Introduction

Lightweight lithium-ion batteries (LIBs) with cathode electrode material LiCoO2 are extensively used in various electronic devices including electric vehicles1,2. LIBs are high-energy storage materials3,4 owing to their low memory effect and long service life5,6. The Demand for LIBs has risen dramatically; the electric cars in China alone surpassed 2 million by 20207 and is predicted to exceed 300 million globally by 20308. Rapid growth will eventually lead to an enormous number of spent LIBs9 thereby causing environmental pollution and the depletion of natural resources10. For sustainability, spent LIBs must be recycled11 since an average of 5 weight% is lithium12.

Salt brine accounts for 69% of the world’s lithium reserves13. Lithium is recovered from various primary resources via several protocols14,15. Recovery processes involving hydrometallurgy and pyrometallurgy have been proposed to recycle valuable materials from waste LIBs16,17.

Exposure to lithium ions causes serious damage to human health18 thereby requiring their removal and recycling from industrial wastewater produced by the disposal of spent LIBs. Cost-effective and environmentally friendly methods must be developed for trapping Li+ selectively from coexisting alkali and alkaline earth metals in the Salt Lake brine. An effective process is essential for extracting low lithium concentrations from seawater and industrial and mining wastewater19.

Crown ethers (CEs) as cyclic polyethers have a hydrophobic ring surrounding a hydrophilic cavity. The unshared electron pairs of oxygen atoms can form host–guest complexes by selectively coordinating the s-block elements20,21. The manufacturing of CEs and their metal ion complexing properties22 focus on the electrostatic attraction between the cation and the oxygen lone pairs of the CEs leading to the metal complexes via the combined effects of chemical and physical sorption of the metal cations. Matching of cavity sizes of 12- to 14-membered CEs in the range of 1.2–1.5 Å and Li+ with an ionic diameter of 1.36 Å leads to its selective complexation23.

Liquid–liquid extraction or liquid membrane systems integrated with CEs utilize toxic organic solvents24,25. They have low to moderate extraction efficiency. However, the solid‒liquid extraction is associated with a high separation factor and can be recycled26. Polymers decorated with CEs as pendants have been utilized in various membrane-based separation applications27,28 including the trapping of Li+ from wastewater29 and have a high adsorption capacity (4.07 mg g−1). A Li+-imprinted material demonstrated a qm value of 16.4 mg g−1 from its 40 ppm Li+ solution30.

In this work, 12-crown-4-derived diallylammonium monomer 6 was synthesized (Scheme 1). Polymers 7 and 8 and novel resins (i.e., cross-linked) 10 and 11 (Schemes 2, 3) bearing motifs of Aliphatic 12-crown-4 have been synthesized via Butler’s cyclopolymerization protocol31. Polymers incorporating five-membered cyclic structures within their main chains have been acknowledged as the eighth fundamental class of synthetic polymer architectures32,33. To our knowledge, the materials used in this work are the first to feature a combination of strongly coordinating ether and amine groups, along with weakly coordinating sulfone (SO₂) groups. Resin 11 was examined for its capability to adsorb lithium ions (Li+) from aqueous solutions via a solid–liquid separation approach. Furthermore, lithium desorption from the cross-linked polymer (resin 11) was investigated to explore the recovery of Li+ ions and the potential recycling of the resin.

Experimental details

Chemicals and materials

2-(Hydroxymethyl)−12-crown-4-ether from TCI Chemicals (Tokyo Chemical Industry), p-toluene sulfonyl chloride (tosyl chloride) from Alfa Aesar, LiCl from BDH Company, HCl and NaCl from Fisher Scientific Company, 2,2′-Azobis(2-methylpropionamidine) dihydrochloride (AMPD), ammonium persulfate (APS), 2,2′-azobis(2-methylpropionitrile) (AIBN), diallylamine, KCl and MgCl2 from Sigma Aldrich were purchased. Tetraallylhexane-1,6-diammonium chloride monomer was synthesized as described previously34. All glassware was cleaned with water (deionized).

