Journal of Cleaner Production 288 (2021) 125648 lable at ScienceDirectContents lists avaiJournal of Cleaner Production journal homepage: www.elsevier .com/locate/ jc leproDevelopment of melamine-impregnated alginate capsule for selective recovery of Pd(II) from a binary metal solution Wei Wei a, Yanzi Qiu a, Yufeng Zhao c, Kai Zhang a, Yajun Ji a, Hui Gao a, John Kwame Bediako d, Yeoung-Sang Yun b, * a Key Laboratory for Synergistic Prevention of Water and Soil Environmental Pollution, Xinyang Normal University, Nanhu Road 237, Xinyang, 464000, China b Division of Semiconductor and Chemical Engineering, Chonbuk National University, Jeonju, Jeonbuk, 54896, Republic of Korea c College of Resource and Environmental Science, South-Central University for Nationalities, Wuhan, 430074, China d School of Engineering Sciences, University of Ghana, Legon, Ghanaa r t i c l e i n f o Article history: Received 21 July 2020 Received in revised form 17 December 2020 Accepted 21 December 2020 Available online 24 December 2020 Handling editor: M.T. Moreira Keywords: Melamine Calcium alginate barrier Pd(II) selectivity* Corresponding author. E-mail address: ysyun@jbnu.ac.kr (Y.-S. Yun). https://doi.org/10.1016/j.jclepro.2020.125648 0959-6526/© 2020 Elsevier Ltd. All rights reserved.a b s t r a c t An “ionic barrier” concept is proposed to introduce selectivity function to adsorbents towards precious metals (PMs) in the present study. As a model system, melamine-impregnated alginate capsule (MIAC) was prepared and used as a Pd(II) selective adsorbent from a binary metal solution containing Pt (IV) and Pd(II). The MIAC exhibited excellent Pd(II) selectivity for a pH around 4.3, where the selectivity coeffi- cient reached a high value of 2190.66. In comparison to the very low Pt (IV) uptake, a high maximum uptake of Pd(II) was estimated of 316.92 ± 9.50 mg/g by the Langmuir isotherm model. Sorption- desorption studies showed that the MIAC had a good reutilization property. The selective process for Pd(II) recovery was proposed as follows: the electroneutral Pd(OH)2 but not anionic PtCl  5first penetrate the outer ionic barrier; then, the penetrated Pd(OH)2 was bound by the inner melamine through che- lation. The ionic barrier-based sorbent thus can be considered as an alternative one for separation and recovery of Pd(II). © 2020 Elsevier Ltd. All rights reserved.1. Introduction Precious metals (PMs) including Au, Ag, Pd and Pt are widely applied in various industries, such as electronic information tech- nologies, catalysts in different chemical processes, and new energy technologies, owing to their particular physicochemical properties (Changmei et al., 2011). Thus, wastewaters from the above in- dustries usually contain certain amounts of various PMs. Consid- ering their limited natural reserves, the recovery of PMs appears to be important from their wastewaters. For this, different kinds of methods have been studied for PMs recovery, including ion ex- change (Gomes et al., 2001), liquid-liquid extraction (Wei et al., 2016a), chemical precipitating process (Behnamfard et al., 2013), and adsorption (Bediako et al., 2020; Wei et al., 2016b). Among these, adsorption appears to be the preferred method, due to its easy operation, simple maintenance, reduced sludge generation, benign performance, potential regeneration and reuse properties.These aspects are why adsorption is considered to be more envi- ronmentally acceptable (Xi et al., 2020; Chai et al., 2020; Yang et al., 2014). A key point in effectively applying adsorption technology for recovery of PMs is to develop high-performance adsorbents (Mao et al., 2020; Won et al., 2014). In the last few years, biopolymers, including alginate, chitosan and cellulose, have attracted interest as starting materials for preparing synthetic adsorbents because of their beneficial properties including natural abundance, non- toxicity and biodegradability (Wei et al., 2016b). However, the metal uptakes of raw biopolymers are relatively low (Bediako et al., 2015, 2016). Generally, the addition of more binding sites to the biopolymers can enhance their sorption capacities. Relevant modification methods have been studied, which include surface coating and composite formation with chelating chemicals which are rich in N and S groups (Kumar et al., 2015; Sharma and Rajesh, 2014). Based on the hard-soft-acids-bases theory (Pearson, 1963), chelating groups containing N and S as donor atoms have high af- finities for soft acids like PMs (Ramesh et al., 2008; Zhou et al., 2009). Based on the above remarks, high sorption capacity can be W. Wei, Y. Qiu, Y. Zhao et al. Journal of Cleaner Production 288 (2021) 125648obtained through the introduction of more chelating groups to the biopolymers. However, PMs generally co-exist in their wastewaters. Thus, the synthetic adsorbent should possess high selectivity to- wards the target PM in order to recover each PM in high purities. For selective recovery of Au(III), several modified biopolymers have been reported on, including N-aminoguanidine functionalized cellulose powder (Gurung et al., 2013), cross-linked cellulose gel by sulfuric acid (Pangeni et al., 2012) and Aliquat-366-impregnated alginate capsules (Wei et al., 2016b). The results showed that each of these developed adsorbents had a high selectivity towards Au(III) in a Au(III)ePt (IV)ePd(II) ternary system (chloride media). Among Au(III), Pt (IV) and Pd(II), N and S based chelating groups generally have a better affinity towards Au(III) (Gurung et al., 2014; Wei et al., 2016a, 2016b). Moreover, the redox potential of Au(III) is also higher than those of Pt (IV) and Pd(II) (Gurung et al., 2014). Thus, the reason for this selectivity is likely to be the preferential binding and easier reduction towards gold ions. Comparatively, the selective adsorption of Pt (IV) or Pd(II) is still problematic. For the acidic chloride-basedmedium, Pt (IV) and Pd(II) are generally in the form of PtCl26 and PdCl24 , respectively (Wei et al., 2016b). Their similar charge density and redox potential values lead to difficulty in their separation. With a view to solve problem, ion-imprinted biopolymer adsorbents recently appeared to be effective candi- dates. The core of this imprinting technology is to prepare an appropriate template for the target metal species. For instance, we recently reported an ion-imprinted fiber form adsorbent and it showed a good selectivity to Pd(II) from a binarymetal solution (Lin et al., 2015). However, one inherent disadvantage of the Pd(II)- imprinted