Figure 3: Different bite angles of the ligands grafted onto KIT-6
(modified after (Florek et al., 2015; Hu et al., 2017, 2019)). DOODA:
3,6-dioxaoctanediamidopropyl, DGA: diglycolylamide FDGA:
furan-2,4-diamidopropyl (Florek et al., 2015, 2020), PA: phthaloyl
diamide (Hu et al., 2017, 2018) PDDA: phenylenedioxy diamide (Hu et al.,
2019).
In other examples of functionalized ordered mesoporous silica used for
REEs uptake, Dy(III) was selectively adsorbed from a solution of Cu(II),
Dy(III), Fe(III), Nd(III), and Zn(II) using amino and carboxylic
functionalized mesoporous silica having different pore sizes (3, 5, 12,
and 22 nm) (Kaneko et al., 2018) and using amino, carboxylic, and
diglycolic anhydride functionalized mesoporous silica with a sheet and
spherical morphologies as well as non-porous stöber silica (Kaneko et
al., 2019). Zhang and their colleagues (Zhang, Avdibegović, et al.,
2017) used titanium(IV) alkylphosphate grafted MCM41 for adsorption in
the binary equimolar solutions of Sc-La and Nd-Dy. Ravi et al. (2018)
functionalized KIT-6 and KCC-1 with EDTA and employed them for Nd(III)
adsorption from single-element solution (Ravi, Zhang, et al., 2018). The
same team, in a different study, used
benzene-1,3,5-triamido-tetraphosphonic acid (BTATPA) modified SBA-15 for
adsorption of Nd(III), Y(III), La(III), and Ce(III) from single element
solution and reported Kd (ml/g) values of
~104,
~104,
~1.5•105, and
~1.04•105 respectively (Ravi, Lee, et
al., 2018). In other studies, SBA-15 functionalized with phosphorus acid
achieved a maximum Gd(II) adsorption capacity of 204.4 mg/g (Gao et al.,
2017), and sulfhydryl MCM-41 (SH-MCM-41) showed maximum adsorption
capacities of 560.56, 467.60, and 540.68 mg/g for La(III), Gd(III), and
Yb(III), respectively (Li et al., 2019).
Mesoporous silica foam (MSF) and fibrous nano-silica KCC-1 were grafted
with poly(amidoamine) dendrimers (PAMAM) using
3-aminopropyltriethoxysilane (APTES) and glutaraldehyde (GTA) as a
cross-linking spacer (Lee et al., 2019). The adsorption capacity of MSF
based adsorbent
PAMAM@GTA-NH2-MSF
was higher than KCC-1 based adsorbent
PAMAM@GTA-NH2-KCC-1 with values of 86, 104, 110, and 132
mg/g for La(III), Ce(III), Nd(III), and Gd(III) respectively for MSF
based adsorbent. Both adsorbents showed selectivity for Gd(III) from a
solution containing Na(I), Ca(II), Mg(II), Al(III), and Fe(III) with the
recovery of 75.8% with PAMAM@GTA-NH2-MSF and 67.2%
with PAMAM@GTA-NH2-KCC-1. The recovery of Gd(III)
increased to 86% with further functionalization of
PAMAM@GTA-NH2-MSF with N-(phosphonomethyl) iminodiacetic
(PMIDA) (Lee et al., 2019).
N-methyl-N-phenyl-1,10-phenanthroline-2-carboxamide (MePhPTA)
functionalized mesoporous silica adsorbed Eu(III) and Sm(III) with
maximum adsorption capacities of 125.63 and 124.38 mg/g, respectively at
pH 4 (Awual et al., 2013). Ordered mesoporous silica functionalized with
4-tert-Octyl-4-((phenyl)diazenyl)phenol (TOPP) reached maximum Yb(III)
adsorption of 139.19 mg/g at pH 5 and was selective against Na(I), K(I),
Ca(II), Mg(II), Zn(II), and Al(III) (Rahman et al., 2020). Hollow
mesoporous silica nanosphere (HMSNs) functionalized with
amino-phosphonic acid had an adsorption capacity of 387.3 mg/g for
Gd(III) at pH 5 (Yin et al., 2020).