Physical methods

A ThermoFisher Scientific™ Nicolet™ iS10 FTIR spectrometer was utilized to obtain the IR spectra. The NMR spectra were measured with a 600-MHz JEOL spectrometer. Thermogravimetric analysis (TGA) under a N2 flow (50 mL/min) and a temperature change of 10 °C/min in the range 20–800 °C was accomplished using an SDT analyzer (TA instruments, Model: Q600). Elemental analyses of the samples were performed in a Perkin Elmer instrument (Series II Model 2400). Viscosity measurements at 30.0 ± 0.1 °C were carried out via a Ubbelohde viscometer (Viscometer Constant = 0.005 cSt/s) in a water bath (CT72/P, SI Analytics, Germany). At each concentration, at least triplicate measurements were taken to maintain < 0.2% error in time. The pHs of the solutions were measured using an Oakton pH meter.

Scanning electron microscopy (SEM) images were captured on a THERMO QUATRRO S (ThermoFisher TM Scientific). For morphological characteristics, the microscope, which was equipped with an energy dispersive X-ray spectrometer (EDX, Oxford Inc.), was operated in backscattered electron mode at different accelerating voltages and working distances. Before analysis, the samples were sprinkled on double-sided adhesive tape and coated with a thin layer of gold for 10 s. ICP‒OES - an Optima 8000 from Perkin Elmer - was used to determine the lithium concentrations. The specifications were as follows: light source: axial view, torch; detector: two-dimensional Echelle semiconductor detector (CCD); wavelength range: 165–800 nm; RF power: 40 MHz; 750–1500 W. An ESCALAB™ 250Xi X-ray photoelectron spectrometer (Thermo Scientific) with monochromatic micro-focused Al Kα as an X-ray source was used for XPS analysis. The C 1 s peak at 284.8 eV was taken as a reference to calibrate the XPS spectra. Organic solvents of HPLC grade and double distilled water (Millipore, 18.2 MΩ·cm) were utilized.

Synthesis of 2-(tosyloxymethyl)−12-crown-4-ether (3)

Tosyl chloride 2 (1.14 g, 6.0 mmol) and KOH (1.12 g, 20 mmol) were added to a solution of 2-(hydroxymethyl)−12-crown-4-ether 1 (1.02 g, 5.0 mmol) in CH2Cl2 (DCM) (40 mL) at 0 °C and stirred at 25 °C for 24 h. After washing the heterogeneous mixture with water (3 × 10 mL), the DCM layer after drying (Na2SO4) and concentration afforded tosylate derivative 3 (colorless oil, 1.9 g, 95%). The above reaction was repeated several times to synthesize more of compound 3. νmax (neat): 2861, 1597, 1495, 1ed450, 1356, 1189, 1175, 1131, 1096, 1073, 1019, 974, 919, 815, 788, 734 and 705 cm−1; δH (CDCl3): 2.43 (3 H, s), 3.43–3.87 (15 H, m), 3.99 (2 H, dq), 7.32 (2 H, d), 7.77 (2 H, d), (CDCl3 at 7.25); δC (CDCl3): 144.98 (1 C), 132.84 (1 C), 129.97 (1 C), 129.91 (1 C), 128.10 (2 C), 77.18 (1 C), 71.13–70.22 (7 C), 69.76 (1 C), 21.76 (1 C). (CDCl3 middle C: 77.14).

Synthesis of 2-(diallylaminomethyl)−12-crown-4-ether (5)

A solution of diallyl amine 4 (14.2 g, 146 mmol) and tosylate derivative 3 (8.6 g, 23.9 mmol) in acetonitrile (AN) (35 mL) in a closed RB flask purged with N2 was stirred at 75 °C (20 h) and 85 °C (24 h). After the removal of AN and excess diallylamine, the residue was taken up in a saturated (NH4)2CO3 (20 mL) and extracted with DCM (3 × 30 mL). After drying (Na2SO4) and concentrating the combined organic layer, the residue was purified by chromatography (silica, 95:5 ether/methanol) to obtain 5 (colorless liquid, 5.7 g, 83%). [Found: C, 62.8; H, 9.7; N, 4.8%. C15H27NO4 requires C, 63.13; H, 9.54; N, 4.91%]. νmax (neat): 3075, 2906, 2856, 1642, 1447, 1418, 1357, 1302, 1253, 1125, 995, 972, 915 and 863 cm−1; δH (CDCl3): 2.34 (dd, 1 H, J 6.6, 13.8 Hz), 2.41 (dd, 1 H, J 5.7, 13.8 Hz), 3.00 (dd, 2 H, J 6.6, 13.8 Hz), 3.06 (dd, 2 H, J 6.0, 13.8 Hz), 3.34 (dd, 1 H, J 8.1, 11.1 Hz), 3.52–3.76 (14 H, m), 5.05 (4 H, m), 5.74 (2 H, m);δC (CDCl3): 135.82 (2 C, CH = CH2), 117.48 (2 C, CH = CH2), 78.15 (1 C), 73.15 (1 C), 71.25–69.62 (6 C), 57.85 (2 C, CH2-CH = CH2), 54.37 (1 C), (CDCl3 middle C: 77.23).

