Study on adsorption behavior of rare earth elements onto magnetic nanocomposites of carboxymethyl chitosan, alginate and novel biodegradable polyamide

Department of Chemical Engineering Study on adsorption behavior of rare earth elements onto magnetic nanocomposites of carboxymethyl chitosan, alginate and novel biodegradable polyamide by Hamedreza Javadian in fulfilment for the award of the degree of DOCTOR by Universitat Politècnica de Catalunya Supervised by: Dr. Ana Maria Sastre Requena Dr. Montserrat Ruiz Planas Barcelona, 2020 Thesis presented by compendium of publications ACKNOWLEDGEMENT I would like to express my sincere gratitude and heartfelt thanks to my supervisor Prof. Ana Maria Sastre Requena for giving me the opportunity to do the research work in Department of Chemical Engineering, Universitat Politècnica de Catalunya, and her sincere guidance and efforts towards the successful completion of the research work, apart from being a great source of support and motivation for me. My heartfelt thanks go to Prof. Montserrat Ruiz for her constant help and providing necessary laboratory facilities to perform the research work. Profound thanks for her kind heart and great constant support to solve my private problems outside the research work. I express my sincere thanks to Prof. Mehdi Taghavi, Department of Chemistry, Faculty of Science, Shahid Chamran University of Ahvaz, 61357-43337, Iran, for his interest to contribute in the research work by his continuous support and valuable suggestions. Last but not the least; I would like to dedicate my thesis to my beloved parents for their constant inspiration and encouragement not only during the research work but also at all the moments of my life. The thesis is developed in the frame of the project entitled “Separacion/recuperacion de tierras raras mediante procesos de sorcion en biopolimeros, composites y membranas. (ReRarE)” (Ref.: CTM2014-52770-R), “Estrategias de reciclado de residuos que contienen tierras raras: procesos de sorción mediante nanocomposites magnéticos y membranas liquidas para su separación y recuperación”(ReciRar) (Ref.: CTM2017-83581-R) and the contract 2015 BES-2015-072506 financed by the Spanish Ministry of Industry and Competitiveness. LIST OF ACRONYMS CA Calcium alginate CCD Central composite design CMC Carboxymethyl chitosan CTAB Cetyltrimethylammonium bromide DMF N,N-dimethylformamide DMSO Dimethyl sulfoxide DW Deionized water EDX Energy-dispersive X-ray spectroscopy FE-SEM Field emission scanning electron microscopy FT-IR Fourier transform infrared HREEs Heavy rare earth elements ICP Inductively coupled plasma ILs Ionic liquids LREEs Light rare earth elements MREEs Middle rare earth elements NMR Nuclear magnetic resonance P(PTA) Poly(pyrimidine-thiophene-amide) PAs Polyamides REEs Rare earth elements REOs Rare earth oxides RSM Response surface methodology TGA Thermogravimetric analysis TMAPD 5,5'-(thiophene-2-ylmethylene)bis(2-aminopyrimidine-4,6diol) TPP Triphenyl phosphite US EPA United States Environmental Protection Agency VSM Vibrating sample magnetometer XRD X-ray diffraction INDEX Page SUMMARY………………………………………………………………………………………2 CHAPTER I: INTRODUCTION………………………………………………………….……4 Importance of REEs………………………………………………………………………...…4 1.1. What are rare earth elements?……………………..……………………………………...4 1.2. Chemical characteristics and solution chemistry of REEs……………………………….9 1.3. Refinement and production of REEs…………………………………………………....12 1.4. Health impacts and environmental effects of REE exposure..........................................14 1.5. Applications of REEs ………………………………………...…..….…….….………..16 1.6. Critical issues related to REEs………………………………….....................................19 1.7. Separation and recovery of REEs……………………………………………………….25 1.8. Separation by adsorption…………………………………………………..……..…..…25 1.8.1. Adsorption by metal oxides………………………….……………………..…..26 1.8.2. Nanotechnology and nano metal oxides…………………………………………………………………………...26 1.8.3. Biopolymers……………………………………………………………...…......27 1.8.4. Synthetic metal chelating polymers……………….………….…………….......32 1.8.5. Synthetic biodegradable polyamide containing chelating groups…………..….32 1.8.6. Nanotechnology and nanocomposites…………………………………….…....33 1.8.6.1. Polymer/metal oxide nanocomposites………………………….….......34 1.8.6.2. Nanotechnology and usage of magnetic separation…………………....34 CHAPTER II: SCOPE OF THE WORK ………………………….………………………....36 2.1. Scope of the work………………………………………………………….……………36 2.2. Objectives ………………………………………………………………………………36 2.2.1. Syntheses and characterization of the magnetic nanocomposites……………….36 2.2.2. Adsorption studies of Nd+3, Tb+3, and Dy+3 REEs from synthetic aqueous solutions………………………………………………………………………..37 CHAPTER III: METHODOLOGY…………………………………………...………..….…38 3.1. Materials and reagents……………………………………………………...………...…38 3.2. Synthesis of the P(PTA)………………………………………………………………...38 4 CHAPTER I INTRODUCTION Importance of REEs In recent decades, the application of rare-earth elements (REEs) has become apparent in numerous technological sectors. The relevance of the REEs was increasing to reach a peak in recent years since these metals play an important role in the emerging clean technologies because of their excellent electronic optic, magnetic and catalytic properties. Currently, there are about 132 million tons of REEs reserves around the world (Gambogi, 2016). Considering the mine production, China is nowadays known as the biggest REEs producer while its mine production rate is about 79%. In the following, Australia is known as the second biggest producer with a mine production rate of 15%. Russia, Brazil, and India with mine production rates of 2.2%, 1.5%, and 1.1% are respectively known as the other countries that have the largest amounts of REEs production (Gambogi, 2016). Since China has reduced its amount of export, the other countries in the world are confronting with a risk of supplying their required REEs. Due to the gap in supplying the required amounts of these elements and fulfilling their increasing demands, as well as the increasing pollution of REEs, it is not only quite necessary to recycle the REEs from secondary sources but also to recover them from waste streams. 1.1. What are Rare Earth Elements? Lanthanides together with scandium and yttrium in the periodic table are known as REEs that are usually assorted into three classes, excluding promethium and scandium. These three classes are as follows: (1) lanthanum (La-57), cerium (Ce-58), praseodymium (Pr-59) and neodymium (Nd-60) as Light Rare Earth Elements (LREEs), (2) samarium (Sm-62), europium (Eu-63) and gadolinium (Gd-64) as middle rare earth elements (MREEs) and (3) the rest of lanthanides and yttrium as heavy rare earth elements (HREEs) (Yanfei et al., 2016). There is another classification that divides REEs to LREEs and HREEs as shown in Fig. 1.1. 5 . Fig. 1.1. Rare earth elements division as LREEs and HREEs (Schuler et al., 2011). Gadolinium and dysprosium are sometimes classified as medium-weight lanthanides because of their physicochemical attributes. Notably, the word “rare” has come from metallurgical chemists around the 1940s (Gupta and Krishnamurthy, 2004). Due to the fact that most of them are commonly sold as oxide compounds, they are also known as "rare earth oxides”. However, it can be claimed that they are hardly found in sufficient abundance in a single place to be economically feasible for mining (Chakhmouradian and Wall, 2012) since most of these elements are not rare with regards to the general amount of these elements in the earth's crust while their levels in the earth’s crust are generally equal to or more than some physiologically important elements like platinum, cobalt, gold, silver, and selenium (Brzyska 1996). Fig. 1.2 indicates the worldwide distribution of REEs while Table 1.1 presents the REEs abundance in the earth’s crust relative to other ordinary metals. Noteworthy, although these abundances from Wedephol (1995) that are presented here are only one of the several interpretations, they can be considered as a general representative. As can be seen, the lanthanides content relative to other REEs in rock-forming minerals is not anywise rare. Moreover, as depicted in Table 1.1, cerium (60 mg/kg), lanthanum (30 mg/kg), neodymium (27 mg/kg), yttrium (24 mg/kg) and scandium (16 mg/kg) can be considered 6 as the most common ones. Lutetium (0.4 mg/kg) and thulium (0.3 mg/kg) are respectively the rarest elements while the concentrations of the remainders are in the range of 0.7 to 6.7 mg/kg. Fig. 1.2. (A) Deposits of REEs in the Greenland and Americas and (B) Deposits of REEs in the rest of the world (Ganguli and Cook, 2018). 7 Table 1.1. The abundance of elements in the Earth’s crust (Wedepohl, 1995) (Bold: Lanthanides, scandium, and yttrium). Elements Abundance (parts per million) Nickel (28Ni) 90 Zinc (30Zn) 79 Copper (29Cu) 68 Cerium (58Ce) 60 Lanthanum (57La) 30 Cobalt (27Co) 30 Neodymium (60Nd) 27 Yttrium (39Y) 24 Scandium (21Sc) 16 Lead (82Pb) 10 Praseodymium (59Pr) 6.7 Thorium (90Th) 6 Samarium (62Sm) 5.3 Gadolinium (64Gd) 4 Dysprosium (66Dy) 3.8 Tin (50Tn) 2.2 Erbium (68Er) 2.1 Ytterbium (70Yb) 2 Europium (63Eu) 1.3 Holmium (67Ho) 0.8 Terbium (65Tb) 0.7 Lutetium (71Lu) 0.4 Thulium (69Tm) 0.3 Silver (47Ag) 0.08 Gold (79Au) 0.0031 Promethium (61Pm) 10-18 Based on the ordinary pattern of the periodic table, it can be stated that the lanthanides with even atomic numbers are generally more typical in nature. Moreover, a pattern in the occurrence and crustal abundance of some lanthanides has been observed by geochemists (McLeod and Shaulis, 2018). It is worth mentioning that lanthanides having lower atomic numbers were not only known as typical ionic constituents in REEs mineral ores but also generally happened in more considerable abundance compared to the lanthanide elements having more atomic numbers (Dostal, 2017). 8 REEs were named “rare” due to the fact that just after they were discovered, it was thought that only small amount of them was present in the Earth’s crust and the term “earths” also refers to the fact that their oxides have an earthy appearance. These metals are generally found together in geologic deposits because of having many identical attributes that are used in an extensive range of applications. For instance, magnets that are made using REEs are not only much more powerful and weigh less but also smaller compared to standard magnets. Moreover, some REEs can withstand severe heat as well as giving off intense white light when heated besides having great electrical conductivity. The REEs elemental forms, extracted from mineral ores as oxides (i.e., REOs), are iron-gray to silvery lustrous metals that are normally malleable, ductile soft and generally reactive, particularly when they are finely divided or at elevated temperatures (Hedrick, 2004). As they are not naturally present as pure elements, refinement processes are needed to separate these elements from ores. However, more than 200 REEsbearing minerals are known. Phosphates (xenotime, monazite, rhabdophane, ningyoite, and florencite) followed by carbonates (synchysite, bastnasite, parasite, and lanthanite) are known as the most common rare elements that contain ores (Oliveira and Inverno, 2014). The main REEs mineral ores that are most practical for the REEs extraction are xenotime monazite and bastnasite according to the following description: Bastnasite, which is the most abundant one compared to the other three REEs mineral ores, is a carbonate mineral that is mostly found in enriched LREEs (like lanthanum, yttrium, and cerium). Bastnasite is mainly found in pegmatites, vein deposits, and contact metamorphic zones and forms in rocks of carbonate-silicate associated to alkaline intrusions (Gupta and Krishnamurthy, 2004). Although LREEs cerium, neodymium, and lanthanum are generally used to enrich monazite, they can also include HREEs, especially yttrium (Ni et al., 1995). The LREEs predominance is because of the lower pressures and crystallization temperature of this mineral while it also has more HREEs compared to the bastnasite ore deposits. It must be noted that it generally happens in metamorphic rocks, acidic igneous rocks (mainly pegmatites) and some vein deposits. Monazite is not only resistant to weathering but also happens in several placer deposits while the host rocks are eroded. Thorium can be also related to monazite in different amounts (Chen et al., 2017). While xenotime is crystallized under higher pressures and temperatures compared to monazite, its crystalline structure can readily accommodate a higher ratio of HREEs (terbium through yttrium and lutetium) compared to the one that is generally found in monazite. It is 9 basically a yttrium phosphate mineral and happens as a minor element of gneissic and granitic rocks. In spite of the fact that it is not always present in remarkable quantities, thorium and uranium can also happen as elements of xenotime (Peiró and Méndez, 2013). Moreover, it has been seen that light rare earths (generally with coordination numbers of 8 to 10) concentrate on phosphates and carbonates. On the other hand, heavy rare earths (generally with coordination numbers of 6 to 8) concentrate on oxides and phosphates (Evans 1997). 1.2. Chemical characteristics and solution chemistry of REEs In order to illustrate the resemblance of the rare earth metals, it is necessary to consider the structure of their atoms. In this regard, all of them have three electrons in their outermost shell while their chemical behavior is also determined using these electrons. Their difference refers to their inner shell -4f, with the systematic filling of f orbital, as well as 5d, 6s and 6p are empty. In spite of cerium that can exist as Ce+4 and Eu+2, happening in both the trivalent and divalent states, all lanthanides happen as Ln+3 in aquatic systems. In contrary to the divalent ions, trivalent ones are enough stable (Topp, 1965). These ions are specified using large ionic radii which means that substitution reactions not only require large cations, like strontium or calcium jointly with their high valence but also tend to be separated from other trivalent ions. Furthermore, the ionic radii of lanthanides reduces from La+3 to Lu+3. These ions are fundamentally spherical and able to form complexes that are similar to alkaline and alkaline earth ions besides being very electropositive. Therefore, their bonding attributes are mostly ionic (Henderson, 1996). The ions have low polarizability due to their high z/r ratio. They are usually found in solids with coordination number 8, while the smaller Sc+3 is found in coordination number of 6. However, Ln(H2O)63+ is the most common lanthanide which tends to bind the water molecules (Brookins, 1989). Although the hydrolysis of these ions is slight, their hydration is increased with the atomic number. In this regard, the hydrated ion size is increased from La to Lu while hydrolysis below pH 5 seems to be insignificant. Consequently, the lanthanide species are not easily hydrolyzed. The Ln(OH)2+, Ln(OH)3 and Ln(OH)4stepwise formational constants are significantly decreased while Ln+3 may be the dominant form in the water at acidic to neutral pH media. Several complexes are important and pH-specific, depending on the ionic media and pH range. MEDUSA software is applied to understand the speciation of metals in dilute solutions 10 (Puigdomenech, 2000). For instance, in Fig. 1.3A, when the solution contains multi-metals in the presence of H2SO4, at the pH near 6, Tb3+ and TbSO4+ are the important ionic species while in the case of Dy3+ and Nd3+, the formation ranges of ionic species are different depending on the type of REE. By increasing pH of the solution from around 6 for Tb3+ and Dy3+ and around 7 for Nd3+, Tb(OH)3 (s), Dy(OH)3 (s) and Nd(OH)3 (s) are formed and their amounts are increased by increasing pH of the solution up to 12. As depicted in Figs. 1.3B and 3C, when the concentration of SO42increases in the solution, Tb3+ is changed to TbSO4+, Tb(SO4)2and Tb(OH)3 depending on the pH of the solution. Similar results are seen for Tb3+ and Dy3+ which demonstrate that the concentration of reagent can affect the types of complexes, their formation ranges, and fractions in the solution. In Figs. 1.3D-F, by changing the reagent to HNO3, different conditions are obtained when other conditions are constant that indicates the type of reagent has an important effect on the type of ions in the solution. It can be concluded that the type of reagent and its concentration is important as it can affect on the complex types and their formation ranges according to the pH value, also their fractions in the solution. It can be observed from the Figs. 1.3A-F that pH ≤ 5.5 and low concentration of NO3prevent the formation of complexes. The importance of the halide complexes of the lanthanides is minor, even at low pH. Generally, at basic pH, hydrolysis becomes more significant for trivalent ions, Ln(OH)3 and Ln(OH)2+ (Topp, 1965). In the terminology of Pearson, the lanthanides are known as hard acids and can preferentially bond with hard bases containing oxygen as donor atoms. The major ligands generally include at least one donor oxygen atom, and kinetically, rare earths react quickly to form complexes. Table 1.2 contains some characteristics of REEs. 11 Fig. 1.3. Species of Nd3+, Tb3+, and Dy3+ (A) in the concentration of 1 mM SO42-, (B) in the concentration of 10 mM SO42-, (C) in the concentration of 100 mM SO42-, (D) in the concentration of 1 mM NO3-, (E) in the concentration of 10 mM NO3-, and (F) in the concentration of 100 mM NO3-. 12 Table 1.2. Some characteristics of REEs. Symbol Element Atomic number Molecular weight Ionic radii (pm) Coordination number=6 Electronic configuration Y Yttrium 39 88.91 104 [Kr]4d15s2 La Lanthanum 57 138.9 117.2 [Xe]5d16s2 Ce Cerium 58 140.1 115 [Xe]4f26s2 Pr Praseodymium 59 140.9 113 [Xe]4f36s2 Nd Neodymium 60 144.2 112.3 [Xe]4f46s2 Sm Samarium 61 147 109.8 [Xe]4f56s2 Pm Promethium 62 150.4 111 [Xe]4f66s2 Eu Europium 63 152 108.7 [Xe]4f76s2 Gd Gadolinium 64 157.3 107.8 [Xe]4f7d16s2 Tb Terbium 65 158.9 106.3 [Xe]4f96s2 Dy Dysprosium 66 162.5 105.2 [Xe]4f106s2 Ho Holmium 67 164.9 104.1 [Xe]4f116s2 Er Erbium 68 167.3 103 [Xe]4f126s2 Tm Thulium 69 168.9 102 [Xe]4f136s2 Yb Ytterbium 70 173 100.8 [Xe]4f146s2 Lu Lutetium 71 175 100.1 [Xe]4f145d16s2 The ionic radii for Ce4+ and Eu2+ is 101 and 131 pm, respectively. 1.3. Refinement and production of REEs REEs are separated as recognizable elements due to the physical similarities in the atomic radius and charge between them. It is worth mentioning that before recent advances that made separation economically practical, no common technological use was found for REEs (Hatch, 2012). Due to the fact that these REEs are low in abundance in rock deposits, their separation is very difficult. Therefore, it can be concluded that besides their low abundance in many rock deposits, REEs were not able to find common use in technologies before recent developments that made separation economically practical. Their similarity signifies that although they can make good substitutes for one another while their applications differ considerably for the overall group (Koerth-Baker, 2012). Generally, production of REEs includes some steps as follows: extraction of REEs that contain mineral, milling, flotation, purification and subsequent processing of the ore (Schüler et al., 2011). Moreover, REEs are principally available as oxidic compounds and the resources are mostly represented as REOs because of their strong affinity with oxygen. Processing REOs into 13 utilizable products is not only a very complicated procedure but also significantly different between deposits. The main factors that affect the selection of treatment processes are mentioned as following (Ferron et al., 1991):  Nature and type of the deposits (like vein type, beach sand, complex ores and igneous) and their complexity.  Nature and type of other precious minerals that are available with REOs.  Nature and type of gangue minerals that are available in the deposit (like clay, slimes and soluble gangue).  Composition and type of the exclusive REO minerals.  The accessability of the process in environmental and social terms. Different chemical techniques and sometimes thousands of steps are required to occur refinement via physical separation of the REEs. Unluckily, all REEs and their particular ores are various and therefore need various chemical methods for refining (according to the vapor pressure and melting point along with other physical attributes of the element) (Tiesman, 2010). Generally, separation and concentration from the host material in alkaline or acidic solutions, the individual REOs reduction into pure metals, and REO separation utilizing ion exchange or solvent extraction are known as the fundamental steps in REOs processing (Gupta and Krishnamurthy, 2004). The primary step commonly contains grinding and crushing where ore is reduced to fine particles and REO is separated using different methods like magnetic, flotation or gravimetric separation. The percentage of REOs in the working material is dramatically increased along with the separation process. The aim of the following steps in the process is to change the concentrated mineral into a more precious chemical that is formed via different chemical and thermal reactions. The mineral concentrates are typically separated into utilizable oxides by employing hydrometallurgy methods (like precipitation, leaching, and extraction). Moreover, the oxides or metal mixtures can be refined into high-purity rare earth metals using methods like the metallothermic reduction for further processing (Suli et al., 2017). Hydrometallurgy is the most typical chemical extraction technique that is used for the separation of individual REOs from the mineral concentrate. Basicity variations between the different rare earths affect the hydrolysis of ions, the solvability of their salts and the creation of complicated species (Gupta and Krishnamurthy, 2004). In this regard, fractional crystallization, 20 environment. Flotation involves chemical beneficiation in ponds, known as tailings. The reminders, like radioactive thorium or uranium other chemicals, are left in the wastewater. Considering the fact that this water is exposed to disruptions or natural environmental conditions, it can pose critical risks of environmental pollution (Schüler et al., 2011). Moreover, although purification is costly and energy-intensive, it is crucial due to the purity of 99% that is almost required (Hatch, 2012). The problem in the production step of REEs is not the only one. Although REEs are extensively utilized in plenty of applications around the world, supplying them is considerably limited to only a few large mining districts (Chakhmouradlan and Wall, 2012). In spite of the fact that REEs occur around the world, the largest mining fields of ores, such as monazite, bastnäsite, and xenotime having essential REEs, are located in the Asian states (Manchen et al., 2019) and among them, China is changed to the first country in mining and extracting REEs due to their most abundant REE deposits. As can be seen in Table 1.7, more than 90 % of mine production of rare earth occurs in China, although this country has less than 40 % of the identified deposits. While China's domestic demand has remarkably increased, it tightened its REEs export quota from 50145 tonnes to only 31130 tonnes from 2009 to 2012 which led to serious problems for REEs users that were outside of China (It was proved by the rare-earth crisis in 2011 with the record of the highest prices) (Binnemans et al., 2013). Table 1.7. World rare earth element production and rare earth stores estimation, 2012 (Lucas et al., 2015). Country Mine Production of REEs (Kilotonnes) Deposits of Rare Earth Elements (Megatonnes) China 75 44 USA 1 13 Australia (mined and concentrated in Australia, extracted in Malaysia) 1 1 India 3 2 Other 33 In 2018, the total world REE deposits were estimated to be around 120 million tons (Table 1.8) (Manchen et al., 2019) which seems to be sufficient for global requirements over hundreds of years (Zhou et al., 2017). However, the REEs world demand was about 142 thousand tons in 2018 (Manchen et al., 2019). 21 Table 1.8. World mine production and reserves of REEs* (U.S. Geological Survey, 2019). Country Mine production Reserves 2017 2018 United States - 15000 1400000 Australia 19000 20000 3400000 Brazil 1700 1000 22000000 Burma (Myanmar) Not available 5000 Not available Burundi - 1000 Not available China 105000 120000 44000000 India 1800 1800 6900000 Malaysia 180 200 30000 Russia 2600 2600 12000000 Thailand 1300 1000 Not available Vietnam 200 400 22000000 Other countries - - 4400000 World total (rounded) 132000 170000 120000000 *Data in metric tons of REEs oxide It is worth mentioning that not only China has the most extensive deposits of REEs but also it has the most complicated processing technologies and facilities for producing the rare earth metals. Each year, China can produce around 120,000 tons of rare earth while the total world production is about 170,000 tons (Manchen et al., 2019, Zhou et al., 2016). Although mining in other countries (like the USA, India, and Australia) is growing, its process is very slow. The peak of all rare earths prices was in 2011 and after that their prices have fallen down which had a discouraging impact on new mining projects (Gambogi, 2013). This situation has also motivated other countries like Japan and most EU Member States that do not have any kind of main rare-earth reserves on their territory to search for secondary and alternative rare earth resources for developing their own rare-earth industry to acquire a source of both heavy and light rare earths (Binnemans et al., 2013). Using natural resources is essential in a sustainable and circular economy. This can only happen by recycling and reusing materials from end-of-life consumer goods. REEs are considered as the most crucial raw materials group with the highest supply risk by the European Commission (European Commission Critical raw materials for the EU). To overcome the challenge of supplying the REEs, a threefold method can be suggested. The first strategy is to substitute crucial rare earths by less crucial metals. Secondly, the risk of supplying REEs can be reduced by investing in sustainable fundamental mining from new or old REEs 22 reserves. Nowadays, mining companies are now extensively looking for new usable rare earth reserves and old mines are being opened again (Humphries, 2012). Prior to the REEs mining boom in China, the global market was dominated by the US. The operations were started by Mountain Pass in California in 1965 (Fig. 1.6) and it was known as the main producer around the world for decades (Barakos, 2017). Notably, because of the contest in China as well as in response to environmental matters in the surrounding area of Mountain Pass, mining activities stopped in 1998 (Mancheri, 2015). The production was restarted in 2012 due to the REEs supply risk. Nonetheless, most of the countries have to invest in technospheric mining due to the absence of operational and/or economic primary reserves on their territory (Johansson et al., 2013). Fig. 1.6. Molycorp Mountain Pass rare earth facility in California's Mojave Desert. Generally, technospheric mining can have many forms. With consideration of crucial metal-having streams, such mining contains (1) direct recycling of pre-consumer manufacturing REEs residues/scrap; (2) urban mining of post-consumer (usually complicated multi-material) End-of-Life products; (3) landfill mining of historic (and future) urban and industrial waste residues having REEs. The focus of urban mining and direct recycling is on resources with very great content of rare earths. However, the total accessible volumes for recycling are relatively low. However, base metals like iron, copper, and aluminum along with valuable metals (silver, gold, and platinum-group metals) achieved high recycling rates. In spite of the fact that there is an extensive literature dealing with (mostly lab-scale) research attempts on recycling REEs, only 23 less than 1% of them were being recycled in 2011. This is generally because of ineffective collection, lack of incentives, and technological difficulties (Binnemans et al., 2013). Generally, neodymium, dysprosium, europium, yttrium, and terbium (Fig. 1.7) are known as the five most critical REEs based on the medium-term criticality matrix of the U.S. Department of Energy (DOE), and their essential applications are in green energy as follows (Binnemans et al., 2013; US Department of Energy, 2011): • Dysprosium and neodymium are utilized in the permanent magnets manufacturing which are employed in wind turbines and many other products like speakers, hard disk drives, and headphones. • Yttrium, europium, and terbium are utilized besides cerium, lanthanum, and gadolinium in phosphors in low-energy fluorescent lamps. • Yttrium, neodymium, lanthanum, cerium, and praseodymium are utilized in the manufacturing of nickel metal hydride (NiMH) batteries for hybrid vehicles. In addition, their hazards are mentioned in Table 1.9. Consequently, recycling REEs from end-user products, such as fluorescent lamps and magnets that present over 70% of the rare-earth market in terms of value (32% for lamp phosphors; 38% for magnets), can provide the opportunity to maintain the supply of these crucial elements and decrease the dependency of UE from other countries. In this context, there is a necessity to develop advanced separation processes for recovery of Nd, Dy, and Tb which are three of the most critical elements. 