Carbon-based supports
Carbon-based adsorbents and
supports have various forms, such as mesoporous carbon, activated
carbon, graphene oxide, and carbon nanotubes. Carbon-based support can
be an alternative to silica supports to overcome the low stability of
silica-based support in acidic conditions (Hu et al., 2018). However,
the surface functionalization of carbon-based supports can be more
challenging than silica-based supports (Hu et al., 2018).
Ordered mesoporous
carbons
Ordered mesoporous carbons (OMCs) can work as better solid support under
acidic conditions because of higher chemical resistance than mesoporous
silica while possessing comparable porosity and surface area (Hu et al.,
2018). However, the synthesis and functionalization of OMCs are more
challenging than OMS. Ordered mesoporous carbon synthesis can be done by
hard or soft template methods.
In the hard template method, an ordered mesoporous solid (mostly ordered
mesoporous silica) is used as a mold, and its pores are impregnated with
a selected carbon precursor (e.g., Sucrose), which is then polymerized
and carbonized (Eftekhari & Fan, 2017). In the end, the template is
dissolved and removed using strong acid (e.g., HF) or base (e.g., NaOH)
(Chuenchom et al., 2012). The porosity and structure of OMC can be tuned
by selecting an appropriate hard template.
On the other hand, in the soft template method, OMC is directly
synthesized by self-assembling a block of copolymer surfactant and a
carbon precursor. In this method, the template is removed by a thermal
process. The pH, temperature, and gel composition can be varied to
control the porosity and morphology of the mesoporous carbon (Ma et al.,
2013). In comparison with the hard template method, OMCs synthesized
through soft templates have a continuous framework with thick pore
walls, hence have more stability against thermal and oxidation treatment
during functionalization (Eftekhari & Fan, 2017).
In adsorption studies involving OMCs, Parsons-Moss et al. (2014)
investigated Eu(III) adsorption using a carbon-silica nanocomposite
(C-CS) type OMC functionalized with -COOH and reported adsorption
capacity (qm) of 138 mg/g at pH 4.0. Perreault et al.
(2017) functionalized CMK-8, a 3D cubic ordered mesoporous carbon, with
different DGA-based ligands. Among these CMK-8 based adsorbents, the
Kd value for different Ln(III) in the presence of
Al(III) and Fe(III) were highest for diglycolylester grafted CMK-8
(CMK-8-DGO) followed by oxidized CMK-8 (CMK-8-O) and chloropropyl
diglycolylamide grafted CMK-8 (CMK-8-PGA) (legend 6-8, Figure 2).
Bertelsen et al. (2019) modified
two OMC materials, MC-l -MSN, and CMK-3 type OMC, with
bis-(2-ethylhexyl) phosphoric acid (HDEHP) and investigated batch
adsorption of Eu(III) as well as column separation of Nd(III) and
Eu(III) using HDEHP-OMC-l- MSN. The HDEHP-OMC showed
qm of 0.35 mmol (or 53.2 mg) Eu(III)/g. Mesoporous
carbon was functionalized with a single stranded oligo consisting of 100
thymine units, and it achieved adsorption of 9.57, 38.27, and 52.15
mg/g, for Lu(III), Dy(III), and La(III), respectively. The adsorbed
amounts of REEs increased with an increase in the metallic radius from
Lu to La (Gismondi et al., 2022; Unsworth et al., 2020).
Activated carbon /Graphene oxide /Carbon nanotube
Activated carbon. Activated carbon (AC) is an environmentally
friendly and low-cost adsorbent with a high surface area but has little
affinity for REE(III) in native form; thus, functionalization with
ligands that have affinity for REEs is required for its application in
REEs adsorption (Asadollahzadeh et al., 2020). The ligands used for AC
functionalization are Schiff’s base derived from 3 diethylenetriamine
and 3,4-dihydroxybenzaldehyde (AC-DETADHBA) (Marwani et al., 2017),
Carboxylic acid (Kilian et al., 2017; Marwani et al., 2017), and EDTA
(Babu et al., 2018). The EDTA functionalization of AC (AC-EDTA)
increased adsorption capacity to 71.4 Nd(III) mg/g from 19.1 Nd(III)
mg/g at pH 5 (Babu et al., 2018). The adsorbent AC-EDTA showed higher
selectivity towards HREEs. Carboxyl functionalized AC (AC-COOH) showed
an adsorption capacity of 89.50 mg/g for La(III) at pH 6.0, whereas
AC-DETADHBA showed a higher adsorption capacity of 144.80 mg/g for
La(III) at the same pH (Marwani et al., 2017). In a different study,
AC-COOH showed very low adsorption of 2.1 mg/g for Sc(III) at pH 2.0
(Kilian et al., 2017), suggesting that the pH plays a vital role in
AC-COOH adsorbents, likely due to the weakly acidic nature of -COOH
group.
