Extractant immobilized materials

Extractants immobilized materials (EIMs) are a class of adsorbents in which the active component or extractant used to bind the target analyte separating it from its original medium is physically immobilized onto a solid support (Hidayah & Abidin, 2017). The active component or the extractant is a compound capable of ionic or complex interaction with the target metal ions. The active component is attached to the solid support’s surface by physical adsorption, or it resides in a thin solvent layer covering the support surface, or it can be part of the bulk solvent filling the pores of the solid support. In literature, solvent impregnated resins (SIRs) is commonly used to identify adsorbent obtained when solvent extractant such as N,N-dimethylformamide (DMF), N,N,N′,N′‐Tetraoctyl diglycolamide (TODGA), and Cyanex-923 are infused onto solid support like silica or resin (Sun et al. 2009; Kabay et al. 2010; Zhu and Chen 2011). Another similar term, supported ionic liquid (SIL), describes adsorbents where ionic liquids are infused into the solid support (Bao et al., 2016; Zhu et al., 2012).
The separation processes involving EIMs have been termed as supported liquid extraction (SLE) or solid-supported liquid-liquid extraction (SLLE) (Florek et al., 2016; Hidayah & Abidin, 2017; Hu et al., 2018). Supported liquid extraction (SLE) is analogous to traditional liquid-liquid extraction (LLE), where target species are separated based on their relative solubilities in two different immiscible liquids, usually water and an organic solvent containing the extractant but instead of shaking the two immiscible phases together, the extractant is supported on a solid support and put in a column (or cartridge). The REEs containing solution is passed through, REEs ion gets adsorbed, and then eluent is used to desorb the adsorbed REEs species. In SLE, it is possible to achieve higher recovery due to the increase in surface area of interaction between the extractant and the target cations and increasing the diffusion rate through control of the porous structure of the support material. Other advantages over traditional LLE are less use of organic compounds, waste reduction, and lack of third phase formation (Florek et al., 2016; Hidayah & Abidin, 2017; Kabay et al., 2010).
The synthesis of EIMs (SIRs and SILs) involves modification of already synthesized polymers or other solid supports like silica or activated carbon using dry or wet methods. In the dry method, the active component is impregnated into support in excess using a volatile solvent then the excess amount is evaporated by heating, thus leaving the active component physically adsorbed to the support surface. In the wet method, support is impregnated with a solution of active component in a non-polar or weakly polar solvent followed by repeated washing with water, leaving a layer of the non-polar solution of the active component (Bao et al., 2016; Ehrlich & Lisichkin, 2017). Based on the types of solid support used for preparing, these adsorbents can be broadly divided into organic polymer-based support and inorganic supports.