24 Fig. 1.7. DOE medium term (2015-2025) criticality matrix, representing the five most crucial rare-earth elements (Nd, Y, Tb, Eu, Dy). Table 1.9. Significant uses of Dy, Nd, and Tb and their toxicological information (Rim et al., 2013). Element Toxicological information Dy Soluble Dy salts, like dysprosium nitrate and chloride, are mildly toxic when ingested. However, the insoluble salts are non-toxic. According to the toxicity of dysprosium chloride to mice, it is determined that the ingestion of 500 g or more might be fatal to a human. Nd Nd compounds are of low to moderate toxicity. Nonetheless, its toxicity has not been explored thoroughly. Neodymium salts are very irritating to the mucous membranes and eyes, and moderately irritating to the skin. Tb It may cause serious irritation to the skin and eyes. 25 1.7. Separation and recovery of REEs Individual REEs purification has obtained remarkable attention in recent years due to the growing requirements for high-purity REEs and their compounds (Anastopoulos et al., 2016). Several techniques are used for the separation and recovery of metal ions from aqueous solution, like chemical precipitation, membrane technology, and extraction of solvent. Among the different separation techniques, solvent extraction is extensively used for aqueous solutions treatment containing metal ions. However, it is costly on a large scale and yields to extensive environmental problems due to the toxic organic diluents and modifiers that are widely used (Yadav et al., 2015). The purification and separation of REEs using solvent extraction need the treatment of a large volume of dangerous volatile organic compound solvents (Florek et al., 2014). Since flammable volatile organic compound solvents are concerned, solvent extraction seems to be dangerous. Chemical precipitation also needs an extensive amount of chemicals to decrease metals to a satisfactory discharge level. Extreme sludge production is known as its other drawbacks which needs more treatment, poor settling, slow precipitation of metal, metal precipitates aggregation, and the long-term environmental effects of sludge disposal (Aziz et al., 2008). High energy consumption related to the high operation pressures, the costs associated with fouling problems, and replacement of membranes can be considered as the main drawbacks of membrane technology. 1.8. Separation by adsorption Adsorption can be considered highly effective, especially because of dilute solutions, cheap and easy techniques and being able to start environmentally without utilizing any organic solvents for recovery of metal ions from aqueous solution. Therefore, it can be considered as a competitive alternative to solvent extraction. Some adsorbents have been used for rare earth metals adsorption, such as Sargassum sp (Oliveira et al, 2011) activated carbon (Murty et al., 1996), titanium dioxide (Liang et al., 2001), cellulose (Zhu et al., 2015), b-cyclodextrin (Zhao et al., 2016), silica (Esser et al., 1994), aminocarboxylic adsorbents (Grebneva et al., 1996), and oxidized multiwalled carbon nanotubes (Koochaki-Mohammadpour et al., 2014), SiO2/UF impregnated with organophosphorus extractant (Naser et al., 2015), activated biochars from cactus fibres (Hadjittofi, 2016), graphene oxide-corn zein composites (Xu et al., 2018) . 26 Due to some disadvantages that are available in the adsorbents which have been used for adsorption of REEs, such as low adsorption capacity, weak mechanical properties, poor chemical resistance, difficult adsorbent separation from the aqueous phase, and high-dose requirement of the adsorbent for complete metal ions removal, it is necessary to perform extensive investigations on the production of novel adsorbents to overcome such defects. The selection of adsorbent relies on the nature of the metal ion, as each type of metal may need a particular adsorbent. On the other hand, adsorption efficiency relies on physicochemical attributes, such as functional groups, porosity, particular surface area, and adsorbent particle size (Chen et al 2003). 1.8.1. Adsorption by metal oxides Generally, the high surface area and high specific affinity of adsorbent are the two key factors to determine the efficiency of metal ion separation from polluted water. Metal oxides are attractive candidates as adsorbents for metal ion separation and recovery (Rittmann et al., 2011). The high surface area of metal oxides may provide rich sites for metal ion adsorption which is normally benefited from porous structures. The high specific affinity is mainly due to the abundant surface hydroxyls on metal oxides (Qu, 2008). Some metal oxides have been used as an adsorbent for adsorption of REEs, such as Al2O3 (Marmier et al., 1997), amorphous silica (Marmier et al., 1999), and α-TiO2 (Ridley et al., 2005). The investigations showed that metal oxides are not only cheap and nontoxic but also chemically stable and environmentally friendly (Chu et al., 2009). Nonetheless, researchers all around the world have been trying to decrease the main disadvantage of metal oxides, such as not being easily dispersed in aqueous solution because of being microsized. Besides, by increasing particle size, the ratio of surface area to volume declines; therefore, the capacity of metal adsorption decreases. 1.8.2. Nanotechnology and nano metal oxides Nanostructured materials with dimensions (grain size, layer shapes or thickness below 100 nm) have yielded to the growing interest in nanotechnology (Hornyak et al., 2009). The eccentric attributes of nanomaterials are related to their nano-dimensions. In fact, nanomaterials are the materials that have at least one dimension which is below 100 nm (Chattopadhyay et al., 2009). 27 During the last two decades, a lot of attempts have been done around the world in both the theory and the empirical research of the development, specification and applications of inorganic nanostructures containing metal oxides, composites and ceramics and they have yielded to a mature and multidisciplinary field. Nanostructured materials are generally known because of their green chemistry, stability, and various technical applications (Bhushan et al., 2010). Amongst inorganic nanostructures, metal oxides nanostructures have received so much attention in recent years. They are applied in versatile applications, such as gas sensors, optical sensors, pressure sensors, electrochemical performance for energy storage, catalytic and photocatalytic, environmental application, etc. Separation of metal ions from polluted waters by the adsorption process is known as one of the most important environmental applications of metal oxides nanostructures that has been investigated by many researchers. Their major benefits include their ability to be simply dispersed in aqueous solution and the existence of a large number of their atoms, known as surface atoms, which not only have high adsorption capacities to many metal ion but also are unsaturated. In addition, the extremely small size of nanoparticles creates a large surface area in relation to their volume and makes them highly reactive in comparison to non-nanoforms of the same adsorbents (Srivastava et al., 2015). The adsorption behavior of nanometal oxides, such as nickel oxide (Saikrishna and Babu, 2015), aluminum oxide (Patra et al., 2012), titanium oxide (Jegadeesan et al., 2010), and zirconium oxide (Hristovski et al., 2008) have been proved by the researchers including the metals that were adsorbed by aluminum oxide, titanium oxide, and zirconium oxide from the papers mentioned above. The primary mechanisms for adsorption of REEs on nanostructured materials, like metal oxides, are surface complexation, precipitation, ion exchange, physical adsorption, and electrostatic attraction. 1.8.3. Biopolymers Biopolymers are polymers produced from biobased materials that are also biodegradable. Biopolymeric materials contain alginates, cellulose, proteins, lignins, carrageenan, chitosan, and chitin derivatives. The salient attribute of biopolymers is that they own a great amount of various functional groups like amines and hydroxyls that can enable metal ions to bind either by chemisorption or physisorption (Saravanan and Sudha, 2014). 28 Amongst the biopolymers investigated for metal ions adsorption, alginate and chitosan as natural carbohydrate biopolymers are industrially and scientifically attractive and have received so much attention as adsorbent and complexing agent by their functional groups that have a strong affinity to heavy metal ions. Due to the fact that natural materials, which are accessible in great quantities, or particular wastes obtained from manufacturing operations, are extensively available, environmentally friendly and practically unexploited resources, they may have a great potential to be utilized as low-cost adsorbents (Spinelli et al., 2004). Alginate Alginate, a natural polysaccharide, is the binary copolymer of (1,4) glycosidically linked α-D-mannuronic acid (M) and β-L-guluronic acid (G) (Fig. 1.8). The abundance of carboxylic and hydroxyl groups gives alginate strong chelating properties for metal ions (Fiset et al., 2008). It is negatively charged in aqueous solution at pH > 3.4 owing to the carboxyl groups available in both M and G subunits, where the carboxyl groups are deprotonated excluding at very low pH (Yu et al., 2013). Polyvalent cations can interact with blocks of M and G residues cooperatively in the gelation process to create ionic cross-links between various polymer chains (Fig. 1.9) that is known as “egg-box” model (Braccini and Pérez, 2001). The polyvalent cations, like calcium cations, can be substituted by ionic adsorbates. Because of their ability to form stable structures, cross-linked alginate has been utilized for the adsorption of heavy metals, like lead (Yakup Arıca et al., 2003), mercury (Yakup Arıca et al., 2004), manganese (Gotoh et al., 2004), and chromium (Ibáñez and Umetsu, 2004). Alginate adsorbents have been also used for the REEs adsorption, such as sodium alginate hydrogel cross-linked with poly-γ-glutamate (Xu et al., 2015), calcium alginate beads (Nayak, 2005), and alginate–poly glutamic acid hybrid gels (Wang et al., 2014). The primary binding mechanism of metal ions to calcium-alginate gel beads contains ion exchange and adsorption. Additionally, new hypotheses of cation binding and alginate crosslinking (Siew et al., 2005) have been explored in recent studies. Based on them, only one or two M and G blocks are involved in creating a binding site. Moreover, Rodrigues and Lagoa proposed that the number of binding sites and binding mechanisms relied on the accessibility of cation in the solution (Rodrigues and Lagoa, 2006). Results showed the successful usability of alginate adsorbents to separate REEs from aqueous solution. 29 Fig. 1.8. Guluronic acid (G) and Mannuronic acid (M) subunits in the chemical structure of alginate at pH>3.4 (Yu et al., 2013). Fig. 1.9. Gelation process by a divalent cation (Braccini and Pérez, 2001). Chitosan Chitosan, a cationic polysaccharide composed of N-acetyl glucosamine and glucosamine subunits (Fig. 1.10), is a potential biopolymer acquired cost-effectively by the derivation of chitin, which is a natural material extensively found in crustacean shells and is proved to have the best chelating properties among other natural polymers (Varma et al., 2004). It has also been extensively taken into consideration during the last decade for heavy metals adsorption from aqueous solution due to its unique properties, especially abundant amine (NH2) and hydroxyl (OH) groups. The amine groups in the structure of chitosan are generally considered primary active sites for adsorption of the metal ion. In addition to amine groups, hydroxyl groups may contribute to adsorb metal ion. Chitosan is positively charged at pH < 6.5 owing to the abundant availability of amine groups that are protonated at low pH (Yu et al., 2013). Moreover, chitosan 36 CHAPTER II SCOPE OF THE WORK 2.1. Scope of the work The objective of this project is to investigate the separation/recovery of the Tb+3, Dy+3, and Nd+3 from aqueous solution by synthesizing novel magnetic nanocomposites using two biopolymers or a biopolymer mixed with a synthetic chelating polymer having different functional groups, and magnetic nanoparticles. For these purposes, calcium alginate (CA), caboxymethyl chitosan (CMC), and a novel synthetic biodegradable poly(pyrimidine-thiopheneamide) (P(PTA)) will be used to produce new nanocomposites that are green, environmentalfriendly, and easy separable from aqueous media by an external magnetic field. Such magnetic polymeric nanocomposites are expected to have some advantages, such as easy separation from aqueous solution, favourable adsorption capacity because of having different functional groups, favourable chemical resistance, and less agglomeration in aqueous solution. The new magnetic nanocomposites will be studied for the separation/recovery of Nd+3, Tb+3, and Dy+3 selected among the REEs that are critical metals with a high risk of supply. 2.2. Objectives 2.2.1. Syntheses and characterization of the magnetic nanocomposites The magnetic nanoparticles (Ni0.2Zn0.2Fe2.6O4) nanocomposites will be synthesized by hydrothermal method to be used for the synthesis of the magnetic P(PTA), a novel synthetic biodegradable polyamide, will be synthesized by polycondensation of a diamine-phenol in 1,3-dipropyl imidazolium bromide ionic liquid as a solvent to avoid the use of the toxic triphenyl phosphite/N-methylpyrolidone/pyridine/LiCl that is required in the conventional direct polycondensation. The syntheses of the CA/CMC/Ni0.2Zn0.2Fe2.6O4, CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, and CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 magnetic nanocomposites will be carried out by gelation technique, and the P(PTA)/Ni0.2Zn0.2Fe2.6O4 will be synthesized by hydrothermal technique. To characterize the materials developed, various techniques will be used, including nuclear magnetic resonance (NMR), X-ray diffraction (XRD), field emission scanning electron 37 microscopy (FE-SEM), Energy-dispersive X-ray spectroscopy (EDX), thermogravimetric analysis (TGA), vibrating sample magnetometer (VSM), and Fourier transform infrared (FT-IR). 2.2.2. Adsorption studies of Nd+3, Tb+3, Dy+3 REEs from synthetic aqueous solutions In the adsorption studies of Nd+3, Tb+3, and Dy+3 by the synthesized nanocomposites, the influence of different parameters such as pH, adsorbent dose, contact time, initial concentration of Nd+3, Tb+3, and Dy+3, ionic strength on the adsorption efficiency will be investigated in single and ternary batch modes. The adsorption kinetic and isotherm models will be applied for fitting the experimental adsorption data. The effect of temperature on the adsorption process of Nd+3, Tb+3, and Dy+3, will be studied and thermodynamic parameters will be determined. In addition, to model the adsorption of the Nd+3, Tb+3, and Dy+3 in the ternary system, RSM-CCD will be applied. To evaluate the reusability of the synthesized nanocomposite, adsorption-desorption cycles will be performed. Finally, Nd+3, Tb+3, and Dy+3 adsorption in a fixed-bed column and fitting the experimental data by the models will be studied. 38 CHAPTER III METHODOLOGY 3.1. Materials and reagents Sodium alginate and CMC were respectively bought from China and PanReac AppliChem and Nantong Chem-Base Co. Dy(NO3)3.5H2O was purchased from Alfa Aesar. Nd(NO3)3.6H2O, Tb(NO3)3·6H2O, Zn(NO3)2.6H2O, Fe(NO3)3.9H2O, Ni(NO3)2·6H2O, glutaraldehyde, cetyltrimethylammonium bromide (CTAB), 2-amino-4,6-dihydroxypyrimidine, 2-thiophenecarboxaldehyde, terephthalic acid, triphenyl phosphite (TPP), dimethyl sulfoxide (DMSO), and methanol were bought from Sigma-Aldrich. CaCl2, HNO3 and HCl were bought from PanReac AppliChem. Since the analytical grade of all chemicals was chosen, they were utilized without further purification. 3.2. Synthesis of the P(PTA) The synthesis of the P(PTA) was performed in two steps. Firstly, a diamine-phenol monomer (TMAPD) was synthesized. Secondly, the polymer was obtained by polycondensation of TMAPD in 1,3-dipropyl imidazolium bromide ionic liquid as a solvent to avoid the use of the toxic triphenyl phosphite/N-methylpyrolidone/pyridine/LiCl that is required in the conventional direct polycondensation. Previously, the 1,3-dipropyl imidazolium bromide ionic liquid (IL) was prepared based on the procedure reported by Vygodskii et al. (2004). The procedures of the syntheses of the TMAPD and P(PTA) were as follows: 3.2.1. Synthesis of the monomer (5,5'-(thiophene-2-ylmethylene)bis(2-aminopyrimidine-4,6diol) (TMAPD)) TMAPD was synthesized according to the following procedure: A mixture of 2.54 g (0.02 mol) 2-amino-4,6-dihydroxypyrimidine, 1 mL (0.01 mol) 2-thiophenecarboxaldehyde, and 20 mL DMSO was stirred for 6 h at 110 °C. While the reaction tested by thin-layer chromatography was completed, the solution temperature was decreased to the room temperature, and the violet powder obtained by pouring the solution into 400 mL of cold DW (-5 °C) was filtered, rinsed many times using DW and then dried using vacuum oven at 100 °C. The reaction yield was 92 % (3.20 g), and 39 the obtained compound did not show a sharp melting point and started to be decomposed at above 300 °C. TMAPD structure is shown in Fig. 3.1. Fig. 3.1. Structure of TMAPD. 3.2.2. Synthesis of the biodegradable P(PTA) by the polycondensation reaction of TMAPD in TPP/IL The P(PTA) was achieved by polycondensation of TMAPD using TPP-IL as catalyst and solvent by the following procedure: A flask of three-necked round-bottomed with the volume of 50 mL was fitted with a mechanical stirrer, a water cooled condenser, and argon gas and then, a mixture containing 1 mmol TMAPD, 1 mmol terephthalic acid, 0.7 g 1,3-dipropyl imidazolium bromide {[1,3-(pr)2im]Br} as IL, and 1.29 mmol TPP was placed. The solution became sticky as the reaction continued at 110 °C for 2.5 h. In the following, the reaction mixture temperature was decreased to the temperature of the room and the precipitation of the obtained P(PTA) was performed using 100 mL of methanol. Then, after the precipitate filtration, the hot water was used for washing it. Afterward, the precipitate was further refined in a Soxhlet apparatus using methanol for 24 h to eliminate the oligomers with low molecular weight. The PA structure is shown in Fig. 3.2. Fig. 3.2. Structure of the P(PTA). 40 3.3. Synthesis of the magnetic nanoparticles Hydrothermal method was used to synthesize the Ni0.2Zn0.2Fe2.6O4 magnetic nanoparticles. A mixed solution of 2.6 M Fe3+, 0.2 M Zn2+ and 0.2 M Ni2+ was made ready in HCl solution and in the following NaOH solution was added into mixed solution under nitrogen gas while the pH value of the mixture was set to 10.5. Thereafter, 0.3 g of CTAB was added to this mixture and it was then located into an autoclave (Teflon-lined stainless steel) for 8 h at 200 ⁰C of an oven to perform hydrothermal treatment. Then, the autoclave temperature was reduced to the temperature of the room naturally, and the precipitate was collected and rinsed many times using deionized water to obtain neutral pH. Consequently, the obtained particles were dried at 50 °C. 3.4. Synthesis of the nanocomposites Four different magnetic nanocomposites, namely CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4, CA/CMC/Ni0.2Zn0.2Fe2.6O4, and P(PTA)/Ni0.2Zn0.2Fe2.6O4, were synthesized according to the procedures described below. 3.4.1. Synthesis of the magnetic CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4 nanocomposite Sodium alginate powder was dissolved in DW with a concentration of 1.5 % (w/v) to prepare a sodium alginate solution. 0.5 g of P (PTA) and 0.7 g of Ni0.2Zn0.2Fe2.6O4 were completely expanded in the sodium alginate solution with severe stirring for 24 h to achieve a homogeneous solution, A. Afterwards, the gelation process was performed by adding the solution A to a solution of CaCl2 (0.05 M) and 2 % glutaraldehyde. Then, the mixture was stirred for 24 h. In the following, the desired product was accumulated by an external magnetic field and rinsed many times using DW until reaching the solution pH of 7 and then dried at 50 °C. Finally, it was powdered. 3.4.2. Synthesis of the magnetic CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 nanocomposite The magnetic nanocomposite was synthesized by the gelation method. CMC powder was dissolved in deionized water (DW) (3 % w/v) under stirring at 150 rpm for 3 h to prepare CMC solution. Then, 0.5 g of the P(PTA) and 0.7 g of Ni0.2Zn0.2Fe2.6O4 were added to CMC solution and fully dispersed with vigorous stirring within 24 h. The gelation process was then performed 41 by adding the mixture to a solution of CaCl2 (0.05 M) and 2 % glutaraldehyde. The mixture was stirred for 24 h and the separation of the obtained nanocomposite was performed using an external magnet. It was also washed using deionized water many times to eliminate all impurities (unreacted GA) and reach the solution pH value of 7. Then, the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 bionanocomposite was dried at 50 °C. Ultimately, the obtained product was powdered. 3.4.3. Synthesis of the magnetic CA/CMC/Ni0.2Zn0.2Fe2.6O4 nanocomposite The procedure of CA/CMC/Ni0.2Zn0.2Fe2.6O4 synthesis is mentioned in the following: 1 g of sodium alginate was dissolved in 80 mL of deionized water at room temperature using a laboratory stirrer. Thereafter, CMC powder (0.5 g) was added into the solution and mixed homogeneously. Then, the Ni0.2Zn0.2Fe2.6O4 (0.7 g) was added to the biopolymers mixture. In order to acquire a homogeneous blend solution, the magnetic and biopolymers particle solution was stirred at the temperature of the room for 24 h. Then, the gelation process was performed by adding the mixture to a solution of calcium chloride 0.05 M and 2 % glutaraldehyde. To separate the resulting bionanocomposite from solution, an external magnetic field was used and then it was washed several times with distilled water for removing any remaining glutaraldehyde and calcium chloride till the pH value of the solution was equal to 7. Finally, the CA/CMC/Ni0.2Zn0.2Fe2.6O4 bionanocomposite was dried at 50 °C. Ultimately, the obtained product was powdered. 3.4.4. Synthesis of the magnetic P(PTA)/Ni0.2Zn0.2Fe2.6O4 nanocomposite The P(PTA)/Ni0.2Zn0.2Fe2.6O4 nanocomposite was synthesized by hydrothermal method as follows: 1.8 g of the P(PTA) was added to the solution of HCl and N,N-Dimethylformamide containing 0.2 M Ni2+, 2.6 M Fe3+, and 0.2 M Zn2+. Thereafter, NaOH solution was added into the mixed solution under nitrogen gas and the pH value of the mixture was set to 10.5. 0.3 g of CTAB and 4 mL glutaraldehyde were added to the mixture and thereafter it was located into an autoclave (Teflon-lined stainless steel) for 8 h at 200 ⁰C of an oven. The autoclave temperature was then naturally reduced to the temperature of the room, and the product was collected and rinsed with DW several times to reach neutral pH. Consequently, the composite was dried at 50 ⁰C. 42 3.5. Instrumentation and characterization KBr pellet was used to record the FT-IR spectra on a PerkinElmer, USA. The GBC MMA tool with CuKα radiation in the 2θ range of 10–70⁰ was used to record the XRD pattern. FE-SEM (Zeiss Neon-40, Germany) was used for specifying the morphology of the products. The magnetic attributes of the products were investigated at the temperature of the room using a VSM (Daghigh Kavir Corporation, Iran). Moreover, TGA was performed on a Mettler TGA/SDTA 851e/LF/1100 thermobalance. The sample temperature was enhanced from the temperature of the room to 1000 ⁰C (rate=10 ⁰C/min) while nitrogen flow was constant. An Agilent 4100 MP-AES Spectrometer was employed to analyze Nd+3, Dy+3, and Tb+3 concentration. 3.6. Adsorption studies of Nd+3, Tb+3, and Dy+3 The adsorption behavior of Nd+3, Tb+3, and Dy+3 by the magnetic nanocomposites was investigated in batch system to analyze the effects of different parameters on the process of adsorption and determaine the optimum conditions for maximum adsorption efficiency of the ions. In addition, the adsorption of the ions was studied in the packed-bed column. 