Synthesis of 2-(diallylaminomethyl)−12-crown-4-ether hydrochloride (6)

After the dropwise addition of HCl (37%, 2.06 g, 20.9 mmol) to the crown ether 5 (4.98 g, 17.4 mmol)/water (15 mL) mixture, it was stirred at 23 °C (30 min) and freeze-dried to afford monomer 6 (colorless liquid, 5.4 g, 96%). [Found: C, 55.6; H, 8.9; N, 4.2%. C15H28ClNO4 requires C, 55.98; H, 8.77; N, 4.35%]. νmax (neat): 3420, 2861, 2360, 1646, 1456, 1358, 1296, 1249, 1125, 1098, 997, 942 and 837 cm−1; δH (D2O): 3.93–3.03 (21 H, m), 5.44 (4 H, m), 5.76 (2 H, m), (D2O at 4.68); δC (D2O): 127.06 (1 C, CH = CH2), 126.86 (1 C, CH = CH2), 125.60 (1 C, CH = CH2), 125.21 (1 C, CH = CH2), 72.87 (1 C), 70.15–68.15 (7 C), 66.60 (dioxane), 56.90 (1 C, CH2-CH = CH2), 54.81 (1 C, CH2-CH = CH2), 53.19 (1 C).

Synthesis of homopolymer (7)

The initiator APS (100.0 mg) was added to monomer 6 (700 mg, 2.17 mmol)/H2O (400 mg) solution in a 5 mL-RB flask at 100 °C. After a 2-min interval, more APS (75 mg) was added and stirred at 100 °C for 20 min. Upon dialysis (Spectra/Por membrane tube: MWCO 1000 Da) of the resulting crude polymer against distilled water (12 h), polymer 7 was obtained by freeze drying as a light brown hygroscopic solid (450 mg, 64%). [Found: C, 55.5; H, 9.0; N, 4.3%. C15H28ClNO4 requires C, 55.98; H, 8.77; N, 4.35%].νmax (KBr): 3387, 2924, 2717, 2081, 1639, 1457, 1364, 1300, 1254, 1133, 915, 842 and 620 cm−1.

Synthesis of copolymer (8) using monomer 6 and SO2

SO2 (234 mg, 3.66 mmol) was absorbed onto a monomer 6 (644 mg, 2.0 mmol)/DMSO (750 mg) solution in a 10 mL-RB flask. After AIBN (15 mg) was introduced, the mixture in the closed flask was stirred (65 °C, 24 h). The resulting polymer was dialyzed in a Spectra/Por membrane tube (MWCO 1000 Da) against distilled water (6 h). Copolymer 8 was obtained by freeze drying as a light brown semisolid (503 mg, 65%). [Found: C, 46.3; H, 7.4; N, 3.5; S, 8.0%. C15H28ClNO6S requires C, 46.69; H, 7.31; N, 3.63; S, 8.31%].νmax (KBr): 3433, 2920, 1644, 1458, 1411, 1363, 1307, 1129, 1030, 915, 859, 618 and 513 cm−1.

