Cellulose Beads as Novel Support of Copper Active Site for the Catalytic Abatement of Water Emerging Contaminant

Different regenerated cellulose (RC) beads were synthesized as supports of copper as the active site for catalytic degradation of emerging contaminants (ECs) in water. Starting from a commercial dissolving pulp, RC beads were prepared from the direct dissolution and from both cellulose carbamate and viscose solutions. Copper was added to the supports by a simple green method. The material characterization by FTIR, TGA, SEM and XPS conrmed the successful incorporation of copper in all the prepared supports. Phenol was adopted as EC test molecule, and catalytic wet peroxide oxidation (CWPO) at 70°C was used to analyse the Cu-cellulosic beads catalytic performance. The novelty and importance of preparing bead-shape catalysts with cellulose reside in the use of an economic, renewable and biodegradable matrix, and the simple separation of the structured catalyst from the heterogeneous solid/liquid reaction media.

aqueous sodium hydroxide at low temperature in combination with different additives (urea, thiourea) has been discussed (Egal, Budtova, & Navard, 2008;Navard, 2012). Nowadays, the regenerated products from dissolved cellulose have gained interest since they are easy to obtain, biocompatible, biodegradable, sustainable, and chemically and thermally stable (Hou et  The main goal of this work consists of studying the synthesis of novel catalysts based on regenerated cellulose (RC) beads supporting copper nanoparticles for the removal of organic contaminants in water. Cellulosic beads obtained from different dissolution methods (NaOH/urea solution, viscose and carbamate solution) were proved for the incorporation of copper. Phenol was proposed as a contaminant test molecule to be decomposed on the synthesized copper/cellulose material.

Production of cellulose solutions
Commercial dissolving pulp (95 % α cellulose, 428 mL/g intrinsic viscosity, 0.7 mmol/g carboxyl group), produced from Eucalyptus (Solucell®) was provided by Bahia Specialty Cellulose S.A. company (Camacari, Brazil). Analytical grade chemical reagents and deionized water were used in all experiments.
Starting from the dissolving pulp, three cellulose solutions were prepared: Direct Dissolution (D), Cellulose Carbamate (C), and Viscose (V). The synthesis protocols are described and schematized in Scheme S1 in the supplementary data.

Direct Dissolution (D)
According to previously reported methods, a mechanical re ning and an enzymatic treatment were performed in order to improve the reactivity and accessibility of the as-received pulp (Grönqvist et al., 2015;Olmos, 2016). The direct dissolution cellulosic pulp (D) was carried out by stirring a mixture of 7 wt.% NaOH (Cicarelli Pro-analysis ACS) and 12 wt.% urea (Cicarelli Pro-analysis) in a thermostatic bath (-5°C) for 30 min. Then the undissolved cellulose was removed by centrifugation during 15 min at 12000 rpm. Finally, the material was stored at 6°C in closed containers until its later use (Egal et

Cellulose Carbamate Solution (C)
Cellulose Carbamate was prepared following previous reports (Fu, Guo, et al., 2014;Fu, Yang, et al., 2014;Guo, Zhou, Song, & Zhang, 2009;Guo, Zhou, Wang, Zhang, & Lin, 2010). A blend composed by cellulose pulp and a solution of 6 w/v % NaOH (Cicarelli Pro-analysis ACS) and 30 w/v % urea (Cicarelli Proanalysis) was stirred during 120 min at 25°C. After that, the obtained material was pressed until obtaining the impregnated cellulose: cellulose pulp ratio of 4.25:1. Then, a drying treatment at 60°C for 20 h was performed. The esteri cation reaction was carried out in a microwave oven (160 W, 15 min). The cellulose carbamate solid was washed with deionized water until neutrality in order to remove residual reactants. Afterwards, it was dried at 60°C for 24 h.
The dried cellulose carbamate was dissolved in a mixture of 8 w/v % NaOH (Cicarelli Pro-analysis ACS) and 0.6 w/v % ZnO (Cicarelli Pro-analysis ACS) at 0°C by vigorous stirring (800 rpm) during 15 min. Then, the solution was centrifuged at 9000 rpm at 0°C for 15 min in order to remove the undissolved cellulose.
In the end, the material was stored at 6°C in closed containers until its later use (Lanieri, 2017).