3.6.1. Batch adsorption experiments Batch adsorption tests were done in 125 mL flasks containing 50 mL solutions with the concentrations in the range of 30-300 mg/L that were made ready from the dilution of 1000 mg/L stock solutions of the metal ions at 180 rpm. The values of pH were regulated via adding a suitable amount of 0.1 M sodium hydroxide or nitric acid solutions and monitored by a pH meter. While the contact time was finished, a magnetic field was used to separate the adsorbent and the residual Nd+3, Tb+3, and Dy+3 in the solution were then analyzed using an Agilent 4100 MP-AES Spectrometer. The adsorption efficiency (%) of Nd+3, Tb+3, and Dy+3 by the adsorbents was computed using the bellowing equation: Adsorption efficiency (%)= (𝐶0−𝐶𝑒 𝐶0)×100 (3.1) Here Ci and Cf respectively refer to the initial and final concentration of Nd+3, Tb+3, and Dy+3 (mg/L). 43 The adsorption capacity (qt (mg/g)) at time t is acquired as bellow: (3.2) Here Ci and Ct (mg/L) respectively show Nd+3, Tb+3, and Dy+3 concentrations at primary and given time t. V (L) also refers to the solution volume and M (g) shows the adsorbent mass. The equilibrium adsorption capacity (qe (mg/g)) was computed using the equation in the following: (3.3) Here Ce shows the concentration of Nd+3, Tb+3, and Dy+3 at equilibrium (mg/L). 3.6.1.1. pH effect As mentioned in section 1.2, the solution pH affects the solution chemistry of REEs. Therefore, the adsorption process is always pH-dependent and the capacity of adsorption is changed with the pH of the medium. The impact of pH on adsorption of Nd+3, Tb+3, and Dy+3 was explored at various levels of 1.5 to 5.5 by adding a given amount of the adsorbents into the Nd+3, Tb+3, and Dy+3 solutions at constant initial concentration and contact time at room temperature. 3.6.1.2. Contact time effect The contact time is important as the possible rapidness of binding and adsorption processes of the metal ion by adsorbent needs to be identified to reach the optimum time for complete adsorption of the metal ion (Mehdinia et al., 2015). Any change in adsorption capacity by the consideration of contact time can be ascribed to the capacity of adsorbent sites and metal concentration gradient. The effect of contact time was performed at different contact times by adding a given quantity of the adsorbent into the Nd+3, Tb+3, and Dy+3 solutions under continuous stirring at constant initial concentration and pH at room temperature. 3.6.1.3. Adsorbent dosage effect As the extent of adsorption is changed by increasing the amount of adsorbent due to the enhancement in the available number of active sites for adsorption of metal ion, the effect of the   M VCC qti t     M VCC qei e   44 adsorbent dosage on the adsorption of Nd+3, Tb+3, and Dy+3 from aqueous solutions was investigated by changing the dosage of the adsorbent at constant initial concentration, time and pH at room temperature. 3.6.1.4. Initial concentration effect The metal ion mass transfer between the solid phase and aqueous is changed by the driving force that is varied at any initial concentration, resulting in adsorption capacity change. To investigate the influence of initial concentration on the adsorption of the metal ions from aqueous solution, the empirical tests were carried out in the solutions having a particular amount of the adsorbent with different concentrations of Nd+3, Tb+3, and Dy+3 ranging from 30 to 300 mg/L at constant pH value and time at room temperature. 3.6.1.5. Ionic strength effect In general, industrial effluents contain various salts at higher levels that decrease the adsorption efficiency owing to the contest with target metal ions to involve adsorption sites. NaNO3, KNO3, and NaCl are the most important ones that have been applied to explore the ionic strength influence on the metal ions adsorption. Among all, NaNO3 can be regarded as an efficient and widely-used salt that has been used by many researchers for this goal. Therefore, the ionic strength effect on the adsorption of Nd+3, Tb+3, and Dy+3 onto the adsorbents was studied by a series of experiments at various concentrations of NaNO3 solutions. 3.6.1.6. Adsorption kinetic models The solute uptake rate of a system is generally described using adsorption kinetics that governs the adsorbate uptake residence time at a solid-solution interface along the diffusion process. It is presumed that no mass transfer resistance (both internal and external) is observed along the whole process of the adsorption. Consequently, the adsorption kinetics can be investigated using the residual metal ion concentration in the solution. Three kinetic models including pseudo-first-order, pseudo-second-order, and intra-particle diffusion were utilized to verify the validity of empirical data regarding Nd+3, Tb+3, and Dy+3 adsorption onto the adsorbents as following: 45 (1) Pseudo-first-order model (Janos et al., 2007): qt = qe (1 − exp−K1t ) (3.4) Where qe (mg/g) refers to the capacity of adsorption at equilibrium, qt is the adsorption capacity at time t, t is time, and K1 (1/min) refers to the pseudo-first-order rate constant. The applicability of the model for adsorption of Nd+3, Tb+3, and Dy+3 onto the adsorbent is investigated by plotting qt vs time ‘t’. (2) Pseudo-second-order model (Smaranda et al., 2010): qt = K2qe2t/1 + K2qet (3.5) The rate constant (K2 (g/mg min)) and the coefficient of determination are calculated from the plot of qt vs t. (3) Intra-particle diffusion (Yousef et al., 2016): qt = Kit0.5 + C (3.6) Here Ki (1/min) is the pseudo-first-order rate constant while C presents information about the thickness of the boundary layer. A higher value of C is related to the boundary layer diffusion effect. Ki and C are also computed using the plot of qt vs t. Error analysis was used to optimize the experimental data with the kinetic models. Chisquare (χ2) was employed in this paper for comparing each model's validity using the equation in the following: χ2 = ∑(𝑞𝑒,𝑒𝑥𝑝−𝑞𝑒,𝑐𝑎𝑙) 𝑞𝑒,𝑐𝑎𝑙 𝑛 𝑖=1 2 (3.7) Here qe,exp and qe,cal respectively present the empirical and calculated adsorbent capacities while n represents the number of data points. 3.6.1.7. Adsorption isotherm models Isotherm models are generally used to represent the relations between the amount of metal adsorbed onto the equivalent adsorption sites and metal concentration in solution while the temperature is constant (Papandreou et al., 2007). In this work, the two most important isotherm models were applied to specify the maximum adsorption capacity of the adsorbent and its relationship with Nd+3, Tb+3, and Dy+3 adsorbed from the solution as follows: 52 4.1.2. Characterization of the Ni0.2Zn0.2Fe2.6O4 The results of FE-SEM showed that the size of nearly spherical Ni0.2Zn0.2Fe2.6O4 particles was less than 100 nm (Fig. 4.2A). In Fig. 4.2B, the peaks at 2θ = 18.13°, 30.07°, 35.50°, 37.08°, 43.07°, 53.95°, 56.96°, and 63.89° are in accordance with the standard pattern of nickel zinc ferrite (JCPDS 08-0234) (Ni et al., 2015; Babua and Tatarchuk, 2018, Albuquerque et al., 2000). The average size of the nanoparticles was 27.68 nm according to the Scherrer equation at the Full Width at Half Maximum (FWHM) of the strongest reflection of the XRD pattern of the Ni0.2Zn0.2Fe2.6O4. Fig. 4.3 indicates that the nanoparticles are gathered and orientated vertically under a magnetic field. Fig. 4.2. (A) FE-SEM image and (B) XRD pattern of Ni0.2Zn0.2Fe2.6O4 nanopartciles. 53 Fig. 4.3. Orientation of the Ni0.2Zn0.2Fe2.6O4 under a magnetic field. 4.1.3. Characterization of the nanocomposites The FE-SEM image of the CA/CMC/Ni0.2Zn0.2Fe2.6O4 in Fig. 4.4 indicates that the Ni0.2Zn0.2Fe2.6O4 was distributed on the biopolymers’ surface or embedded with the biopolymers that confirmed the effective synthesis of the magnetic nanocomposites. Such a result was obtained for the CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4 and CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4. The FE-SEM image of the P(PTA)/Ni0.2Zn0.2Fe2.6O4 in Fig. 4.5 shows that the Ni0.2Zn0.2Fe2.6O4 particles are also synthesized nearly spherically in the presence of the P(PTA) in the solution with the average particle size of 31.62 nm. Similar peaks were obtained in the XRD pattern of the P(PTA)/Ni0.2Zn0.2Fe2.6O4 in comparison to the XRD pattern of the Ni0.2Zn0.2Fe2.6O4. FT-IR results obtained for the nanocomposites confirmed their successful synthesis in comparison to the FT-IR results of the CA, CMC, Ni0.2Zn0.2Fe2.6O4, and P(PTA). The EDX spectra of the biopolymer nanocomposites showed the successful crosslinking reaction of sodium alginate with calcium as the peak of Na+ was not seen in the spectra, and the expected elements were seen in the EDX of the biopolymer nanocomposites. In the case of the P(PTA)/Ni0.2Zn0.2Fe2.6O4, the EDX spectrum showed N, C, O, Ni, Zn, Fe and S peaks that confirmed the creation of the P(PTA)/Ni0.2Zn0.2Fe2.6O4. 54 Fig. 4.4. FE-SEM image of the CA/CMC/ Ni0.2Zn0.2Fe2.6O4. Fig. 4.5. FE-SEM image of the P(PTA)/Ni0.2Zn0.2Fe2.6O4. 55 According to the obtained results of VSM, the magnetic nanoparticles showed superparamagnetic behavior while the favorable magnetic saturation value was about 45.78 emu/g. Since the magnetic nanoparticles were combined with other materials, lower magnetic saturation (14.14 emu/g for the CA/CMC/Ni0.2Zn0.2Fe2.6O4, 15.28 emu/g for the CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, and 14.88 emu/g for the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4) was obtained for all synthesized composites as expected. The obtained decrease did not affect the separation of the nanocomposites and they were easily separated by a magnetic field as shown in Fig. 4.6. In the case of the P(PTA)/Ni0.2Zn0.2Fe2.6O4, the magnetic saturation value (50.49 emu/g) was close to that obtained for the magnetic nanoparticles, indicating that the magnetic saturation was not affected under the condition of the solution and in the presence of the dissolved P(PTA) in the solution. Thus, the P(PTA)/Ni0.2Zn0.2Fe2.6O4 was the strongest magnetic adsorbent in comparison with others. Fig. 4.6. Photo of magnetic separation. According to the TGA analysis obtained for the P(PTA), favorable thermal stability as 17.53% weight loss was seen by increasing temperature from RT to 330 ◦C. The weight loss of the P(PTA) was 61.69 % by increasing temperature up to ⁓ 600 ◦C. Enhancing the temperature up to 1000 ◦C caused the complete decomposition of the P(PTA). By combining the P(PTA) with 56 the Ni0.2Zn0.2Fe2.6O4, CA, and CMC, degradation temperature was shifted to a higher temperature. The thermal stability of the synthesized products in the range of room temperature to 1000 ◦C was in the order of P(PTA)/Ni0.2Zn0.2Fe2.6O4 (43.21 % weight loss) ˃ CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4 (64.1 % weight loss) ˃ CA/CMC/Ni0.2Zn0.2Fe2.6O4 (66.8 % weight loss) ˃ CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 (70.65 % weight loss). 4.2. Adsorption of Nd+3, Tb+3, and Dy+3 from aqueous solution The adsorption of Nd+3, Tb+3, and Dy+3 from aqueous solution was studied in single, ternary, and column systems. 4.2.1. Investigation of general ability of the P(PTA) for Nd+3, Tb+3, and Dy+3 adsorption The P(PTA) was used for the sorption of rare earth metals in primary tests. For this purpose, 0.1 g of the P(PTA) was added into 50 mL solution having 30 mg/L of Dy3+ at pH of 5.3, and in the following the solution was stirred for 14 h. The adsorption efficiency was approximately 100 %. Similar result of adsorption efficiency was achieved by adding 0.1 g of the P(PTA) into the multi-metal solution containing 10 mg/L of Dy+, Tb+3, and Nd+3. The results confirmed the adsorption ability of the P(PTA) for adsorption of Nd+3, Tb+3, and Dy+3. Moreover, the presence of pyrimidine and thiophene rings, hydroxyl groups and amide linkages in the backbones of the P(PTA) could act as hosts for the formation and adsorption of the complex with the metal ions. Mechanism of metal adsorption by the P(PTA) is offered (depicted) in Fig. 4.7. Fig. 4.7. Mechanism offered for the separation of metal ions. 57 4.2.2. Nd+3, Tb+3, and Dy+3 adsorption in the single system The adsorption of Nd+3, Tb+3, and Dy+3 in the single system was studied as a function of different parameters including pH, contact time, adsorbent dosage, initial ion concentration, ionic strength, and temperature. 4.2.2.1. pH effect on Nd+3, Tb+3, and Dy+3 adsorption pH is generally known as an essential controlling parameter along the adsorption process. In this regard, hydrogen ion concentration role was studied for the adsorption of Nd+3, Tb+3, and Dy+3 at various pH, particularly in the range of 1.5 to 5.5 (1.5, 2.5, 3.5, 4.5, and 5.5) with various prepared materials at room temperature of 25 °C by batch mode adsorption studies at 30 mg/L of the ions with 0.03 g of the CA/CMC/Ni0.2Zn0.2Fe2.6O4, CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4, CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 nanocomposites for 20 min. The experiments were not studied at a pH higher than 5.5 because the formation of the ions hydroxides may have taken place (Aghayan et al., 2013). The observed experimental results of the impact of pH on the ions adsorption from the aqueous solution using the adsorbents presented that the adsorption efficiency of the ions increased from 0 % to a maximum value for all the adsorbents, for a variation of pH from pH = 1.5 to pH = 5.5 within 20 min of adsorption study. Based on the obtained results, it was clear that there was practically no adsorption at pH = 1.5 due to the highly acidic solution, preventing the ions from further adsorption onto the surface of the adsorbents. The maximum adsorption efficiency was achieved at pH = 5.5 for all the adsorbents with initial ions concentration of 30 mg/L, which indicated their maximum adsorption efficiency at a higher pH value. However, while the pH value is low, electrostatic repulsion will be high along with the metal ions uptake which can be because of the high positive charge density and the protons on the surface active sites. Moreover, there is a competition between the metal ions and H+ ions for occupying the same site which results in lower efficiency of elimination. On the other hand, by increasing the pH value, electrostatic repulsion will be reduced which can be due to the positive charge density reduction on the adsorption sites which results in an increase in the metal ions adsorption. Several earlier studies had also supported the mentioned fact about the influence of pH on adsorption (Koochaki-Mohammadpour et al., 2014, Akkaya, 2014). Therefore, the optimum value of the solution pH was 5.5. In the case of the 58 P(PTA)/Ni0.2Zn0.2Fe2.6O4 the influence of pH was not conducted and the solution pH for adsorption studies was 5.5. 4.2.2.2. Contact time effect on Nd+3, Tb+3, and Dy+3 adsorption Nd+3, Tb+3, and Dy+3 adsorption efficiency also depends on their contact time with the adsorbents. The ions adsorption at various contact times was investigated for initial ions concentration of 30 mg/L at pH = 5.5 keeping all other parameters constant. Batch adsorption tests were carried out at 25 °C by changing the contact time from 2.5 to 70 minutes for the CA/CMC/Ni0.2Zn0.2Fe2.6O4 and CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, 2.5 to 120 min for the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4, and 2.5 to 150 min for the P(PTA)/Ni0.2Zn0.2Fe2.6O4. Based on the obtained results, more than 65% of adsorption efficiency took place within the first 5 min for the CA/CMC/Ni0.2Zn0.2Fe2.6O4 and CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4 and equilibriums were respectively established after 40 min and 50 min. In the case of the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4, the adsorption efficiency of more than 65% occurred at the contact time of 90 min. Similar result of adsorption efficiency was obtained at 50 mg/L of the ions with 0.13 g of the P(PTA)/Ni0.2Zn0.2Fe2.6O4 while a contact time of 30 min was applied. The value of adsorption efficiency was found to increase to more than 90% for the CA/CMC/Ni0.2Zn0.2Fe2.6O4 and CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4 and 70% for the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 by enhancing the contact time to the equilibrium time. For the P(PTA)/Ni0.2Zn0.2Fe2.6O4, the adsorption efficiency was found to increase to more than 80% with a contact time of 130 min as equilibrium for initial ions concentration of 50 mg/L. The change in the adsorption efficiency rate might refer to the fact that all adsorbent sites are initially empty and the solute concentration gradient is high. Therefore, the adsorption rate is high too. Thereafter, the ions uptake rate by adsorbent decreases owing to the reduction in adsorption sites besides the ions concentration. By passing the time, the number of sites on the adsorbent packed with the ions enhances too (Farghali et al., 2013). Notably, the adsorption rate is equal to the desorption rate at equilibrium while all the sites are packed. Therefore, it is observed that there is not any enhancement in adsorption efficiency by increasing the contact time after equilibrium. Decreased rate of removal, especially towards the end of experiments, indicates the feasible monolayer creation of the ions on the outer surface of the adsorbents. In the current study, steady growth in adsorption efficiency was seen up to a contact time of 40 min for the CA/CMC/Ni0.2Zn0.2Fe2.6O4, 50 min for the CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, 90 min for the 59 CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4, and 130 min for the P(PTA)/Ni0.2Zn0.2Fe2.6O4, while there was not any further enhancement in adsorption effectiveness with an enhancement in time. 4.2.2.3. Adsorbent dosage effect on Nd+3, Tb+3, and Dy+3 adsorption Batch adsorption investigations were conducted at 25 °C at pH = 5.5 and the optimum time for each adsorbent. The influence of different adsorbent dosage on the adsorption efficiency of the ions from aqueous solution with adsorbents utilized in this work presented that the ions adsorption increased by enhancing the adsorbent dosage for all the adsorbents. It was seen that the adsorption efficiency of ions respectively increased from 53.08 to 97.75 %, 51.4 to 96.83 %, and 59.04 to 97.88 % for Nd+3, Tb+3, and Dy+3, by using 0.01 to 0.04 g of the CA/CMC/Ni0.2Zn0.2Fe2.6O4 and initial ions concentration of 30 mg/L. Under the same conditions, the results were 51.14 to 96.73 %, 49.33 to 94.82 %, and 50.66 to 97.58 % for Nd+3, Tb+3, and Dy+3, respectively, by using the CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4. The adsorption efficiency increased from 26.63 to 88.24 %, 23.13 to 82.03 %, and 28.96 to 91.27 % for Nd+3, Tb+3, and Dy+3 respectively, using the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 in the range of 0.01 to 0.04 g that was lower than those of the CA/CMC/Ni0.2Zn0.2Fe2.6O4 and CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4. By increasing the adsorbent dosage of the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 to 0.06 g, 98.15, 97.6, and 99.42 % adsorption efficiencies were respectively obtained for Nd+3, Tb+3, and Dy+3. As for the P(PTA)/Ni0.2Zn0.2Fe2.6O4, the obtained results were 26.34 to 95.67 %, 31.21 to 97.48 %, and 32.08 to 98.41 % for Nd+3, Tb+3, and Dy+3, respectively at 50 mg/L of the ions in the dosage range of 0.05 to 0.15 g. Such results were expected for a constant initial concentration of the ions because an increase in adsorbent dosage provides larger adsorption sites or surface area which results in a higher adsorption efficiency (Das and Das, 2013). However, it was seen that after upper value of the dosage for each adsorbent, there was not any considerable modification in the adsorption efficiency of the ions that might be owing to the active sites overlapping at a higher dosage. Therefore, there was not any considerable enhancement in the efficient surface area because of the adsorbent particles conglomeration (Pathania et al., 2017). 4.2.2.4. Initial concentration effect on Nd3+, Tb+3, and Dy+3 adsorption The initial metal ions concentration plays a significant role in adsorption due to the fact that only a fixed amount of metal ions can be adsorbed by the given mass of adsorbent material. The initial concentration of the ions solution was ranging from 30 to 300 mg/L (except as for the 60 adsoption of the ions by the P(PTA)/Ni0.2Zn0.2Fe2.6O4 that the lower initial ions concentration value was 50 mg/L), and batch mode tests were carried out using the optimum dosage of the adsorbents at room temperature of 25 °C to explore the impact of initial concentration of the ions on their adsorption by the magnetic adsorbents. The experimental data presented that the ions adsorption efficiency reduced by enhancing the initial concentration. Adsorption efficiencies of the CA/CMC/Ni0.2Zn0.2Fe2.6O4 for Nd+3, Tb+3, and Dy+3 respectively decreased from 97.3 to 19.4 %, 96.83 to 27.53 %, and 97.85 to 31.51 % which indicated that increasing initial concentration had more effect on the adsorption of in comparison to Tb+3 and Dy+3. The obtained values of the CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4 were approximately similar to those of the CA/CMC/Ni0.2Zn0.2Fe2.6O4. At the optimum dosage of the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 (0.06 g), the obtained adsorption efficiencies at the initial concentration of 300 mg/L were 16.74, 21.45, and 21.76 % for Nd+3, Tb+3, and Dy+3, respectively, that were lower than those of CA/CMC/Ni0.2Zn0.2Fe2.6O4 and CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4 in the investigated range of initial concentration. The reason for the observed decrease by increasing initial concentration is due to the fact that the amount of ions that can be adsorbed for a given mass of adsorbent material is fixed. Therefore, the higher concentration of the ions leads to a smaller quantity of adsorption efficiency. While the concentration is low, there are large vacant active sites on the surface of the adsorbents and by increasing the initial ions concentration, the number of active sites that are required for adsorption is decreased. Nonetheless, the real amount of ions adsorbed per unit mass of adsorbent enhanced by increasing the ions concentration. Consequently, it can be due to the high driving force to overcome the mass transfer stability between the aqueous and solid phase at a high initial ions concentration. Actually, as the initial concentration of metal ions is increased, the interaction between adsorbent and adsorbate also enhances (Giese and Jordão, 2019). However, the adsorption efficiency of the ions is maximum at lower initial concentrations. Therefore, the process of the adsorption greatly relies on the initial concentration of the ions. 4.2.2.5. Adsorption kinetics From fitting the data of kinetic of the adsorption of Nd+3, Tb+3, and Dy+3 by the adsorbents (Tables 4.1-4.4), it was found that the R2 values of the ions for the pseudo-first-order model ˂ 0.91 for the CA/CMC/Ni0.2Zn0.2Fe2.6O4, ˂ 0.90 for the CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, ˂ 0.89 for the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4, and ˂ 0.91 for the P(PTA)/Ni0.2Zn0.2Fe2.6O4 were 61 lower than those acquired by pseudo-second-order model (˃ 0.98 for the CA/CMC/Ni0.2Zn0.2Fe2.6O4 and CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, ˃ 0.95 for the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4, and ˃ 0.97 for the P(PTA)/Ni0.2Zn0.2Fe2.6O4, implying that the adsorption of the ions was not described with pseudo–first–order model. Based on the obtained results, the calculated values of adsorption capacity (qe,cal) of the metal ions onto the adsorbents at equilibrium by the pseudo–second–order were in good compromise with the experimental values (qe,exp) of adsorption capacities of ions. The results were also confirmed by lower values of χ2 obtained by pseudo–second–order fitting in comparison with those of the pseudo-first-order model. The stability of the empirical data with the pseudo–second–order kinetic model indicated that the ions adsorption onto the adsorbents was supervised by chemical adsorption (chemisorption). The pseudo–first–order and pseudo–second–order equations were not able to distinguish the diffusion method. The data were then fitted by the intra-particle diffusion technique. Based on the model, the curve of qt vs. t0.5 can be linear if intra-particle diffusion is related to the adsorption. If the plots go through the origin, then intra-particle diffusion is the rate-determining step (Iftekhar et al., 2018). The data fitting by the intra-particle diffusion model showed a multi-steps adsorption process for each metal by the CA/CMC/Ni0.2Zn0.2Fe2.6O4, CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, and P(PTA)/Ni0.2Zn0.2Fe2.6O4. The first sharper portion was attributed to the diffusion of the ions through the solution to the external surface of the adsorbents (external diffusion). The second linear portion explained the gradual adsorption stage where the intra-particle diffusion is ratelimiting. Finally, the third linear portion denoted the establishment of the equilibrium step where the intra-particle diffusion started to decelerate owing to the remarkably low concentration of the ions in the solution (intraparticle diffusion) (Song et al., 2013). The curves did not go through the origin (intercept > 0) that might be indicative of some degree of effect and control by boundary layer. This might be representative of some degree of boundary layer control. Moreover, it also intimates that the intra-particle diffusion does not only contribute to the rate-determining step but also the adsorption rate may be simultaneously controlled by other processes (Das et al., 2017). In the case of the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4, the R2 values of 0.9737, 0.9782, and 0.9676 along with the χ2 values of 1.63, 1.31, and 2.03 respectively obtained by fitting with the intraparticle diffusion for Nd+3, Tb+3, and Dy+3, showed that this model can also explain the ions adsorption. 68 4.2.2.7 Ionic strength effect The presence of salts in the solution can cause a disturbance in the adsorption performance of target ions with competition for interaction with the active sites of the adsorbent. In the present study, the influence of NaNO3 in the solution on the Nd+3, Tb+3, and Dy+3 adsorption was studied at pH = 5.5 with 30 mg/L concentration of the ions and NaNO3 concentration in the range of 0.02 to 1 M at optimum conditions obtained for the dosage of the adsorbents and contact time. It was found that by increasing the concentration of NaNO3, a negative influence on the adsorption efficiency of the adsorbents was obtained. The adsorption efficiency of the CA/CMC/Ni0.2Zn0.2Fe2.6O4 decreased from 97.75 to 85.7, 96.83 to 84.73, and 97.85 to 93.5 % for Nd+3, Tb+3, and Dy+3, respectively. The negative effect of NaNO3 on the adsorption the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 was approximately the same as the results obtained for the CA/CMC/Ni0.2Zn0.2Fe2.6O4 with the values of 98.15 to 88.39, 97.6 to 82.4, and 99.42 to 94.1 % that were respectively obtained for Nd+3, Tb+3, and Dy+3. The values of adsorption efficiency obtained by the P(PTA)/Ni0.2Zn0.2Fe2.6O4 expressed reduction from 95.67 to 67.42 for Nd+3, 97.48 to 81.7 for Tb+3, and 98.41 to 87.41 % for Dy+3, respectively. In the case of the CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, the presence of NaNO3 in the solution had the highest negative effects on the adsorption of the ions with the values of 96.73 to 56.8, 94.82 to 63.23, and 97.58 to 63.96 % respectively obtained for Nd+3, Tb+3, and Dy+3. The results might be due to the competition of Na+ ions in the solution with the metal ions for interacting with the active sites of the adsorbents and reduction in the activity coefficient of the metal ions that causes the limitation of the ions transfer to the adsorbent surface. Similar effect of NaNO3 for the adsorption of REEs using adsorbents has been reported by the researchers (Šolić et al., 2020). 