Synthesis of cross-linked resin (10)

A solution of 6 (805 mg, 2.5 mmol) and the cross-linker tetraallylhexane-1,6-diammonium chloride monomer 9 (154 mg, 0.44 mmol) in H2O (562 mg) in a 10 mL-RB flask was heated to 100 °C; thereafter, APS (155 mg) was added. After 2 min, an additional amount of APS (100 mg) was added, and the mixture was stirred at 100 °C (6 h) under N2. After washing with water, resin 10 was dried in vacuo at 55 °C (light brown hygroscopic solid, 594 mg, 61%). [Found: C, 56.1; H, 9.1; N, 4.7; S, 8.0%. Monomers 9 and 6 incorporated in an 85:15 ratio require C, 56.92; H, 8.94; N, 4.94%]. νmax (KBr): 3390, 2923, 2863, 2511, 1644, 1455, 1359, 1026, 837 and 606 cm−1.

Synthesis of cross-linked resin (11)

After absorbing SO2 (655 mg, 10.2 mmol) onto a solution of 6 (1.25 g, 3.88 mmol) and 9 (241 mg, 0.69 mmol) in DMSO (2.0 g, 27.4 mmol) in a 50 mL-RB flask and adding AIBN (70 mg), the mixture was heated in the closed flask at 65 °C (24 h) and 75 °C (6 h). The mixture was dialyzed in a Spectra/Por membrane tube (MWCO 1000 Da) against distilled water (12 h). The separated resin was dried in vacuo at 55 °C to obtain resin 11 (creamy white solid) (1.5 g, 80%). [Found: C, 45.7; H, 7.4; N, 3.9; S, 8.9%. Monomers 9 and 6 and SO2 incorporated at an 85:15:115 mol ratio require C, 46.43; H, 7.29; N, 4.03; S, 9.23%]. νmax (KBr): 3390, 2917, 2858, 2444, 1704, 1644, 1452, 1360, 1300, 1120, 1022, 837 and 605 cm−1.

Evaluation of adsorption performance

Adsorption of Li+ ions

Several concentrations at the ppm level were prepared from a 100 ppm Li⁺ ion stock solution (LiCl). The adsorption capacities (qe) in mg g−1 were determined via Eq. (1), where Co and Ce in mg L−1 represent the initial and equilibrium concentrations of Li+ ions, respectively; the solution volume is V in L; and m is the mass of the resin (g).

$$\:{q}_{e}=\:\left({C}_{o}-\:{C}_{e}\right)\times\:\:\frac{V}{m}$$
(1)

The extraction efficiency (E%) was determined via Eq. (2).

$$\:E\%=\:\frac{({C}_{o\:}-\:{C}_{e})}{{C}_{o}}\:\times\:100$$
(2)

Adsorption experiments involved resin 11 (50 mg) and 20 ppm Li⁺ solution (10 mL) contained in screw-capped glass vials. The mixtures were continuously stirred at 200 rpm (1 h, 23 °C). The pH was adjusted to 7.0 using either 0.01 N NH₄OH or 0.01 M HCl. After treatment, the solutions were filtered, and the lithium-ion concentrations were analyzed via ICP-OES (Perkin Elmer).

Adsorption kinetics

Kinetics experiments were performed with resin 11 (50 mg) in 10 mL of 20 ppm Li+ solution (pH: 7.0, 23 °C). The supernatant at adsorption times of 5, 15, 30 and 60 min was analyzed via ICP-OES. The qt at various times was calculated via Eq. (1).30,35

Adsorption isotherms

Effect of initial Li+ concentration (Co)

Several experiments were carried out by stirring resin 11 (50 mg) in a Li+ solution (10 mL, pH = 7.0, T = 23 °C) with lithium-ion Co concentrations of 20, 40, 60, 80 and 100 ppm for 30 min. The qt values were calculated via Eq. (1).30,36

Effect of adsorbent amount

Adsorption experiments were conducted with various doses (25, 50, 75 and 100 mg) of resin 11 in Li+ solution (10 mL, 20 ppm, pH = 7.0, T = 23 °C). The supernatant at an adsorption time of 30 min was analyzed via ICP‒OES to obtain qt via Eq. (1).30,36

Effect of pH

The qe values at various pH values (3, 5, 7 and 9) (30 min, T = 23 °C) for adsorption were determined using resin 11 (50 mg), 20 ppm Li+ solution (10 mL) and Eq. (1).30,36

