on the lter paper was leached. The lter paper was relatively stable at pHs 5-11 and no metal leaching was detected and the catalytic lter paper can be used with condence in this pH range. Schiff base bond hydrolysis at very acidic and highly alkaline pHs was the main cause of the high leaching rate observed at these pHs. However, this stability was much higher in the acidic environment than in the basic, so that at pH 14 only 4.2 % Wt Pd leaching was detected.
In order to study the effect of successive drying-wetting of the lter paper on the shrinkage and its swelling rate, the catalytic lter paper was washed and dried for 5 consecutive times by EtOH : H 2 O (2 : 1) solution and its shrinkage and swelling rate in each cycle was measured. The results show that the physical properties of the cellulosic bers were affected by successive drying-wettings, 56 and these properties can be changed by proper functionalization of the bers. 57,58 Fig. 7 shows the results of this study. According to the results, the amount of swelling in the catalytic lter paper did not change signicantly during successive drying-wetting cycles, which reects the high stability of the lter paper arising from the proper modication with organic groups, especially silica supported groups.
The results were completely in agreement with the thermal behavior of the silylated lter paper (Fig. 2b). Also, the amount of shrinkage remained constant until the end of the 5th dryingwetting cycle and there was no signicant reduction in the paper diameter (Fig. 7).
The amount of shrinkage observed in each run was negligible and at the end of the 5th run, only 1 mm decrease in the size of the lter paper diameter was observed. This stability in the lter paper can be directly attributed to the presence of silica groups as well as the Pd-salen complex immobilized on the lter paper. The study of shrinkage and swelling amounts on the plain and the silylated lter paper conrmed the effect of these two factors on maintaining the properties of lter paper. 19 According to the results, the plain lter paper suffers from 7 mm shrinkage, and the swelling rate was reduced to 8.6 mL g; While these values were equal to 2 mm and 12 mL g À1 respectively, for the silylated lter paper. The results showed a correlation between these two parameters that with the suitable surface functionalization, both improve. 19 Comparison of activity of FP 7 with literature reports Table S9 † shows the performance and catalytic properties of FP@Si-Pd II -Salen-[IM]OH lter paper compared to the previously reported coordinated Pd-based heterogeneous catalysts for the preparation of 10a (coupling reaction of phenylacetylene with iodobenzene). [59][60][61][62][63][64] As shown in Table S9, † FP@Si-Pd II -Salen-[IM]OH was superior to most previously reported heterogeneous catalysts in terms of ease of work-up, environmentally friendly, and the product efficiency. FP@Si-Pd II -Salen-[IM]OH was prepared on a cheap and plain cellulose lter paper using the available raw materials and was portable. Unlike nanomaterials-based heterogeneous catalysts, FP@Si-Pd II -Salen-[IM]OH lter paper can be recovered without metal leaching for several consecutive cycles. Heterogeneous catalysts based on nanomaterials and transition metal complexes oen have metal leaching that allows toxic metals such as Pd to enter the environment. In addition, the recovery of FP@Si-Pd II -Salen-[IM]OH was done with just a simple washing and, therefore, no mass was lost from the catalyst, while the recovery of nanomaterials-based heterogeneous catalysts is oen done by centrifugation which the loss of catalyst during successive recovery cycles due to their nano-dimensions is inevitable and causes a decrease in catalytic properties. As shown in Table S9, † the recovery of heterogeneous catalysts is usually done using methods such as a magnet, ltration, extraction, and centrifuge, which either cause the catalyst to be lost or contaminated. [59][60][61][62][63][64] In addition, contamination of heterogeneous catalysts with raw materials and products in each cycle requires tedious and continuous washing. [59][60][61][62][63][64] These disadvantages also limit the use of nanomaterial-based heterogeneous catalysts for use at higher scales. The FP@Si-Pd II -Salen-[IM]OH potable catalyst can be used anywhere, even by non-experts. FP@Si-Pd II -Salen-[IM]OH has a multiple function, which owes it to the immobilization of the Pd-salen complex, so that the Sonogashira reaction takes place in the absence of Cu salt (as a cocatalyst), a toxic and expensive phosphine ligand, and a base agent. These factors make the lter paper more environmentally friendly. FP@Si-Pd II -Salen-[IM]OH as a portable catalyst, eliminates most of the disadvantages associated with heterogeneous catalysts, not only providing better efficiency than heterogeneous catalysts, but also easy recovery and no metal leaching, sustaining the environment.