4.2.2.8. Temperature effect and the evaluation of thermodynamic parameters Investigating the effect of temperature on the ions adsorption efficiency using the synthesized nanocomposites was performed at different temperatures of 25, 35, and 45 °C with 90 mg/L of metal ion solutions at optimum conditions of adsorbent dosage and contact time. It was observed that the adsorption efficiency towards Nd+3, Tb+3, and Dy+3 was positively affected by increasing the temperature (Tables 4.10-4.13). As stated in section 3.6.1.8, the viscosity of metal solution decreased at the higher temperature which resulted in an increase in the diffusion of adsorbate molecules across the external boundary layer and the internal pores of the adsorbent particle. The values of 69 thermodynamic parameters in Tables 4.10-4.13 show that the values of ΔH° were positive, confirming the endothermic process of the ions adsorption. The positive values of ΔS° were a result of an increase in randomness on the liquid-solid interface. As for ΔG°, the obtained values for the CA/CMC/Ni0.2Zn0.2Fe2.6O4 and CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4 were negative at all temperatures, indicating that the process was naturally spontaneous and feasible while the obtained values for the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 showed the spontaneity of the process at a higher temperature. In the case of P(PTA)/Ni0.2Zn0.2Fe2.6O4, the values of ΔG0 were obtained to be positive at all temperatures in the case of Nd+3, suggesting that the adsorption of Nd+3 onto the P(PTA)/Ni0.2Zn0.2Fe2.6O4 the was non-spontaneous, while the values obtained for Tb+3 and Dy+3 showed the spontaneity of the adsorption process at higher temperatures. Table 4.10. Effect of temperature on the adsorption of the ions at 90 mg/L by the CA/CMC/Ni0.2Zn0.2Fe2.6O4 and thermodynamic parameters. Temperature (ºC) Adsorption efficiency (%) Nd+3 Tb+3 Dy+3 25 61.72 68.51 71.95 35 65.95 74.61 77.31 45 71.81 80.59 83.84 Thermodynamic parameters Nd+3 Tb+3 Dy+3 ΔHº (kJ/mol) 19.00 26.86 29.27 ΔSº (kJ/mol K) 0.068 0.097 0.106 ΔGº (kJ/mol) Temperature (ºC) 25 35 45 Nd+3 Tb+3 Dy+3 Nd+3 Tb+3 Dy+3 Nd+3 Tb+3 Dy+3 -1.737 -2.478 -2.887 -2.263 -3.331 -3.710 -3.062 -4.353 -4.943 70 Table 4.11. Effect of temperature on the adsorption of the ions at 90 mg/L by the CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4 and thermodynamic parameters. Temperature (ºC) Adsorption efficiency (%) Nd+3 Tb+3 Dy+3 25 61.71 66.79 69.33 35 65.10 71.78 73.90 45 69.15 77.25 79.73 Thermodynamic parameters Nd+3 Tb+3 Dy+3 ΔHº (kJ/mol) 13.72 21.78 23.03 ΔSº (kJ/mol K) 0.051 0.079 0.084 ΔGº (kJ/mol) Temperature (ºC) 25 35 45 Nd+3 Tb+3 Dy+3 Nd+3 Tb+3 Dy+3 Nd+3 Tb+3 Dy+3 -1.736 -2.284 -2.574 -2.168 -2.961 -3.236 -2.725 -3.823 -4.211 Table 4.12. Effect of temperature on the adsorption of the ions at 90 mg/L by the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 and thermodynamic parameters. Temperature (ºC) Adsorption efficiency (%) Nd+3 Tb+3 Dy+3 25 45.72 50.58 52.61 35 51.05 57.51 59.89 45 58.08 64.8 67.5 Thermodynamic parameters Nd+3 Tb+3 Dy+3 ΔHº (kJ/mol) 20.66 24.40 25.97 ΔSº (kJ/mol K) 0.065 0.079 0.085 ΔGº (kJ/mol) Temperature (ºC) 25 35 45 Nd+3 Tb+3 Dy+3 Nd+3 Tb+3 Dy+3 Nd+3 Tb+3 Dy+3 0.877 0.394 0.193 0.359 -0.308 -0.559 -0.861 -1.131 -1.446 71 Table 4.13. Effect of temperature on the adsorption of the ions at 90 mg/L by the P(PTA)/Ni0.2Zn0.2Fe2.6O4 and thermodynamic parameters. Temperature (ºC) Adsorption efficiency (%) Nd+3 Tb+3 Dy+3 25 60.22 68.83 73.94 35 63.53 71.54 77.32 45 69.16 79.41 85.34 Thermodynamic parameters Nd+3 Tb+3 Dy+3 ΔHº (kJ/mol) 16.32 23.18 29.88 ΔSº (kJ/mol K) 0.048 0.074 0.098 ΔGº (kJ/mol) Temperature (ºC) 25 35 45 Nd+3 Tb+3 Dy+3 Nd+3 Tb+3 Dy+3 Nd+3 Tb+3 Dy+3 1.694 0.759 0.137 1.391 0.452 -0.328 0.770 -0.664 -1.753 4.2.2.9. Reusability studies To study the reusability of the adsorbents, the optimum dosage of each adsorbent (0.04 g of the CA/CMC/Ni0.2Zn0.2Fe2.6O4 and CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, 0.06 g of the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4, and 0.15 g of the P(PTA)/Ni0.2Zn0.2Fe2.6O4) was shaken with 50 mL solution of 30 mg/L of Nd+3, Tb+3, and Dy+3 (except for the P(PTA)/Ni0.2Zn0.2Fe2.6O4 with 50 mg/L of Nd+3, Tb+3, and Dy+3) on a shaker at 180 rpm. The pH of the solution was adjusted to the desired value by adding HNO3 or NaOH solutions. After reaching equilibrium at the optimum time, the magnetized adsorbents were separated by an external magnet and the concentrations of the Nd+3, Tb+3, and Dy+3 in the solutions were measured. Then, the adsorbents were washed with deionized water to remove the unadsorbed metal ions. For desorption of the ions from the adsorbents, 50 mL of 0.2 M HNO3 was used. The desorption process was performed for 2 h. The regenerated adsorbent was employed for adsorption of the Nd+3, Tb+3, and Dy+3 at the optimized conditions. The desorption efficiency values of more than 95, 96, and 99 % were respectively obtained for Nd+3, Tb+3, and Dy+3 using the CA/CMC/Ni0.2Zn0.2Fe2.6O4. The desorption efficiencies of the Nd+3, Tb+3, and Dy+3 were respectively more than 89, 91, and 95 % for the CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, more than 82, 84, and 88 % for the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4, and 82, 86, and 90 % for the P(PTA)/Ni0.2Zn0.2Fe2.6O4. It was observed that the adsorption efficiencies of the adsorbents were decreased after the fourth cycle. After the fourth cycle, the decreases in adsorption efficiency of Nd+3, Tb+3, and Dy+3 were 72 respectively more than 95, 94, and 95% using the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4 that were lower than those of other adsorbents. After the fourth cycle of adsorption, adsorption efficiencies of Nd+3, Tb+3, and Dy+3 were reduced to 91.78, 90.36, and 93.56 % using the CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, 86.43, 84.61, and 87.52 % using the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4, and 84.64, 86.37, and 88.75 % using the P(PTA)/Ni0.2Zn0.2Fe2.6O4, which showed higher efficiency of the CA/CMC/Ni0.2Zn0.2Fe2.6O4 in comparison with other adsorbents. The reduction in the efficiency of adsorption might be due to the chemically bonding of the ions with the functional groups that cannot be desorbed easily, decrease in the functional groups during acid treatment, and decrease in the weight of adsorbent during cycles. Nevertheless, the obtained results revealed that the adsorbents can be potentially used for the adsorption of the Nd+3, Tb+3, and Dy+3. 4.2.2.10. Competitive adsorption The optimum dosage of the adsorbents obtained in the single system was used to evaluate the adsorption of Nd+3, Tb+3, and Dy+3 ions competitively in the ternary system of 30 mg/L of the ions while the ratio of the ions was 1:1:1. The results showed that the adsorption efficiencies for Nd+3, Tb+3, and Dy+3 ions reduced to 28.25, 48.62, and 50.58 % using the CA/CMC/Ni0.2Zn0.2Fe2.6O4, 22.3, 48.05, and 50.28 % using the CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, 19.4, 37.6, and 40.9 % using the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4, and 14.32, 32.46, and 34.64 % using the P(PTA)/Ni0.2Zn0.2Fe2.6O4. In a multi-component system, antagonism, synergism, and non-interaction effects can occur with values of qmix/q0 ˂ 1, qmix/q0 ˃ 1, and qmix/q0 = 1 (Wang et al., 2017), respectively, where qmix and q0 are respectively the adsorption capacities of each ion in the mixture and single systems. Results showed an antagonism effect of each ion on the adsorption of other ions in a multi-component system with the qmix/q0 values for Nd+3, Tb+3 and Dy+3 that were respectively 0.27, 0.5, and 0.52 using CA/CMC/Ni0.2Zn0.2Fe2.6O4, 0.23, 0.51, and 0.51 using the CA/P(PTA)/Ni0.2Zn0.2Fe2.6O4, 0.2, 0.38, and 0.41 using the CMC/P(PTA)/Ni0.2Zn0.2Fe2.6O4. 0.15, 0.33, and 0.35 for Nd+3, Tb+3, and Dy+3 using the P(PTA)/Ni0.2Zn0.2Fe2.6O4. EDX spectrum was also recorded for each ions loaded-adsorbents after the adsorption process and the obtained results showed the existence of Nd+3, Tb+3, and Dy+3 peaks in the spectra. 73 4.2.3. Simultaneous adsorption of Nd3+, Tb+3, and Dy+3 in the ternary system Although all the synthesized adsorbent showed their favorable efficacy for the adsorption of Nd3+, Tb+3, and Dy+3, the CA/CMC/Ni0.2Zn0.2Fe2.6O4 was found to be more efficient than other adsorbents, according to the results obtained by the single adsorption studies. Therefore, it was applied for studying Nd+3, Tb+3, and Dy+3 adsorption in a multi-component system. First, the effect of pH on the adsorption of Nd+3, Tb+3, and Dy+3 was investigated. The results were the same as those obtained in the single system as following: minimum adsorption efficiency at pH 1.5 due to the highly acidic solution and maximum adsorption efficiency at pH 5.5 because of reduction in H+ ions. Then, to decrease the number of experiments, obtain useful information about the effect of independent parameters individually and/or interactively that leads to a decrease in experimental error, and model the process of Nd+3, Tb+3, and Dy+3 adsorption by the CA/CMC/Ni0.2Zn0.2Fe2.6O4, Response Surface Methodology (RSM) was applied. RSM approach usually includes three stages: design and experiments, response surface modeling by regression, and optimization. By considering adsorbent dosage, contact time, initial concentration, response (ions adsorption efficiency), and CCD at five levels of -α (-2), low (-1), central (0), high (+1) and +α (+2), 32 individual experimental runs were proposed by Design Expert 10.0. The quadratic polynomial model for response versus the independent variables was presented as follows (Srivastava, 2015): 2 01 1 1 1 k k k k i i ij i j ii i i i j i Y x x x x                  (4.2) Where Y refers to the predicted response (adsorption efficiency), and 0  , i  , ii  , and ij  respectively refer to the constant coefficient, linear coefficient, quadratic coefficient, and interaction coefficient. Notably, xi and xj are the independent variables, k shows the number of the independent variables, and  is the residual error. Analysis of variance (ANOVA) was used to obtain information about the adequacy of the models by evaluating coefficient of determination (R2), lack of fit, and the Fisher test (F-value) values (Markandeya et al., 2017). By applying quadratic regression modeling between the response and independent variables for each metal, the equations obtained for adsorption efficiency (%) were as follows: Nd (III) adsorption efficiency (%) = 45.421 + 1.866 X1 + 514.024 X2 – 0.571 X3 – 1.322 X4 + 0.175 X5 – 0.0043 X1X3 + 0.00312 X1X4 – 0.00255 X1X5 – 2.27 X2X1 + 1.796 X2X3 + 1.351 74 X2X4 + 0.759 X2X5 + 0.00712 X3X4 -0.00258 X3X5 + 0.00469 X4X5 – 0.0126 X12 – 1237.07 X22 – 0.000393 X32 + 0.00122 X42 – 0.00378 X52 (4.3) Tb (III) adsorption efficiency (%) = 51.793 + 1.663 X1 + 410.645 X2 – 0.502 X3 – 0.603 X4 + 0.24 X5 – 7.549 X1X3 + 1.073 X1X4 – 1.132 X1X5 – 2.253 X2X1 + 1.651 X2X3 + 1.185 X2X4 + 1.283 X2X5 + 0.00328 X3X4 – 0.00178 X3 X5 + 0.00243 X4X5 -0.0124 X12 -1059.01 X22 – 0.0001808 X32 + 0.000553 X42 – 0.00103X52 (4.4) Dy (III) adsorption efficiency (%) = 58.691 + 1.632 X1 + 405.813 X2 - 0.354 X3 – 0.778 X4 – 0.315 X5 – 0.00156 X1X3 + 0.00241X1X4 – 0.000818 X1X5 – 2.405 X2X1 + 1.486 X2X3 + 1.371 X2X4 + 1.272 X2X5 + 0.00183 X3X4 – 0.00243 X3X5 + 0.00253 X4X5 – 0.0122 X12 – 1039.17 X22 + 0.000237 X32 + 0.000437 X42 – 0.0000959 X52 (4.5) ANOVA results showed R2 and Radj2 of 0.9951 and 0.9862 for Nd+3, 0.9948 and 0.9853 for Tb+3, and 0.9938 and 0.9826 for Dy+3, indicating a good agreement between the predicted and experimental data. The F-values with a very low probability value of 0.0001 showed that the predicted models are statistically significant. The ‘Adequate precision’’ values for Nd+3, Tb+3, and Dy+3 obtained by ANOVA were favorable due to a value > 4 (Soltani et al., 2013). Additionally, the coefficient of variation values (C.V. %) were obtained to be 2.1, 1.87, and 1.97 for Nd+3, Tb+3, and Dy+3 respectively, that are low and show the reliability of the modeling. Under the conditions of pH = 5.5, the adsorbent dosage of 0.1 g, initial concentration of 30 mg/L, and contact time of 53 min were predicted by RSM while the adsorption efficiencies of Nd+3, Tb+3, and Dy+3 were respectively equal to 95.72, 96.17and 99.44 %. 4.2.3.1. Batch adsorption kinetic and isotherm studies Using 50 mL of the solutions at 30 mg/L of Nd+3, Tb+3, and Dy+3 ions, kinetic studies were performed by contacting the ions with 0.09 g of the CA/CMC/Ni0.2Zn0.2Fe2.6O4 at different times. The CA/CMC/Ni0.2Zn0.2Fe2.6O4 was used at the dosage of 0.09 g contacting with 50 mL of Nd+3, Tb+3, and Dy+3 solutions at different concentrations in the range of 30-180 mg/L at pH = 5.5. The results of kinetic studies showed that the main mechanism for the adsorption of the ions is chemisorption due to the highest values of R2 (0.9927 for Nd+3, 0.9933 for Tb+3, and 0.9929 for 75 Dy+3) and lowest values of ꭓ2 (0.045 for Nd+3, 0.043 for Tb+3, and 0.048 for Dy+3) obtained by PSO as compared with PFO (R2 values of 0.9617, 0.9585, and 0.9587, and ꭓ2 values of 0.24, 0.268, and 0.279 respectively for Nd+3, Tb+3, and Dy+3) and IPD (R2 values of 0.8157, 0.8225, and 0.8222, and ꭓ2 values of 1.16, 1.15, and 1.20 respectively for Nd+3, Tb+3, and Dy+3) models. From the fitting results obtained by the isotherm models, Freundlich isotherm showed to be more suitable for the description of the adsorption of Nd+3, Tb+3, and Dy+3 by the CA/CMC/Ni0.2Zn0.2Fe2.6O4 according to the higher values of R2 (0.9879, 0.9654, and 0.9633 respectively for Nd+3, Tb+3, and Dy+3) and lower values of ꭓ2 (0.159, 0.55, and 0.675 respectively for Nd+3, Tb+3, and Dy+3) in comparison with the Langmuir model (R2 values of 0.8773, 0.9238, and 0.9125, and ꭓ2 values of 1.62, 1.21, and 1.61 respectively for Nd+3, Tb+3, and Dy+3). In addition, the values of RL for Nd+3, Tb+3, and Dy+3 were between 0 and 1 (0.007, 0.006, and 0.004 respectively for Nd+3, Tb+3, and Dy+3), suggesting favorable adsorption of Nd+3, Tb+3, and Dy+3 by the CA/CMC/Ni0.2Zn0.2Fe2.6O4. The values of 9, 9.26, and 9.71 that were obtained for n showed a strong interaction between the CA/CMC/Ni0.2Zn0.2Fe2.6O4 and the metal ions. 4.2.3.2. Ionic strength effect The results of ionic strength presented a negative effect of NaNO3 on the adsorption of Nd+3, Tb+3, and Dy+3. As for Nd+3, the negative effect was greater than those obtained for Tb+3 and Dy+3. Greater negative effect of NaNO3 on the adsorption efficiency of Nd (III) was found in comparison with Tb (III) and Dy (III). The adsorption efficiencies for Nd (III), Tb (III), and Dy (III) without the presence of NaNO3 were 92.33, 93.91, and 96.25 %. In the presence of 0.1 M NaNO3, adsorption efficiencies decreased to 77.12, 85.6, and 91.43 % respectively for Nd+3, Tb+3, and Dy+3 that might be due to the competition of sodium ions with Nd+3, Tb+3, and Dy+3 for interacting with the active adsorption sites of the adsorbent and a decrease in the adsorption sites of the adsorbent as a result of an increase in aggregation of the adsorbent by an enhancement in ionic strength. 4.2.3.3. Temperature effect and the evaluation of thermodynamic parameters 50 mL of the ions at the concentration of 90 mg/L was contacted with 0.09 g of the adsorbent to investigate the influence of various temperatures (25, 35, and 45 °C) on the adsorption efficiency of the ions. A positive effect of temperature was seen for the process of 76 ions adsorption while adsorption efficiencies increased from 44.31, 47.14, and 49.21 to 52.64, 55.99, and 58.42 for Nd+3, Tb+3, and Dy+3, respectively. The positive values of ΔH° (13.9, 14.76, and 15.45 respectively for Nd+3, Tb+3, and Dy+3) showed an endothermic process. The positive values of ∆G◦ at all temperatures revealed a non-spontaneous process of the ions adsorption at the studied concentration. ΔS° values were also positive that indicated an increase in randomness at the interface of the solid–solution was obtained during the metal ions fixation on the adsorbent surface. 4.2.3.4. Reusability studies Reusability test of the adsorbent was performed by the adsorbent loaded with the ions at the conditions of initial concentration = 30 mg/L and pH = 5.5 with 0.1 g of the adsorbent for 53 min. Desorption process by 0.2 M HNO3 showed that desorption efficiencies were ˃ 93, 96, and 97 % for Nd+3, Tb+3, and Dy+3, respectively. Using the adsorbent in four cycles showed an insignificant decrease in the adsorption efficiency as the difference between the adsorption efficiency of the first cycle and the fourth cycle which were about 2.54, 1.63, and 1.16 % respectively for Nd+3, Tb+3, and Dy+3. This result might be due to the reasons mentioned in section 4.2.2.9. The reusability of the adsorbent was concluded according to the obtained results. 4.2.3.5. Column mode The adsorption of Nd+3, Tb+3, and Dy+3 was studied using a glass column (D: 0.5 cm and L:12 cm) packed with 0.3 g of the adsorbent while mixed with 1.8 g acid-cleaned sand for decreasing pressure drop. First, a flow rate of 1 mL/min of DW provided by a peristaltic pump was passed upward for flushing the column for an hour. Then, a ternary solution of 30 mg/L of Nd+3, Tb+3, and Dy+3 was injected to the column for 520 min while a fraction collector was used to collect the effluent every 10 min. Finally, Agilent 4100 MP-AES Spectrometer was used to determine the concentration of the ions in the effluent. The times for breakthrough (Ce/C0=0.05) and exhaustion (Ce/C0=0.95) were obtained to be 95 and 410 min for Nd+3, 105 and 430 min for Tb+3, and 120 and 440 min for Dy+3. Adsorption efficiencies of Nd+3, Tb+3, and Dy+3 were respectively 46.33, 47.07, and 49.11 % while the obtained adsorption capacities were equal to 22.70, 24.00, and 25.54 mg/g for Nd+3, Tb+3, and Dy+3, respectively. Both Thomas and Yan 77 models were fitted well the experimental data of Nd+3, Tb+3, and Dy+3 column adsorption. The results of column adsorption indicated successful practical usage of the adsorbent. 84 REFERENCES  Aghayan, H., Mahjoub, A.R., Khanchi, A.R., Samarium and dysprosium removal using 11molybdo-vanadophosphoric acid supported on Zr modified mesoporous silica SBA-15, Chemical Engineering Journal, 225 (2013) 509–519.  Akkaya, R., Synthesis and characterization of a new low-cost composite for the adsorption of rare earth ions from aqueous solutions, Chemical Engineering Journal, 200 (2012) 186–191.  Akkaya, R., Terbium adsorption onto polyhydroxyethylmethacrylate–hydroxyapatite composite and its modified composition by phytic acid, Desalination and Water Treatment, 52 (2014) 1440-1447.  Albuquerque, A.S., Ardisson, J.D., Macedo, W.A.A., Alves, M.C.M., Nanosized powders of NiZn ferrite: Synthesis, structure, and magnetism, Journal of Applied Physics, 87, 4352 (2000).  Anastopoulos, I., Bhatnagar, A., Lima, E.C. Adsorption of rare earth metals: A review of recent literature, Journal of Molecular Liquids, 221 (2016) 954-962.  Anitha, A., Divya Rani, V.V., Krishna, R., Sreeja, V., Selvamurugan, N., Nair S.V., Tamura H., Jayakumar, R., Synthesis, characterization, cytotoxicity and antibacterial studies of chitosan, O-carboxymethyl and N,O-carboxymethyl chitosan nanoparticles, Carbohydrate Polymers, 78 (2009) 672–677 .  Ashour, R.M., El-Sayed, R., Abdel-Magied, A.F., Abdel-Khalek, A.A., Ali, M.M., Forsberg, K., Uheida, A., Muhammed, M., Dutta, J., Selective separation of rare earth ions from aqueous solution using functionalized magnetite nanoparticles: kinetic and thermodynamic studies, Chemical Engineering Journal, 327 (2017) 286–296.  Awual, M.R., Alharthi, N.H., Okamoto, Y., Karim, M.R., Halim, M.E., Hasan, M.M., Rahman, M.M., Islam, M.M., Khaleque, M.A., Sheikh, M.C., Ligand field effect for Dysprosium(III) and Lutetium(III) adsorption and EXAFS coordination with novel composite nanomaterials, Chemical Engineering Journal, 320 (2017) 427-435.  Aziz H.A., Adlan M.N., Ariffin K.S., Heavy metals (Cd, Pb, Zn, Ni, Cu and Cr(III)) removal from water in Malaysia: post treatment by high quality limestone, Bioresource Technology, 99 (2008) 1578-1583. 85  Babu, B.R., Tatarchuk, T., Elastic properties and antistructural modeling for Nickel-Zinc ferrite-aluminates, Materials Chemistry and Physics, 207 (2018) 534-541.  Balaram, V., Rare earth elements: A review of applications, occurrence, exploration, analysis, recycling, and environmental impact, Geoscience Frontiers, 10 (2019) 1285-1303.  Balaram, V., Recent advances in the determination of elemental impurities in pharmaceuticals - Status, challenges and moving frontiers, Trends in Analytical Chemistry, 80 (2016) 83-95.  Barakos, G., An assessment tool for the mineability of rare earth element deposits. Ph.D. Diss. Technical University Bergakademie Freiberg, Germany, 2017.  Barros, Ó., Costa, L., Costa, F., Lago, A., Rocha, V., Vipotnik, Z., Silva, B., Tavares, T., Recovery of rare earth elements from wastewater towards a circular economy, Molecules, 24 (2019) 1005.  Baybaş, D., Ulusoy, U., Polyacrylamide–clinoptilolite/Y-zeolite composites: Characterization and adsorptive features for terbium, Journal of Hazardous Materials, 187 (2011) 241–249.  Bhushan, B., Handbook of Nanotechnology, publisher Springer, (2010) 1-5.  Binnemans, K., Tom Jones, P., Blanpain B., Van Gerven, T., Yang, Yongxiang., Walton Allan., Buchert, M., Recycling of rare earths: a critical review, Journal of Cleaner Production, 51 (2013) 1-22.  Braccini, I., Pérez S., Molecular basis of Ca2+-induced gelation in alginates and pectins: the egg-box model revisited, Biomacromolecules, 2 (2001) 1089-1096.  Brookins, D.G., Aqueous geochemistry of rare earth elements. Geochemistry and Mineralogy of Rare Earth Elements, B.R. Lipin and G.A. McKay, eds. Washington, DC, The Mineralogical Society of America, 21 (1989) 201-212.  Brzyska, W., Lanthanides and Actinides, Wydawnictwa Naukowo-Techniczne Poland, (1996) 40-41.  Chakhmouradlan, A.R., Wall, F., Rare earth elements: Minerals, mines, magnets (and more), Elements, 8 (2012) 333–340.  Chattopadhyay, K.K., Banerjee, A.N., Introduction to nanoscience and nanotechnology, Publisher PHI learnings pvt ltd, Chap 1 (2009) 1-5. 86  Chen, W., Honghui, H., Bai, T., Jiang, S., Geochemistry of monazite within carbonatite related REE deposits, Resources, 6 (2017) 51.  Chen, X.G., Park, H.J., Chemical characteristics of O-carboxymethyl chitosans related to the preparation conditions, Carbohydrate Polymers, 53 (2003) 355-359.  Chu, L., Yan, S., Xing, X.H., Sun, X., Jurcik, B., Progress and perspectives of sludge ozonation as a powerful pretreatment method for minimization of excess sludge production, Water Research, 43 (2009) 1811-1822.  Das, N., Das, D., Recovery of rare earth metals through biosorption: An overview, Journal of Rare Earths, 31 (2013) 933-943.  Das, R., Giri, S., Muliwa, A.M., Maity, A., High-performance Hg(II) removal using thiolfunctionalized polypyrrole (PPy/MAA) composite and effective catalytic activity of Hg(II)- adsorbed waste material, ACS Sustainable Chemistry & Engineering, 5 (2017) 7524-7536.  de Abreu, F.R., Campana-Filho, S.P., Characteristics and properties of carboxymethyl chitosan, Carbohydrate Polymers, 75 (2009) 214-221.  Dostal, J., Rare earth element deposits of alkaline igneous rocks, Resources, 6 (2017) 34.  Dou, J., Liu, Z., A simple and efficient synthetic method for poly(ethylene terephthalate): phenylalkyl pyrrolidinium ionic liquid as polycondensation medium, Green Chemistry, 14 (2012) 2305-2313.  Dupont, D., Brullot W., Bloemen, M., Verbiest, T., Binnemans, K., Selective uptake of rare earths from aqueous solutions by EDTA-functionalized magnetic and Nonmagnetic nanoparticles, ACS Applied Materials & Interfaces, 6 (2014a) 4980−4988.  Dupont, David., Luyten, J., Bloemen, M., Verbiest, T., Binnemans, K., Acid-stable magnetic core−shell nanoparticles for the separation of rare earths, Industrial & Engineering Chemistry Research, 53 (2014b) 15222−15229.  Esmat, M., Farghali, A.A., Khedr, M.H., El-Sherbiny I.M., Alginate-based nanocomposites for efficient removal of heavy metal ions, International Journal of Biological Macromolecules, 102 (2017) 272-283.  Esser, B.K., Volpe, A., Kenneally, J.M., Smith, D.K., Preconcentration and purification of rare earth elements in natural waters using silica-immobilized 8-hydroxyquinoline and a supported organophosphorus extractant, Analytical Chemistry, 66 (1994) 1736-1742. 87  European Commission Critical raw materials for the EU, Report of the Ad-hoc Working Group on defining critical raw materials, http://ec.europa.eu/enterprise/policies/rawmaterials/critical/index_en.htm.  Evans, E.1., Fellows, J., Coffer, A., Wood, R.D., Open complex formation around a lesion during nucleotide excision repair provides a structure for cleavage by human XPG protein, The EMBO Journal, 16 (1997) 625-638.  Fagundes-Klen M.R., Ferri, P., Martins, T.D., Tavares, C.R.G., Silva, E.A., Equilibrium study of the binary mixture of cadmium–zinc ions biosorption by the Sargassum filipendula species using adsorption isotherms models and neural network, Biochemical Engineering Journal, 34 (2007) 136-146.  Farghali, A.A., Bahgat, M., Allah, A.E., Khedr, M.H., Adsorption of Pb(II) ions from aqueous solutions using copper oxide nanostructures, Beni-Suef University Journal of Basic and Applied Sciences, 2 (2013) 61-71.  Ferron, C.J., Bulatovic, S.M., Salter, R.S., Beneficiation of rare earth oxide minerals. Materials Science Forum. Trans Tech Publications, Switzerland, (1991).  Fiset, J.F, Blais, J.F, Riveros, P.A., Review on the removal of metal ions from effluents using seaweeds, alginate derivatives and other sorbents, Journal of Water Science, 21 (2008) 283308.  Florek, J., Chalifour, F., Bilodeau, F., Larivière, D., Kleitz, F., Nanostructured hybrid materials for the selective recovery and enrichment of rare earth elements, Advanced Functional Materials, 24 (2014) 2668-2676.  Gabor, A.E. Davidescu, C.M., Negrea, A., Ciopec, M., Butnariu, M., Ianasi, C., Muntean, C., Negrea, P., Lanthanum separation from aqueous solutions using magnesium silicate functionalized with tetrabutylammonium dihydrogen phosphate, Journal of Chemical & Engineering Data, 61 )2016 ( 535−542.  Gambogi, J., 2016. Rare Earths. U.S. Geological Survey. Mineral Commodity Summaries.  