Selective adsorption

The competitive ions Na+, K+ and Mg2+ were examined to determine the recognition selectivity of resin 11 (50 mg) for Li+. Adsorption experiments involving binary mixtures (10 mL, pH = 7.0) of Li+/Na+, Li+/K+ and Li+/Mg2+ pairs with 20 ppm concentrations of each ion were performed at 23 °C for 30 min. Another experiment was carried out using a solution containing Li+ (20 ppm) and Mg2+ (40 ppm). After the qe and Ce values of the various ions were determined, the distribution coefficient Kd (L g−1) was calculated via Eq. (3):37,38

$$\:{K}_{d}=\:\frac{{q}_{e}}{{C}_{e}}$$
(3)

The selectivity coefficient k of resin 11 toward Li+ versus the competitive ion Mn+ was determined via Eq. (4):37,38

$$\:k=\:\frac{{K}_{d\left({Li}^{+}\right)}}{{K}_{d\left({M}^{n+}\right)}}$$
(4)

Regeneration: adsorption/desorption

Resin 11 (50 mg) was stirred with a 20 ppm Li+ solution (10 mL) for 30 min to determine qe. Then, the Li+-resin 11 complex was placed in 0.5 M HCl (5 mL) for 30 min at 45 °C. The desorption efficiency was determined by analyzing the supernatants. The cycle was repeated five times.

Results

Synthesis and characterization of monomer 6, polymers 7 and 8, and resins 10 and 11

Before we proceeded to synthesize cross-linked resins 10 and 11, a feasibility study was essential to synthesize linear polymers 7 and 8. The successful synthesis of the linear polymers paved the way for the generation of resins 10 and 11. Resin 11 was chosen for the adsorption study involving Li+ ions since it was obtained in excellent yield (vide infra).

CE 1 was converted to its tosylate derivative 3 which was exposed to substitution of the tosyl group with diallyl amine 4 to produce 2-(diallylaminomethyl)−12-crown-4-ether (5) in very good yield (Scheme 1). Upon treatment with 1 equivalent of HCl, tertiary amine 5 afforded amine salt monomer 6 in almost quantitative yield (Scheme 1).

Scheme 1
scheme 1

Preparation of a diallyl amine salt Functionalized with a methyl 12-crown-4 ether pendant group (compound 6).

Full size image

The NMR spectra of 1, 3, and 5 are displayed in Figs. 1 and 2. Owing to a chiral center at the carbon marked ‘e’, two Hs attached to each carbon at ‘c’ and ‘d’ in 5 become diastereotopic. As a result, CH2 protons (CHAHM) are split into an AM doublet whereby each signal is further split into doublets owing to the presence of an adjacent HX to give signals of AMX pattern as a doublet of doublets (Fig. 1c). The protons marked ‘a’ and ‘b’ appeared in the olefinic region. The 13C spectrum of 5 in Fig. 2c displays the usual chemical shifts for the alkene carbons along with the chemical shifts for the remaining carbons, confirming the structure of tertiary amine 5.

Fig. 1
figure 1

1H spectra in CDCl3 of (a) CE 1, (b) tosylate derivative 3, and (c) diallyl methyl-12-crown-4-ether amine 5.

Full size image
Fig. 2
figure 2

13C spectra of (a) CE 1, (b) tosylate derivative 3, and (c) diallyl methyl-12-crown-4-ether amine 5 in CDCl3.

Full size image

As outlined in Scheme 2, monomer 6 underwent APS-initiated cyclopolymerization, affording homopolymer 7, whereas AIBN-initiated copolymerization of 6 with SO2 led to alternate copolymer 8. Reactivity ratios of r1 ≈ 0 and r2 ≈ 0 for the diallylamine salt/SO2 pair dictate the formation of the alternate copolymer39,40.

Scheme 2
scheme 2

Synthesis of homopolymer 7 and the copolymer with SO2 8.

Full size image

The NMR spectra of 6–8 are displayed in Figs. 3  and 4. Note that the two allyl groups are diastereotopic in nature, and as such, they become magnetically non-equivalent, which leads to separate signals of equal intensity for each carbon marked ‘a’, ‘b’, and ‘c’ (see the expanded spectrum in the inset (Fig. 4 a) in the range ~ δ125-127 ppm). The alkene proton signals at δ5.3 and 5.9 ppm (Fig.  3a) disappeared in the polymer spectra (Fig. 3b, c), confirming the formation of the polymers. Likewise, the alkene carbon signals marked ‘a’ and ‘b’ (Fig. 4a) were not present in the polymer spectra (Fig. 4b, c). The absence of residual alkene signals points towards the involvement of chain transfer to the monomer41 and coupling42 in the termination process. The major and minor signals for the carbons marked ‘a’ and ‘b’ are attributed to the substituents in the cis- and trans- dispositions in an approximate ratio of 3:1 (Fig. 4c)43,44.