adsorbent is the high cost of the material due to the use of the expensive Pd(II) chemical. Moreover, in order to make a matched template, the preparation procedure is complex. Thus, more effective Pd(II)-selective adsorbents with low costs and simpler preparation are still required. Biopolymer alginate belongs to the linear polysaccharide, and it usually can be obtained from the seaweed. Owing to its benign properties including low-cost, non-toxicity, biodegradability and abundant carboxylate and hydroxyl functional groups, many modified alginate adsorbents have been studied for removal/re- covery of metal ions (Chen et al., 2013; Wei et al., 2016b). For instance, alginate and calcium ion can easily form calcium alginate hydrogel through a simple cross-linking reaction. To enhance the adsorptive capacity of calcium alginate hydrogel, many researchers studied on various modification methods including surface coating and composite formation with ionic polymers (Gokila et al., 2017; Jiang et al., 2012; Song et al., 2018). The above reported technolo- gies mainly concentrate on increasing the binding sites to the hydrogel by introducing more functional groups. However, less attention focus on using the original charge property of the calcium alginate hydrogel to fabricate ionic barrier based selective adsor- bents towards PMs. Therefore, a new Pd(II)-selective adsorbent, melamine- impregnated alginate capsule (MIAC), was successfully developed through an encapsulated process in this study. The MIAC is composed of an ionic barrier i.e. calcium alginate shell and inside mixture containing melamine (used as the chelating agent) and carboxymethyl cellulose (used as the viscous reagent). Melamine was chosen as the inner chelating agent mainly owing to its amino groups, which could be used as binding sites for the penetrated Pd(OH)2. The preparation scheme of the MIAC is principally on account of the charge differences between electroneutral Pd(OH)2 and negatively charged PtCl-5. The selective behaviors of MIAC were evaluated by a series of batch sorption tests. The morphological features of the MIAC were conducted by microscopy, Fourier transform infrared spectroscopy (FTIR) and Scanning Electron Microscope-Energy Dispersive Spectrometer (SEM-EDS). By using2 X-ray photoelectron spectroscopy (XPS), the Pd(II)-selective sorp- tion mechanism was studied and is discussed in detail.2. Materials and methods 2.1. Materials PdCl2 and H2PtCl65.5H2O were obtained from Kojima Chem- icals (Japan). Sodium alginate was employed as an encapsulation chemical and purchased from Showa Chemical (Korea). Carbox- ymethyl cellulose (CMC) and melamine were supplied by Sigma Aldrich (Korea). CaCl2 was employed as a crosslinking reagent and supplied by Samchun Pure Chemical (Korea). All the other used chemicals were of analytical grade.2.2. Preparation of MIAC The MIAC fabrication was accomplished as follows: 4 g of mel- amine was added into 50 mL of mixture solution containing 1.5% (w/v) CaCl2 and 1.5% (w/v) CMC and stirred well with a magnetic stirrer to give high viscous liquid. This mixed solution was trans- ferred dropwise through a needle into 0.6% (w/v) sodium alginate solution to form capsules. Then, the obtained capsules were sub- merged in 100 mL of 2% (w/v) CaCl2 solution for 30 min to reinforce the alginate layer. Finally, the obtained MIACs were rinsed with deionized water for three times, and then they were stored at 4 C. The preparing procedure of MIAC is concisely shown in Fig. 1(a). The loaded amount of melamine in the capsule was calculated based on the mass balance. Total 4 g of melamine was used to prepare the capsules (MIACs). The number of obtained capsules were around 1395. Thus, each capsule contains around 2.86 mg of melamine.2.3. Preparation of calcium alginate barrier and its penetrating test To better understand the penetration phenomenon of Pt (IV) and Pd(II) through calcium alginate barrier, the penetrating prop- erty of the barrier was conducted by the following two steps. First, preparation of the barrier: 2.5 mL of sodium alginate aqueous so- lution (0.6% (w/v)) was poured onto a filtrator carrier taking along a rubber band; then, 2.5 mL of 2% (w/v) CaCl2 was mixed into the alginate solution; after around 20 min, the calcium alginate barrier was formed by the gelation reaction between calcium ions and alginate. Second, penetrating test of the barrier: a series of Pd(II)e Pt (IV) binary solutions with the pH ranges between 1 and 4 were separately dropped onto the barrier; then, the penetrated droplets were collected through the gravity and the metal concentration of the droplets were evaluated using ICP-AES.2.4. Characterization of MIAC The physical structure of MIAC was characterized by a micro- scope (Nikon SMZ 1500, Japan). The main functional groups of the MIAC were measured by FTIR spectra (Spectrum GX, PerkinElmer, USA) through KBr discs method within the range from 400 to 4000 cm1. SEM-EDS (S 4800, Japan) analysis was carried out to study the morphology and element component of MIAC. To more accurately understand the binding mechanism of MIAC towards Pd(II), the XPS spectrum (Axis-Nova spectrometer, Kratos Analyt- ical, Ltd., UK) was employed to analyze the atomic valence state information of N and Pd (before and after sorption). The detecting condition of Axis-Nova spectrometer is by the monochromatic Al Ka radiation with 1486.7 eV of photons. W. Wei, Y. Qiu, Y. Zhao et al. Journal of Cleaner Production 288 (2021) 125648 Fig. 1. Preparation procedure of MIAC (a), Micrograph of MIAC (b), and FTIR spectra of MIAC (c).2.5. Sorption experiments To evaluate the selective recovery performance of MIAC, a sequence of batch sorption experiments were conducted, including effect of pH, sorption kinetic and sorption isotherm. A 1000 mg/L of Pd(II)ePt (IV) binary stock solution (each concentration of about 1000 mg/L) was firstly prepared in 0.01 M HCl solution, and then it was used for the above experiments through appropriate dilution. According to previous study (Morisada et al., 2011), chloride ions had significant effect for metal species of Pd(II) and Pt (IV) in aqueous solutions. In the medium of 0.01 M HCl, water soluble species of both Pd(II) and Pt (IV) were present in the current pH range between 2 and 5. The adsorbent dosage used in this study was 0.03 g (dry weight) per 30 mL of Pd(II)ePt (IV) binary metal solution in the conical tube (volume of 50 mL). All experiments were carried out in duplicate by a shaker at 25 ± 2 C (100 rpm for around 24 h), and