Gambogi, J., Rare Earths, United States Geological Survey, Mineral Commodity Summaries, (2017).  Gambogi, J., Rare Earths, United States Geological Survey, Minerals Yearbook, (2011).  Gambogi, J., Rare Earths. USGS Geological Survey, Mineral Commodity Summaries, USGS, Reston, VA, 2013. 88  Ganguli, R., Cook, D.R., Rare earths: A review of the landscape, MRS Energy & Sustainability, (2018).  Gao, Z., Bandosz, T.J., Zhao, Z., Han, M., Qiu, J., Investigation of factors affecting adsorption of transition metals on oxidized carbon nanotubes, Journal of Hazardous Materials, 167 (2009) 357–365.  García, JM., García, F.C., Serna, F., de la Peña, J.L., High-performance aromatic polyamides, Progress in Polymer Science, 35 (2010) 623–686.  Giese, E.C., Jordão, C.S., Biosorption of lanthanum and samarium by chemically modified free Bacillus subtilis cells, Applied Water Science, 9 (2019) 182.  Gomez-Valdemoro, A., San-Jose, N., Clemente Garcia, F., De La Pena, J.L., Serna, F., Miguel Garcia, J., Novel aromatic polyamides with main chain and pendant 1,2,4-triazole moieties and their application to the extraction/elimination of mercury cations from aqueous media, Polymer Chemistry, 1 (2010) 1291–1301.  Gotoh, T., Matsushima, K., Kikuchi, K.I., Adsorption of Cu and Mn on covalently crosslinked alginate gel beads, Chemosphere, 55 (2004) 57-64.  Grebneva, O.N., Kuz'min, N.M., Tsysin, G.I., Zolotov, Y.A., On-line-sorption preconcentration and inductively coupled plasma atomic emission spectrometry determination of rare earth elements, Spectrochimica Acta Part B: Atomic Spectroscopy, 51 (1996) 1417-1423.  Guibal, E., Interactions of metal ions with chitosan-based sorbents: a review, Seperation and Purification Technology, 38 (2004) 43-74.  Gupta, C.K., Krishnamurthy, Extractive metallurgy of rare earths, CRC Press, (2004).  Hadjittofi, L., Charalambous, S., Pashalidis, I., Removal of trivalent samarium from aqueous solutions by activated biochar derived from cactus fibres, Journal of Rare Earths, 34 (2016) 99–104.  Hanbali, M., Holail, H., Hammud, H., Remediation of lead by pretreated red algae: adsorption isotherm, kinetic, column modeling and simulation studies, Green Chemistry Letters and Reviews, 7 (2014) 342–358.  Hatch, G.P., Dynamics in the global market for rare earths. Elements, 8(5) (2012) 341-346.  Hearle, JW., High-performance fibres, Woodhead Publishing Ltd, Cambridge, England (2001). 89  Hedrick, J.B., Rare earths, Us. geological survey minerals yearbook, (2004).  Henderson, P., The rare earth elements: introduction and review. In: Jones, A.P., Wall, F., Williams, C.T. (Eds.) Rare earth minerals: Chemistry, origin and ore deposits. Chapman & Hall, London, (1996) 1–19.  Hornyak, L., Tibbals, H.F., Dutta, J., Moore, J.J., Introduction to nanoscience and nanotechnology Taylor and Francis CRC Press ISBN: 978-1-4200-4779-0.  Horsfall Jr, M., Spiff, A.I., Adsorption of transition metals in aqueous solutions by fluted pumpkin (Telfairia occidentalis Hook f) waste, Chemistry & Biodiversity, 2 (2005).  Hristovski, K.D., Westerhoff, P.K., Crittenden, J.C., Olson, L.W., Arsenate removal by nanostructured ZrO2 spheres, Environmental Science & Technology, 42 (2008) 3786-3790.  Humphries, M., Rare Earth Elements: The Global Supply Chain, CRS Report for Congress, R41347 (www.crs.gov), 2012.  Ibáñez, J.P., Umetsu, Y., Uptake of trivalent chromium from aqueous solutions using protonated dry alginate beads, Hydrometallurgy, 72 (2004) 327-334.  Iftekhar, S., Srivastava, V., Casas, A., Sillanpää, M., Synthesis of novel GA-g-PAM/SiO2 nanocomposite for the recovery of rare earth elements (REE) ions from aqueous solution, Journal of Cleaner Production, 170 (2018a) 251-259.  Iftekhar, S., Srivastava, V., Hammouda, S.B., Sillanpää, M., Fabrication of novel metal ion imprinted xanthan gum-layered double hydroxide nanocomposite for adsorption of rare earth elements, Carbohydrate Polymers, 194 (2018b) 274-284.  Jalali, M., Moradi, F., Competitive sorption of Cd, Cu, Mn, Ni, Pb and Zn in polluted and unpolluted calcareous soils, Environmental Monitoring and Assessment, 185 (2013) 8831– 8846.  Janos, P., Sypecka, J., Mlckovska, P., Kuran, P., Pilarova, V., Removal of metal ions from aqueous solutions by sorption onto untreated low-rank coal (oxihumolite), Separation and Purification Technology, 53 (2007) 322-329.  Jegadeesan, G., Al-Abed, S.R., Sundaram, V., Choi, H., Scheckel, K.G., Dionysiou, D.D., Arsenic sorption on TiO2 nanoparticles: size and crystallinity effects, Water Research, 44 (2010) 965-973. 90  Johansson, N., Krook, J., Eklund, M., Berglund, B., An integrated review of concepts and initiatives for mining the technosphere: towards a new taxonomy, Journal of Cleaner Production, 55 (2013) 35-44.  Kaneko, T., Nagata, F., Kugimiya, S., Kato, K., Optimization of carboxyl-functionalized mesoporous silica for the selective adsorption of dysprosium, Journal of Environmental Chemical Engineering, 6 (2018) 5990-5998.  Kegl, T., Ban, I., Lobnik, A., Košak, A., Synthesis and characterization of novel γ-Fe2O3NH4OH@SiO2(APTMS) nanoparticles for dysprosium adsorption, Journal of Hazardous Materials, 378 (2019) 120764.  Koerth-Baker, M., 4 rare earth elements that will only get more important, (2012).  Koochaki-Mohammadpour, S.M.A., Torab-Mostaedi, M., Talebizadeh-Rafsanjani, A., Naderi-Behdani, F., Adsorption isotherm, kinetic, thermodynamic, and desorption studies of lanthanum and dysprosium on oxidized multiwalled carbon nanotubes. Journal of Dispersion Science and Technology, 35 (2014) 244−254.  Lee, Y.R., Yu, K., Ravi, S., Ahn, W.S., Selective adsorption of rare earth elements over functionalized Cr-MIL-101, ACS Applied Materials & Interfaces, 10 (2018) 23918-23927.  Li, H., XIAO, D., HE, H., Lin, R.; Zuo, P.; Adsorption behavior and adsorption mechanism of Cu(II) ions on amino-functionalized magnetic nanoparticles, The Transactions of Nonferrous Metals Society of China, 23 (2013) 2657−2665.  Liang, P., Hu, B., Jiang, Z., Qin, Y., Peng, T., Nanometer-sized titanium dioxide microcolumn on-line preconcentration of La, Y, Yb, Eu, Dy and their determination by inductively coupled plasma atomic emission spectrometry, Journal of Analytical Atomic Spectrometry, 16 (2001) 863-866.  Lim, A.P., Aris A.Z., Continuous fixed-bed column study and adsorption modeling: Removal of cadmium (II) and lead (II) ions in aqueous solution by dead calcareous skeletons, Biochemical Engineering Journal, 87 (2014) 50-61.  Lin C.C., Lin, C.W., Preparation of N,O-carboxymethyl chitosan nanoparticles as an insulin carrier, Drug Delivery, 16 (2009) 458–464.  Liou, GS., Hsiao, SH., Ishida, M., Kakimoto, M., Imai, Y., Synthesis and characterization of novel soluble triphenylamine-containing aromatic polyamides based on N,N’-bis(4- 91 aminophenyl)-N,N’-diphenyl-1,4-phenylenediamine, Journal of Polymer Science Part A: Polymer Chemistry, 40 (2002) 2810−2818.  Liu, E., Zheng, X., Xu, X., Zhang, F., Liu, E., Wang, Y., Li, C., Yan, Y., Preparation of diethylenetriamine-modified magnetic chitosan nanoparticles for adsorption of rare-earth metal ions, New Journal of Chemistry, 41 (2017) 7739-7750.  Long, K.R., Van Gosen, B.S., Foley, N.K., Cordier, D., The principal rare earth elements deposits of the United States—A summary of domestic deposits and a global perspective: U.S. Geological Survey Scientific Investigations Report, 2010–5220, 96 p.  Lucas, J., Lucas, P., Le Mercier, T., Rollat, A., Davenport, W., Rare earths: Science, technology, production and use, Elsevier Inc, (2014).  Ma, Wei., Ya, F.Q., Han, M., Wang, R., Characteristics of equilibrium, kinetics studies for adsorption of fluoride on magnetic-chitosan particle, Journal of Hazardous Materials, 143 (2007) 296-302.  Madbouly, H.A., El-Hefny, N.E., El-Nadi, Y.A., Adsorption and separation of terbium(III) and gadolinium(III) from aqueous nitrate medium using solid extractant, Separation Science and Technology, (2019) 1-13.  Manchen, N.A., Sprecher, B., Bailey, G., Ge, J., Tukker, A., Effect of Chinese policies on rare earth supply chain resilience, Resources, Conservation & Recycling, 142 (2019) 101– 112.  Mancheri, N.A., World trade in rare earths, Chinese export restrictions, and implications, Resource Policy, 46 (2015) 262-271.  Mansoori, Y., Mohsenzadeh, R., Designed polyamides based on 1,4-bis[(4-aminophenyl)- 1,3,4-oxadiazolyl]phenylene (BAPO) for removal of Cu(II) and Co(II), Designed Monomers and Polymers, 18 (2015) 1-10.  Mansur, H.S., Mansur, A.A.P., Curti, E., De Almeida, M.V., Functionalized chitosan/ quantum dots nano-hybrids for nanomedicine applications: towards biolabeling and biosorbing phosphate metabolites, Journal of Materials Chemistry B, 1 (2013) 1696-1711.  Markandeya, Dhiman, N., Shukla, S.P., Kisku, G.C., Statistical optimization of process parameters for removal of dyes from wastewater on chitosan cenospheres nanocomposite using response surface methodology, Journal of Cleaner Production, 149 (2017) 597-606. 92  Marmier, N., Delise´e, J.A., Fromage, F., Surface complexation modeling of Yb(III) and Cs(I) sorption on silica, Journal of Colloid and Interface Science, 212 (1999) 228–233.  Marmier, N., Dumonceau, A.J., Fromage, F., Surface complexation modeling of Yb(III) sorption and desorption on hematite and alumina, Journal of Contaminant Hydrology, 26 (1997) 159–167.  McLeod, C.L., Shaulis, B.J., Rare earth elements in planetary crusts: Insights from chemically evolved igneous suites on earth and the moon, Minerals, 8 )2018( 455.  Mehdinia, A., Shegefti, S., Shemirani, F., Removal of Lead(II), Copper(II) and Zinc(II) ions from aqueous solutions using magnetic amine-functionalized mesoporous silica nanocomposites, Journal of the Brazilian Chemical Society, 26 (2015) 2249-2257.  Melnyk, I.V., Goncharyk, V.P., Kozhara, L.I., Yurchenko, G.R., Matkovsky, A.K., Zub, Y.L., Alonso, B., Sorption properties of porous spray-dried microspheres functionalized by phosphonic acid groups, Microporous and Mesoporous Materials, 153 (2012) 171–177.  MEP (Ministry of Environmental Protection). Pollutant discharge standards for the Rare Earth Industry (in Chinese), (2011a).  Meyer, L., Bras, B., Rare earth metal recycling. Sustainable Systems and Technology, (2011) IEEE International Symposium, 16-18 May, Chicago Il.  Mondal N.K., Kar, S., Potentiality of banana peel for removal of Congo red dye from aqueous solution: isotherm, kinetics and thermodynamics studies, Applied Water Science, 8 (2018) 157.  Murty, D.S.R., Chakrapani, G., Preconcentration of rare earth elements on activated carbon and its application to groundwater and sea-water analysis, Journal of Analytical Atomic Spectrometry, 11 (1996) 815820.  Naser, A., El-deen, G.S., Bhran, A.A., Metwally, S., El-Kamash, A., Elaboration of impregnated composite for sorption of europiumand neodymium ions from aqueous solutions, Journal of Industrial and Engineering Chemistry, 32 (2015) 264–272.  Nayak, D., Lahiri, S., Biosorption of toxic, heavy, no-carrier-added radionuclides by calcium alginate beads, Journal of Radioanalytical and Nuclear Chemistry, 267 (2005) 59–65.  Ng, J.C.Y., Cheung, W.H., McKay, G., Equilibrium studies for the sorption of lead from effluents using chitosan, Chemosphere, 52 (2003) 1021-1030. 93  Ng, J.C.Y., Cheung, W.H., McKay, G., Equilibrium studies of the sorption of Cu(II) ions onto chitosan, Journal of Colloid and Interface Science, 255 (2002) 64-74.  Ni, D., Lin, Z., Xiaoling, P., Xinqing, W., Hongliang, G., Preparation and characterization of nickel-zinc ferrites by a solvothermal method, Rare Metal Materials and Engineering, 44 (2015) 2126-2131.  Ni, Y., Hughes, J.M., Mariano, A.N., Crystal chemistry of the monazite and xenotime structures, American Mineralogist, 80 (1995) 21-26.  Nobahar, S., Parvini, M., Eisazadeh, H., Removal of copper ions from aqueous solutions using polypyrrole and its Nanocomposites, Journal of Heat and Mass Transfer Research, 1 (2014) 101-106.  Ogata, T., Narita, H., Tanaka, M., Rapid and selective recovery of heavy rare earths by using an adsorbent with diglycol amic acid group, Hydrometallurgy, 155 (2015) 105-109.  Oliveira, D., Inverno, C., Classification of rare earth deposit and occurrence types: what’s new?, Comunicações Geológicas, 101 (2014) 821-823.  Oliveira, R.C., Jouannin, C., Guibal, E., Garcia Jr, O., Samarium(III) and praseodymium(III) biosorption on Sargassum sp.: Batch study. Process Biochemistry, 46 (3) (2011) 736–744.  Orha, C., Manea, F., Ratiu, C., Burtica, G., Iovi, A., Obtaining and characterization of romanian zeolite supporting silver ions, Environmental Engineering and Management Journal, 6 (2007) 541-544.  Pan, B.J., Pan, B.C., Zhang, W.M., Lv, L., Zhang, Q.X., Zheng, S.R., Development of polymeric and polymer-based hybrid adsorbents for pollutants removal from waters, Chemical engineering Journal, 151 (2009) 19–29.  Papandreou, A., Stournaras, C., Panias, D., Copper and cadmium adsorption on pellets made from fired coal fly ash, Journal of Hazardous Materials, 148 (2007) 538-547.  Pathania, D., Sharma S., Singh P., Removal of methylene blue by adsorption onto activated carbon developed from Ficus carica bast, Arabian Journal of Chemistry, 10 (2017) S1445S1451.  Patra, A.K., Dutta, A., Bhaumik, A., Self-assembled mesoporous g-Al2O3 spherical nanoparticles and their eficiency for the removal of arsenic from water. Journal of Hazardous Materials, 201-202 (2012) 170-177. 100  Zhang, M., Zhang, Y., Helleur, Robert., Selective adsorption of Ag+ by ion-imprinted Ocarboxymethyl chitosan beads grafted with thiourea–glutaraldehyde, Chemical Engineering Journal, 264 (2015) 56–65.  Zhao, F., Repo E., Meng, Y., Wang, X., Yin, D., Sillanpää, M., An EDTA-β-cyclodextrin material for the adsorption of rare earth elements and its application in preconcentration of rare earth elements in seawater, Journal of Colloid and Interface Science, 465 (2016) 215– 224.  Zhao, Z., Sun, X., Dong, Y., Wang, Y., Synergistic effect of acid–base coupling bifunctional ionic liquids in impregnated resin for rare earth adsorption, ACS Sustainable Chemistry & Engineering, 4 (2) (2016) 616–624.  Zheng, X., Liu, E., Zhang, F., Yan, Y., Pan, J., Efficient adsorption and separation of dysprosium from NdFeB magnets in an acidic system by ion imprinted mesoporous silica sealed in a dialysis bag, Green Chemistry, 18 (2016) 5031–5040.  Zheng, X., Zhang, Y., Bian, T., Zhang, Y., Li, Z., Pan, J., Oxidized carbon materials cooperative construct ionic imprinted cellulose nanocrystals films for efficient adsorption of Dy(III), Chemical Engineering Journal, 381 (2020) 122669.  Zhou, B., Li, Z., Chen, C., Global potential of rare earth resources and rare earth demand from clean technologies, Minerals, 7 (2017) 203.  Zhou, B., Li, Z., Zhao, Y., Zhang, C., Wei, Y., Rare earth elements supply vs. clean energy technologies: New problems to be solve, Gospodarka Surowcami Mineralnymi, 32 (2016) 29–44.  Zhu Y., Zheng, Y., Wang, A., A simple approach to fabricate granular adsorbent for adsorption of rare elements, International journal of biological macromolecules, 72 (2015) 410-420.  Zhu, Y., Zheng, Y., Wang, A., A simple approach to fabricate granular adsorbent for adsorption of rare elements, International Journal of Biological Macromolecules, 72 (2015) 410-420. 101 102 LIST OF PUBLICATIONS  Hamedreza Javadian, Montserrat Ruiz, Tawfik A. Saleh, Ana Maria Sastre, Caalginate/carboxymethyl chitosan/Ni0.2Zn0.2Fe2.6O4 magnetic bionanocomposite: Synthesis, characterization and application for single adsorption of Nd+3, Tb+3, and Dy+3 rare earth elements from aqueous media, Journal of Molecular Liquids, 306 (2020) 112760. https://doi.org/10.1016/j.molliq.2020.112760  Hamedreza Javadian, Montserrat Ruiz, Ana Maria Sastre, Response surface methodology based on central composite design for simultaneous adsorption of rare earth elements using nanoporous calcium alginate/carboxymethyl chitosan microbiocomposite powder containing Ni0.2Zn0.2Fe2.6O4 magnetic nanoparticles: Batch and column studies, International Journal of Biological Macromolecules, 154 (2020) 937-953. https://doi.org/10.1016/j.ijbiomac.2020.03.131  Hamedreza Javadian, Montserrat Ruiz, Mehdi Taghavi, Ana Maria Sastre, Synthesis of magnetic CMC bionanocomposite containing a novel biodegradable nanoporous polyamide selectively synthesized in ionic liquid as green media: Investigation on Nd+3, Tb+3, and Dy+3 rare earths adsorption, Journal of Molecular Liquids, 308 (2020) 113017. https://doi.org/10.1016/j.molliq.2020.113017  Hamedreza Javadian, Montserrat Ruiz, Mehdi Taghavi, Ana Maria Sastre, Synthesis of calcium alginate/novel selectively synthesized biodegradable poly(pyrimidine-thiopheneamide) with free hydroxyl groups in ionic liquid as green media /Ni0.2Zn0.2Fe2.6O4 magnetic bionanocomposite powder: Adsorption properties towards rare earth elements, Under Review in Microchemical Journal.  Hamedreza Javadian, Montserrat Ruiz, Mehdi Taghavi, Ana Maria Sastre, One-step hydrothermal synthesis of green synthesized poly(pyrimidine-thiophene-amide) as novel polyamide in imidazolium based ionic liquid as green media/Ni0.2Zn0.2Fe2.6O4 nanocomposite: Investigation on Nd+3, Tb+3, and Dy+3 adsorption, Under Review in Applied Organometallic Chemistry. 103 104 ANNEX I Ca-alginate/carboxymethyl chitosan/Ni0.2Zn0.2Fe2.6O4 magnetic bionanocomposite: Synthesis, characterization and application for single adsorption of Nd+3, Tb+3, and Dy+3 rare earth elements from aqueous media. Hamedreza Javadian, Montserrat Ruiz, Tawfik A. Saleh, Ana Maria Sastre. Journal of Molecular Liquids, 306 (2020) 112760. https://doi.org/10.1016/j.molliq.2020.112760 105 Ca-alginate/carboxymethyl chitosan/Ni 0.2 Zn 0.2 Fe 2.6 O 4 magnetic bionanocomposite: Synthesis, characterization and application for single adsorption of Nd +3 ,Tb +3 ,andDy +3 rare earth elements from aqueous media Hamedreza Javadian a, ⁎, Montserrat Ruiz b ,TawfikA.Saleh c, ⁎, Ana Maria Sastre a a Department of Chemical Engineering, ETSEIB, Universitat Politècnica de Catalunya, Diagonal 647, 08028 Barcelona, Spain b Department of Chemical Engineering, EPSEVG, Universitat Politècnica de Catalunya, Av. Víctor Balaguer, s/n, 08800 Vilanova i la Geltrú, Spain c Chemistry Department, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia abstractarticle info Article history: Received 22 November 2019 Received in revised form 18 February 2020 Accepted 20 February 2020 Available online 21 February 2020 Keywords: Carboxymethyl chitosan Calcium alginate Ni 0.2 Zn 0.2 Fe 2.6 O 4 Adsorption Rare earth elements This study aims to research the adsorption of Nd +3 ,Tb +3 ,andDy +3 from aqueous media onto the magnetic calcium alginate/carboxymethyl chitosan/Ni 0.2 Zn 0.2 Fe 2.6 O 4 (CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 ) bionanocomposite in a single system. FE-SEM, FT-IR, EDX, VSM, and TGA were applied to characterize the product. The VSM result showedthesaturationmagnetizationvaluesof 45.87 and14.14 emu/gfor thebareNi 0.2 Zn 0.2 Fe 2.6 O 4 nanoparticles and CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 , respectively. Theadsorption results showed that at optimum conditions ofcontact time of 40 min, pH of 5.5, and 0.8 g/L, the adsorption efficiency of the adsorbent for Nd +3 ,Tb +3 ,andDy +3 was 97.75, 96.83, and 97.85%, respectively. The ions adsorption kinetic onto the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 was in accordance with pseudo-second-order (PSO) model. The evaluation of equilibrium data was performed by the isotherm models of Langmuir and Freundlich. Fitting the experimental data of Tb +3 and Dy +3 was done better with Freunlich model than Langmuir model,while fitting tests for Nd +3 adsorption data showed better coverage using Langmuir model with a maximum adsorption capacity of 73.37 mg/g. The results of the parameters of thermodynamic showed the endothermic and spontaneous properties of the process. Additionally, the efficacy of the adsorbent was studied using 0.2 M HNO 3 in four adsorptions–desorption cycles. Overall, the obtained results demonstrated that the environmentally friendly magnetic bionanocomposite adsorbent can be applied effectively for Nd +3 ,Tb +3 ,andDy +3 adsorption with favorable adsorption efficiency. © 2020 Published by Elsevier B.V. 1. Introduction Rare earth elements (REEs) are getting impressive considerations and progressively requested in innovative industries in view of their novel properties [1]. As of late, they have been discovered broad demands in batteries, electronics, and chemical engineering [2,3]. Because of enormous and expanding local requests, China reduced its amount of REEs send out from 50,145 tons in 2009 to just 31,130 tons in 2012. These fare quantities may create difficult issues for REE applicants outside of China, as proved by the crisis of REEs in 2011, recording high costs of these elements [4]. This circumstance has additionally animated many nations,for example, Japanand most EUMember States that don't have anysort of essential REEs stores ontheir region to search for option and auxiliary resources of REEs and to extend their own REEs industry so as to get a wellspring of REEs, especially heavy REEs [5]. In this manner, a productive method is expected to overcome all difficult issues with respect to REEs. The conventional methodologies applied for REEs recovery are chemical precipitation, membrane separation, ion-exchange, reverse osmosis, extraction, and adsorption [6–8]. However, each method has its advantages and disadvantages. For instance, chemical precipitation has the advantage of low-cost and simple operation, but large amounts of chemical products are produced, resulting in landfill problems [9]. Membrane separation method has theadvantages of high separation efficiency for heavy metal ions, yet low economic feasibility and high maintenance cost restrict its application on a large scale. Adsorption technology, which is easy to perform, highly effective, and low-cost, is considered as a fast and relatively inexpensive approach for metal ions adsorption [10]. Adsorption of REEs has been investigated by some materials suchas biosorbents [11–13], carbon [14,15], silicaother inorganic [16–18], and polymeric materials [19,20]. Journal of Molecular Liquids 306 (2020) 112760 ⁎Corresponding authors. E-mail addresses: [email protected],[email protected] (H. Javadian), tawfi[email protected] (T.A. Saleh). https://doi.org/10.1016/j.molliq.2020.112760 0167-7322/© 2020 Published by Elsevier B.V. Contents lists available at ScienceDirect Journal of Molecular Liquids journal homepage: www.elsevier.com/locate/molliq Alginate, as a valuable natural polymer, has pulled in extreme consideration. It is an ordinary polysaccharide that is made up of the residues of mannuronic (M) and guluronic (G) acid (linear copolymer of β-D-mannuronic acid and α-L-guluronic acid units with(1–4)linkages) [21]. This environmentally friendly polymer has highlights of cheapness, plentiful sources, biocompatibility, and hydrophilicity. More often than not, industrially accessible alginates are extraction of brown algae cell wall [22]. It has been broadly utilized in immobilization studies owing to easy preparation, hydrophilicity and efficient adsorption of target contaminations. Chitosan (CS) as a natural polysaccharide is generally made using the deacetylation of chitin. Its derivatives can be generated by the functional groups' modification, for instance, reactive hydroxyl, amino, and N-acetyl groups [23]. Carboxymethyl chitosan (CMC) is considered as the most significant derivatives amongst others. It is the result of the carboxylation of chitosan that has carboxymethyl substituents on amino and hydroxyl groups of the glucosamine units [24]. Because of having special characteristics such as nontoxicity, hydrophilicity, biodegradability, environmentally friendly, and metal-chelating capacity, it is viewed as a potential candidate for bioadsorption. Nevertheless, it couldn't be used for the recovery of ions because of being waterdissolvable and having weak chemical stability [25]. So as to defeat this issue, modification of CMC by other biopolymers such as alginate and inorganic nanoparticles can be regarded as one of the best techniques to increase its hydrolysis resistance. The purpose of this study was to synthesize the Ni 0.2 Zn 0.2 Fe 2.6 O 4 magnetic nanoparticles by the hydrothermal technique and using the synthesized nanoparticles for the synthesis of a bionanocomposite in gelation process of sodium alginate and carboxymethyl chitosan biopolymers in a medium of CaCl 2 and glutaraldehyde. After characterizing the product with various techniques including FE-SEM, EDX, XRD, FT-IR, and VSM, it was used as an adsorbent to investigate the adsorption efficiency of Nd +3 ,Tb +3 , and Dy +3 depending on adsorbent dosage, pH, contact time, initial metal ion concentration by performing a series of batch experiments. Various kinetic and isotherm models were tested for fitting the experimental data. Thermodynamic parameters (ΔS ° , ΔG ° , and ΔH ° ) were also evaluated to find the property of adsorption process. To the best of our knowledge, the utilization of the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 has not been studied for Nd +3 ,Tb +3 ,andDy +3 adsorption. 2. Materials and methods 2.1. Materials and reagents Carboxymethyl chitosan and sodium alginate were purchased from Nantong Chem-Base Co, China, and PanReac AppliChem, respectively. Nd(NO 3 ) 3 .6H 2 O, Tb(NO 3 ) 3 ·6H 2 O, Dy(NO 3 ).5H 2 O, Zn(NO 3 ) 2 , Fe(NO 3 ) 3 .6H 2 O, Ni(NO 3 ) 2 ·6H 2 O, and glutaraldehyde were bought from SigmaAldrich. Since the analytical grade of all chemicals was chosen, they were utilized without further purification. The experiment solutions of Nd +3 ,Tb +3 , and Dy +3 were prepared by dilution of 1000 mg/L of ions. The values of pH were regulated via adding a suitable amount of 0.1 M sodium hydroxide or nitric acid solutionsand monitored by a pH meter. 2.2. Instrumentation and characterization FT-IR spectra were recorded on a PerkinElmer, USA, by KBr pellet. XRD pattern was recorded by a GBC MMA instrument with CuK α radiation in the 2θrange of 10–70 0 . A FE-SEM (Zeiss Neon-40, Germany) was also utilized for characterizing the products morphology. The magnetic properties of the products were explored at the room temperature (RT) by employing a VSM (Daghigh Kavir Corporation, Iran). TGA was done on a Mettler TGA/SDTA 851e/LF/1100 thermobalance. The temperature of the sample was increased from RT to 1000 °C (rate = 10 °C/min) under constant nitrogen flow. For analyzing Nd +3 ,Tb +3 ,andDy +3 concentration, an Agilent 4100 MP-AES Spectrometer was used. 2.3. Synthesis of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 magnetic nanoparticles The Ni 0.2 Zn 0.2 Fe 2.6 O 4 magnetic nanoparticles were synthesized by the hydrothermal method. A mixed solution of 0.2 M Ni 2+ ,0.2M Zn 2+ and 2.6 M Fe 3+ was prepared in HCl solution, and then NaOH solution was introduced under nitrogen gas and the value of the pH of the mixture was regulated to 10.5. To this mixture, 0.3 g of CTAB was added, and then it was placed into an autoclave (Teflon-lined stainless steel) at 200 °C of an oven for 8 h for hydrothermal treatment. In the following, the temperature of the autoclave was naturally decreased to RT. The precipitate was then collected and washed with deionized water (DW) several times to reach pH = 7. Finally, the obtained particles were dried at 50 °C. 2.4. Synthesis of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 magnetic bionanocomposite The synthesis procedure of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 wasasfollows: Sodium alginate (1 g) was dissolved in 80 mL of DW atRT by a laboratory stirrer. 