Fig. 3
figure 3

1H spectra of (a) monomer 6, (b) homopolymer 7, and (c) the copolymer with SO2 8.

Full size image
Fig. 4
figure 4

13C spectra of (a) Monomer 6, (b) homopolymer 7 and (c) the copolymer with SO2 8.

Full size image

As outlined in Scheme 3, a solution of 6 (85 mol%) and cross-linker 9 (15 mol%) underwent APS-initiated polymerization to yield water-insoluble copolymeric resin 10, whereas AIBN-initiated terpolymerization of 6/9/SO2 in an 85:15:115 mol ratio afforded cross-linked water-insoluble terpolymeric resin 11. Compared with copolymer 10, terpolymer 11 was obtained in better yield.

If monomers 6 and 9 exhibit equal reactivity, their feed ratio is expected to be reflected in the polymer composition at high conversion. In polymers 10 and 11, the amine salt monomers are distributed randomly throughout the network, whereas in polymer 11, the SO₂ units are incorporated in an alternating sequence.

Scheme 3
scheme 3

Synthesis of cross-linked resin bearing residues of 12-crown-4-ether.

Full size image

Viscosity measurements

The viscosity plots for the PEs, displayed in Fig. 5, become concave upward, as expected, in salt-free water, whereas the plots become linear in 0.1 M NaCl since Cl− ions shield the positive charges on the macromolecules. The respective intrinsic viscosities [η] of 7 and 8, as determined via the relationship ηsp/C = [η] + k [η2C], were 0.0661 and 0.261 dL g−1, respectively. The low [η] value for homopolymer 7 is indicative of its lower molar mass. GPC analysis did not provide acceptable molar masses for the polymers; erratic results were obtained, presumably owing to the crown ether motifs being attached to the column materials.

Fig. 5
figure 5

Viscosity plots of 7 and 8 at 30 °C.

Full size image

TGA of cyclopolymers 8, 9, 10, and 11

The TGA curves (Fig. 6) revealed that polymers 8 and 9 and resins 10 and 11 are stable up to 200 °C. The weight loss for 9 and 11 occurring in the range 220–280 °C is accounted for the loss of SO2 motifs, whereas in their absence, polymers 8 and 10 were more stable.

Fig. 6
figure 6

TGA results for homopolymer 8, SO2 copolymer 9, resin 10 and SO2 resin 11.

Full size image

FTIR analysis

The FTIR spectra of unloaded and lithium-loaded resin 11 are displayed in Fig. 7. The symmetric and asymmetric stretching bands of SO₂ appeared at ~ 1120 and 1305 cm−1, respectively. There was a perceptible change in the vibration band of the crown ether C-O-C at approximately 1000–1100 cm⁻¹. A new band at 917 cm−1 is assigned to the complex of Li+ ions with ether motifs45.

Fig. 7
figure 7

FTIR spectra of unloaded and loaded cross-linked polymer 11 with lithium ions.

Full size image

BET analyses of cross-linked resin 11

The textural properties of CLR 11 were determined through BET analysis (Table 1 ; Fig. 8 a). The material exhibited a significantly greater surface area (410.47 m²/g) than conventional ionic resins46,47. This increased surface area is likely due to the expanded morphology resulting from the incorporation of crown ether moieties. Desorption isotherms were analyzed via the BJH method to determine the pore size distribution of the resin (Fig. 8 b). Average pore diameter of 2.9 nm reflects mesoporous nature of resin 11.

Table 1 BET surface area analysis of cross-linked resin 11.
Full size table
Fig. 8
figure 8

For resin 11, (a) nitrogen adsorption–desorption isotherms and (b) pore size distributions as determined via the BJH method.