Viscose Solution (V)
The viscose solution was obtained following a similar protocol as proposed by Whistler et al. (Whistler, 1962). The dissolving pulp was rstly treated in a NaOH (Cicarelli Pro-analysis ACS) solution 18 w/v % to obtain swelled Alkali Cellulose (AC). After that, it was pressed until achieving an AC: cellulose pulp ratio of 4.25:1. Then, in order to control and reduce the cellulose polymerization degree the material was stored 48 h at 23°C. Afterwards, the xanthation reaction was conducted by the addition of CS 2 (Cicarelli Proanalysis ACS) to the aged AC and this mixture was stirred for 3 h at 23°C. Finally, the cellulose xanthate was dissolved at 0°C in a 6 w/v % NaOH solution and it was stored at 6°C in closed containers until its later use (Kotek, 2007;Lanieri, 2017; Lanieri, Olmos, Alberini, & Maximino, 2014).

Preparation Of Regenerated Cellulose Hydrogel Beads (Rc)
The obtained cellulosic solutions (D, C and V) were used to prepare the bead-shape supports. The solutions were extruded through a syringe pump with a ow rate of 30 mL/h. The regenerated cellulose beads were prepared by coagulation followed by regeneration of the cellulose in a mixture of 5 v/v% H 2 SO 4 (Cicarelli 95-98 % Pro-analysis ACS) and 5 w/v % anhydrous Na 2 SO 4 (Cicarelli Pro-analysis). The beads were washed with deionized water until neutrality and stored in a refrigerator (
The aqueous mixture was magnetically stirred in a reaction vessel, which was placed into a thermostatic bath (25°C or 80°C), while the pH was adjusted to 7.0 ± 0.1 (Han, Liu, Singhal, Wang, & Gao, 2012;Liu, Yasunami, Kuruda, & Okido, 2012). Then 20 mL of 0.037 M L-ascorbic acid solution was pumped dropwise into the reaction vessel using a syringe pump. After 4 hours under reaction, the pH resulted c.a. 3, and the bead's colour turned reddish, suggesting that metallic copper nanoparticles were formed on the cellulosic surface. Finally, the suspension containing the catalytic beads was ltered, and they were carefully washed with deionized water and stored in a refrigerator until their use. The copper nominal content was xed at 5 wt. %. The catalysts prepared were labelled as Cu-T/RC, where T refers to the synthesis temperature (25°C or 80°C) and RC describes the starting regenerated cellulose (i.e. D, C or V).

Characterization Of The Synthesized Materials
Beads average diameters were determined from the images taken with an optic microscopy (OLYMPUS BX51) with an OLYMPUS E330 camera. The average value, standard deviation, and coe cient of variation were calculated. The diameter was expressed taking into account the statistical value according to the t distribution at 95 % con dence level.
Total porosity, de ned as the total liquid contained in a particle related to the as-prepared cellulose bead dry mass, was calculated with the equation proposed by Ettenauer et al. (Ettenauer et al., 2011).
Copper content in the catalysts was determined by thermogravimetric analysis (TGA-SDTA) in a Mettler Toledo STAR instrument with a TGA/SDTA851e module. A 70 µL alumina pan containing 30-40 mg of sample was heated from 30°C to 600°C in air ow (50 mL/min). A 10°C min − 1 heating rate was xed. The derivative form of TGA (DTG) was obtained using TGA differentials data. The residual weight after the experiment was related to CuO content in the sample. The TGA analyses of the supports without copper were also performed in order to verify the total combustion of the cellulosic material in the experimental conditions. FTIR spectra of samples were obtained with a Shimadzu IR Prestige-21 spectrometer, in the 4000-400 cm − 1 range, with a resolution of 4 cm − 1 . The dried supports and catalysts were powdered and mixed with BrK, and pressed to form the corresponding wafers.
The surface morphology of the material was studied using scanning electron microscopy (SEM benchtop Phenom ProX) working at 5-15 kV. Measurements were performed in a holder with controlled cooling. The samples were cooled at -20°C and normal pressure. After that, they were introduced into the microscope measurement chamber where water was removed by sublimation at high vacuum.
The nature of the copper species on the surface catalyst was analyzed with a multi-technique system (SPECS) equipped with a dual Mg/Al X-ray source and a PHOIBOS150 hemispherical analyzer, operating in the xed analyzer transmission mode (FAT). The spectra were obtained with a passing energy of 30 eV and the AlKα X-ray source was operated at 100 W. The analysis chamber pressure was less than 2 × 10 − 9 kPa. The XPS spectra of Cu 2p and C 1s were recorded and all the binding energies were referenced to the C 1s peak surface at 284.6 eV. The kinetic energy (KE) in the region of the Cu LMM Auger transition was also measured. The analysis of the peaks was performed with the CasaXPS software, using a weighted sum of the Lorentzian and Gaussian components after Shirley background subtraction. For the quanti cation of the elements, sensitivity factors provided by the manufacturer were used.