0.50 g of the carboxymethyl chitosan was introduced into the solution and homogeneously mixed. 0.7 g of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 wasadded to the mixture of the biopolymers. Toobtain a homogeneous blend solution, the mixture of biopolymers and magnetic particles was stirred at the RT for 24 h. In the following, it was added into the solution of calcium chloride 0.05 M and 2% glutaraldehyde for gelation process. After the completion of gelation process, an external magnetic field was utilized for the separation of the resulting bionanocomposite, and then it was washed using DW several times to eliminate any remaining calcium chloride and glutaraldehyde until the pH of the solution reached 7. The washed CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 bionanocomposite was dried at 50 °C. Finally, it was powdered. 2.5. Batch adsorption and reusability studies The stock solutions of metal ions were prepared by dissolving Tb (NO 3 ) 3 ·6H 2 O, Nd(NO 3 ) 3 .6H 2 O, and Dy(NO 3 ).5H 2 O separately in DW to achieve 1000 mg/L of each ion, and all the experiment solutions containing 50mL of single ion prepared by dilution of each stock solution to the required concentration were agitated at 180 rpm by a laboratory shaker. Equal concentration of the ions was applied in single batch adsorption studies. For studying the influences of pH and adsorbent dosage on the ions adsorption efficiency, batch adsorption experiments were performed in the pH range of 1.5–5.5 and dosage range of 0.01–0.06 g, respectively. For the kinetic evaluation, the adsorbent was added into the solutions with 30 mg/L initial concentration, and the tests were performed at different contact time (2.5–70 min). Initial metal concentration in the range of 30–300 mg/L at the optimum time was used to investigate the isotherm of the adsorption. The effect of ionic strength was studied with NaNO 3 solution at various concentrations of 0.02, 0.04, 0.06, 0.08, and 0.1 M. To evaluate thermodynamic parameters, the experiments were carried out at three various temperatures of 25, 35 and 45 °C at a constant initial concentration of 90 mg/L. The ions concentration in the solution was measured by an Agilent 4100 MP-AES Spectrometer. The adsorption efficiency (%) and capacity of metal adsorption by the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 were computed using the equations in the following: Adsorption efficiency %ðÞ¼C0−Ce ðÞ=Co100 ð1Þ qe¼C0−Ce ðÞV=mð2Þ qt¼C0−Ct ðÞV=mð3Þ 2H. Javadian et al. / Journal of Molecular Liquids 306 (2020) 112760 where the q e and q t (mg/g) refer to the quantities of metal ionadsorbed at equilibrium and adsorption time t in min, respectively. C 0 and C e show the initialand equilibriumconcentrationsof metal ion in mg/L, respectively. C t refers to the concentration of a metal ion in solution at time t (min). Moreover, m refers to the adsorbent weight (g), and V refers to the solution volume in L. To investigate the reusability of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 ,a given amount of the adsorbent was treated with 50 mL of 30 mg/L metal solution at pH of 5.5 by a shaker at speed of 180 rpm to obtain the exhausted adsorbent. The ions-loaded adsorbent was separated with an external magnetic field, washed by DW to eliminate the unadsorbed ions and then agitated for 2 h by 50 mL HNO 3 (0.2 M) eluent solution. Subsequently, the regenerated CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 was separated and washed several times with DW until the solution pH reached 7. The regenerated adsorbent was applied in four cycles of adsorption with the same regeneration procedure. 3. Results and discussion 3.1. Analyses of the products Fig. 1 indicates the XRD pattern of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 . The peaks at 2θ= 18.13°, 30.07 ° ,35.50 ° ,37.08 ° ,43.07 ° ,53.95 ° ,56.96 ° , and 63.89 ° are in agreement with the standard pattern of nickel zinc ferrite (JCPDS 080234) [26]. Full Width at Half Maximum (FWHM) of the strongest reflection of the XRD pattern was utilized to estimate the average crystal size based on the Scherrer equation as following [27]: D¼kλ=βcos θð4Þ Where k shows the shape function, 0.89, λrefers to the X-ray radiation wavelength, βrefers to the FWHM at 2θ=35.50 ° ,andθshows the diffraction angle. Basedon the equation of Scherrer, thecalculated value of D was 27.68 nm. The Ni 0.2 Zn 0.2 Fe 2.6 O 4 FE-SEM image in Fig. 2A indicates that the synthesized particles are nearly spherical in shape and homogenous in distribution with a diameter of b100 nm. Fig. 2B shows the distribution of the magnetic nanoparticles on the surface of the CA/CMC or embedding with the CA/CMC that confirms the successful synthesis of the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 magnetic bionanocomposite. Fig. 3 indicates the FT-IR spectra of CA, CMC, Ni 0.2 Zn 0.2 Fe 2.6 O 4 , and CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 . The FT-IR spectra of CA and CMC respectively in Fig. 3AandBshowO\\H stretching vibration at 3389 (CA) and 3436 (CMC) cm −1 , carboxylic groups asymmetrical stretching at 1622 (CA) and 1631 (CMC) cm −1 , carboxylic groups symmetrical stretching at 1423 (CA) and 1411 (CMC) cm −1 and C-O-C stretching at 1052 (CA) and 1061 (CMC) cm −1 [28,29]. The FT-IR spectrum of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 in Fig. 3C shows a broad absorption band with a value of 3424 cm −1 and less intensive band at 1633 cm −1 related to the O\\H groups stretching vibration [30]. The bands at 2925 and 2853 cm −1 respectively correspondto the anti-symmetric and symmetric C\\HvibrationsofCTAB[31]. The band at 567 cm −1 relates to the inherent metal stretching vibrations at the tetrahedral site (Fe\\O), and the value of 478 cm −1 is related to the octahedral metal (M-O) stretching [30]. The comparison of the spectrum in Fig. 3D with other spectra expresses the successful synthesis of the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 . EDX was recorded to analyze the elements of the products (Fig. 4). Fig. 4A shows Ni, Zn, Fe, and O peaks that confirm the formation of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 . The elemental analysis of the nanocomposite in Fig. 4B represents similar peaks available in Fig. 4A along with the new peaks for N and Ca because of combining the nanoparticles with CA and CMC. Sodium peak is not seen in the spectrum of the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 , suggesting that sodium ions were released completely from the matrix of sodium alginate into the solution during the crosslinking reaction process of sodium alginate with calcium. The CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 weight loss curve recorded in the range of RT to 1000 °C is demonstrated in Fig. 5A. As it is seen, there are three different weight-loss steps in the TGA curve of the CA/CMC/ Fig. 1. (A) XRD pattern of Ni 0.2 Zn 0.2 Fe 2.6 O 4 nanopartciles; Photo of Ni 0.2 Zn 0.2 Fe 2.6 O 4 nanoparticles (B) before drying and (C) in the solution under magnetic field after drying. 3H. Javadian et al. / Journal of Molecular Liquids 306 (2020) 112760 Ni 0.2 Zn 0.2 Fe 2.6 O 4 .Obviously,thefirst step (around 190 °C) with a weight loss of 8.77% can be attributed to trapped and physisorbed water evaporation. The second step between around 190 and 550 °C is the largest weight loss with the amount of 35.08% that could be due to the sorption and degradation of CA and CMC. The last step with 22.95% weight loss at a temperature beyond 550 °C could be related to the further decomposition of CA and CMC and their conversion to CO 2 and H 2 O. At the end of the process, the residue percentage is about 33.2% that is principally assigned to the presence of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 . An important issue related to the magnetic bionanocomposite is that it should possess sufficient magnetic properties for its practical application. According to the magnetic hysteresis loops in Fig. 5B, thesaturation magnetizationvalue for Ni 0.2 Zn 0.2 Fe 2.6 O 4 is about 45.87emu/g that indicates superparamagnetic behavior of the synthesized product. It is obvious from Fig. 5C that the process of the synthesis of the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 results in a decrease of saturation magnetization to the value of 14.14 emu/g. This decline is due to the combination of the magnetic nanoparticles with CA and CMC. Despite this difference, the ions-loaded CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 could be easily separated by applying an external magnetic field from aqueous solution to avoid secondary pollution as indicated in Fig. 5D. 3.2. pH effect pH of the solution as a key parameter of the adsorption process affects solution chemistry, metal speciation, adsorption capacity, the activity of adsorbent functional groups, and mechanism of adsorption. It is associated directly with H + competition with ions for the occupation of the surface active sites of the adsorbent. Adsorption efficiency values of Nd +3 ,Tb +3 ,andDy +3 onto the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 as a pH function are demonstrated in Fig. 6A. Electrostatic interaction could have a key role on Nd +3 ,Tb +3 ,andDy +3 adsorption onto the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 at different pH values. It can be interpreted that at lower pH values, the adsorption efficiency of the adsorbent is low and increase with increasing solution pH. At pH = 1.5 that is highly acidic, the protonation of the adsorbent functional groups blocks the metal ions approach to the binding sites of the adsorbent. The metal ions and H + ionscompeteforthesame binding sites of the adsorbent, leading to decrease in adsorption efficiency [32]. At pH = 1.5, adsorption efficiency value is zero for all ions. When pH values are adjusted between 2.5 and 5.5, adsorption efficiency for Nd +3 ,Tb +3 ,andDy +3 onto the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 increases with pH increase owing to the reduction in competition between H + ions and metal ions. The maximum adsorption efficiency for all metal ions occurs at pH = 5.5. pH increase beyond 5.5 was not investigated to prohibit the precipitation of the ions in the form of hydroxide. Further experiments were carried out at pH = 5.5 as optimum pH. 3.3. Contact time effect It is essential to consider the adsorption rate in designing the batch experiments. The influence of contact time on Nd +3 ,Tb +3 , and Dy +3 adsorption onto the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 bionanocomposite is shown in Fig. 6B. As can be observed, adsorption efficiency of the adsorbent increases rapidly during the first period and then increases slowly until reaching equilibrium state. The experimental outcomes indicate that Nd +3 ,Tb +3 ,andDy +3 adsorption can be split into two definite parts: an extremely rapid initial adsorption occurs in the first 10 min, and then much slower adsorption is seen for higher contact time. In general, approximately 80% of the metal ions contact quickly in the first 10 min because of the presence of largely accessible active sites of the adsorbent and then slowly increase owing to a gradual decrease in the active sites and weakness of the driving force and finally adsorption reaches equilibrium [33]. The contact time of 40 min was taken as an optimum time for adsorption of Nd +3 , Tb +3 ,andDy +3 . Fig. 3. FT-IR spectra of (A) CA, (B) CMC, (C) Ni 0.2 Zn 0.2 Fe 2.6 O 4 ,and(D)CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 . Fig. 2. FE-SEM images of (A) Ni 0.2 Zn 0.2 Fe 2.6 O 4 and (B) CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 . 4H. Javadian et al. / Journal of Molecular Liquids 306 (2020) 112760 [43] I.V. Melnyk, V.P. Goncharyk, L.I. Kozhara, G.R. Yurchenko, A.K. Matkovsky, Y.L. Zub, B. Alonso, Sorption properties of porous spray-dried microspheres functionalized by phosphonic acid groups, Microporous Mesoporous Mater. 153 (2012) 171–177. [44] T. Kegl, I. Ban, A. Lobnik, A. Košak, Synthesis and characterization of novel γ-Fe 2 O 3 - NH 4 OH@SiO 2 (APTMS) nanoparticles for dysprosium adsorption, J. Hazard. Mater. 378 (2019), 120764. [45] A. Negrea, A. Gabor, C.M. Davidescu, M. Ciopec, P. Negrea, N. Duteanu, A. Barbulescu, Rare earth elements removal from water using natural polymers, Sci. Rep. 8 (2018) 316. [46] Y.R. Lee, K. Yu, S. Ravi, W.S. Ahn, Selective adsorption of rare earth elements over functionalized Cr-MIL-101, Appl. Mater. Interfaces 10 (2018) 23918–23927. [47] X.-H. Qi, K.-Z. Du, M.-L. Feng, Y.-J. Gao, X.-Y. Huang, M.G. Kanatzidis, Layered A 2 Sn 3 S 7 ·1.25H 2 O(A= organic cation) as efficient ion-exchanger for rare earth element recovery, J. Am. Chem. Soc. 139 (12) (2017) 4314–4317. [48] S.M.A. Koochaki-Mohammadpour, M. Torab-Mostaedi, A. Talebizadeh-Rafsanjani, F. Naderi-Behdani, Adsorption isotherm, kinetic, thermodynamic, and desorption studies of lanthanum and dysprosium on oxidized multiwalled carbon nanotubes, J. Dispers. Sci. Technol. 35 (2014) 244–254. [49] M.M. Yusoff, N.R. Nik Mostapa, Md. Shaheen Sarkar, T.K. Biswas, Md. Lutfor Rahman, S.E. Arshad, M.S. Sarjadi, A.D. Kulkarni, Synthesis of ion imprinted polymers for selective recognition and separation of rare earth metals, J. Rare Earth 35 (2017) 177–186. [50] H.A. Madbouly, N.E. El-Hefny, Y.A. El-Nadi, Adsorption and separation of terbium (III) and gadolinium(III) from aqueous nitrate medium using solid extractant, Sep. Sci. Technol. (2019) 1–13. [51] S. Tong, S. Zhao, W. Zhou, R. Li, Q. Jia, Modification of multi-walled carbon nanotubes with tannic acid for the adsorption of La, Tb and Lu ions, Microchim. Acta 174 (2011) 257–264. [52] D. Baybaş, U. Ulusoy, Polyacrylamide–clinoptilolite/Y-zeolite composites: characterization and adsorptive features for terbium, J. Hazard. Mater. 187 (1) (2011) 241–249. [53] R. Akkaya, Synthesis and characterization of a new low-cost composite for the adsorption of rare earth ions from aqueous solutions, Chem. Eng. J. 200 (2012) 186–191. [54] R. Akkaya, Terbium adsorption onto polyhydroxyethylmethacrylate–hydroxyapatite composite and its modified composition by phytic acid, Desalin. Water Treat. 52 (2014) 1440–1447. [55] F.A. Alakhras, K.A. Dari, M.S. Mubarak, Synthesis and chelating properties of some poly (Amidoxime-hydroxamic acid) resins toward some trivalent lanthanide metal ions, J. Appl. Polym. Sci. 97 (2) (2005) 691–696. [56] M.A. Attia, S.I. Moussa, R.R. Sheha, H.H. Someda, E.A. Saad, Hydroxyapatite/NiFe 2 O 4 superparamagnetic composite: facile synthesis and adsorption of rare elements, Appl. Radiat. Isot. 145 (2019) 85–94. [57] N.S. Reddy, K.M. Rao, S.V. Krishna, C.S. Ha, Synthesis of 1–acryloyl–3–phenyl thiourea based pH sensitive hydrogels for removal of samarium and terbium, Macromol. Res. 24 (2016) 494–501. [58] M.R. Awual, N.H. Alharthi, Y. Okamoto, M.R. Karim, M.E. Halim, M.M. Hasan, M.M. Rahman, M.M. Islam, M.A. Khaleque, M.C. Sheikh, Ligand field effect for Dysprosium(III) and lutetium(III) adsorption and EXAFS coordination with novel composite nanomaterials, Chem. Eng. J. 320 (2017) 427–435. [59] R.M. Ashour, R. El-Sayed, A.F. Abdel-Magied, A.A. Abdel-Khalek, M.M. Ali, K. Forsberg, A. Uheida, M. Muhammed, J. Dutta, Selective separation of rare earth ions from aqueous solution using functionalized magnetite nanoparticles: kinetic and thermodynamic studies, Chem. Eng. J. 327 (2017) 286–296. [60] H. Aghayan, A. Mahjoub, A. Khanchi, Samarium and dysprosium removal using 11molybdo-vanadophosphoric acid supported on Zr modified mesoporous silica SBA15, Chem. Eng. J. 225 (2013) 509–519. [61] T. Kaneko, F. Nagata, S. Kugimiya, K. Katob, Optimization of carboxyl-functionalized mesoporous silica for the selective adsorption of dysprosium, J Environmen Chem Eng 6 (2018) 5990–5998. [62] X. Zheng, Y. Zhang, T. Bian, Y. Zhang, Z. Li, J. Pan, Oxidized carbon materials cooperative construct ionic imprinted cellulose nanocrystals films for efficient adsorption of Dy(III), Chem. Eng. J. 381 (2020), 122669. [63] X. Zheng, E. Liu, F. Zhang, Y. Yan, J. Pan, Efficient adsorption and separation of dysprosium from NdFeB magnets in an acidic system by ion imprinted mesoporous silica sealed in a dialysis bag, Green Chem. 18 (2016) 5031–5040. [64] İ. Kara, D. Yilmazer, S. Tunali Akar, Metakaolin based geopolymer as an effective adsorbent for adsorption of zinc(II) and nickel(II) ions from aqueous solutions, Appl. Clay Sci. 139 (2017) 54–63. [65] D. Alipour, A.R. Keshtkar, M.A. Moosavian, Adsorption of thorium (IV) from simulated radioactive solutions using a novel electrospun PVA/TiO 2 /ZnO nanofiber adsorbent functionalized with mercapto groups: study in single and multicomponent systems, Appl. Surf. Sci. 366 (2016) 19–29. [66] M. Rajabi, K. Mahanpoor, O. Moradi, Preparation of PMMA/GO and PMMA/GO-Fe 3 O 4 nanocomposites for malachite green dye adsorption: kinetic and thermodynamic studies, Compos Part B Eng 167 (2019) 544–555. [67] H. Javadian, P. Vahedian, M. Toosi, Adsorption characteristics of Ni(II) from aqueous solution and industrial wastewater onto Polyaniline/HMS nanocomposite powder, Appl. Surf. Sci. 284 (2013) 13–22. [68] N. Subedi, A. Lähde, E. Abu-Danso, J. Iqbal, A. Bhatnagar, A comparative study of magnetic chitosan (Chi@Fe 3 O 4 ) and graphene oxide modified magnetic chitosan (Chi@Fe 3 O 4 GO) nanocomposites for efficient removal of Cr(VI) from water, Int. J. Biol. Macromol. 137 (2019) 948–959. 11H. Javadian et al. / Journal of Molecular Liquids 306 (2020) 112760 117 118 ANNEX II Response surface methodology based on central composite design for simultaneous adsorption of rare earth elements using nanoporous calcium alginate/carboxymethyl chitosan microbiocomposite powder containing Ni0.2Zn0.2Fe2.6O4 magnetic nanoparticles: Batch and column studies. Hamedreza Javadian, Montserrat Ruiz, Ana Maria Sastre, International Journal of Biological Macromolecules, 154 (2020) 937-953. https://doi.org/10.1016/j.ijbiomac.2020.03.131 119 Response surface methodology based on central composite design for simultaneous adsorption of rare earth elements using nanoporous calcium alginate/carboxymethyl chitosan microbiocomposite powder containing Ni 0.2 Zn 0.2 Fe 2.6 O 4 magnetic nanoparticles: Batch and column studies Hamedreza Javadian a, ⁎, Montserrat Ruiz b , Ana Maria Sastre a a Department of Chemical Engineering, ETSEIB, Universitat Politècnica de Catalunya, Diagonal 647, 08028 Barcelona, Spain b Department of Chemical Engineering, EPSEVG, Universitat Politècnica de Catalunya, Av. Víctor Balaguer, s/n, 08800 Vilanova i la Geltrú, Spain abstractarticle info Article history: Received 22 November 2019 Received in revised form 20 February 2020 Accepted 14 March 2020 Available online xxxx Keywords: Calcium alginate Carboxymethyl chitosan Ni 0.2 Zn 0.2 Fe 2.6 O 4 Adsorption Rare earth elements RSM In this research paper, the utilization of the magnetic calcium alginate/carboxymethyl chitosan/Ni 0.2 Zn 0.2 Fe 2.6 O 4 (CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 ) was investigated for the simultaneous aqueous adsorption of Nd (III), Tb (III), and Dy (III). The magnetic products were characterized by FE-SEM, EDX, XRD, FT-IR, TGA, and VSM techniques. The saturation magnetization value for Ni 0.2 Zn 0.2 Fe 2.6 O 4 and CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 was found to be 45.87 and 14.14 emu/g, respectively. Using RSM, a quadratic polynomial equation was obtained to predict the adsorption efficiency of each ion. Under the conditions of pH = 5.5, adsorbent dosage of 0.1 g, initial concentration of 30 mg/L, and contact time of 53 min predicted by RSM, the adsorption efficiencies of Nd (III), Tb (III), and Dy (III) were respectively 95.72, 96.17, and 99.44%. The isotherm and kinetic data were respectively fitted well with Freundlich and pseudo-second-order (PSO) models. The desorption of the loaded ions was effectively carried out by 0.2 M HNO 3 , and the adsorbent was consecutively utilized with 2.54, 1.63, and 1.16% decrease in adsorption efficiency for Nd (III), Tb (III), and Dy (III), respectively, after the forth cycle. Additionally, the adsorption behavior of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 towards Nd (III), Tb (III), and Dy (III) was studied by using a fixed-bed column technique. © 2020 Published by Elsevier B.V. 1. Introduction Rare earth elements (REEs) are regularly alluded in terms of “seeds of technology”due to their utilization in electronic devices, high strength lasting magnets, green energy sectors, lasers, automotive catalytic converters, fiber superconductors/optics, etc. [1,2]. Because of the progressing advancement in new trend-setting innovations, there is an over-expanding interest for REEs in the universal markets, with an accentuation on distinguishing new origins to guarantee satisfactory supply for utilizing in the present and future. This issue becomes more important since N90% of mine production of rare earth occurs in China, and its REEs export was decreased 19,015 tonsfrom 2009 to 2012, leading to serious problems for REE users outside of China. Therefore, REEs recovery from wastes has been one of the most incredible worries in the ongoing years [3]. A few techniques, such as solvent extraction, chemical precipitation, ion exchange, membrane separation, adsorption and so on, have been applied for the REEs recovery [4,5]. Investigators have noted adsorption as a standout technique because of being easy, cost-effective, and environmentally friendly for REEs recovery in comparison with the regular techniques [6]. Alginate, as a natural biopolymer, is extracted from brown algae. Some of its benefits, such as biodegradability, biocompatibility, being cheap and nontoxic, make it a great potential material to be broadly and effectively utilized in water treatment [7–10]. It tends to be utilized to produce hydrogels under conditions of moderate pH and temperatures. Alginate can likewise be altered via physicochemical procedures to enhance its chemical and mechanical strength [11]. In this manner, its adsorption behavior can beincreased byraising itsadsorption capacity [12]. The utilization of alginate in theform of hydrogel beads is a typical technique to enhance its adsorption capacity [13]. International Journal of Biological Macromolecules 154 (2020) 937–953 ⁎Corresponding author at: Department of Chemical Engineering, ETSEIB, Universitat Politècnica de Catalunya, Diagonal 647, 08028 Barcelona, Spain E-mail address: [email protected] (H. Javadian). https://doi.org/10.1016/j.ijbiomac.2020.03.131 0141-8130/© 2020 Published by Elsevier B.V. Contents lists available at ScienceDirect International Journal of Biological Macromolecules journal homepage: http://www.elsevier.com/locate/ijbiomac Chitosan-based adsorbent materials are broadly applied for the adsorption removal of contaminations aqueous solution [14]. Chitosan possesses valuable characteristics, for example, biodegradability, hydrophilicity, nontoxicity, biocompatibility, high mechanical strength, film preparation, and antibacterial characteristics [15].Its chemical structure contains amino(-NH 2 ) and hydroxyl (-OH) groups as major active functional groups for adsorption of metal ions from aqueous media [16,17]. The dissolvability of chitosan can be enhanced by modifying its structure with –COOH groups without influencing on the mentioned characteristics [18]. Besides, the carboxyl gathering presented in carboxymethyl chitosan (CMC) is additionally useful for the adsorption of metal ions. It is necessary to easily separate adsorbents that are applied in the form of ultrafine powder for the separation of metal ions from aqueous media. For this purpose, centrifugation and filtration methods are not efficient to completely separate such a powder from aqueous media, while adsorbents having magnetic properties can be easily separated using an external magnetic field [19]. The active surface and small size of nanoparticles lead to their easy aggregation in aqueous media [20]. To solve this issue, amending the stability, increasing the application of magnetic nanoparticles, and their combining with biopolymers such as alginate and carboxymethyl chitosan could be considered as effective methods. The goal of this study was to synthesize the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 as a magnetic adsorbent for simultaneous adsorption of Nd (III), Tb (III), and Dy (III) ions from aqueous solution. The adsorbent was analyzed by FE-SEM, EDX, XRD, FT-IR, TGA, and VSM techniques. The influences of adsorbent dosage, contact time, and initial concentration as main parameters were studied, and RSM-CCD was used to optimize them. The kinetic and isotherm models were applied for fitting the experimental data. The performance of the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 was also evaluated in a fixed-bed column. 2. Materials and methods 2.1. Materials and reagents Carboxymethyl chitosan and sodium alginate were purchased from Nantong Chem-Base Co, China, and PanReac AppliChem, respectively. Dy(NO 3 ) 3 .5H 2 O was purchased from Alfa Aesar. Nd(NO 3 ) 3 .6H 2 O, Tb (NO 3 ) 3 ·6H 2 O, Zn(NO 3 ) 2 .6H 2 O, Fe(NO 3 ) 3 .9H 2 O, Ni(NO 3 ) 2 ·6H 2 O, CaCl 2 , and glutaraldehyde were bought from Sigma-Aldrich. Analytical grade materials were used without further purification. The experiment solutions of Nd (III), Tb (III), and Dy (III) ions were made by dilution of 1000 mg/L of ions. HNO 3 or NaOH solution with the molarity of 0.1 was utilized to carefully adjust the pH of the solutions. Table 1 CCD levels, experimental design and the responses. Symbol −Unit Levels −−α(−2) Low (−1) Central (0) High (1) +α(+2) X 1 Contact time min 2 19 36 53 70 X 2 Adsorbent dosage g 0.05 0.125 0.2 0.275 0.35 X 3 Nd (III) concentration mg/L 30 45 60 75 90 X 4 Tb (III) concentration mg/L 30 45 60 75 90 X 5 Dy (III) concentration mg/L 30 45 60 75 90 Run Factors Nd adsorption efficiency (%) Tb adsorption efficiency (%) Dy adsorption efficiency (%) X1 X2 X3 X4 X5 Observed Predicated Observed Predicated Observed Predicated 1 36 0.2 60 60 30 90.76 90.92 95.85 96.19 97.53 98.38 2 53 0.275 75 75 75 94.56 95.81 96.87 97.99 98.11 99.49 3 36 0.2 90 60 60 87.96 86.97 90.53 89.74 92.89 92.17 4(C) 36 0.2 60 60 60 91.45 91.56 94.03 93.83 95.75 95.59 5 19 0.125 45 45 75 67.08 65.91 70.16 68.76 73.17 71.86 6 36 0.2 60 90 60 89.7 90.18 91.64 91.83 92.85 92.90 7 19 0.125 75 75 75 54.56 53.65 57.87 57.07 59.06 58.32 8 19 0.125 75 45 45 62.65 61.90 67.59 66.71 72.45 71.31 9 70 0.2 60 60 60 88.12 86.43 90.56 88.55 92.92 90.59 10 53 0.125 75 75 45 71.98 71.84 77.64 77.97 79.51 79.74 11(C a ) 36 0.2 60 60 60 91.97 91.56 94.46 93.83 96.01 95.59 12 2 0.2 60 60 60 65.78 67.47 67.73 70.34 69.45 72.25 13 53 0.275 75 45 45 92.56 93.97 95.54 96.59 97.98 98.96 14 19 0.275 75 45 75 89.1 89.74 91.94 91.86 95.7 95.71 15 53 0.125 45 45 45 87.6 87.20 90.19 89.92 92.85 92.51 16 36 0.2 30 60 60 94.46 95.45 96.21 97.60 98.25 99.45 17(C) 36 0.2 60 60 60 91.72 91.56 94.09 93.83 95.71 95.59 18(C) 36 0.2 60 60 60 90.96 91.56 93.52 93.83 95.35 95.59 19(C) 36 0.2 60 60 60 91.58 91.56 93.85 93.83 95.32 95.59 20 36 0.2 60 30 60 95.64 95.16 94.43 94.84 98.65 99.07 21 19 0.275 45 45 45 92.75 92.82 94.74 94.09 97.54 96.76 22 53 0.275 45 75 45 94.56 95.23 96.44 97.00 98.45 99.05 23 19 0.275 75 75 45 92.49 92.81 94.63 94.58 96.2 95.99 24 53 0.125 45 75 75 76.01 75.45 79.06 78.86 81.63 81.69 25 19 0.125 45 75 45 59.07 57.58 68.72 67.35 69.38 67.85 26(C) 36 0.2 60 60 60 91.7 91.56 93.65 93.83 95.9 95.59 27 53 0.125 75 45 75 65.46 65.63 72.26 72.55 76.16 76.61 28 19 0.275 45 75 75 90.07 89.98 94.78 94.21 96.59 96.21 29 53 0.275 45 45 75 94.11 95.10 96.07 96.59 98.5 99.31 30 36 0.05 60 60 60 35.32 37.94 45.19 47.04 47.59 49.51 31 36 0.35 60 60 60 92.14 89.52 94.22 92.97 96.36 94.91 32 36 0.2 60 60 90 85.56 85.40 89.35 89.61 93.01 92.63 a Central point. 