Graphene oxide. Graphene oxide (GO) contains hydroxyl, epoxide,
carbonyl, and carboxyl groups, essential sites for functionalization
with ligands. Moreover, due to the presence of these groups, GO can
adsorb REEs without functionalization. A colloidal graphene oxide
suspension had Gd(III) adsorption capacity of 286.86 mg/g at pH 5.9
(Chen et al., 2014). Graphene oxide nanosheets (GONS) obtained maximum
adsorption of 175.44 mg/g for Eu(III) at pH 6 (Sun et al., 2012). The
maximum Eu(III) adsorption decreased slightly to 167.16 mg/g at pH 2
(Sun et al., 2012). In another study with GO nanosheet, the maximum
adsorption capacities were in the order of
Gd(III)>Nd(III)>Y(III)>La(III)
with values 225.5, 188.6, 135.7, and 85.7 mg/g at pH 6, respectively
(Ashour et al., 2017). Unmodified Graphene oxide showed an adsorption
capacity of 39.7 mg/g for Sc(III) at pH 4, which was better than -COOH
functionalized AC and chelating resin Chelex 100 (contains iminodiacetic
acid) but lower than -COOH modified MWCNT (Kilian et al., 2017). Higher
adsorption of 89.7 mg/g and 70.2 mg/g for Eu(III) using GO and magnetic
GO (by modifying with Fe3O4) was
achieved at pH 7 ( Li et al., 2015).
GO was used to produce amine-functionalized mesoporous graphene (AMG)
followed by functionalization with carbamoyl phosphine oxide moiety
(CMPO, triethylphosphonoacetate) and it reached an adsorption capacity
of 26.9 mg/g for La(III) (Kim et al., 2019).
Fe3O4 and MnO2 modified
GO nanocomposites,
Fe3O4/MnO2/rGO (FMG),
provided high maximum adsorption capacities of 1016 and 981 mg/g for
La(III) and Ce(III), respectively, at pH 7 (Liu et al., 2020) and is
among the adsorbents with the highest capacity for REE(III) (Table 1).
Another adsorbent, polyaniline functionalized GO, achieved 250.74 mg/g
Eu(III) adsorption (Sun et al., 2013). Polyurethane sponge was used as a
support polymer for titanium phosphate with graphene oxide
(GO@TiP-Sponge) and showed a maximum capacity of 576.17 mg/g for Dy(III)
(Peng et al., 2020). In comparison, GO@TiP had a maximum Dy(III)
adsorption capacity of 316.75 mg/g (Peng et al., 2020). GO-
tris(4-aminophenyl)amine composite had a Langmuir adsorption capacity of
46.35 mg/g for Yb(III) at pH 6 (Zhao et al., 2021). Another GO-based
adsorbent, graphene oxide/poly (N-isopropyl acrylamide-maleic acid)
[GO/P(NIPAM-MA)] cryogel, adsorb La(III) with equilibrium adsorption
capacity of 33.1 mg/g at pH 5 with separation factors of 2.86, 7.57,
8.00, 6.69, and 1.45 for La over Cu, Co, Ni, Nd, and Yb, respectively
(Yang et al., 2020). GO functionalized with 3-[2-(2-aminoethylamino)
ethylamino] propyl-trimethoxysilane (GO-APTS) showed adsorption
capacities of 110.0, 93.4, 103.2, 83.7, 97.2, 48.3, and 92.8 mg/g for
Ho(III), Er(III), Eu(III), Lu(III), Tm(III), Y(III), and Yb(III),
respectively (Bao et al., 2022). The adsorbent was selective for Ho(III)
over alkali or alkaline-earth metal ions with SFs of above 600. A
further modified adsorbent combining GO-ATPS with
Fe3O4 resulted in lower capacities of
72.1, 67.1, 65.2, 65.0, 73.2, 36.3, and 68.7 mg/g for Ho(III), Er(III),
Eu(III), Lu(III), Tm(III), Y(III), and Yb(III), respectively, due to
lower APTS fraction in the composite (Bao et al., 2022).