Conclusion
A Pd II -salen complex-functionalized cellulose FP was used as an efficient portable catalytic lter paper for the construction of C-C bonds via the Heck, Suzuki and Sonogashira reactions based on the ltration and controlled passage of the reactants. The modied FP was characterized by ATR, EDX, XPS, TGA, XRD, and FESEM analyses. High to excellent yields were obtained for the coupling reactions with an average of 3-8 ltration cycles for 54-180 min. The Pd II -salen complex was covalently bonded to a pre-silylated lter paper, which were responsible for high stability and revocability by maintaining the catalytic activity with no leaching and shrinkage during successive wetting-drying cycles. The lter paper was stable in a wide range of acidic pHs (5, 6) and alkalis (8)(9)(10)(11) and thus can be used in a wide range of organic reactions. Sustainability, portable, easy work-up, clean prole, reusability, high stability, high versatility, high catalytic activity and selectivity (as much as known heterogeneous catalytic systems), etc. were some of the present novel protocol, which make it as a promising candidate and alternative than traditional heterogeneous catalytic systems. As best of our knowledge, this is the rst report of coupling reactions catalyzed by ltration, and further studies based on various modications on the lter paper could create a new world for portable catalysts.

Materials and methods
A circle Whatman 1201-320, grade 1V cellulose lter paper (folded), with a 32 cm in diameter, 150 s/100 mL speed (Herzberg) 0.2 mm thickness, 8-10 micron, with basic weight of 120 g m À2 was provided from Sigma. To control of ltration rate (the duration of contact time), the reaction set up was equipped by a N 2 -inlet (low pressure: 0-1 bar) V-121K091 R-21 regulator (inlet connection W24,32). Orbital Shaker, 10 mm Orbit from BTLabSystems was used for shaking. ATR (in the case of lter papers) and FTIR analyses were taken on a JASCO FT/IR 4600 instrument. The NMR ( 1 H and 13 C) spectra were recorded on a Bruker AVANCE III 300 MHz spectrometer in deuterated solvents (CDCl 3 and DMSO-d 6 ). NexION 2000B ICP Mass Spectrometer was used for ICP analyses. X-ray diffraction (XRD) of the FPs was studied on a Rigaku Smart-Lab X-ray diffractometer with Cu Ka (l ¼ 1.5418 nm) radiation. Field emission scanning electron microscopy (FESEM) of the FPs were taken using a Tescan MIRA3 microscope. Energy-dispersive X-ray spectroscopy (EDX) was conducted on a JEOL 7600F eld emission scanning electron microscope, equipped with a spectrometer of energy dispersion of X-ray from Oxford Instruments. Statistical studies on the coupling parameters and also subsequent regression analyses of the experimental data were studied by a Design-Expert statistical soware version 11, Stat-Ease Inc. Minneapolis, MN, USA.
The change of the halide substituent from Cl to I was performed based on a part of the Hoffman elimination and the formation of the hydroxide counter ions on the ionic moiety. 17,44 Experiments have shown that the formation of chloride groups (imidazolium chloride) in the presence of Ag 2 O does not produce any efficiency for the imidazolium chloride and its subsequent conversion to the hydroxide counter ions, but in the presence of iodine counter ions, the reaction proceeds well, in complete agreement with the previous published articles. 17,44 Characterization data for SA-5-IM (2) Note: Different synthetic routes were evaluated for the synthesis of 5, and the routes shown in Scheme 1 was chosen as the most promising and repeatable with the highest efficiency for the preparation of catalytic lter paper 7.
Synthesis of [N,N 0 -(4-OH-phenylenebis(SA-5-ImI))-Pd (II) complex, 5]. Complexation of Pd ion to N,N 0 -(4-OH-phenylenebis(SA-5-ImI)) ionic ligand was performed simply by dissolution of 2.0 mmol of ligand 4 and 1.0 mmol of PdCl 2 to ETOH (20.0 mL) at ambient temperature. The reaction was stirred for 2.0 h. The resulting brown solid was ltered, washed with deionized water as well as cooled EtOH, then isolated as a stable light brown powder at room temperature for the next step.