the standard errors were kept less than 5%. For pH edge experiment, the pH was controlled by 1 M NaOH, and the selectivity coefficient (Wei et al., 2016b) was chosen to evaluate the selectivity property towards Pd(II). The related equation was expressed as follows: . ðqPd ¼ .C Þ Sel fPdPd ð (1)qPt C ÞfPt where qPd and qPt are the uptakes of these two metals; CfPd and CfPt represent the final metal concentrations. Moreover, the sampling times of kinetic experiments (pH around 4.3) were conducted from 0 to 24 h, and two typical kinetic models (i.e. pseudo-first-order and pseudo-second-order models) were used to fit the data (Ho and McKay, 1999, S, 1898). The initial3 concentrations for isotherm experiments (pH around 4.3) were from 0 to 1000 mg/L, and the isotherm experimental data were analyzed by Langmuir and Freundlich models (Freundlich, 1906; Langmuir, 1916). The metal concentrations of all samples were quantized by inductively coupled plasma-atomic emission spec- trometry (ICPS-7500, Shimadzu, Japan). In addition, the mass bal- ance equation was employed to calculate the metal uptake and it was shown as below: ¼CiVi  Cq f Vf (2) M here, Ci and Cf (mg/L) stands for the initial and final Pt (IV) or Pd(II) concentrations; Vi and Vf (L) separately denote the initial and final working volumes; M (g) represents the usage amount of the adsorbent.2.6. Reutilization experiments A 100 mg/L of the binary solution was used for the sorption- desorption experiment. After sorption of metal ions, the desorp- tion process for the adsorbedMIAC was carried out by using several eluants including 0.5M HCl, 1M HCl, 0.5M thiourea, 1M thiourea and acidic thiourea (1M HCl þ 1M thiourea). The desorption effi- ciency (DF) toward MIAC was evaluate by the equation as shown below:   ¼ Desorbed metalðmgÞDF % ð Þ  100% (3)Adsorbed metal mg To study the reuse property of MIAC, the sorption-desorption experiments were repeated five times by using acidic thiourea (1 M HCl þ 1 M thiourea). W. Wei, Y. Qiu, Y. Zhao et al. Journal of Cleaner Production 288 (2021) 125648 Fig. 2. pH effect on the performance of MIAC in the binary solution: (a) Metal uptake and (b) SelPd. (Initial concentration ¼ 100 mg/L, sorbent dosage ¼ 0.05 g/30 mL, temperature ¼ 25 ± 2 C, and contact time ¼ ~24 h).3. Results and discussion 3.1. Characterization studies of microscope and FTIR To understand the morphology feature of MIAC, a microscope was used and its internal and external structure was shown in Fig. 1(b). The MIAC was nearly spherical with an outer calcium alginate shell and an inner solution of melamine and CMC mixture. The thickness of the outer shell and the diameter of MIAC were separately measured about 0.16 and 3.65 mm. To further confirm that melamine was successfully entrapped within the core shell of the MIAC, FTIR analysis was conducted including CMC, melamine and MIAC (shown in Fig. (c)). In the case of CMC, the peak at 3363 cm1 corresponds to OH stretching vibration. The bands at 1430 and 1022 cm1 are assigned to stretching vibrations of COO and CeOeC (Wei et al., 2015, 2016b). In addition, the peaks belonging to melamine are as follows (Ming et al., 2016; Mircescu et al., 2012): a series of peaks observed between 3000 and 3500 cm1 appertain to NH stretching vibrations. The observed peaks of 1653 and 1620 cm1 were corresponding to bending vi- brations for NH2. The peaks from the 1, 3, 5-triazine skeleton of melamine were found at 1547 and 813 cm1. The band at 1455 cm1 was assigned to CN stretching vibration. Comparing to the results of CMC and melamine, MIAC possessed typical peaks belonging to melamine (i.e. NH, NH2, 1, 3, 5-triazine skeleton and CN) and CMC (i.e. COO and CeOeC). FTIR results further revealed that the shell well wrapped up the melamine. For SEM-EDS anal- ysis, the wet MIAC (before sorption) was firstly freeze dried, and then cut into halves (shown in Fig. S1). The component of elements from inside of MIAC (a small region mark in red) was listed in Table S1. The presence of C and O was from inside CMC. The exis- tence of N was attributed to inner melamine. In addition, Ca and Cl were from CaCl2, which was used to fabricate MIAC. The results of SEM-EDS thus proved the successful synthesis of MIAC. 3.2. pH effect and mechanism of selective sorption Generally, the solution pH can influence the adsorptive process through changing the activity of functional group in the adsorbent and metal species while in the aqueous solution. Moreover, the variation of the pH significantly affects the equilibrium and dynamics features for the adsorbent (Gurgel and Gil, 2009). Hence, the experiment of pH effect was investigated for MIAC and the results were illustrated in Fig. 2(a). Both uptakes of Pd(II) and Pt (IV) were almost nil until pH 3. Interestingly, a pH level greater than pH 3, the Pd(II) uptake drastically increased by only a small change of the pH. Whereas, the Pt (IV) uptake remained negligible throughout the pH range investigated. Furthermore, from Fig. 2(b), the selectivity co- efficients towards Pd(II) approached to zero when pH< 3, because of its extremely low uptake. After pH 3, the value rapidly increased and then gradually achieved a high value of about 2190.66. Considering the importance of melamine dosage, effect of melamine amount experiments were conducted including 0 g (M0 g), 1 g (M1 g), 3 g (M3 g) and 4 g (M4 g) for fabricating MIACs. As shown in Fig. S1, the capsule without melamine had the least metal uptakes. However, the Pd(II) uptakes significantly increased with increasing melamine amounts of the capsules. It indicated that the present melamine (providing binding sites of amino groups) in the sorbent played dominant role for sorption of Pd(II). The MIAC (prepared by 4 g melamine) exhibited best Pd(II) selectivity with 58.12 mg/g of Pd(II) uptake and 0.83 mg/g of Pt (IV) uptake. By careful examination of related literature, it can be noted that the forms of themetal complexes can be vary as functions of pH and chloride concentration in solution. Briefly, the solution pH has little influence on the Pt (IV) species and its main form is a stable anionic4 complex, PtCl-5 (Morisada et al., 2011) at the pH ranging from 1 to 5. On the other hand, PdCl3 is the major species for Pd(II) when the pH is less than 3. Starting from pH 3, Pd(OH)2 as an electroneutral molecule appears and gradually becomes the main form when the pH is from 4 to 4.5 (Ho Kim and Nakano, 2005; Morisada et al., 2011). In addition to metal speciation, another important factor for the Pd(II) selective sorption