938 H. Javadian et al. / International Journal of Biological Macromolecules 154 (2020) 937–953 2.2. Instrumentation and characterization The record of FT-IR spectra were performed from 4000 to 450 cm −1 on a PerkinElmer, USA, by the KBr disk method. To identify the crystalline structure of the product, XRD pattern was taken down on a GBC MMA instrument with CuK α radiation (wavelength λ= 0.154 nm) in the 2θrange of 10–70 0 . The morphological structure and particle size of the products were determined by FE-SEM (Zeiss Neon-40, Germany). The magnetic characteristics of the products were studied using VSM (Daghigh Kavir Corporation, Iran) at room temperature (RT). Thermal analyses were done on a Mettler TGA/SDTA 851e/LF/ 1100 thermobalance. The temperature of the samples was increased from RT to 1000 °C at rate = 10 °C/min under a constant flow of N 2 . For analyzing the concentration of Nd (III), Tb (III), and Dy (III), an Agilent 4100 MP-AES Spectrometer was used. The Design Expert software, version 10, was utilized to define the experimental design by CCD and analyze the regression of the experimental data. 2.3. Synthesis of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 magnetic nanoparticles The Ni 0.2 Zn 0.2 Fe 2.6 O 4 magneticnanoparticles were synthesized using hydrothermal method. A mixture of 0.2 M Ni 2+ , 0.2 M Zn 2+ and 2.6 M Fe 3+ was prepared in HCl solution, and then NaOH solution was added into mixed solution under nitrogen gas and the mixture pH value was adjusted to 10.5. 0.3 g of CTAB was added to the mixture, and then it was placed into an autoclave (Teflon-lined stainless steel) and maintained at 200°C ofan oven for 8 h for hydrothermal treatment. The temperature of the autoclave was naturally decreased to RT, and the precipitate wascollected and rinsed several times with deionized water to reach neutral pH. Finally, the obtained particles were dried at 50 °C. 2.4. Synthesis of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 magnetic bionanocomposite The synthesis procedure of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 wasasfollows: Sodium alginate (1 g) was dissolved in 80 mL of deionized water at RT using a laboratory stirrer. 0.5 g of the carboxymethyl chitosan powder was added into the solution and homogeneously mixed. 0.7 g of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 was added to the mixture of the biopolymers. To obtain a homogeneous blend solution, the mixture of biopolymers and magnetic particles was stirred at RT for 24 h. Then, it was added to the solution of calcium chloride 0.05 M and 2% glutaraldehyde for the gelation process. An external magnetic field was used to separate the resulting bionanocomposite from solution, and then it was washed several times with deionized water for removing remaining calcium chlorideand glutaraldehyde until the pHvalue thesolution was reached 7. The CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 bionanocomposite was dried at 50 °C. Finally, it was powdered. 2.5. Batch adsorption The adsorption experiments of Nd (III), Tb (III), and Dy (III) by the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 bionanocomposite were done in 125 mL flasks containing 50 mL solutions prepared from the dilution of 1000 mg/L stock solutions at different pHs, adsorbent dosages, contact times, and initial concentrations with the ratio of 1:1:1. The agitation of the flasks was carried out on a laboratory shaker (rpm = 180). Fig. 1. (A) XRD pattern of Ni 0.2 Zn 0.2 Fe 2.6 O 4 nanoparticles; photo of Ni 0.2 Zn 0.2 Fe 2.6 O 4 nanoparticles (B) before drying and (C) in the solution under magnetic field after drying. 939H. Javadian et al. / International Journal of Biological Macromolecules 154 (2020) 937–953 Fig. 2. FE-SEM images of (A) Ni 0.2 Zn 0.2 Fe 2.6 O 4 and (B) CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 . 940 H. Javadian et al. / International Journal of Biological Macromolecules 154 (2020) 937–953 After adsorption process for a predefined time, the adsorbent was externally separated by a magnetic field, and the adsorption efficiency and adsorption capacity of the metal ions by the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 were computed using the equations as following: Adsorption efficiency %ðÞ¼ C0Ce C0100 ð1Þ qe¼ðC0CeÞV mð2Þ qt¼ðC0CtÞV mð3Þ where q e and q t (mg/g) respectively refer to the quantities of metal ion adsorbed on the adsorbent at equilibrium and adsorption time t (min), C 0 (mg/L) is the initial concentration of metal ion, and C e (mg/L) is the equilibrium concentration of metal ion. Moreover, C t refers to the concentration of metal ion in solution at time t, V is the volume (L) of solution, and m refers to the weight (g) of the adsorbent. 2.6. Central composite design Experimental design is applied by investigators to decrease the number of experiments in the adsorption process. It also presents helpful information about the effect of independent parameters individually and/or interactively that leads to a decrease in experimental error [21]. RSM was utilized to model the adsorption process of the ions by investigating the independent variables including adsorbent dosage, contact time, initial concentration, and response (ions adsorption efficiency). For simultaneous adsorption of Nd (III), Tb (III), and Dy (III) ions by the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 by using batch mode, CCD was utilized at five levels (−α(−2), low (−1), central (0), high (+1) and +α (+2)) (Table 1). Analysis of variance (ANOVA) was used to obtain information about the adequacy of the models by evaluating coefficient of determination (R[2]), lack of fit,and theFisher test (F-value) values [22]. The quadratic polynomial model for response versus the independent variables was presented as follows [23]: Y¼β0þX k i¼1 βixiþX k i¼1X k j¼1 βijxixjþX k i¼1 βiix2 iþεð4Þ where Y refers to the predicted response (adsorption efficiency), and β 0 , β i ,β ii ,andβ ij , respectively refer to the constant coefficient, linear coefficient, quadratic coefficient, and interaction coefficient. Notably, x i and x j are the independent variables, k shows the number of the independent variables, and εis the residual error. 2.7. Column mode Fixed-bed column investigation is necessary to successfully design a process and study the behavior of adsorbent in a large-scale utilization. To investigate the fixed-bed column, a columnmade of glass with an internal diameter of 0.5 cm and a length of 12 cm was used. 0.3 g of the adsorbent was mixed with 1.8 g of acid-cleaned sand, and the mixture was packed in the column between two layers of glass wool. Sand was applied to decrease the pressure drop. The final height of the mixture in the column was about 6.5 cm. An upward flow rate (1 mL/min) was provided by a peristaltic pump to flush the column with deionized water for 1 h. Then, the column was fed with a ternary solution of the ions at 30 mg/L initial concentration as an influent for 520 min. The collection of the effluent was performed every 10 min by a fraction collector, and the concentration of the ions was determined by Agilent 4100 MP-AES Spectrometer. The breakthrough curve is usually shown by C t /C 0 versus time. The amount of ion adsorbed (q total ,mg)wasobtainedby calculating the curve of breakthrough (upper area) by using the following equation: qtotal ¼Q 1000Zt0 ttotal 1−Ct C0  ð5Þ where Q is the flow rate (mL/min) that was determined via division of the effluent volume (V eff ,mL)tothetotaltime(t total , min): Q¼Veff ttotal ð6Þ The entire quantity of the metals passed through the column (mg) was obtained by the following equation: mtotal ¼C0Qttotal 1000 ð7Þ The total metal adsorption efficiency was calculated from the ratio of the entire quantity of the metals sent to the column (q total )tothemetalmassadsorbed(m total ) by the following equation: Adsorption efficiency ¼qtotal mtotal 100 ð8Þ The capacity of equilibrium adsorption (q e (mg/g)) and the equilibrium metal concentration (C e (mg/L)) were respectively computed using Eqs. (9) and (10) as following: qe¼qtotal mð9Þ Ce¼mtotal−qtotal Veff 100 ð10Þ where m shows the adsorbent mass (g). 2.8. Error analysis Error analysis was used to optimize the fitness of the experimental data obtained from the non-linear approach. In this study, Chi-square (χ 2 ) wasemployed to compare the validityof each model bythefollowing equation: χ2¼X n i¼1 qe;exp−qe;cal  qe;cal 2 ð11Þ where n shows the number of data points, q e,exp is the experimental capacity of the adsorbent, and q e,cal refers to the computed capacity of the adsorbent. 3. Results and discussion 3.1. Products characterization Fig. 1 indicates the XRD pattern of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 . The peaks at 2θ= 18.13°, 30.07°, 35.50°, 37.08°,43.07°, 53.95°, 56.96° and 63.89° are in agreement with the standard pattern of nickel zinc ferrite (JCPDS 080234) [24]. Full Width at Half Maximum (FWHM) of the strongest reflection of the XRD pattern was used to estimate the average crystal size based on the Scherrer equation as following [25]: D¼kλ=βcos θð12Þ 941H. Javadian et al. / International Journal of Biological Macromolecules 154 (2020) 937–953 where k refers to the function of shape (k = 0.89), and λrefers to the radiation X-ray wavelength. Moreover, βand θrespectively refer to the Full Width at Half Maximum (FWHM) at 2θ= 35.50 ° , and the diffraction angle. Based on the Scherrer equation, the calculated value of Dwas27.68nm. The FE-SEM image of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 in Fig. 2A indicates that the synthesized particles are nearly spherical in shape and homogenous in distribution with a diameter b100 nm. Fig. 2B shows the distribution of the magnetic nanoparticles on the surface of the CA/CMC or embedding with the CA/CMC that confirms the successful synthesis of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 magnetic bionanocomposite. Fig. 3 indicates the FT-IR spectrum for CA, CMC, Ni 0.2 Zn 0.2 Fe 2.6 O 4 and CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 . The FT-IR spectrum for CA and CMC in Fig. 3A and B, respectively, shows O\\H stretching vibration at 3389 (CA) and 3436 (CMC) cm −1 , carboxyl groups asymmetrical stretching at 1622 (CA) and 1631 (CMC) cm −1 , carboxyl groups symmetrical stretching at 1423 (CA) and 1411 (CMC) cm −1 and C-O-C stretching at 1052 (CA) and 1061 (CMC) cm −1 [26,27]. The FT-IR spectrum of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 in Fig. 3C shows a broad band at around 3424 cm −1 and less intensive band at 1633 cm −1 that are related to the O\\H groups stretching vibration [28]. The bands at 2925 and 2853 cm −1 are assigned to the anti-symmetric and symmetric C\\H vibrations of CTAB [29]. The band at 567 cm −1 attributes to intrinsic metal stretching vibrations at the tetrahedral site (Fe\\O), and octahedral metal stretching(M-O)isseenataround478cm −1 [28]. The successful synthesis of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 can be confirmed by comparing its spectrum shown in Fig. 3Dwith other spectra. N 2 adsorption–desorption isotherm and the corresponding BarrettJoyner-Halenda (BJH) pore size distribution of the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 are shown in Fig. 4. The nitrogen adsorptiondesorption isotherm for the sample in Fig. 4A is assigned to type IV indicating the presence of mesoporous structure. The Brunauer–Emmett– Teller (BET) surface area, pore volume, and pore size (obtained by the BJH method (Fig. 4B)) were calculated to be 7.1143 m 2 /g, 0.034971 cm 3 /g, and 19.0379 nm, respectively. The results clearly demonstrate the formation of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 with nano size pores. EDX was recorded to analyze the elements of the products, and the results are presented in Fig. 5.Fig. 5A shows Ni, Zn, Fe, and O peaks that confirm the formation of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 . The elemental analysis of the nanocomposite in Fig. 5B represents similar peaks available in Fig. 5A along with the new peaks for N and Ca because of combining the nanoparticles with CA and CMC. Sodium peak is not seen in the spectrum of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 ,suggesting that sodium ions were released completely from the matrix of sodium alginate into the solution during the crosslinking reaction process of sodium alginate with calcium. EDX spectrum was also recorded after the adsorption process and the result is presented in Fig. 5C. The existence of Nd (III), Tb (III), and Dy (III) in the spectrum strongly confirms the successful adsorption of these ions by the CA/ CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 . The CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 weight loss curve recorded in the range of RT to 1000 °C is demonstrated in Fig. 6. As it is seen, there are three different weight-loss steps in the TGA curve of the CA/ CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 . Obviously, the first step (around 190 °C) with a weight loss of 8.77% can be attributed to trapped and physisorbed water evaporation. The second step between around 190 and 550 °C is the largest weight loss with the amount of 35.08% that could be due to sorption and degradation of CA and CMC. The last step with 22.95% weight loss at temperature beyond 550 °C could be related to the further decomposition of CA and CMC and their conversion to CO 2 and H 2 O. At the end of the process, the residue percentage is about 33.2% that is principally assigned to the presence of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 . According to the magnetic hysteresis loops in Fig. 7A, the magnetic saturation value for Ni 0.2 Zn 0.2 Fe 2.6 O 4 is about 45.87 emu/g that indicates the superparamagnetic behavior of the synthesized product. Based on Fig. 7B, it is obvious that the process of the synthesis of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 results in a decrease of saturation magnetization to the value of 14.14 emu/g. This decline is due to combining the magnetic nanoparticles by CA and CMC. Despite this difference, the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 can be easily separated from aqueous solution using an external magnetic field to avoid secondary pollution. Fig. 7C shows the easy separation of the metal ionsloaded adsorbent from the solution by applying an external magnetic field. 3.2. Effect of pH Solution pH is considered as an essential parameter in the process of adsorption owing to its effect on metal ionssolubility, counterions concentration on the adsorbent functional groups, and the adsorbate ionization degree. In this study, the influence of pH value on the process of adsorption was considered from 1.5 to 5.5. The pH ZPC of the CA/ CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 was found to be 6. Since the precipitation of the ions may occur at pH values higher than 5.5, the experiments were carried out near to pH ZPC ; therefore, adsorption of the ions were not studied at pH ˃5.5. As it is indicated in Fig. 8, the adsorption efficiency for the ions at pH = 1.5 is zero that indicates a highly acidic solution strongly affects the ions adsorption. At acidic solution, H + concentration and its mobilityare high that lead to strong competition with the ions to occupy the active sites. Actually, the protonation of the active sites occurs in a low value of pH, leading to electrostatic repulsion between positively charged cations and positively charged active sites; therefore, the value of adsorption efficiency is low. As the pH of solution increases, the amount of H + being available in the solutiondecreases; hence, more negatively charged sites are available that facilitate higher uptake of the ions by electrostatic attraction [30]. According to the obtained results, further adsorption studies were performed at pH = 5.5 as an optimum value. Fig. 3. FT-IR spectra of (A) CA, (B) CMC, (C) Ni 0.2 Zn 0.2 Fe 2.6 O 4 ,and(D)CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 . 942 H. Javadian et al. / International Journal of Biological Macromolecules 154 (2020) 937–953 conditions given by CCD was compared with the value predicted by RSM. As can be observed from the obtained results presented in Table 3, difference between the values obtained from the experiments and predicted by the model results in error percentage in the range of 0.54–1.87, 0.58–2.28, and 0.61–2.11% for Nd (III), Tb (III), and Dy (III), respectively. AtpH = 5.5, theadsorptionefficiencyundertheconditions given by the model (adsorbent dosage = 0.1 g, contact time = 53 min, and initial concentration = 30 mg/L) was calculated to be 95.72, 96.17, and 99.44% with 1.17, 0.58, and 0.61% error for Nd (III), Tb (III), and Dy (III), respectively, indicating its agreement with the experimental value. The errors show the capability of the RSM model for the prediction of values that are favorably in accordance with the experimental data. 3.6. Batch adsorption kinetic studies Kinetic investigations were carried out using the solutions with the volume of 50 mL prepared at 30 mg/L of Nd (III), Tb (III), and Dy (III) ions that were contacted with 0.09 g of the adsorbent at various times. Non-linear pseudo-first-order (PFO), PSO, and intra-particle diffusion (IPD) models were applied to model the kinetic data of Nd (III), Tb (III), and Dy (III) ions adsorption. The equations are as follows [36,37]: qt¼qe1−exp:−K1t  PFO ð16Þ qt¼K2qe2t=1þK2qetPSO ð17Þ qt¼Kit0:5þCIPD ð18Þ where K 1 (1/min), K 2 (g/mg min), and K i (1/min) respectively refer to the PFO rate constant, PSO rate constant, and the rate constant of IPD. Moreover, C provides information about the thickness of the boundary layer: higher value of C is related to the boundary layer diffusion influence. The initial rate of adsorption (h) can be computed using K 2 and q e values by the following equation: h¼K2qe2ð19Þ The values of kinetic parameters are shown in Table 4.Asitisobvious from the results, the highest values of R 2 and the lowest values of χ 2 obtained by PSO shows that the main mechanism for controlling the adsorption of Nd (III), Tb (III), and Dy (III) ions onto the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 is chemisorption. The values of R 2 obtained by IPD model show that the adsorption of the ions is a multi-stage process. The stages were related to the strong electrostatic forces of attractions between the ions and the functional groups of the adsorbent, and gradualadsorption by theionsdiffusion into the poresof the adsorbent until the occupation of most or all of the active sites. In addition, IPD model was not the sole rate-limiting step (the related plots do not pass through the origin). 3.7. Batch adsorption isotherm studies Batch isotherm experiments were performed using 50 mL of metal ions solutions at different concentrations in the range of 30–180 mg/L contacting with 0.09 g of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 at pH = 5.5. The data obtained at equilibrium were modeled by Langmuir and Freundlich models that are respectively related to monolayer and multilayer adsorption. The nonlinear Langmuir and Freundlich models were used according to following equations [38,39]: qe ¼bqmCe 1þbCeðÞ Langmuir ð20Þ qe¼KC e1=nFreundlich ð21Þ where q e and q m (mg/g) respectively refer to the equilibrium adsorption and maximum adsorption capacities, and C e (mg/L) shows the adsorbate equilibrium concentration. Moreover, b (L/mg) and K (mg 1–1/n L 1/n /g) respectively refer to Langmuir and Freundlich constants, and n shows adsorption intensity. Adsorption is favorable if nN1. The coefficient of determination (R [2]) values and the corresponding parameters obtained by the models are presented in Table 5. The n values are 9, 9.26, and 9.71 for Nd (III), Tb (III), and Dy (III), respectively, showing a strong interaction between the CA/ CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 and the metal ions. According to the values of R 2 and χ 2 , it is obvious that Freundlich model better fits the experimental data than Langmuir model for Nd (III), Tb (III), and Dy (III) ions adsorption. Consequently, the adsorption of the ions is multilayer adsorption, and the adsorption takes place on a non-uniform surface. 3.8. Ionic strength effect The ions adsorption can be affected by the co-ions that are available in the solution. The influence of salt concentration, known as ionic strength, on the adsorption efficiency of the ions by the CA/ Table 4 Kinetic constants for adsorption of Nd (III), Tb (III), and Dy (III) by the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 . Nd (III) Tb (III) Dy (III) PFO K 1 (1/min) 0.2905 0.290 0.289 q e (mg/g) 15.07 15.30 15.61 R 2 0.9617 0.9585 0.9587 χ 2 0.240 0.268 0.279 PSO K 2 (g/mg min) × 10 2 2.99 2.93 2.86 q e (mg/g) 16.15 16.40 16.74 h (mg/g min) × 10 –2s 7.79 7.88 8.01 R 2 0.9927 0.9933 0.9929 χ 2 0.045 0.043 0.048 IPD K i (1/min) 3.88 3.96 4.05 R 2 0.8157 0.8225 0.8222 χ 2 1.16 1.15 1.20 Table 3 Confirmation experiments for the ions adsorption onto the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 . No. Condition Nd (III) Tb (III) Dy (III) Error (%) Time (min) Adsorbent dosage (g) Initial concentration (mg/L) Observed Predicted Observed Predicted Observed Predicted Nd (III) Tb (III) Dy (III) Nd (III) Tb (III) Dy (III) 1 53 0.1 30 30 30 94.61 95.72 96.73 96.17 98.83 99.44 1.17 0.58 0.61 2 19 0.125 75 45 45 63.12 61.9 68.27 66.71 72.85 71.31 1.93 2.28 2.11 3 36 0.2 90 60 60 87.44 86.97 90.86 89.74 93.73 92.17 0.54 1.23 1.66 4 53 0.275 45 75 45 93.48 95.23 96.18 97 97.54 99.05 1.87 0.85 1.54 949H. Javadian et al. / International Journal of Biological Macromolecules 154 (2020) 937–953 CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 was evaluated using NaNO 3 at various concentrations ranging from 0.02 to 0.1 M, and the results are presented in Fig. 13A. As it is illustrated in Fig. 13A, the presence of NaNO 3 has a greater negative effect on the adsorption efficiency of Nd (III) in comparison with Tb (III) and Dy (III). The adsorption efficiency for Nd (III), Tb (III), and Dy (III) respectively decreases from 92.33 to 77.12, 93.91 to 85.6, and 96.25 to 91.43% by an increase in NaNO 3 concentration. This phenomenon can be related to the competition between the metal ions and sodium ions for the available active adsorption sites of the adsorbent [40]. In addition, the aggregation of adsorbent could be heightened by enhancing ionic strength that results in a decrease in adsorption sites of adsorbent [41]. 3.9. Thermodynamic parameters Thermodynamic parameters are considered as key factors to realize theoptimalcondition andgive furtherinformationregardingchangesin inherent energetic related to adsorption process. The adsorption process was conducted at various temperatures (25, 35, and 45 °C) to obtain ΔS ° and ΔH ° values based on the following equation: LnKd¼ΔS° R−ΔH° RT ð22Þ whereR,T,andK d respectively refer to the gas constant (8.314 J/mol K), temperature (K), and distribution coefficient that was obtained by the equation as following: Kd¼qe Ce ð23Þ where C e refer to the equilibrium concentration in the solution (mg/L). ΔH ° value for each metal is calculated from the slope of Ln K d versus 1/T plot, and ΔS ° value is computed from its intercept (Fig. 13B). The ΔG ° values were also calculated at different temperatures by using the following equation: ΔG°¼−RT Ln Kdð24Þ Table 6 shows the values of thermodynamic parameters. The ΔG ° valuesfor the ions are positive at all temperatures that show the process is non-spontaneous, and the adsorption of the ions onto the adsorbent requires additional energy from an external source. The lower values of ΔG ° at higher temperatures mean that an increase in temperature leads to an increase in the tendency of spontaneous reaction. The ΔH ° value ˃zero shows the endothermic adsorption of the metal ions, and the ΔS ° value ˃zero expresses the increase in randomness at the interface of solid–solution during the metal ions fixation on the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 surface [39]. 3.10. Reusability studies The synthesized CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 was utilized in four consecutive adsorption-desorption cycles to investigate its reusability. For this purpose, the adsorption of the ions by the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 was performed at initial concentration = 30 mg/L and pH = 5.5 with 0.1 g of the adsorbent for 53 min. The batch flask containing the CA/ CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 and 50 mL of 0.2 M HNO 3 as eluent was shaken for 2 h for desorption of the ions loaded onto the adsorbents. Then, the adsorbent was separated by an external magnetic field from the HNO 3 solution, and the functional groups wereneutralized by NaOH solution. The neutralized CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 was then used for the adsorption of 30 mg/L of the ions in the next cycle. Nd+3, Tb+3, and Dy+3 ions were desorbed from the adsorbent > 93, 96, and 97 %, respectively. The results of adsorption efficiency in 4 cycles are shown in Fig. 13C. Due to the decrease in the release of the ions and number of active sites by acid treatment during the cycles, the adsorption efficiency for Nd (III), Tb (III), and Dy (III) respectively decreases from 94.18 to 91.64, 96.45 to 94.82, and 98.33 to 97.17%. After the last cycle, the adsorption efficiency of the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 indicates 2.54, 1.63, and 1.16% lose for Nd (III), Tb (III), and Dy, respectively, in comparison with the first cycle. The results also indicated that the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 remained magnetic during the process of adsorptiondesorption. According to the results, the adsorbent suitability for a practical application can be concluded. 