Carbon nanotubes. Carbon nanotubes (CNTs) are tubes of graphite
sheets and can be divided into two types: Single-walled carbon nanotubes
(SWCNTs) and multi-walled carbon nanotubes (MWCNTs) with a set of
concentric nanotubes. The examples of CNTs based adsorbent studied for
REEs separation include oxidized MWCNTs (Alguacil et al., 2020; Behdani
et al., 2013; Fan et al., 2009; Koochaki-Mohammadpour et al., 2014; K.
Li et al., 2015), CNT modified with -COOH (Kilian et al., 2017), MWCNT
modified with tannic acid (Tong et al., 2011), and
MWCNT/Fe3O4 composites (Chen et al.,
2009; Fan et al., 2009). The CNT modified with -COOH showed higher
adsorption (37.9 mg/g) for Sc(III) than AC-COOH (2.1 mg/g) at pH 2.0
(Kilian et al., 2017). The adsorption of Sc(III) onto CNT-COOH increased
to 42.5 mg/g at pH of 4.0. Oxidized MWCNTs showed good adsorption
(80-100%) for target REE(III) in the presence of competitive ions, and
the adsorption increases with pH since at higher pH, the surface is
negatively charged, leading to higher sorption of positively charged
REE(III) cations (Cardoso et al., 2019). A CNTs/GO hybrid hydrosols had
Gd(III) adsorption capacity of 534.76 mg/g at pH 5.9 (Lanyu Guo et al.,
2018). CNT modified with furfuryl amine and further functionalized with
poly(acrylic acid) reached a maximum equilibrium Eu(III) adsorption
capacity of 130.8 mg/g as opposed to 46.67 mg/g for pristine CNT at pH 7
(Linru Guo et al., 2020).
Other supports/adsorbents
Magnetic nanomaterials
Researchers have explored
adsorbents obtained after coating Fe3O4nanoparticles with SiO2 or TiO2 shell
and then the functionalization of the surface of the coated layer. The
advantage of magnetic nano-adsorbent is the convenient
collection/separation of adsorbents by using magnets after adsorption,
avoiding the issue involved with the isolation of nanoparticles, which
tends to form stable sols in batch reactors. The coating with
SiO2 or TiO2 provides higher surface -OH
density, which results in higher surface ligand density and easier
functionalization. The coating also protects the
Fe3O4 nanoparticle from oxidation.
Polido Legaria et al. (2017) used EDTA, DTPA, and
triethylenetetraminehexaacetic acid (TTHA) functionalized
SiO2 nano-adsorbents, both magnetic and nonmagnetic, and
reported uptake up to 300 mg REE(III)/g. REE uptake order was Dy(III)
> Nd(III) > La(III) for EDTA-functionalized,
Nd(III) > La(III) > Dy(III) for
DTPA-functionalized, and La(III) > Nd(III) >
Dy(III) for TTHA-functionalized adsorbent. Kostenko et al. (2019)
synthesized Fe3O4 magnetic nanoparticles
(MNPs) with and without a coating of silica and functionalized them with
aminomethylenephosphonic (MNPs/SiO2-AMPA and MNPs/AMPA).
The adsorption capacities for Eu(III) were 69 and 77 mg/g for
MNPs/SiO2-AMPA and MNPs/AMPA at pH 7.0, respectively.
Magnetic Fe2O3 coated with silica and
functionalized with APTMS,
(γ-Fe2O3-NH4OH@SiO2(APTMS))
achieved maximum Dy(III) adsorption of 23.3 mg at pH 7 (Kegl et al.,
2019). Magnetic nanoparticle functionalized with poly(aminoethylene
N-methyl 1-formic acid, 1-phosphonic acid) (PAEMFP) reached 25 mg/g
Nd(III) adsorption capacity at PH 6 (Miraoui et al., 2016). Other
research involving coated magnetic nano-adsorbents for REEs adsorption
include silica-coated magnetic particle grafted with 2-ethylhexyl
phosphonic acid mono-2-ethylhexyl (P507) using
3-chloropropyltryethosysilane as silane-coupling agent (Wu et al., 2013)
and magnetic Fe3O4 coated with
SiO2 and TiO2 and modified with
N-[(3-trimethoxysilyl) propyl] EDTA (TMS-EDTA) (Dupont et al.,
2014).