is the outer layer (calcium alginate shell) of the developed MIAC. The alginate gel of the shell was formed by the well-known crosslinking reaction between Ca2þ and the carboxylates of alginate. The structure of the shell can be briefly presented as “eCOO… Ca2þ … OOCe“. Thus, it is proposed that the shell should be negatively charged owing to its electronegative carboxyl group. The pHpzc was determined by the method of pre- vious report (Bian et al., 2015) in order to further confirm that the surface layer of MIAC possessed negative charges during the cur- rent studied pH range between 2 and 5. From Fig. S3, the pHpzc of MIAC was found to be around 1.97, indicating that MIAC was positively charged at pH less than pHpzc, but was negatively charged at pH greater than pHpzc. It thus verified that the developed MIAC had a certain amount of negative charges at the present pH range from 2 to 5. Based on the above remarks, we characterize the sorption behavior of MIAC towards Pd(II) and Pt (IV) as follows: when pH< 3, the twometal ionsmainly exist in their anionic forms, PtCl-5 and PdCl3 . Thus, due to the repulsive force of slightly negative charge of the outer calcium alginate barrier, these negative W. Wei, Y. Qiu, Y. Zhao et al. Journal of Cleaner Production 288 (2021) 125648complexes are likely difficult for penetration of the barrier and it brings about no chance to be in touch with the interior amino groups, which leads to negligible metal uptakes. On the other hand, when the pH is greater than 3, the major form of Pd(II) becomes to Pd(OH) -2, but Pt (IV) still maintains the form of PtCl5. As such, the outer alginate shell does not electrostatically hinder the electro Pd(OH)2, unlike the anion of PtCl-5. Since Pd(OH)2 penetrated the shell, melamine in the inner part is thought to have functioned as binding sites for the sorption of Pd(OH)2 (shown in Fig. 3(a)). Subsequently, calcium alginate hydrogel film was prepared to verify its barrier property. The inset figure in Fig. 3(b) showed the preparation process and appearance of the fabricated calcium alginate film. From Fig. 3(b), before pH 3, the concentrations of Pt (IV) and Pd(II) in the penetrate droplets both were very low, below 1 mg/L. Differently, the concentrations of penetrable Pd(II) were separately improved to 11.62 mg/L and 16.78 mg/L with pH at 3.5 and 4.0, while the Pt (IV) concentrations were still less than 1 mg/L. Thus, the above findings prove that calcium alginate hydrogel has barrier effects to obstruct the penetrating behaviors towards negative ions of PdCl3 and PtCl-5 apart from electroneutral Pd(OH)2. SEM-EDS analysis of MIAC (after sorption) was carried out to confirm that Pd(II) was adsorbed in the inner of the capsule. The SEM image of one half of freeze-dried capsule was shown in Fig. S4. The EDS result of a small region from inside of the capsule (after sorption) was listed in Table S2. The existence of Pd element from the innerMIACwas thus verified that Pd(II) was bound inside of theFig. 3. Proposed selective sorption process of MIAC (a), penetration results of the calcium alginate barrier (b). (Initial concentration ¼ 100 mg/L, temperature ¼ 25 ± 2 C, and pH ¼ 1e4). 5 capsule. In addition, the absence of Pt element was further proved the good selectivity for Pd(II) byMIAC. XPSwas employed to further study the sorption mechanism of the inner melamine towards the penetrated Pd(OH)2. From Fig. 4(a) (N1s spectrum), before sorption, the existence of two peaks indicates two different forms based on N inside the MIAC. The peaks of 398.1 and 399.3 eV separately belong to ¼ N‒ and eNH2 of melamine (Baskar et al., 2013). The bands belonging to imine and amine groups separately increased to 398.7 and 399.8 eV after sorption. These observations confirm that the inner melamine worked as the binding site for the penetrated Pd(OH)2. More precisely, the inside amino groups should be responsible for the coordination of Pd(OH)2. At a pH between 3 andFig. 4. XPS analysis of (a) N1s and (b) Pd3d. W. Wei, Y. Qiu, Y. Zhao et al. Journal of Cleaner Production 288 (2021) 125648 Fig. 5. Sorption kinetics in the binary solution by MIAC: pseudo first and second ki- netic modeling (a) and intra-particle diffusion modeling (b). (Initial concentration ¼ 100 mg/L, sorbent dosage ¼ 0.17 g/100 mL, temperature ¼ 25 ± 2 C, and pH ¼ ~4.3).5, melamine is supposed to exhibit partial positive charge owing to its pKa of 5.11 (Fashi et al., 2015). However, the triazine molecule of melamine, which is a highly electron-rich ring, participates in resonance with the amine groups and may cause them act as a neutral compound (as a whole molecule). According to HSAB the- ory, the affinities those exist between soft bases and soft acids will be more stable and the probable interactions are covalent in nature (Morisada et al., 2011, 2012). In other words, the Pd(II) is a soft metal and N atom is more softer than O atom. Therefore, as shown in Fig. 3(a), the amine group (soft base) of melamine provides electrons to Pd(II) and displaces the OH group (hard base) of Pd(OH)2. In addition, the imine group at the ring of melamine can also donate electrons to Pd(II), making it a more stable complex by chelation. It was found that the solution pH increased during the pH control process, which supports the ligand substitution mecha- nism. For the Pd3d spectrum obtained from XPS after sorption is shown in Fig. 4(b), indicating two distinct valences towards Pd element. The peaks at 335.8 eV (Pd3d 5/2) and 341.1 eV (Pd3d 3/2) are consistent with zero-valent palladium (Pd0), while the bands at 337.2 eV (Pd3d 5/2) and 342.5 eV (Pd3d 3/2) correspond to Pd2þ from Pd(OH)2 (Yi et al., 2016). Based on the calculation by using the software of XPSPEAK Version 4.1, the proportions of Pd2þ and Pd0 are 61.2% and 38.8%, respectively. This demonstrates that the adsorbed palladium existed in a way that Pd2þ dominated and Pd0 was secondary. Previous reports suggested that the hydroxyl group as a benign reductant could be utilized for reduction of the Au3þ and Pd2þ to their zero valences (Pangeni et al., 2012; Yi et al., 2016). Besides melamine, the inner viscous solution of MIAC also includes CMCwhich has abundant hydroxyl groups (Wei et al., 2015, 2016b). Thus, the Pd0 present was probably due to the redox reaction be- tween Pd2þ and ReOH. The reaction is likely expressed by the following equation: Pd2þ þ2R OH/Pd0 þ 2R ¼ Oþ 2Hþ (4) Based on the above analysis, the selective sorption mechanism towards Pd(II) by MIAC is proposed as follows: (1) electroneutral