Organic polymer-based supports

Many researchers have used organic polymers as the solid support for solvent impregnated resins (SIRs) used for REEs uptake from solution. Relevant examples of polymer based SIRs include N,N,N′N′ tetraoctyldiglycolamide (TODGA) and N,N,N′N′ tetrakis‐2‐ethylhexyldiglycolamide (TEHDGA) impregnated Amberchrom CG‐71 for REEs separation; tributyl phosphate infused Styrene–divinylbenzene to copolymer separate Am and Eu (Louis & Duyckaerts, 1984, 1985); 2-ethylhexyl phosphonic acid mono-2-ethylhexyl ester (PC88A) infused Amberlite XAD-7 for separation of Gd and Y mixtures (WAKUI et al., 1988); 8-Quinolinol (oxine) and 2-(2-(5 chloropyridylazo)-5-dimethylamino)-phenol (5ClDMPAP) immobilized Amberlite XAD-4 and XAD-7 each for preconcentration of Ce, La, and Pr (Masi & Olsina, 1993); Amberlite XAD-7 impregnated with binary mix of cobalt dicarbollide and dibutyl-N,N-diethylcarbamoylmethyl phosphonate (DBDECMP) for Eu sorption (Svoboda et al., 1997); bis(2,4,4-trimethyl pentyl) phosphinic acid (Cyanex 272) (Wang et al., 1998), 1-hexyl-4-ethyloctyl hydrogen isopropylphosphonic acid (HEOPA) (Wang et al., 2002) , and bis(2,4,4-trimethylpentyl) monothiophosphinic acid (CL302, HL) (Jia et al., 2004) infused resin for HREEs separation.
Additional examples include PC88A infused styrene–divinyl benzene copolymer for Gd(III) and Tb(III) chromatographic separation (PARK et al., 2005), Octacarboxymethyl-C-methylcalix 4-resorcinarene impregnated Amberlite XAD-16 for La(III), Ce(III), and Y(III) preconcentration (Gok et al., 2007); XAD-4 infused with tri-n-octylmethylammonium chloride (Aliquat-336) in benzene for Ln(III) and Gd(III) (Elsofany, 2008); microporous polymer infused with phosphorus podands bearing two Ph2P(O)CH2C(O)NH for REEs (Turanov et al., 2008); Cyanex272- 2-ethylhexyl phosphoric acid-2-ethylhexyl ester (P507) impregnated resin for HREEs (Liao et al., 2010); TOPS-99 infused Amberlite XAD-4  to extracted Tb (Kumar et al., 2011), β-glycerophosphate impregnated Amberlite XAD 7 for La(III) (Gabor et al., 2016); di-(2-ethylhexyl)phosphinic acid (P227) impregnated XAD-7HP for HREEs (Yang et al., 2020); and HDEHP impregnated XAD-7 for REEs (Sert et al., 2021).
Lee et al. (2009) used trialkylphosphine-based (Cyanex 923), PC88A, and HDEHP for selective separation of La from Ce, Pr, Nd, and Sm, as well as from other concomitant metals in chloride medium. HDEHP provided the best separation with separation factors of 2.57, 3.63, 4.72, and 32.06 for Ce, Pr, Nd, and Sm over La (Lee et al., 2009, 2010b). The same adsorbent was used, followed by anion exchange and oxalate precipitation to obtain La2O3 of more than 99.9998% purity (Lee et al., 2010a). Amberlite XAD-7 coated with β-glycerophosphate had qm of 33.8 mg/g for La (Gabor et al., 2016). Mondal et al. (2019) used N, N,N′,N′-tetrakis-2-ethylhexyldiglycolamide (TEHDGA) impregnated XAD-7 resin to selectively adsorb REEs from coal fly ash solution containing Fe, Ca, Al, Mg, and Si with Kd of 200-520 ml/g for REEs and 0.2-2.9 ml/g for impurities. Cyanex 272 impregnated Amberlite XAD-7 resin showed distribution coefficient in order of Gd > Eu > Sm > Nd > Pr > La with Kd value for Gd and Eu as 208.1 and 156.4 ml/g (İnan et al., 2018). The SIR P227 infused XAD-7HP showed selectivity series of Fe > Lu > Tm > Zn > Mg > Ca > Ho > Co > Ni > Cu > Al (B. Yang et al., 2020). Another adsorbent, HDEHP-XAD-7 showed selectivity series Gd > Eu > Sm > Nd > Pr > La with separation factors achieved Gd/La separation factors of >140 (Sert et al., 2021).
The examples SILs with organic polymer include 1-octyl-3-methylimidazolium hexafluorophosphate (C8mim+PF6-) containing Cyanex923 immobilized on XAD-7 for REE extraction (Sun et al., 2008), trialkylmethylammonium sec-nonylphenoxyacetate ([A336][CA-100]) impregnated on Amberlite XAD-7 (Sun et al., 2009), Trihexyl(tetradecyl) phosphonium mono-(2-ethylhexyl) 2-ethylhexyl phosphonate ([P66614][EHEHP]) and trioctylmethylammonium bis(2,4,4-trimethylpentyl) phosphonate ([N1888][BTMPP]) impregnated Amberlite XAD-7 (Zhao et al., 2016), betainium sulfonyl(trifluoromethanesulfonylimide) poly(styrene-co-divinylbenzene) [Hbet–STFSI–PS–DVB] impregnated Amberlite XAD-16 (Avdibegović et al., 2017), 1-butyl-3-methylimidazoliumbis(trifluoromethylsulfonyl)imide ([Bmim][NTf2]) combined with N,N-dioctyldiglycolamic acid (DODGAA) and impregnated into Amberlite XAD7HP (Friend et al., 2020), and trihexyl (tetradecyl) phosphine bis (2,4, 4-trimethyl-amyl)–phosphonate (Cyphos IL 104) impregnated polymer membrane (Wang et al., 2020).