Some studies have directly functionalized the magnetic nano-particles
(Chen et al., 2012; Gaete et al., 2021; Ngomsik et al., 2012; Yang et
al., 2012). Citric acid-functionalized maghemite nanoparticles had
qm of 74.4 mg/g and 71 mg/g for Eu(III) and La(III),
respectively (Ngomsik et al., 2012), Magnetic nano-particles
Fe3O4 with humic acid coating
(Fe3O4@ HA MNPs) synthesized through
chemical coprecipitation was able to remove ∼99% of Eu(III) at pH 8.5
(Yang et al., 2012). In another study, bentonite was used as a support
to prepare di (2-thylhexly) phosphoric acid-immobilized magnetic GMZ
bentonite and was used to adsorb Eu(III) (capacity of 48.02 mg/g at pH
3.05) (Chen et al., 2012). Another adsorbent, phosphonic
acid-functionalized magnetite nanoparticle (PA-MNP), showed a capacity
of 18.4, 17.6, and 23.9 mg/g for La(III), Pr(III), and Sm(III),
respectively at pH 4 (Gaete et al., 2021).
Metal-organic frameworks
Metal-organic frameworks (MOFs) have ordered crystalline porous
structures composed of a three-dimensional network of metal or metal
oxide clusters held together by multidentate organic ligands (i.e.,
linker), e.g., aromatic polycarboxylates. MOFs have received
considerable attention as support for grafting various REEs selective
ligands. The desirability of MOFs as support lies in their high surface
area, high porosity, adjustable pore size, and tunable surface
functionality (Baumann et al., 2019).
The MOF, HKUST-1, showed an adsorption capacity of 353 mg/g for Ce(III)
at pH 6 (Zhao et al., 2019). MOF ZIF-8 showed higher adsorption capacity
(385 mg/g) for La(III) than another MOF ZIF-90 (qm = 168
mg/g for La(III)) that had a different linker (Jiang et al., 2016). MOF
Zeolitic imidazolate frameworks-8 nanoparticle (ZIF-8 NP) has adsorption
capacities of 28.8 mg/g, 281.1 mg/g, and 430.4 mg/g for La(III),
Sm(III), and Dy(III), respectively at pH 7 (Abdel-Magied et al., 2019).
MOF ZIF-8 functionalized with -COOH showed an adsorption capacity of 175
mg/g for Nd(III), significantly higher than unfunctionalized ZIF-8 (57
mg Nd(III)/g)) (Ahmed et al., 2021). Another
MOFsUiO-66-(COOH)2 functionalized with polyacrylonitrile
had maximum adsorption capacities of 214.1 mg/g and 191.9 mg/g for
Tb(III) and Eu(III), respectively at pH 6 (Hua et al., 2019). Similar
MOF UiO-66 dual functionalized with -COOH and -NH2obtained equilibrium adsorption capacity of 79 mg/g for Gd(III),
significantly higher than pristine UiO-66 with 16 mg Gd(III)/g (Ahmed,
Lee, et al., 2019). A MOF, ZnGA, synthesized by the reaction of zinc
acetate and glutaric and then functionalized with polymer-based on
p-chlorocresol and piperazine had Y(III) adsorption capacity of 377.02
mg/g at pH 6 (Mahmoud et al., 2019).
In several different studies, MIL-101 was functionalized with different
functional groups and used for REEs adsorption
(de Decker et al., 2016; Kavun et
al., 2021; Y.-R. Lee et al., 2018; Lou et al., 2019; Ryu et al., 2021).
Among MIL-101 functionalized with -NH2, ethylenediamine
(ED), diethylenetriamine (DETA), and N-(phosphonomethyl)iminodiacetic
acid (PMIDA), the REE adsorption capacity increased in the order of
MIL-101 < MIL-101-NH2 < MIL-101-ED
< MIL-101-DETA < MIL-101-PMIDA with the highest
being 90.0 mg Gd(III)/g onto MIL-101-PMIDA at pH 4.5 (Lee et al., 2018).