Pd(OH)2 but not anionic PtCl5 penetrates the alginate shell (ionic barrier); (2) after penetration, the inner amino groups exchange the OH of Pd(OH)2 and bonded Pd2þ through chelation; (3) a portion of the adsorbed Pd2þ is reduced to Pd0with the help of the hydroxyl groups. For further evaluation of the sorption performance of MIAC, the following experiments were carried out to examine various properties including sorption kinetics, isotherm and reusability.3.3. Sorption kinetics With regard to a sorbent, it is important to study the property of sorption kinetics for better understanding of the sorption process (Lin et al., 2019; Wei et al., 2016b). The kinetic experiment for MIAC thus was carried out in the binary metal solution. From Fig. 5(a), we can see a significant difference in sorption capacities towards Pd(II) and Pt (IV). For the uptake of Pd(II), it improvedwith the increase of contact time finally reaching an equilibrium state, but Pt (IV) was hardly adsorbed by MIAC within the experimental time. Moreover, it took a relatively long time (~24 h) for the Pd(II) sorption to attain equilibrium. Because unlike direct sorption on the sorbent surface, the present Pd(II) sorption process goes through two steps: the Pd(OH)2 firstly passed through the outer layer of calcium alginate, and then the inner amino groups could bind the penetrated Pd(OH)2 by the reaction of ligand substitution. Since penetrated Pd(OH)2 molecules need time to gradually occupy the inside binding sites, it requires a relatively long time for Pd(II) sorption by MIAC. The parameters and coefficient of determination (R2) of the two6 kinetic models were summarized in Table 1. The pseudo-second- order model exhibits a better representation towards Pd(II) owing to its higher value of R2 (0.991) and more reasonable prediction value of the equilibrium uptake (53.51mg/g). According to previous studies, if the kinetic process is more aligned with the pseudo- second-order kinetic model, chemisorption should be considered as themainmechanism (Wang et al., 2014). For this study, the Pd(II) binding mechanism is mainly on account of the ligand substitution reaction between the amino groups and Pd(OH)2. Furthermore, the measured values of R2 for Pt (IV) were both less than 0.6 by these two kinetic models. The extremely small Pt (IV) uptake is likely to be the reason for this poor response. To further explore the kinetic diffusion mechanism, the intra- particle diffusion (IPD) model (Bediako et al., 2020) was employed to fit the kinetic data. According to IPD modeling results (shown in Fig. 5(b)), the sorption process of Pd(OH)2 was characterized by three steps including fast sorption stage (steep line), slow sorption stage (subdued line) and final equilibrium stage (less steep line). The IPD rate constants in Table 1 follow the order of k 2i,1st > ki,2nd > ki,3rd. In addition, R values of these three stages were all greater than 0.9, indicating a good fitting by IPD model. In the W. Wei, Y. Qiu, Y. Zhao et al. Journal of Cleaner Production 288 (2021) 125648 Table 1 Kinetics equations and fitting parameters by MIAC. Kinetics model Equation Parameter value R2 value Pseudo-first-order kinetics qt ¼ q1ð1 expðk1tÞÞ q1 ¼ 49.83 ± 1.01 0.956 k1 ¼ 0.0237 ± 0.0028 Pseudo-second-order kinetics q2 q ¼ 53.50 ± 0.5359 0.991 q ¼ 2k2t 2t 1þ q k t k2 ¼ 0.0006 ± 0.000052 2 IPD model qt ¼ ki$t0:5 þ C ki,1st ¼ 4.24 ± 0.26 0.985 ki,2nd ¼ 1.07 ± 0.10 0.972 ki,3rd ¼ 0.16 ± 0.01 0.914 Here, q1 and q2 (mg/g) are on behalf of equilibrium uptakes; qt (mg/g) stands for the uptake towards arbitrary sorption time t; k 11 (min ) and k2 (g/(mgmin)) are rate constants of the two kinetics models; ki (mg/(g$min0.5)) and C are intra-particle diffusion rate constant and the thickness of the boundary layer, respectively.beginning, Pd(OH)2 could easily penetrate the barrier and be bound by the outer melamine from the internal of the capsule, leading to a fast sorption stage. Conversely, the relatively low sorption rate of the second stage probably was due to the decreasing concentration gradients which slowed down the diffusion rate of Pd(OH)2. After most of binding sites from the outer and middle melamine inside the capsule were occupied, Pd(OH)2 molecule was more difficult to diffuse and then to be bound by the innermost melamine, resulting a gradually equilibrium stage. It is thus confirmed that IPD model could elucidate the true sorption and diffusion behaviors of the developed capsule.Fig. 6. Isotherm modeling at 298 K (a) and Langmuir modeling at different tempera- tures of 288 K, 298 K and 308 K (b). (Initial concentration ¼ 0e1000 mg/L, sorbent dosage ¼ 0.05 g/30 mL, pH ¼ ~4.3 and contact time ¼ ~24 h).3.4. Sorption isotherms In general, knowledge of sorption isotherms is helpful to the design of sorption process for the industrial application (Bediako et al., 2019). According to the above studies, it was clear to know that the prepared MIAC could be used as a good candidate for se- lective sorption of Pd(II). To further evaluate the maximum Pd(II) uptake by MIAC, sorption isotherm tests were conducted in the binary metal solution with the initial metal concentrations ranging between 0 and 500 mg/L. The sorption isotherm plots of Pd(II) and Pt (IV) by MIAC are shown in Fig. 6(a). For Pd(II), the amount of adsorbed Pd(II) significantly increased along with increasing equilibrium metal concentrations and gradually became stable. Comparatively, the Pt (IV) uptakes were very small over the whole range of metal con- centrations tested. Table 2 summarized the parameters of Langmuir and Freundlich isotherm models. By contrast, the Langmuir model possessed a better fitting towards Pd(II) owing to its more higher R2 value of 0.989. In the current Pd(II)ePt (IV) binary solution, the predicted maximum uptake of Pd(II) was 316.92 ± 9.50mg/g. It was around 47 times higher than the uptake of Pt (IV). Moreover, the maximum uptake and selectivity coefficient towards Pd(II) were chosen as comparative factors for comparison of the sorption per- formance. As listed in Table 3, MIAC possessed higher uptake and better selectivity than majority of previously reported Pd(II) adsorbents. To better explore the sorption mechanism, Isotherm experi- ments for thermodynamic study were conducted at different temperatures of 288 K, 298 K and 308 K. The fitting lines and pa- rameters by Langmuir model were shown in Fig. 6(b) and Table 4, respectively. The maximum Pd(II) uptake increased with increasing the sorption temperature. Thermodynamic study usually used to reveal