Catalytic Activity Of The Synthesized Catalysts
Catalytic wet peroxide oxidation (CWPO) of phenol was used as a reaction test to analyse the catalytic performance of the copper containing beads. The experiments were performed in a batch reactor equipped with magnetic stirring. The nal catalyst system contained 50 mg of Cu in 50 mL of 1000 ppm phenol (Ciccarelli 99 %) solution. As soon as the phenol solution reached 70°C, the catalyst was added and, nally 5 mL of 8.8 M H 2 O 2 (Cicarelli Pro-analysis) were incorporated. That moment was considered as the initial reaction time. During the total time reaction (2 h) several aliquots of the reactive solution were taken in order to analyse the phenol and total organic carbon (TOC) conversion. The phenol concentration was analysed using a colorimetric method according to assay 5530-D (Clesceri, Greenberg, & Eaton, 2012). Additionally, at the nal reaction TOC was determined according to ISO-8245 (1999) using a TOC Analyzer (Elementar, Vario TOC Cube/Trace). For TOC conversion calculation, the original phenol solution was assumed as the initial organic carbon source.

Material Characterization
The properties of the regenerated bead-shaped products prepared from cellulose direct dissolution (RC/D), cellulose carbamate solution (RC/C) and viscose solution (RC/V), are shown in Table 1. It could be observed that in all the bead types, the average diameter and the total porosity resulted similarly. The coe cients of variation (CV %) were below 10 % indicating a homogeneous diameter distribution in all cases. During the incorporation of copper nanoparticles into the supports, several reaction steps were observed. The different paths are shown in the Supplementary Data. Table 2 shows the copper content in the prepared catalysts measured by TGA, and the copper addition yield referred to the initial concentration of the precursor solution (5 wt.%). The e ciency of Cu incorporation on RC/C and RC/V resulted higher when the reaction temperature was 25°C, when the temperature was 80°C, the yield decreased 60 % and 35 % respectively. However, the temperature did not greatly affect the incorporation of copper on RC/D beads.
The thermal degradation of the pure supports and catalysts were or was tested by TGA and DTG analysis (Fig. 1). The TGA thermograms showed the complete weight loss of the supports at 600°C, and no char residue was observed. However, in the presence of copper a residual weight was observed in all cases.
This residual weight associated with copper oxide formed during the organic support combustion revealed the Cu content in the cellulosic matrix (Table 2). During the heating treatment in owing air, three steps could be observed: (1)  In all cases, the initial temperature of the second step decreased after the copper addition in the cellulosic matrix. On the other hand, the highest weight loss temperature decreased after the incorporation of the metal. The shifts of the degradation temperatures compared with that of bare beads depend on the support nature and on the Cu content. The higher Cu content, the lower temperature shifts ( Table 2). As a result, the inclusion of the metallic NPs reduced the catalyst thermostability of the catalysts. That effect is lower on the samples prepared with the RC/D support. (1) Catalyst copper content determined by TGA, (2) Copper addition yield related to the initial precursor concentration. (3) Difference between the observed maximum DTG temperature in the pure support and the corresponding T max observed in the catalyst.