Acrylic acid-functionalized MIL-101 had adsorption capacities of 90.21,
104.59, 58.29, and 74.94 mg/g for Sc(III), Nd(III), Gd(III), and
Er(III), respectively, at pH 5.5, and was selective for Sc(III) in the
presence of other REEs and competitive cations (Lou et al., 2019). In a
different study involving MIL-101 functionalization with tributyl
phosphate (TBP), HDEHP, and bis(2,4,4-trimethylpentyl) phosphinic acid
(Cyanex 272), the adsorption capacity increased in the order of
MIL-101-TBP < MIL-101-Cyanex 272 < MIL-HDEHP with a
maximum adsorption capacity of 57.47 mg/g at pH 5.5 for HDEHP
functionalized adsorbent (Kavun et
al., 2021). The MIL-101-HDEHP showed selectivity factors of 22.8 and 7.7
for Er(III) over Nd(III) and Gd(III) in Er(III)-Nd(III)-Gd(III) mix
solution with all three adsorbent showing high selectivity
(>90%) for Er(III) over Co(II), Ni(II), Cu(II), and
Zn(II).
Metal-based supports
A metal-based support Mesoporous Zirconium Titanate (ZrTi-0.33) was
functionalized with methylphosphonic acid (MeP), amino
trismethylenephosphonic acid (ATMP), phosphono-imido-dicarboxylic acid
(PIDC), 4-amino,1-hydroxy,1,1-bis-phosphonic acid butane (HABDP), and
1-hydroxylethylene-1,1-bis-phosphonic acid (HEDP) bis-phosphonic acids,
1,4-diphosphonic acid butane (BuDP) and 1,4-diphosphonic acid
benzene(BenDP) and used for 153Gd(III) adsorption
(Griffith et al., 2010). ATMP based adsorbent achieved the highest
Kd (>10000 in 10-3 M and
10-5 M HNO3).
Similar support, mesoporous TiO2 particle, was
functionalized with dimethylphosphato-ethyltriethoxysilane and reached
Ce(III) adsorption capacity of 92.6 mg/g while completely separating
Ce(III) from Sr(II) and Cs(I) (Moloney et al., 2014). Mg-Fe hydrotalcite
modified with Cyanex 272 (bis (2,4,4-trimethylpentyl) phosphinic acid)
showed separation factor (SF) >6.3 for La(III)/Nd(III)
(Gasser & Aly, 2013) and >3.6 for Ce(III)/Eu(III) (Gasser
et al., 2017).
Composite supports
A composite support of nano-silica (∼12 nm particle size) and AC (0.8
mm pellets) was modified with 1-(2-Pyridylazo) 2-naphthol (PAN)
(Ramasamy, Puhakka, Repo, Ben Hammouda, et al., 2018). The adsorbent
obtained adsorption capacities of 103.5, 112.7, and 84.1 mg/g for
La(III), Sc(III), and Y(III), respectively. Another composite support
consisting of nano-silica grafted carbon nanotubes (both SWCNT and
MWCNT) functionalized with 1-(2-pyridylazo)-2-naphthol (PAN) were
selective for Sc(III) in mixed REE solution (Ramasamy et al., 2019).
However, the adsorption capacities for Sc(III) were lower than La(III)
and Y(III) in mono-element solution with adsorption capacities of 12.68,
80.68, and 48.34 mg/g for Sc(III), La(III), and Y(III), respectively, in
the SWCNT and 103.2, 32.92, and 68.78 mg/g, respectively with the MWCNT
(Ramasamy et al., 2020). MWCNT embedded with
Fe3O4 nanoparticles were functionalized
with carbon disulfide and achieved 23.23 mg/g La(III) adsorption (Huang
et al., 2021). A nano porous graphene and zinc-trimesic acid (Zn-BTC)
MOF composite showed selectivity for Ce(III) with SFs 12,081.04, 20.05,
and 328.80 for Ce(III) over Lu(III), La(III), and Pr (III), respectively
(Wu et al., 2021). The SFs between adjacent lanthanide Nd/Pr, Sm/Eu,
Gd/Tb, and Tb/Dy reached 9.80, 3.11, 2.20, and 1.68. The maximum
capacity of the adsorbent was about 300 mg REEs/g. Another composite
consisting of nano porous graphene oxide and zinc- terephthalic acid
(Zn-BDC) MOF reached a maximum capacity of 344.48 mg REEs/g with high
SFs for Sc, i.e., Sc/Tm ≈ 529.57, Sc/Er ≈ 461.91, and Sc/Y ≈ 445.70
(Chen et al., 2022). It was also selective for Tm and Er with SFs for
Tm/Eu ≈ 4.55, Tm/Pr ≈ 4.20, Tm/Nd ≈ 3.96, and Er/Eu ≈ 3.9. A composite
obtained via the introduction of a MOF MIL-101 shell over a magnetite
(Fe3O4) core and functionalized with
-SO3 and diethylenetriamine (DETA) extracted REEs from
aqueous and brine solution with efficiency up to 99.99% (Elsaidi et
al., 2018).