3.11. Column mode Fig. 14A presents the Nd (III), Tb (III), and Dy (III) ions adsorption breakthrough curves obtained from the fixed-bed column packed with Fig. 13. (A) Effect of ionic strength on the adsorption of Nd (III), Tb (III), and Dy (III), (B) Ln K d versus 1/T for calculation of enthalpy and entropy changes, and (C) Reusability of CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 for adsorption of the ions. Table 5 Isotherm constants for adsorption of Nd (III), Tb (III), and Dy (III) by the CA/CMC/ Ni 0.2 Zn 0.2 Fe 2.6 O 4 . Nd (III) Tb (III) Dy (III) Langmuir b (L/mg) 0.78 0.89 1.27 q m (mg/g) 23.15 24.41 25.24 R 2 0.8773 0.9238 0.9125 χ 2 1.62 1.21 1.61 Freundlich K (mg 1–1/n L 1/n /g) 14.40 15.46 16.49 n 9 9.26 9.71 R 2 0.9879 0.9654 0.9633 χ 2 0.159 0.55 0.675 950 H. Javadian et al. / International Journal of Biological Macromolecules 154 (2020) 937–953 the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 . According to the obtained results, the breakthrough and exhaustion times that respectively correspond to C e /C 0 = 0.05 and 0.95 are about 95 and 410 min for Nd (III), 105 and 430 min for Tb (III), and 120 and 440 min for Dy (III). Exhaust volume (V eff ) for Nd (III), Tb (III), and Dy (III) is respectively 410, 430, and 440 mL. Fig. 14. (A) Experimental data of the column adsorption and (B–D) Modeling of the experimental data with Thomas and Yan models. Table 6 Effect of temperature on the adsorption of Nd (III), Tb (III), and Dy (III) at 90 mg/L and thermodynamic parameters. Adsorption efficiency (%) Temperature (°C) Nd (III) Tb (III) Dy (III) 25 44.31 47.14 49.21 35 48.23 50.48 53.45 45 52.64 55.99 58.42 Thermodynamic parameters Nd (III) Tb (III) Dy (III) ΔH ° (kJ/mol) 13.9 14.76 15.45 ΔS ° (kJ/mol K) 0.039 0.043 0.046 Temperature ( º C) ΔG ° (kJ/mol) 25 2.022 1.739 1.534 35 1.686 1.408 1.151 45 1.274 0.918 0.655 951H. Javadian et al. / International Journal of Biological Macromolecules 154 (2020) 937–953 Nonlinear Thomas and Yan models were utilized to fit the experimental data of the fixed-bed column by Eq. (25) and Eq. (26), respectively. Ct C0¼1 1þexp KThqex Q−KThC0t  ð25Þ Ct C0¼1−1 1þC0Qt qex  að26Þ where K Th shows the rate constant (mL/min mg), a is a constant coefficientandq e is the maximum adsorptioncapacity(mg/g),xis the mass of adsorbent (g), C 0 is the initial concentration (mg/L) of the ions, C t is outlet ions concentration (mg/L), t is the contact time (min), and Qis the flow rate (mL/min). K Th ,a,andq e values are computed using the slope and intercept of the plot of Ct C0against t. The parameters obtained by the models (Figs. 14B-D) are indicated in Table 7. The adsorption capacity (q e ) for Nd (III), Tb (III), and Dy (III), calculated by Eq. (9), is respectively 22.70, 24.00, and 25.54 mg/g under the studied conditions. According to the values of R [2], both models can fit the experimental data well but Yan model presents higher values of R 2 in comparison with Thomas model. The rate constant value of Thomas (K Th ) for Nd (III) is higher than those of Tb (III), and Dy (III), showing higher intensity of Nd (III) adsorption onto the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 , while the amount of maximum adsorption capacity (q e ) for Dy (III) is greater than the value obtained for Nd (III) and Tb (III). This is in conformity with the results achieved from experiments of batch adsorption. 4. Conclusion In this paper,the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 was synthesized successfully by the gelation process of CA/CMC in the presence of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 . RSM-CCD was applied to explore the influences of experimental parameters on Nd (III), Tb (III), and Dy (III) ions adsorption. The conditions predicted by RSM for optimum adsorption of 30 mg/L of the ions were 0.1 g of the adsorbent and 53 min contact time at pH = 5.5. The experimental data were fitted by isotherm and kinetic models. PSO kinetic model fitted the data better compared with IPD and PFO models. The data of equilibrium were fitted well with Freundlich model. The values of ΔH ° revealed the endothermic adsorption process of the metal ions. 0.2 M HNO 3 was used for regeneration of the ionloaded adsorbent, and the adsorbent was repeatedly used in four cycles with N91, 94, and 97% adsorption efficiency for Nd (III), Tb (III), and Dy (III), respectively, after the fourth cycle. Besides, the ions were successfully adsorbed in a continuous process by applying a packed-bed column, and the data were found to be fitted well by Thomas and Yan models. The results showed that the CA/CMC/Ni 0.2 Zn 0.2 Fe 2.6 O 4 can be applied as a potential adsorbent in both adsorption modes (batch and column) for Nd (III), Tb (III), and Dy (III) ions adsorption. CRediT authorship contribution statement Hamedreza Javadian: Investigation, Data curation, Formal analysis, Writing - original draft, Writing - review & editing. Montserrat Ruiz: Data curation, Supervision. Ana Maria Sastre: Conceptualization, Supervision. Acknowledgments This work has been supported by the Spanish Ministry of Economy and Competitiveness (Ref. CTM2017-83581-R).HamedrezaJavadianacknowledges the financial support received (Ref. BES-2015-072506). References [1] G.A. Moldoveanu, V.G. Papangelakis, Recovery of rare earth elements adsorbed on clay minerals. I. Desorption mechanism, Hydrometallurgy 117–118 (2012) 71–78. [2] J. Ponou, L.P. Wang, G. Dodbiba, K. Okaya, T. Fujita, K. Mitsuhashi, T. Atarashi, G. Satoh, M. Noda, Recovery of rare earth elements from aqueous solution obtained from Vietnamese clay minerals using dried and carbonized parachlorella, J. Environ. Chem. Eng. 2 (2014) 1070–1081. [3] W. Franus, M.M. Wiatros-Motyka, M. Wdowin, Coal fly ash as a resource for rare earth elements, Environ. Sci. Pollut. Res. 22 (12) (2015) 9464–9474. [4] D. Sadovsky, A. Brenner, B. Astrachan, B. Asaf, R. Gonen, Biosorption potential of cerium ions using Spirulina biomass, J. Rare Earths 34 (2016) 644–652. [5] F. Zhao, E. Repo, Y. Meng, X. Wang, D. Yin,M. Sillanpää, An EDTA-β-cyclodextrin material for the adsorption of rare earth elements and its application in preconcentration of rare earth elements in seawater, J. Colloid Interface Sci. 465 (2016) 215–224. [6] K. Vijayaraghavan, R. Balasubramanian, Single and binary biosorption of cerium and europium onto crab shell particles, Chem. Eng. J. 163 (2010) 337–343. [7] J.L. Wang, C. Chen, Biosorbents for heavy metals removal and their future, Biotechnol. Adv. 27 (2) (2009) 195–226. [8] H.A. Shawky, Improvement of water quality using alginate/montmorillonite composite beads, J. Appl. Polym. Sci. 119 (4) (2010) 2371–2378. [9] Y. Ma, P. Qi, J. Ju, Q. Wang, L. Hao, R. Wang, K. Sui, Y. Tan, Gelatin/alginate composite nanofiber membranes for effective and even adsorption of cationic dyes, Compos Part B Eng 162 (2019) 671–677. [10] F. Ahmadpoor, S.A. Shojaosadati, S.Z. Mousavi, Magnetic silica coated iron carbide/ alginate beads: synthesis and application for adsorption of cu (II) from aqueous solutions, Int. J. Biol. Macromol. 128 (2019) 941–947. [11] T. Guo, Y. Pei, K. Tang, X. He, J. Huang, F. Wang, Mechanical and drug release properties of alginate beads reinforced with cellulose, J. Appl. Polym. Sci. 134 (8) (2016) 44495. [12] Z.H. Hu, A.M. Omer, X.K. Ouyang, D. Yu, Fabrication of carboxylated cellulose nanocrystal/sodium alginate hydrogel beads for adsorption of Pb(II) from aqueous solution, Int. J. Biol. Macromol. 108 (2018) 149–157. [13] R. Ahmad, A. Mirza, Adsorption of Pb (II) and cu (II) by alginate-au-mica bionanocomposite: kinetic, isotherm and thermodynamic studies, Process Saf Environ 109 (2017) 1–10. [14] L. Zhang, Y. Zeng, Z. Cheng, Removal of heavy metal ions using chitosan and modified chitosan: a review, J. Mol. Liq. 214 (2016) 175–191. [15] M. Li, J. Xu, R. Li, D. Wang, T. Li, M. Yuan, J. Wang, Simple preparation of aminothiourea-modified chitosan as corrosion inhibitor and heavy metal ion adsorbent, J. Colloid Interface Sci. 417 (3) (2014) 131–136. [16] D.W. Cho, B.H. Jeon, C.M. Chon, Y. Kim, F.W. Schwartz, E.S. Lee, H. Song, A novel chitosan/clay/magnetite composite for adsorption of cu(II) and as(V), Chem. Eng. J. 200–202 (34) (2012) 654–662. [17] L. Fan, C. Luo, Z. Lv, F. Lu, H. Qiu, Removal of Ag + from water environment using a novel magnetic thiourea-chitosan imprinted Ag + , J. Hazard. Mater. 194 (11) (2011) 193–201. [18] P.O. Boamah, Y. Huang, M. Hua, Q. Zhang, J. Wu, J. Onumah, L.K. Sam-Amoah, P.O. Boamah, Sorption of heavy metal ions onto carboxylate chitosan derivatives—a mini-review, Ecotoxicol. Environ. Saf. 116 (2015) 113–120. [19] H. Yang, H. Li, J. Zhai, L. Sun, Y. Zhao, H. Yu, Magnetic prussian blue/graphene oxide nanocomposites caged in calcium alginate microbeads for elimination of cesium ions from water and soil, Chem. Eng. J. 246 (2014) 10–19. [20] G. Reiss, A. Hütten, Magnetic nanoparticles: applications beyond data storage, Nat. Mater. 4 (2005) 725–726. Table 7 Parameters of breakthrough and the values predicted by Thomas and Yan models. Metal Breakthrough analysis Thomas model Yan model q e Adsorption efficiency (%) C e K Th ×10 4 q e R 2 q e aR 2 Nd (III) 22.70 46.33 16.10 5.65 22.18 0.9945 21.10 3.74 0.9957 Tb (III) 22.70 47.07 15.88 5.52 23.44 0.9937 22.32 3.83 0.9962 Dy (III) 25.54 49.11 15.26 5.46 24.96 0.9940 23.92 4.01 0.9970 952 H. Javadian et al. / International Journal of Biological Macromolecules 154 (2020) 937–953 [21] D. Gusain, F. Bux, Y. Chandra Sharma, Abatement of chromium by adsorption on nanocrystalline zirconia using response surface methodology, J. Mol. Liq. 197 (2014) 131–141. [22] N. Markandeya, S. Dhiman, P. Shukla, G.C. Kisku, Statistical optimization of process parameters for removal of dyes from wastewater on chitosan cenospheres nanocomposite using response surface methodology, J. Clean. Prod. 149 (2017) 597–606. [23] V. Srivastava, Y.C. Sharma, M. Sillanpää, Application of response surface methodology for optimization of co(II) removal from synthetic wastewater by adsorption on NiO nanoparticles, J. Mol. Liq. 211 (2015) 613–620. [24] D. Ni, Z. Lin, P. Xiaoling, W. Xinqing, G. Hongliang, Preparation and characterization of nickel-zinc ferrites by a solvothermal method, Rare Metal Mat Eng 44 (9) (2015) 2126–2131. [25] C. Tamez, R. Hernandez, J.G. Parsons, Removal of cu (II) and Pb (II) from aqueous solution using engineered iron oxide nanoparticles, Microchem. J. 125 (2016) 97–104. [26] R. Fabryanty, C. Valencia, F. Edi Soetaredjo, J. Nyoo Putro, S. Permatasari Santoso, A. Kurniawan, Y.H. Ju, S. Ismadji, Removal of crystal violet dye by adsorption using bentonite-alginate composite, J. Environ. Chem. Eng. 5 (2017) 5677–5687. [27] S.P. Wu, X.Z. Dai, J.R. Kan, F.D. Shilong, M.Y. Zhu, Fabrication of carboxymethyl chitosan–hemicellulose resin for adsorptive removal of heavy metals from wastewater, Chin. Chem. Lett. 28 (2017) 625–632. [28] A.S.A. Bakr, Y.M. Moustafa, E.A. Motawea, M.M. Yehia, M.M.H. Khalil, Removal of ferrous ions from their aqueous solutions onto NiFe 2 O 4 –alginate composite beads, J. Environ. Chem. Eng. 3 (2015) 1486–1496. [29] S.A. Elfeky, S. Ebrahim Mahmoud, A. Fahmy Youssef, Applications of CTAB modified magnetic nanoparticles for removal of chromium (VI) from contaminated water, J. Adv. Res. 8 (4) (2017) 435–443. [30] R. Darvishi Cheshmeh Soltani, G. Shams Khorramabadi, A.R. Khataee, S. Jorfi,Silica nanopowders/alginate composite for adsorption of lead (II) ions in aqueous solutions, J. Taiwan Inst. Chem. Eng. 45 (2014) 973–980. [31] R.D.C. Soltani, A. Rezaee, H. Godini, A.R. Khataee, A. Hasanbeiki, Photoelectrochemical treatment of ammonium using seawater as a natural supporting electrolyte, Chem. Ecol. 29 (2013) 72–85. [32] M. El kady, H. Shokry, H. Hamad, Effect of superparamagnetic nanoparticles on the physicochemical properties of nano hydroxyapatite for groundwater treatment: adsorption mechanism of Fe(II) and Mn (II), RSC Adv. 6 (2016) 82244–82259. [33] A. Mohseni-Bandpi, B. Kakavandi, R.R. Kalantary, A. Azari, A. Keramati, Development of a novel magnetite-chitosan composite for the removal of fluoride from drinking water: adsorption modeling and optimization, RSC Adv. 5 (2015) 73279–73289. [34] H. Shekari, M.H. Sayadi, M.R. Rezaei, A. Allahresani, Synthesis of nickel ferrite/titanium oxide magnetic nanocomposite and its use to remove hexavalent chromium from aqueous solutions, Surf. Interface. 8 (2017) 199–205. [35] K.S. Padmavathy, G. Madhu, P.V. Haseena, A study on effects of pH, adsorbent dosage, time, initial concentration and adsorption isotherm study for the removal of hexavalent chromium (Cr (VI)) from wastewater by magnetite nanoparticles, Procedia Technol 24 (2016) 585–594. [36] S.C.W. Sakti, Y. Narita, T. Sasaki, S. Tanaka Nuryono, A novel pyridinium functionalized magnetic chitosan with pH-independent and rapid adsorption kinetics for magnetic separation of Cr(VI), J. Environ. Chem. Eng. 3 (2015) 1953–1961. [37] H. Javadian, P. Vahedian, M. Toosi, Adsorption characteristics of Ni(II) from aqueous solution and industrial wastewater onto Polyaniline/HMS nanocomposite powder, Appl. Surf. Sci. 284 (2013) 13–22. [38] V.S. Munagapati, D.S. Kim, Equilibrium isotherms, kinetics, and thermodynamics studies for Congo red adsorption using calcium alginate beads impregnated with nano-goethite, Ecotoxicol. Environ. Saf. 141 (2017) 226–234. [39] H. Javadian, M. Torabi Angaji, M. Naushad, Synthesis and characterization of polyaniline/γ-alumina nanocomposite: a comparative study for the adsorption of three different anionic dyes, J. Ind. Eng. Chem. 20 (2014) 3890–3900. [40] S. Malamis, E. Katsou, A review on zinc and nickel adsorption on natural and modified zeolite, bentonite and vermiculite: examination of process parameters, kinetics and isotherms, J. Hazard. Mater. 252 (2013) 428–461. [41] F.F. Zeng, Y. He, Z.H. Lian, J.M. Xu, The impact of solution chemistry of electrolyte on the sorption of pentachlorophenol and phenanthrene by natural hematite nanoparticles, Sci. Total Environ. 466 (2014) 577–585. 953H. Javadian et al. / International Journal of Biological Macromolecules 154 (2020) 937–953 137 138 ANNEX III Synthesis of magnetic CMC bionanocomposite containing a novel biodegradable nanoporous polyamide selectively synthesized in ionic liquid as green media: Investigation on Nd+3, Tb+3, and Dy+3 rare earths adsorption. Hamedreza Javadian, Montserrat Ruiz, Mehdi Taghavi, Ana Maria Sastre, Journal of Molecular Liquids, 308 (2020) 113017. https://doi.org/10.1016/j.molliq.2020.113017 139 Synthesis of magnetic CMC bionanocomposite containing a novel biodegradable nanoporous polyamide selectively synthesized in ionic liquid as green media: Investigation on Nd +3 ,Tb +3 ,andDy +3 rare earth elements adsorption Hamedreza Javadian a, ⁎,Montserrat Ruiz b , Mehdi Taghavi c , Ana Maria Sastre a a Department of Chemical Engineering, ETSEIB, Universitat Politècnica de Catalunya, Diagonal 647, 08028 Barcelona, Spain b Department of Chemical Engineering, EPSEVG, Universitat Politècnica de Catalunya, Av. Víctor Balaguer, s/n, 08800 Vilanova i la Geltrú, Spain c Department of Chemistry, Faculty of Science, Shahid Chamran University of Ahvaz, 61357-43337, Iran abstractarticle info Article history: Received 27 October 2019 Received in revised form 20 March 2020 Accepted 28 March 2020 Available online 31 March 2020 Keywords: Carboxymethyl chitosan Poly(pyrimidine-thiophene-amide) Magnetic Bionanocomposite Adsorption Rare earth elements In this research study, the carboxymethyl chitosan/poly(pyrimidine-thiophene-amide)/Ni 0.2 Zn 0.2 Fe 2.6 O 4 (CMC/P (PTA)/Ni 0.2 Zn 0.2 Fe 2.6 O 4 ) was prepared as a novel magnetic bionanocomposite adsorbent. FE-SEM, EDX, NMR, XRD, FT-IR, and VSM techniques were applied for the analyses of the products. The adsorption behavior of the prepared bionanocomposite was investigated towards Nd +3 ,Tb +3 ,andDy +3 as adsorbates. The adsorption process was evaluated considering the influence of independent parameters including pH of the solution, contact time, adsorbent dosage, initial metal ions concentration, and ionic strength. The adsorption efficiency values of 98.15, 97.6, and 99.42% were respectively obtained for Nd +3 ,Tb +3 , and Dy +3 at optimum conditions of pH = 5.5, 30 mg/L of the ions, adsorbent dosage of 0.06 g, and contact time of 90 min. The data of the adsorption equilibrium of the ions were fitted well by Freundlich model. Kinetic studies showed that Nd +3 ,Tb +3 ,andDy +3 adsorption followed both pseudo-second-order (PSO) and intra-particle diffusion (IPD) kinetic models. The values of ΔH ° indicated that the ions adsorption process onto the bionanocomposite was endothermic, and the ΔG ° values revealed that it was spontaneous at higher temperature. The CMC/P(PTA)/Ni 0.2 Zn 0.2 Fe 2.6 O 4 could be regenerated by 0.2 M HNO 3 and its separation was viable utilizing a magnetic field with the saturation magnetization value of 14.88 emu/g. © 2020 Published by Elsevier B.V. 1. Introduction Rare earth elements (REEs) contain 17 components of the periodic table that include 15 lanthanides together with yttrium and scandium [1–3]. Light and heavy REEs are the additional subdivision of REEs based on atomic number. They possess exceptional properties and are regularly named as “seeds of technology”[4]. They are extensively utilized in various fields, for example, electronics, metallurgy, catalysis, alloys, superconductors, lasers, fertilizers, chemical reagents, nuclear energy, and magnets [5,6]. The total demand for REEs was reported 128,000 tons in 2011, and this amount was raised up to 170,000 tons in 2015. It is also anticipated that it would increase up to 255,000 tons in 2020 with around 6–10%/year growth rate [7]. This anticipated large demand for REEs is as a result of their broad usage in numerous fields of human life.According tothis matter, their recovery from wastes seems to be necessary. Different techniques, such as precipitation, ion exchange, solvent extraction, and adsorption have been utilized for REEs recovery from aqueous media [8–11]. Numerous investigations have shown adsorption as a high-efficient, cost-effective and simple technique for REEs recovery from aqueous media amongst the physicochemical treatment techniques. Chitosan is a derivation of chitin, a natural polysaccharide that comprises β(1–4)-2-amino-2-deoxy-D-glucan units and possesses superb adsorption performance for metal ions, chiefly owing to containing a large amount of amino and hydroxyl groups [12]. Its structure is like a crystal with hydrogen bonds. Nonetheless, the protonation of the amino groups in the acidic medium leads to losing structural strength by the formation of a gel-like solution that results in significant restrictions of its applications [13]. So as to solve this issue, the modification of chitosan has been carried out by sulfonation, nitration, hydroxyalkylation, quarternarization, hydroxylation, polyethyleneglycol-grafting, carboxymethylation, and so on [14,15]. Amongst the derivatives of chitosan, carboxymethyl chitosan Journal of Molecular Liquids 308 (2020) 113017 ⁎Corresponding author. E-mail address: [email protected] (H. Javadian). https://doi.org/10.1016/j.molliq.2020.113017 0167-7322/© 2020 Published by Elsevier B.V. Contents lists available at ScienceDirect Journal of Molecular Liquids journal homepage: www.elsevier.com/locate/molliq (CMC) is an amphoteric ether derivative that each molecule comprises theactivegroupsof\\COOH,\\OH, and\\NH 2 . The carboxymethylation of chitosan causes it to be dissolved in water. Thus, the problem related to losing the structural stability that takes place by\\NH 2 protonation in the acidic medium can be solved [16]. In addition, carboxyl groups are able to increase metal ions adsorption [17]. Whereas chitosan is a natural polymer, its derivatives like CMC also have some advantages such as being biocompatible and low toxic. Recently, numerous polycondensation reactions have been conducted in ionic liquids (ILs) at room temperature (RT) as a substitute to several volatile toxic solvents [18]. In industry, it is necessary to substitute a great number of organic solvents with environmentally, nonvolatile and green solvents for preparing high molecular weight polymers. The attraction of ambient temperature imidazolium based ILs for substantial commercial productions and applications as solvents and catalysts in polymerization, extraction, and as alternatives for ordinary volatile organic solvents has been confirmed. Reviewing the ILs application in polymer fabrication verifies their performance. Consequently, different processes of polycondensation in the ionic medium have been effectively conducted [19]. In recent years, increasing attention in the studies of synthetic polymer-polymer composites owing to the unique combination of beneficial properties and construction of multifunctional structures of each component has been reported. These composites display potentially greater thermal, mechanical, and electrical properties than theuniquepolymer[20]. Synthetic PAs are generally known as the first engineering plastics and are still considered as one of the best and most prominent classes of these types of materials. Hence, these polymers introduction in the natural polymers chemical structures is able to supply enhanced properties for individual usages, for which polymer having these sequences have shown incredibly excellent properties [21]. In this study, CMC, a novel biodegradable nanoporous polyamide (poly(pyrimidine-thiophene-amide)) synthesized by polycondensation reaction of 5,5′-(thiophen-2-ylmethylene)bis(2-aminopyrimidine-4,6diol) (TMAPD) with terephthalic acid in 1,3-dipropyl imidazolium bromide {[1,3-(pr) 2 im]Br} ionic liquid as green media, and the hydrothermally synthesized Ni 0.2 Zn 0.2 Fe 2.6 O 4 were applied to produce a novel bionanocomposite (CMC/P(PTA)/Ni 0.2 Zn 0.2 Fe 2.6 O 4 ) by the gelation process. FE-SEM, EDX, NMR, XRD, FT-IR, and VSM were utilized to confirm the fabrication of the products. In the following, it was used as an adsorbent to investigate its effectiveness for Nd +3 ,Tb +3 ,andDy +3 ions adsorption. The influences of different factors including solution pH, contact time, adsorbent dosage, initial metal ions concentration, and ionic strengthon the adsorption efficiency were considered.The models of adsorption kinetic and isotherm were employed to reach the best fitting of the experimental data. The reusability of the adsorbent was also investigated. 2. Materials and methods 2.1. Materials and reagents Carboxymethyl chitosan was purchased from Nantong Chem-Base Co, China. Dy(NO 3 )·5H 2 O was purchased from Alfa Aesar. Nd(NO 3 ) 3 ·6H 2 O, Tb(NO 3 ) 3 ·6H 2 O, Zn(NO 3 ) 2 ·6H 2 O, Fe(NO 3 ) 3 ·9H 2 O, Ni(NO 3 ) 2 ·6H 2 O, glutaraldehyde, 2-amino-4,6-dihydroxypyrimidine, 2thiophenecarboxaldehyde, terephthalic acid, triphenyl phosphite (TPP), dimethyl sulfoxide (DMSO), and methanol were bought from Sigma-Aldrich. All chemicals chosen in this study were at analytical grade and were utilized as received without further purification.Theexperiment solutions of Nd +3 ,Tb +3 ,andDy +3 ions were made by the dilution of 1000 mg/L of ions. To adjust the initial value of pH in the test solutions to the desired value, appropriate molarity of HNO 3 or NaOH was used. 2.2. Instrumentation and characterization The XRD pattern was recorded by an X-ray diffractometer (GBC MMA), and the samples were scanned from 2θ= 10° to 70°. A spectrophotometer (PerkinElmer, USA) was also used to record FT-IR spectra. Bruker Advance DRX was employed to record 1 HNMRand 13 CNMR spectra at 400 MHz and 100MHz, respectively, by DMSO‑d 6 as a solvent. The nitrogen adsorption-desorption isotherm was measured at −196 °C using a Micromeritics Tristar 3000 apparatus. The morphology of the products was explored by a FE-SEM (Zeiss Neon-40, Germany). TGA measurements were performed by Mettler TGA/SDTA 851e/LF/ 1100 thermobalance under the atmosphere of N 2 from RT to 1000 °C with a rate of 10 °C/min. Magnetic measurements were done using a vibrating sample magnetometer (VSM, Daghigh Kavir Corporation, Iran). For analyzing the concentration of Nd +3 ,Tb +3 ,andDy +3 , an Agilent 4100 MP-AES Spectrometer was used. 2.3. Synthesis of the Ni 0.2 Zn 0.2 Fe 2.6 O 4 magnetic nanoparticles The Ni 0.2 Zn 0.2 Fe 2.6 O 4 magnetic nanoparticles were synthesized by the hydrothermal method. A mixed solution of 0.2 M Ni 2+ ,0.2M Zn 2+ ,and2.6MFe 3+ was prepared in HCl solution, and then NaOH solution was added into the mixed solution under nitrogen gas, and the mixture pH value was set to 10.5. 0.3 g of CTAB was added to this mixture, and then it was placed into an autoclave (Teflon-lined stainless steel) at 200 °C of an oven. After 8 h of hydrothermal treatment, the temperature of the autoclave was naturally decreased to RT, and the precipitate was collected and washed several times with deionized water (DW) to reach neutral pH. Finally, the obtained particles were dried at 50 °C. 2.4. Synthesis of the ionic liquid (1,3-dipropyl imidazolium bromide) and the monomer (5,5′-(thiophen-2-ylmethylene)bis(2-aminopyrimidine4,6-diol (TMAPD)) The RT ionic liquid (IL) was synthesized based on the process presented in the literature [22]. TMAPD was synthesized according to the following procedure: A mixture of 2.54 g (0.02 mol) 2-amino-4,6dihydroxypyrimidine, 1 mL (0.01 mol) 2-thiophenecarboxaldehyde, and 20 mL DMSO was stirred for 6 h at 110 °C. After completion of the reaction tested by thin-layer chromatography, the temperature of the solution was decreased to RT, and the violet powder obtained by pouring the solution into 400 mL of cold DW (−5°C)wasfiltered, rinsed several times using DW and then dried using vacuum oven at 100°C. The reaction yield was 92% (3.20g), and the obtained compound has not shown sharp melting point and started tobe decomposed above 300 °C. FT-IR (KBr, cm −1 ): 3153–3477 (stretching of O\\H and NH 2 ), 3049 (stretching of C\\H aromatic), 2944 (stretching of C\\Haliphatic), 1651 (stretching of C_N), 1586 (stretching of C_C), 1232 (C\\N) and 1163 (C\\O). 1 H NMR (DMSO‑d 6 ,δin ppm) (Fig. 1): 5.33 (s, 1H, CH), 6.61 (s, 4H,\\NH 2 ), 6.78–6.80 (d, 1H, Ar\\H, J= 5.6 Hz), 6.93–6.94 (d, 1H, Ar\\H, J= 5.6 Hz), 7.42–7.44 (d, 1H, Ar\\H, J=5.2Hz), 10.95–11.28 (m, 4H, broad, hydroxy pyrimidine). 13 C NMR (100 MHz, DMSO‑d 6 ,δin ppm) (Fig. 2): 30.98, 115.41, 128.27, 128.55, 137.86, 140.13, 143.76, 170.77. 2.5. Synthesis of the biodegradable nanoporous P(PTA) by polycondensation reaction of TMAPD in TPP/IL The synthesis of the biodegradable nanoporous P(PTA) was carried out from a compound containing multi polar thiophene, amine, and free hydroxyl chelating groups. It was particularly synthesized from the diamine-phenol compound in 1,3-dipropyl imidazolium bromide as an ionic liquid without using toxic triphenyl phosphite/Nmethylpyrolidone/pyridine/LiCl that is needed in the ordinary direct polycondensation. The P(PTA) was achieved by polycondensation of 2H. Javadian et al. / Journal of Molecular Liquids 308 (2020) 113017