Inorganic supports

Inorganic supports mostly include silica-based polymers. In some cases, ordered mesoporous silica and carbon nanotubes have been used. Extractant bis(2- ethylhexyl) hydrogen phosphate (HDEHP) coated C18- hydrophobized silica was used for chromatographic separation of lanthanides (Sivaraman et al., 2002); 2-ethylhexylphosphonic acid mono-2-ethylhexyl ester impregnated silica-based urea–formaldehyde (SiO2/UF) for Eu(III) and Nd(III) (Naser et al., 2015); HDEHP impregnated Polyethersulfone (PES) embedded with polyvinyl alcohol (PVA) and multiwalled carbon nanotubes (MWCNT) for Y(III) sorption (Yadav et al., 2015); and tetraoctyldiglycolamide impregnated Carbon inverse opals (C-IOP) for REEs (Turanov et al., 2015).
Cyanex 272 impregnated mesoporous silica MCM 41 was used for the extraction of lanthanides. The selectivity between light lanthanide (La(III)), medium lanthanide (Eu(III)), and heavy lanthanide (Lu(III)) followed the order of Lu(III) > Eu(III) > La(III) with separation factor of 61, 61 and 3504 for Lu(III)/Eu(III), Eu(III)/La(III) and Lu(III)/La(III), respectively (Mohammedi et al., 2019). Silica gel modified with 1-(-2-pyridylazo) naphthol (PAN) and acetylacetone (Acac) through solvent evaporation process was selective for Sc(III) against Fe(III) with selectivity factors of 5.79 and 8.95, respectively at pH 6 (Ramasamy, Puhakka, Repo, & Sillanpää, 2018). Magnetite nanoparticle was used as a support for coating di(2-ethylhexyl)phosphonate (HDEHP), bis(2,4,4-trimethylpentyl)phosphinic acid (CYANEX 272), and bis(2,4,4-trimethylpentyl)dithiophosphinic acid (CYANEX 301) (Molina et al., 2019). These coated magnetite nanoparticles had qmof 6.6-8.3 mg/g for La(III), 7.7-8.7 mg/g for Pr(III), and 4.8-8.9 mg/g for Nd(III). The same extractant, HDEHP was infused into styrene-divinylbenzene copolymer immobilized silica particles (HDEHP-SiO2-P) for REEs adsorption (Shu et al., 2017, 2018; Zhang et al., 2019). The adsorption capacity (qm) of HDEHP-SiO2-P was 38.95 mg/g for Ce(III) and 52.84 mg/g for Gd(III) (Shu et al., 2017). In a different study, the HDEHP-SiO2-P had qm of 14.1 mg/g for Sc(III) with Sc(III) selectivity factor of >50 over other REEs (Zhang et al., 2019).
Inorganic solid supports infused with ionic liquids are mesoporous silica doped with binary IL mixtures (C8mim+PF6-/Cyphos IL 104 or C4mim+PF6-/Cyphos IL 104) (Liu et al., 2009), silica gel impregnated with N-PhenacylPyrNTf2 ionic liquid (Marwani & Alsafrani, 2013), silica doped with bifunctional ionic liquid trioctylmethylammonium 1-phenyl-3-methyl-4-benzoylpyrazol-5-onate ([A336]+[L]−) (Turanov et al., 2016), and MCM-41 silica impregnated with [Hbet–STFSI–PS–DVB] (Avdibegović et al., 2017). Among the SILs, trihexyl(tetradecyl)phosphonium bis(2,4,4-trimethylpentyl)phosphinate (Cyphos) entrapped in silica reached maximum adsorption capacities of 14.7 mg/g for Nd(III) and 19.8 mg/g at pH 4.0 (Mohamed et al., 2017). Ionic liquid (IL), 1-butyl-3-methylimidazolium bromide ([C4mim]+[Br]) impregnated metal-organic framework (MOF) UiO-66 had a maximum Gd(III) adsorption capacity of 65 mg/g at pH 6 with Gd(III) selectivity in the presence of other metal ions (Na(I), Ca(II), Ma(II), Al(III), and Fe(III)) (Ahmed, Adhikary, et al., 2019).
The main limitation of SLEs is the gradual washout of the active component during operation (Bao et al., 2016). Possible solutions include the selection of highly hydrophobic and low soluble active components (Muraviev et al., 1998; Sparfel & Cote, 2004) and/or thorough washing of weakly adsorbed active components after their application onto the solid surface (Muraviev, 1998; Muraviev et al., 1998). Another method to prevent the leaking of the extractant from adsorbent grain is to coat a layer of polymer on the adsorbent, which acts as a semipermeable membrane or protective barrier for extractant while allowing the metal ion diffusion (Bao et al., 2016; Muraviev et al., 1998; Nishihama et al., 2013; Trochimczuk et al., 2004).
A protective barrier around the SIR was formed using polysulfone in dimethylformamide (DMF); however, loss of the adsorbed extractant [bis(2-ethylhexyl) hydrogen phosphate, bis- (2-ethylhexyl) hydrogen dithiophosphate, or bis(3- propylphenyl) dithiophosphate] still occurred (Muraviev et al., 1998). Better results were obtained with the coating of water-soluble polymer poly(vinyl alcohol) (PVA) precipitate onto SIR (Trochimczuk et al., 2004). The stability of the coating layer can be increased by cross-linking PVA with vinyl sulfone (Trochimczuk et al. 2004), boric acid (Yuan et al., 2010), and glutaraldehyde (Nishihama et al., 2013).
In general, EIMs (SIRs and SILs) can improve the performance of extractants by avoiding drawbacks of LLE, such as poor contact between the organic and aqueous phase, requirement of a large amount of extractant, emulsion formation, and formation of the third phase. The issue with SLEs is the stability of the active layer, which can wash out during operation. Moreover, the extractant layer may also be not uniform throughout the support surface, e.g., capillary forces may accumulate the extractant in narrow pores, a process enhanced in contact with an aqueous solution (Shenxu Bao et al., 2016). In addition, the complex nature of real-life REEs sources such as very acidic solutions can put extra pressure on these adsorbents, which can limit their industrial applications.
The adsorbents obtained through ligand functionalization to the surface of solid support using coupling agents provide an alternative to EIMs. The surface-functionalized adsorbents can be an alternative with better control of ligand distribution on the surface and better ligand and surface bonding and thus better stability, but however, they can be more resource intensive to produce.