whether the sorption process is spontaneous or not. The thermodynamic parameters including Gibbs free energy change (△Go), standard enthalpy change (△Ho) and standard entropy change (△So) were calculated based on the following Van’t Hoff equations (Kumar et al., 2015): DGo ¼  RT ln Ko (5)7 DHo DSo ln Ko ¼ þ (6)RT R where R stands for the universal gas constant (8.314 J/mol$K), T (K) is the sorption temperature, and Ko means sorption equilibrium constant. The values of Ko were equal to Langmuir constant of b. △Ho and△So could be obtained from the slope and intercept of the linear fitting plot between lnKo and 1/T. The thermodynamic parameters were listed in Table 5. The W. Wei, Y. Qiu, Y. Zhao et al. Journal of Cleaner Production 288 (2021) 125648 Table 2 The fitting parameters of these two isotherm models. Isotherm model Equation Parameter value R2 value Langmuir isotherm ¼ qmbCf qm ¼ 316.92 ± 9.50 0.989qe 1þ bCf b ¼ 0.0215 ± 0.0023 Freundlich isotherm qe ¼ K C 1=n KF ¼ 37.72 ± 6.03 n ¼ 2.8841 ± 0.2481 0.968F f Here, qe (mg/g) means the equilibrium uptake; b (L/mg) means the affinity constant; qm (mg/g) represents the maximum metal uptake; Cf (mg/L) denotes the final metal concentration; KF ((mg/g) (L/g)1/n) means the Freundlich constant; n stands for the Freundlich exponent without unit. Table 3 Sorption performances of MIAC compared to the other related adsorbents. Sorbents pH best CCl qm SelPd Ion imprinted porous polymer particles (Jiang and Kim, 2013) 0.5 e 38.9 6.91 Ion-imprinted modified silica gel (Zheng et al., 2007) 5.0 e 26.7 50 Ion imprinted polymer (Daniel et al., 2006) 4.0 e 41.6 11.8 Precipitation based Pd(II) imprinted polymer (Daniel et al., 2005) e e 20.16 1100.0 Suspension based Pd(II) imprinted polymer (Daniel et al., 2005) e e 18.76 110.0 Ion-imprinted chitosan fiber (Lin et al., 2015) 1.0 0.1 324.6 20.82 Nanopore based Pd(II)-imprinted polymer (Daniel et al., 2003) e e 21.5 521.0 Ion-imprinted modified chitosan resin (Monier et al., 2016) 5.0 e 275.0 5.85 Escherichia coli biomass (Kim et al., 2015) 1.0 0.1 38.87 e Grafting chitosan on persimmon tannin extract (Zhou et al., 2015) 5.0 0.01 330.0 e Magnetic cross-linking chitosan nanoparticles (Zhou et al., 2010) 2.0 e 138.0 e Dimethylamine-modified waste paper (Adhikari et al., 2008) 0 1 223.48 e Silica-based adsorbent (Bai et al., 2013) 1.0 e 83.0 e Lewatit MonoPlus TP 214 (Won and Yun, 2013) 1.0 0.1 241.1 e Aliquat-336 impregnated chitosan (Kumar et al., 2015) 3.5 e 187.61 e Valonea tannin resin (Can et al., 2013) 3.0 e 74.43 e Hierarchically flower-like WS2 microcrystals (Wang et al., 2020) 3.20 e 67.29 e Calcium alginate gel beads (Cataldo et al., 2016) 3.0 0.025 127 e MIAC (This study) 4.3 0.01 316.92 2190.66 Units of CCl and qm separately are mol/L and mg/g. Table 4 Langmuir parameters by MIAC at different temperature. Temperature (K) qm (mg/g) b (L/mg) R2 288 293.46 ± 8.76 0.0188 ± 0.0020 0.990 298 316.92 ± 9.50 0.0215 ± 0.0023 0.989 308 325.32 ± 10.10 0.0242 ± 0.0028 0.988negative values of△Go demonstrated a favorable and spontaneous sorption for Pd(II). The positive △Ho value indicated that the sorption process was endothermic, which was in a good agreement with the result that higher temperature was beneficial for Pd(II) uptake. In addition, the positive △So value revealed that the randomness increased at solid/liquid interface during the sorption process (Chen et al., 2020). Pd(II) was transferred from the bulk solution when it was adsorbed by the solid sorbent, leading to increasing the entropy of the system.3.5. Regeneration and reutilization For the perspective of cost-effectiveness, a benign reusability property is indispensable for a good adsorbent. Five cycles of sorption-desorption experiments thus were carried out to evaluate two properties namely reutilization and stable selectivity of MIACTable 5 Thermodynamic parameters by MIAC. Temperature (K) △Go (kJ$mol1) △Ho (kJ$mol1) △So (J$mol1$K1) 288 18.83 9.25 95.32 298 19.17 308 20.11 8 towards Pd(II). The desorption efficiency of Pd(II) was first evalu- ated using several desorption solutions before reutilization exper- iments. From Fig. 7(a), it was found that a mixture of 1M HCl þ 1M thiourea was the best candidate solution for desorption of Pd(II) owing to the high desorption efficiency (close to 100%). As stated, reusability experiments were performed using the acidic thiourea solution (1M HCl þ 1M thiourea). From Fig. 7(b), MIAC exhibited a good reusability, and simultaneously maintained the high selec- tivity for Pd(II). This demonstrates that MIAC merely lost a negli- gible amount of melamine within the five cycles, probably because of the benign encapsulation ability of the calcium alginate barrier. Finally, a simple separation process for these two metals was employed in order to recover high purity of each metal solution. From left to right (illustrated in Fig. 8), the developed capsules of MIAC can be clearly seen in the first bottle. The second one shows the appearance after Pd(II) selective sorption from the binary so- lution. In this way, Pd(II)-enrichedMIACs and a Pt (IV) solutionwith a purity of around 98.25% both were collected. The third one re- flected that a Pd(II) solution with a high purity around 99.12% was obtained by the desorption experiment using 1M HCl þ 1M thiourea.4. Conclusions MIAC was successfully fabricated through a facile synthesis process at the present study. The designed MIAC showed excellent Pd(II) selectivity in the binary metal solution. The outer calcium alginate shell of MIAC likely works as an “ionic barrier” to obstruct the penetration of anionic PtCl-5, but not of electroneutral Pd(OH)2 molecules. This is likely the first, crucial step for Pd(II) selective sorption through the use of MIAC. Therefore, this work probably suggests a new fabrication method of designing ionic barrier-based selective adsorbents towards PMs. W. Wei, Y. Qiu, Y. Zhao et al. Journal of Cleaner Production 288 (2021) 125648 Fig. 7. Desorption test using different eluents (a) and five sorption-desorption cycles by acidic thiourea (b). (Initial concentration ¼ 100 mg/L, sorbent dosage ¼ 0.05 g/ 30 mL, temperature ¼ 25 ± 2 C, and contact time ¼ ~24 h). Fig. 8. Recovery of Pd(II) and Pt (IV) by MIAC. (Initial concentration ¼ 100 mg/L, sorbent dosage ¼ 0.05 g/30 mL, temperature ¼ 25 ± 2 C, and desorption reagent by 1M HCl þ 1M thiourea).CRediT authorship contribution statement Wei Wei: Conceptualization, Methodology, Investigation, Soft- ware, Writing - original draft. Yanzi Qiu: Validation, Visualization,9 Investigation. Yufeng Zhao: Validation, Visualization. Kai Zhang: Validation, Visualization. Yajun Ji: Visualization. Hui Gao: Visuali- zation. John Kwame Bediako: Visualization. Yeoung-Sang Yun: Writing - review & editing. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments This work was supported by the National Natural Science Foundation of China (51808477), the Korean Government through NRF (2017R1A2A1A05001207) grants, the Key Scientific Research Project of Colleges and Universities of Henan Province in China (19A610009), the Science and Technology Project of Henan Prov- ince in China (182102311018), the Open Fund of Key Laboratory for Synergistic Prevention of Water and Soil Environmental Pollution (KLSPWSEP-A07), and the Nanhu Scholars Program for Young Scholars of XYNU in China. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.jclepro.2020.125648. References Adhikari, C.R., Parajuli, D., Kawakita, H., Inoue, K., Ohto, K., Harada, H., 2008. Dimethylamine-modified waste paper for the recovery of precious metals. Environ. Sci. Technol. 42, 5486e5491. S, L., 1898. Zur theorie der sogenannten adsorption gelo€ster stoffe. K. - Sven. Vetenskapsakademiens Handl. 24, 1e39. Bai, F., Ye, G., Chen, G., Wei, J., Wang, J., Chen, J., 2013. Highly selective recovery of palladium by a new silica-based adsorbent functionalized with macrocyclic ligand. Separ. Purif. Technol. 106, 38e46. Baskar, S., Liao, C.-W., Chang, J.-L., Zen, J.-M., 2013. Electrochemical synthesis of electroactive poly(melamine) with mechanistic explanation and its applica- bility to functionalize carbon surface to prepare nanotubeenanoparticles hybrid. Electrochim. Acta 88, 1e5. Bediako, J.K., Wei, W., Kim, S., Yun, Y.-S., 2015. Removal of heavy metals from aqueous phases using chemically modified waste Lyocell fiber. J. Hazard Mater. 299, 550e561. Bediako, J.K., Kim, S., Wei, W., Yun, Y.-S., 2016. Adsorptive separation of Pb(II) and Cu(II) from aqueous solutions using as-prepared carboxymethylated waste Lyocell fiber. Int. J. Environ. Sci. Te. 13, 875e886. Bediako, J.K., Sarkar, A.K., Lin, S., Zhao, Y., Song, M.-H., Choi, J.-W., Cho, C.-W., Yun, Y.- S., 2019. Characterization of the residual biochemical components of sequen- tially extracted banana peel biomasses and their environmental remediation applications. Waste Manag. 89, 141e153. Bediako, J.K., Lin, S., Sarkar, A.K., Zhao, Y., Choi, J.-W., Song, M.-H., Wei, W., Reddy, D.H.K., Cho, C.-W., Yun, Y.-S., 2020. Benignly-fabricated crosslinked polyethylenimine/calcium-alginate fibers as high-performance adsorbents for effective recovery of gold. J. Clean. Prod. 252, 119389. Behnamfard, A., Behnamfard, A., Veglio, F., 2013. Process development for recovery of copper and precious metals from waste printed circuit boards with emphasize on palladium and gold leaching and precipitation. Waste Manag. 33, 2354e2363. Bian, Y., Bian, Z.-Y., Zhang, J.-X., Ding, A.-Z., Liu, S.-L., Wang, H., 2015. Effect of the oxygen-containing functional group of graphene oxide on the aqueous cad- mium ions removal. Appl. Surf. Sci. 329, 269e275. Can, M., Bulut, E., O€rnek, A., O€zacar, M., 2013. Synthesis and characterization of valonea tannin resin and its interaction with palladium (II), rhodium (III) chloro complexes. Chem. Eng. J. 221, 146e158. Cataldo, S., Gianguzza, A., Pettignano, A., 2016. Sorption of Pd(II) ion by calcium alginate gel beads at different chloride concentrations and pH. A kinetic and equilibrium study. Arabian J. Chem. 9, 656e667. Chai, F., Wang, R.K., Yan, L.L., Li, G.H., Cai, Y.Y., Xi, C.Y., 2020. Facile fabrication of pH- sensitive nanoparticles based on nanocellulose forfast and efficient As(V) removal. Carbohydr. Polym. 245, 116511. Changmei, S., Guanghua, Z., Chunhua, W., Rongjun, Q., Ying, Z., Quanyun, G., 2011. A resin with high adsorption selectivity for Au (III): preparation, characteriza- tion and adsorption properties. Chem. Eng. J. 172, 713e720. W. Wei, Y. Qiu, Y. Zhao et al. Journal of Cleaner Production 288 (2021) 125648Chen, J.H., Xing, H.T., Guo, H.X., Li, G.P., Weng, W., Hu, S.R., 2013. Preparation, characterization and adsorption properties of a novel 3- aminopropyltriethoxysilane functionalized sodium alginate porous membrane adsorbent for Cr(III) ions. J. Hazard Mater. 285e294. Chen, B., Cao, Y.R., Zhao, H.N., Long, F.X., Feng, X.F., Li, J., Pan, X.J., 2020. A novel Fe3þ- stabilized magnetic polydopamine composite for enhanced selective adsorption and separation of Methylene blue from complex wastewater. J. Hazard Mater. 392, 122263. Daniel, S., Gladis, J.M., Rao, T.P., 2003. Synthesis of imprinted polymer material with palladium ion nanopores and its analytical application. Anal. Chim. Acta 488, 173e182. Daniel, S., Prabhakara Rao, P., Prasada Rao, T., 2005. Investigation of different polymerization methods on the analytical performance of palladium(II) ion imprinted polymer materials. Anal. Chim. Acta 536, 197e206. Daniel, S., Praveen, R.S., Rao, T.P., 2006. Ternary ion-association complex based ion imprinted polymers (IIPs) for trace determination of palladium(II) in environ- mental samples. Anal. Chim. Acta 570, 79e87. Fashi, A., Yaftian, M.R., Zamani, A., 2015. Determination of melamine in dairy products using electromembraneeLPME followed by HPLC. Food Chem. 188, 92e98. Freundlich, H.M.F., 1906. Uber die adsorption in losungen (Adsorption in solution). Z. Phys. Chem. 57, 385e470. Gokila, S., Gomathi, T., Sudha, P.N., Anil, S., 2017. Removal of the heavy metal ion chromiuim(VI) using Chitosan and Alginate nanocomposites. Int. J. Biol. Mac- romol. 104, 1459e1468. Gomes, C.P., Almeida, M.F., Loureiro, J.M., 2001. Gold recovery with ion exchange used resins. Separ. Purif. Technol. 24, 35e57. Gurgel, L.V.A., Gil, L.F., 2009. Adsorption of Cu(II), Cd(II), and Pb(II) from aqueous single metal solutions by succinylated mercerized cellulose modified with triethylenetetramine. Carbohydr. Polym. 77, 142e149. Gurung, M., Adhikari, B.B., Morisada, S., Kawakita, H., Ohto, K., Inoue, K., Alam, S., 2013. N-aminoguanidine modified persimmon tannin: a new sustainable ma- terial for selective adsorption, preconcentration and recovery of precious metals from acidic chloride solution. Bioresour. Technol. 129, 108e117. Gurung, M., Adhikari, B.B., Gao, X.P., Alam, S., Inoue, K., 2014. Sustainability in the metallurgical industry: chemically modified cellulose for selective biosorption of gold from mixtures of base metals in chloride media. Ind. Eng. Chem. Res. 53, 8565e8576. Ho, Y.S., McKay, G., 1999. Pseudo-second order model for sorption processes. Pro- cess Biochem. 