Silylation of cellulose lter paper. Cellulose FP silylation was performed using a solution method (in water) that was described in our previous work in detail. 19,28 Preparation of 4-OH-phenylenebis(SA-5-ImI)-Pd (II) complexembedded lter paper (FP@Si-Pd II -Salophen). Initially, 0.2 mmol Ag 2 O was added to an ethanolic ammonia solution (15 mL) in a 6 cm diameter glassy petri dish. The role of Ag 2 O was in the converting iodide ions to hydroxides as a part of the Hoffman elimination. 44 Next, 0.1 mmol 4-OH-phenylenebis(SA-5-ImI)-Pd (II) complex (200 mg) was added to the mixture in a petri dish, and the SiCFP was slowly inserted into the solution in the petri dish. To covalent functionalization of the lter paper with the palladium complex, the solution was shacked at ambient temperature for 2 days. The FP@Si-Pd II -Salophen was dried and stored in a vacuum desiccator containing P 2 O 5 .
Typical procedure for FP@Si-Pd II -Salen-[IM]OH catalyzed C-C cross-coupling reactions. As a typical procedure for the FP@Si-Pd II -Salen-[IM]OH catalyzed C-C Sonogashira crosscoupling reaction, phenylacetylene (1.0 mmol) and iodobenzene (1.2 mmol) was dissolved in a solution of EtOH : H 2 O (2 : 1, v/v, 5 mL) at ambient temperature. FP@Pd II -Salen was placed on a glass funnel, that was set-up on a vacuum Erlenmeyer ask equipped with a N 2 inlet (0.3 bar) (Scheme 2). The reaction set up was placed on a water bath adjusted at 50 C. The reaction was performed by ltration of the reactants. In each ltration run, the reaction mixture was poured on the lter paper in one step, and the ltration time was calculated from the moment of pouring the reaction mixture on the funnel to the ltration of the last drop, and the total time was considered as the reaction time. The reaction progress was monitored by GC and/or TLC from the ltrate. Finally, the used FP was washed with hot absolute ethanol (10 mL) and the residue was added to the mixture to calculate the reaction efficiency.
Swelling measurements of lter papers. The swelling amounts of the FPs were determined by weight changes in the lm in the solvents 65,66 using eqn (1): where V s , W 2 , W 1 and d s are swelling amount of the lter paper (mL g À1 ), weight of swollen network (gr), weight of dry lter paper (gr) and density of the solvent, respectively. Other procedures. Two other protocols were also examined for set-up of the reaction to evaluate the catalytic activity of the FP via the ltration method including, of (1) suspending the lter paper inside the reaction mixture; 2 and (2) use of the cutted FP species in the reaction mixture (like a heterogeneous catalyst).
(1) In the rst set up, the FP was suspended inside the reaction mixture using a clamp according to a previously described procedure by Nishikata et al. 2 Given that the Pd loading was depends on the surface area of the FP, one complete sheet of the FP (with diameter of 5.5 cm) was used for this test to make a logical comparison with the other methods. For this purpose, a sheet of the FP was divided into 8 equal cone-shaped parts with a 2.75 cm in rim and then, suspended into the mixture, and the mixture was stirred magnetically and the reaction progress was monitored by TLC or GC continuously. Aer the reaction completion, the recovered FP pieces were shaken in absolute EtOH (15 mL) to achieve the highest efficiency, and the resulting solution was added to the nal mixture to calculate the reaction efficiency. 19 (2) In the second set up, the FP was cut into several square pieces with 1 cm in diameters and added to the reaction mixture. The mixture was stirred by a magnet and its progress was monitored by TLC or GC analysis. Upon the reaction completion, the FP species were separated from the reaction mixture by forceps and shaken for 5 h along with a 15 mL of EtOH, and the resulting ethanolic mixture was added to the main mixture to calculate the reaction conversion.

Author contributions
All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Indah Raya, Svetlana Danshina, Wanich Suksatan, Mustafa Z. Mahmoud, Ali B. Roomi, Yasser Fakri Mustafa. The rst dra of the manuscript was written by Milad Kazemnejadi and Abduladheem Turki Jalil and all authors commented on previous versions of the manuscript. All authors read and approved the nal manuscript.

Conflicts of interest
There are no conicts to declare.