A silica/polymer (SiO2-P) composite support synthesized
by immobilizing styrene-divinylbenzene copolymer (SDB) in porous silica
(SiO2) was functionalized with
2,6-bis(5,6,7,8-tetrahydro-5,8,9,9-tetramethyl-5,8-methano-1,2,4-benzotriazin-3-yl)pyridine
(Me2-CA-BTP) and was used for Ln(III) adsorption.
Similar SiO2-P supports were functionalized with a mixed
trialkyl phosphine oxide (TRPO) (where R1,
R2, and R3 are different alkyl groups)
(Yu et al., 2018) and functionalized with HDEHP for REEs(III) adsorption
(Zhang et al., 2019).
An organic-inorganic hybrid hydrogel,
polyethylenimine–acrylamide/SiO2, showed selectivity
factors of 105-450 for REEs over other metals (Wang et al., 2017). A
porous carboxymethyl cellulose hydrogel containing polyacrylic acid had
a monolayer adsorption capacity (qm) of 381.72 mg/g for
La(III) at pH 5.58 and 320.47 mg/g for Ce(III) at pH 5.83 (Roosen et
al., 2016; Zhu et al., 2016). Chitosan containing acrylic
acid-co-styrene sulfonate/smectite clay hybrid granular hydrogel
(CTS-g-(AA-co-SS)/ISC) had adsorption capacities of 232.97 mg/g for
Gd(III) and 185.47 mg/g for Ce(III) at pH 5 (WANG et al., 2017). A
magnetic
MnFe2O4@SiO2-chitosan
composite achieved one of the highest adsorption capacities of 1030 mg/g
and 1020 mg/g for La(III) and Ce(III), respectively, (Table 1) (Liu et
al., 2021).
Similar to chitosan containing hydrogel, many composite supports are
based on composites of biological materials with inorganic materials or
polymers (Iftekhar, Ramasamy, et al., 2018). REEs adsorption has been
studied with silica-chitosan hybrid beads functionalized with EDTA, DTPA
(Roosen et al., 2014), PAN, and acetylacetone (Ramasamy, Puhakka,
Iftekhar, et al., 2018), Zn/Al Layered double hydroxide (LDH)
intercalated cellulose (CL) (Iftekhar et al., 2017b), cellulose-silica
nanocomposite (Iftekhar et al., 2017a), and Gum Arabic
(GA)-polyacrylamide-silica nanocomposite (Iftekhar, Srivastava, et al.,
2018). Among these adsorbents, silica-chitosan-PAN showed the highest
adsorption capacities of 199.8, 198.8, and 123.4 mg/g for La(III),
Sc(III), and Y(III), respectively, at pH 5. The LDH-CL adsorbent had
qm values of 102.25, 92.51, and 96.25 mg/g for Y(III),
La(III), and Ce(III), respectively at pH 7 (Iftekhar et al., 2017b). The
cellulose-silica nanocomposite and the GA-polyacrylamide-silica
nanocomposite reached qm of 29.48 mg/g and 7.90 mg/g for
La(III), respectively and 23.76 mg/g and 11.05 mg/g for Sc(III),
respectively at pH 6 (Iftekhar et al., 2017a; Iftekhar, Srivastava, et
al., 2018).The EDTA and PDTA functionalized silica-chitosan had
adsorption capacity (qm) of 0.27 mmol/g (or 38.9 mg/g)
for Nd(III) at pH 6 (Roosen et al., 2014), and equilibrium capacity of
6.74 mg/g and 8.99 mg/g for Sc(III) at pH 2.