34, 451e465. Ho Kim, Y., Nakano, Y., 2005. Adsorption mechanism of palladium by redox within condensed-tannin gel. Water Res. 39, 1324e1330. Jiang, Y., Kim, D., 2013. Synthesis and selective adsorption behavior of Pd(II)- imprinted porous polymer particles. Chem. Eng. J. 232, 503e509. Jiang, N., Xu, Y., Dai, Y., Luo, W., Dai, L., 2012. Polyaniline nanofibers assembled on alginate microsphere for Cu2þ and Pb2þ uptake. J. Hazard Mater. 215e216, 17e24. Kim, S., Song, M.-H., Wei, W., Yun, Y.-S., 2015. Selective biosorption behavior of Escherichia coli biomass toward Pd(II) in Pt(IV)ePd(II) binary solution. J. Hazard Mater. 283, 657e662. Kumar, A.S.K., Sharma, S., Reddy, R.S., Barathi, M., Rajesh, N., 2015. Comprehending the interaction between chitosan and ionic liquid for the adsorption of palla- dium. Int. J. Biol. Macromol. 72, 633e639. Langmuir, I., 1916. The constitution and fundamental properties of solids and liq- uids: Part I solids. J. Am. Chem. Soc. 38, 2221e2295. Lin, S., Wei, W., Wu, X., Zhou, T., Mao, J., Yun, Y.-S., 2015. Selective recovery of Pd(II) from extremely acidic solution using ion-imprinted chitosan fiber: adsorption performance and mechanisms. J. Hazard Mater. 299, 10e17. Lin, S., Zhao, Y., Bediako, J.K., Cho, C.-W., Sarkar, A.K., Lim, C.-R., Yun, Y.-S., 2019. Structure-controlled recovery of palladium(II) from acidic aqueous solution using metal-organic frameworks of MOF-802, UiO-66 and MOF-808. Chem. Eng. J. 362, 280e286. Mao, J., Lin, S., Lu, X.J., Wu, X.H., Zhou, T., Yun, Y.-S., 2020. Ion-imprinted chitosan fiber for recovery of Pd(II): obtaining high selectivity through selective adsorption and two-step desorption. Environ. Res. 182, 108995. Ming, G., Duan, H., Meng, X., Sun, G., Sun, W., Liu, Y., Lucia, L., 2016. A novel fabrication of monodisperse melamineeformaldehyde resin microspheres to adsorb lead(II). Chem. Eng. J. 288, 745e757. Mircescu, N.E., Oltean, M., Chiş, V., Leopold, N., 2012. FTIR, FT-Raman, SERS and DFT10study on melamine. Vib. Spectrosc. 62, 165e171. Monier, M., Abdel-Latif, D.A., Abou El-Reash, Y.G., 2016. Ion-imprinted modified chitosan resin for selective removal of Pd(II) ions. J. Colloid Interface Sci. 469, 344e354. Morisada, S., Kim, Y.-H., Ogata, T., Marutani, Y., Nakano, Y., 2011. Improved adsorption behaviors of amine-modified tannin gel for palladium and platinum ions in acidic chloride solutions. Ind. Eng. Chem. Res. 50, 1875e1880. Morisada, S., Kim, Y.-H., Yakuwa, S., Ogata, T., Nakano, Y., 2012. Improved adsorption and separation of palladium(II) and platinum(IV) in strong hydrochloric acid solutions using thiocyanate-retaining tannin gel. J. Appl. Polym. Sci. 126, E478eE483. Pangeni, B., Paudyal, H., Abe, M., Inoue, K., Kawakita, H., Ohto, K., Adhikari, B.B., Alam, S., 2012. Selective recovery of gold using some cross-linked poly- saccharide gels. Green Chem. 14, 1917e1927. Pearson, G.R., 1963. Hard and soft acids and bases. J. Am. Chem. Soc. 85, 3533e3539. Ramesh, A., Hasegawa, H., Sugimoto, W., Maki, T., Ueda, K., 2008. Adsorption of gold(III), platinum(IV) and palladium(II) onto glycine modified crosslinked chitosan resin. Bioresour. Technol. 99, 3801e3809. Sharma, S., Rajesh, N., 2014. 2-Mercaptobenzothiazole impregnated cellulose pre- pared by ultrasonication for the effective adsorption of precious metal palla- dium. Chem. Eng. J. 241, 112e121. Song, M.-H., Kim, J.A., Wei, W., Kim, S., Yun, Y.-S., 2018. Polyethylenimine-coated biomass-chitosan composite fibers for recovery of ruthenium from industrial effluents: effects of chitosan molecular weight and drying method. Hydro- metallurgy 182, 114e120. Wang, F.C., Zhao, J.M., Wei, X.T., Huo, F., Li, W.S., Liu, H.z., 2014. Adsorption of rare earths (III) by calcium alginateepoly glutamic acid hybrid gels. J. Chem. Technol. Biotechnol. 89, 969e977. Wang, L., Wang, K.M., Huang, R.H., Qin, Y.C., Tong, S.S., 2020. Hierarchically flower- like WS2 microcrystals for capture and recovery of Au (III), Ag (I) and Pd (II). Chemosphere 252, 126578. Wei, W., Kim, S., Song, M.H., Bediako, J.K., Yun, Y.S., 2015. Carboxymethyl cellulose fiber as a fast binding and biodegradable adsorbent of heavy metals. J. Taiwan Inst. Chem. E. 57, 104e110. Wei, W., Cho, C.-W., Kim, S., Song, M.-H., Bediako, J.K., Yun, Y.-S., 2016a. Selective recovery of Au(III), Pt(IV), and Pd(II) from aqueous solutions by liquideliquid extraction using ionic liquid Aliquat-336. J. Mol. Liq. 216, 18e24. Wei, W., Reddy, D.H.K., Bediako, J.K., Yun, Y.-S., 2016b. Aliquat-336-impregnated alginate capsule as a green sorbent for selective recovery of gold from metal mixtures. Chem. Eng. J. 289, 413e422. Won, S.W., Yun, Y.-S., 2013. Recovery of metallic palladium from hydrochloric acid solutions by a combined method of adsorption and incineration. Chem. Eng. J. 218, 303e308. Won, S.W., Kotte, P., Wei, W., Lim, A., Yun, Y.-S., 2014. Biosorbents for recovery of precious metals. Bioresour. Technol. 160, 203e212. Xi, C.Y., Wang, R.K., Rao, P.H., Zhang, W.Q., Yan, L.L., Li, G.H., Chai, F., Cai, Y.Y., Luo, T.T., Zhou, X.Y., 2020. The fabrication and arsenic removal performance of cellulose nanocrystal-containing absorbents based on the “bridge joint” effect of iron ions. Carbohydr. Polym. 237, 116129. Yang, J., Zhang, H., Yu, M., Emmanuelawati, I., Zou, J., Yuan, Z., Yu, C., 2014. High- content, well-dispersed g-Fe2O3 nanoparticles encapsulated in macroporous silica with superior arsenic removal performance. Adv. Funct. Mater. 24, 1354e1363. Yi, Q., Fan, R., Xie, F., Zhang, Q., Luo, Z., 2016. Recovery of Palladium(II) from nitric acid medium using a natural resin prepared from persimmon dropped fruits residues. J. Taiwan Inst. Chem. E. 61, 299e305. Zheng, H., De, Z., Wang, W.Y., Fan, Y.Q., Li, J., Han, H.P., 2007. Highly selective determination of palladium(II) after preconcentration using Pd(II)-imprinted functionalized silica gel sorbent prepared by a surface imprinting technique. Microchim. Acta 157, 7e11. Zhou, L., Liu, J., Liu, Z., 2009. Adsorption of platinum(IV) and palladium(II) from aqueous solution by thiourea-modified chitosan microspheres. J. Hazard Mater. 172, 439e446. Zhou, L., Xu, J., Liang, X., Liu, Z., 2010. Adsorption of platinum(IV) and palladium(II) from aqueous solution by magnetic cross-linking chitosan nanoparticles modified with ethylenediamine. J. Hazard Mater. 182, 518e524. Zhou, Z., Liu, F., Huang, Y., Wang, Z., Li, G., 2015. Biosorption of palladium(II) from aqueous solution by grafting chitosan on persimmon tannin extract. Int. J. Biol. Macromol. 77, 336e343.