The chemical state of the copper species supported on the surface of the RC beads was analysed by XPS. After scanning a wide spectra region, the Na 1s signal was slightly detected in the beads with lower Cu content (Cu-25/D and Cu-80/C), whereas in the samples with higher Cu content, Na signal was not observed. This is in agreement with Li et al. who reported that the Na 1s signal corresponding to Nacounterions on the structural carboxylates, disappeared after the CuNPs addition (Li, Messele, Boluk, & El-Din, 2019). This phenomenon could contribute to copper stabilization on the cellulosic matrix through coordination with -COO terminals, giving rise to more stable chemical links.
The Cu 2p XPS region presents two main peaks at around 932 eV and 952 eV, which correspond to Cu 2p3/2 and Cu 2p1/2 respectively. The shakeup located at 10 eV, which is higher than the Cu 2p3/2 signal, suggests the presence of Cu 2+ . In order to differentiate Cu 0 entities from Cu + species, X-ray excited Auger  Table 3, while Figure S2 in the Supplementary data displays the spectra used for such integrations and calculations. It could be seen that all spectra presented two components in the Cu 2p3/2 region. The lowest BEs 932.4 to 933.7 eV (Fig. S2. A)    The intensity of the band at c.a. 1640 cm −1 remained stable after copper addition in all cases, regardless of the synthesis temperature. Nevertheless, some differences on the signals could be remarked when comparing the catalyst spectra with their corresponding bare supports. These differences suggest that copper entities were successfully joined on the cellulose chains. In the spectra of Cu-RC/V and Cu-RC/D catalysts, the band at c.a. 1723 cm −1 assigned to carbonyl groups added to the cellulose structure due to the presence of ascorbic acid. In all cases, the synthesis temperature increment led to 1723 cm −1 lower intensity signal, suggesting the loss of C=O terminals. The appearance of a slight peak at 612 cm −1 could also con rm the presence of copper nanoparticles on the cellulose matrix (Prenesti & Berto, 2002). After copper addition, the C-OH band decreased, in agreement with Dong et al. who proposed that these groups could immobilize the copper nanoparticles through electrostatic interactions (Dong & Hinestroza, 2009).
The optical images of supports and catalysts are shown in Figs. 3A and 3B. Figure 3C presents the copper impregnated beads after the cryo-congelation performed previously on the SEM observations. It could be seen that the beads preserved their shape and surface morphology. As a result, the morphological data obtained from SEM observations could be considered for the no cryo-treated beads.
SEM micrographs of the three types of regenerated cellulose and catalysts are shown in Figs. 4 and 5. RC/C support (Fig. 4A) showed a non-homogeneous pore structure with different shapes and sizes in a smooth surface. The RC/D sample (Fig. 4B) presented a surface with wavy cavities and big pores. In contrast, in the RC/V support (Fig. 4C) a highly porous smoothly wavy surface could be observed.
The presence of copper is evidenced in the surface of all the prepared supports (Fig. 5). The distribution of the copper particles was greatly in uenced by the RC morphology; however, the metallic phase addition did not modify the initial support structure. It could be also suggested that the greater the material porosity, the more possibility of metallic entities deposition on the surface of the pores. The metal particles on the RC/C support appeared to be supported as platelet arrangement (Fig. 4). This conformation is present in all supports but, in the RC/C samples, in which the surface seems to be softer and more closed the particles are notably bigger (Sone, Diallo, Fuku, Gurib-Fakim, & Maaza, 2020). This could be observed by comparison of Fig. 5A to D. On the other hand, it could be seen that the higher the synthesis temperature the more aggregation of copper particles. The EDS results con rmed the presence of copper in all the prepared catalysts.

Catalytic tests
In order to prove the activity and e ciency for ECs elimination of the synthesized Cu/RC catalysts, the catalytic degradation of phenol with H 2 O 2 was used as a reaction test.
Several supports were studied for the proposed catalytic reaction with copper as the active site (Table  S1) Table S1) showed that the synthesis of Cu/SiO 2 and Cu/Al 2 O 3 catalysts starting from pre-formed copper nanoparticles with controlled particle size (Gioria et al., 2019) with lower Cu loadings (lower than 1.0 wt.%) converted 100 % of phenol with a more than 50 % of mineralization (Table S1). As a result, cellulose was proposed as an interesting biodegradable low-cost organic material for the same reaction.
Firstly, in order to analyse the cellulosic support (RC/D, RC/C and RC/V) activity for phenol degradation, blank experiments were performed using the supports without metal active phase. The experiments were carried out under the same experimental conditions of temperature, time, and peroxide concentrations. As a result, phenol conversion was not detected.
Then the cellulosic supported catalysts were tested for the phenol CWPO reaction. Figure 6 shows the catalytic results of phenol conversion during 2 h reaction time. It could be observed that the catalysts prepared at 80°C presented more than 90% of phenol conversion. Whereas the catalyst synthesized at 25°C on RC/D and RC/V supports presented lower conversions. In contrast, the conversion of phenol with cellulose carbamate-based catalysts (Cu-RC/C) did not present differences in the catalytic activity (Fig.  6C).
The mineralization process (TOC) measured after 120 min reaction time (Fig. 6) suggested that the phenol was not totally degraded to CO 2 , resulting in a colourless organic by-product mixture. It could be observed, that: (i) in Cu-RC/D and Cu-RC/V solids, the higher phenol conversion the higher the mineralization process and (ii) that the catalysts prepared at 80°C showed similar TOC conversions and high phenol conversion after 120 min.
Clearly, it is not possible to establish a direct relationship between the solid catalytic performances with their total copper content. However, the XPS results could help drawing some relevant hypotheses. In fact, RC/D and RC/C catalysts contain a higher proportion of surface Cu 2+ (Table 3), while the Cu 2+ / (Cu + +Cu 0 ) ratio calculated in RC/V was lower than the other catalysts, suggesting a higher proportion of copper remained as reduced species.
It is well known that the organic contaminant present in the reaction medium could reduce the Cu +  Catalyst morphology (surface area and porosity) plays an important role in catalytic activity together with the active site size and its distribution into the support. The combination of these catalyst properties could give rise to different availability and accessibility of the active sites into the catalytic matrix. As a result, it is di cult to establish a direct relationship between the catalytic results and the physicochemical characteristics of the catalysts due to the observed disparate morphological characteristics of the RC supports. Even so, to the best of our knowledge, the use of copper-based catalysts on this eco-friendly organic matrix has not been reported, at least for CEs degradation in water.
For this reaction, some advantages of the Cu-cellulosic beads could be remarked In comparison to Cubased catalysts prepared on large speci c area commonly used supports: (i) in the low speci c area cellulosic material lower content of copper conduces to similar catalytic yield, (ii) the synthesis of cellulosic supports is more economic and simpler than the synthesis of other more complex materials, (ii) the protocol for the catalytic site incorporation on the cellulose matrix resulted facile and suitable, (iii) facile recovery of the catalyst from the reaction vessel.

Conclusions
Looking for "greener" synthesis routes of catalysts, a protocol was developed to obtain Cu/cellulose hybrid materials for wastewater treatment. As a result, different supports based on regenerated cellulose showed good properties to anchor copper nanoparticles. Applying these bead shaped catalysts for liquid phase heterogeneous catalytic reactions allowed the easy separation of the solid/liquid media after the reaction processes.
The characterization by FTIR, TGA, SEM and XPS of the synthesized materials con rmed the successful incorporation of copper on the supports prepared by the direct dissolution and from both solutions of viscose and carbamate. TGA analysis showed that copper incorporation affects the support thermal stability, being this effect more evident in the catalysts prepared at 25°C. Morphology observations showed that the synthesis protocols of the supports conducted to particular surface properties and consequently to different Cu particles size on the cellulosic material.
Among the prepared materials the Cu-80/D solid showed a slightly better catalytic performance: conversion 96.7 % and mineralization 31.7 %. This behaviour could be attributed to the bigger Cu particles size compared with the observed on the other catalysts. Based on these results, it could be asserted that it is necessary to improve the metallic phase NPs incorporation on the biodegradable support in order to obtain an optimal particle size. Even though all the catalysts presented activity for the removal phenol, further insights are needed so as to enhance the catalytic behaviour.
In view of these promising results, several future studies have been proposed in order to improve the use of cellulosic materials. Among them, the optimization of the active site addition protocol to produce well dispersed Cu nanoparticles, and the use of cellulosic bers and/or nanocellulose for the subsequent incorporation of metallic catalytic sites for low temperature aqueous reactions. Figure 1 TGA and DTG analyses of as-prepared supports and the corresponding catalysts. TGA: A, B and C. DTG: A1, B1 and C1.Inserted graph: remaining weight for the calculation of the copper content.