Full size image

SEM and EDX analysis

The surface morphology of unloaded and metal-ion-loaded resin 11 was examined by SEM after gold coating. The unloaded resin exhibited a smooth surface with flat crystalline structures. In contrast, the loaded resin showed noticeable changes in surface roughness and morphology, indicating the deposition of metal ions. EDX analysis confirmed the presence of C, O, N, Cl, and S, which is consistent with the composition of resin 11. However, in the case of the Li⁺-loaded resin, lithium was not detected in the EDX spectrum alongside C, O, S, and N. This is likely because lithium is difficult to detect via EDX because of several factors, such as its low atomic number, which emits X-rays with very low energy (approximately 54 eV), weak characteristic X-ray emission, and the limited sensitivity of Detectors for elements lighter than boron. The Decrease in chloride ions in the loaded sample resulted from the adjustment of the pH to 7 with NH4OH, whereby Cl− ions may have been leached out as NH4Cl. The morphology changes suggest the adsorption of Li+ ions on resin 11 (Fig. 9)48,49. The EDX analysis is not known to detect Li; conventional EDX detectors are optimized for elements with atomic number ≥ 4.

Fig. 9
figure 9

SEM and EDX analysis of (a) unloaded 11, (b) Li+-loaded 11, and (c) elemental mapping images of unloaded 11.

Full size image

XPS analysis

The XPS survey scan for the unloaded and loaded resin 11 compositions revealed the presence of carbon, oxygen, sulfur and nitrogen atoms within the resin surface structure, having highest percentage of carbon contents (Tables 2 and 3; Figs. 10 a5 and b5). The XPS deconvoluted profiles of the C 1s spectrum for unloaded resin 11 showed a two-peak profile (Fig. 10a1); the less intense peak at 284.76 eV was attributed to C–C Aliphatic bonds, whereas the intense peak at 286.24 eV indicated the presence of C–O, C–S and C–N bonds. For the loaded resin 11, the C 1 s spectrum showed two peaks, like the unloaded sample, but the only difference was the Decreased intensity of the peak at 285.87 cm−1 (Fig. 10b1).

The presence of O 1 s peaks for the unloaded resin 11 at 532.46, 532.41 and 531.98 eV were attributed to O = S, O–C and O2, respectively, whereas the loaded resin 11 had the same three peaks with approximately the same binding Energy in the range of 531–533 eV (Fig. 10a2 and b2). The XPS spectrum of N 1 s for unloaded resin 11 presented two peaks related to the C‒N bond; the less intense peak at 399.25 eV and the highly intense peak at 401.97 eV were indicative of the presence of trivalent unprotonated and quaternary protonated nitrogen (Fig. 10 a3 and b3). On the other hand, loaded resin 11 showed the opposite trend in intensity. The S 2p XPS spectra of the unloaded and loaded resins 11 were similar, as both had two peaks at 168.31, 169.37 eV attributed to S = O and S-C bonds (Fig. 10a4 and b4). Unfortunately, the presence of Li+ ions was not detected in loaded resin 11, where the initial concentration before adsorption was 20 ppm50,51. The determination of low content Li + using XPS was either difficult or inaccurate50.

Table 2 XPS survey scan composition of the unloaded resin 11.
Full size table
Table 3 XPS survey scan composition of Li+-loaded resin 11.
Full size table
Fig. 10
figure 10

XPS analysis of (a) unloaded resin 11 and (b) Li+-loaded resin 11.

Full size image

Discussion

Adsorption

In a simplified form, the process of Li+ extraction from its aqueous solution into the solid phase of the resin can be described as follows:

$$\text{CE}_\text{s}+\text{Li}^+_\text{aq}+\text{Cl}^-_\text{aq}\rightleftharpoons\text{CE}\cdot\text{Li}\text{}\text{Cl}_\text{suspended}.$$

The equilibrium concentration (Ce) and adsorption capacity (qe) were calculated via Eq. (1) (Table 4) for resin 11, with initial lithium concentrations (Co) ranging from 20 to 100 ppm at pH 7.0, for a Duration of 60 min at 23 °C. The percentage of lithium removal efficiency (E%) by resin 11 was determined via Eq. (2). For a solution of 20 ppm Li+ at various pH values, the plot of qe versus pH (Fig.  11a) reveals the maximum adsorption capacity at pH 7; at lower pH values, competition between H+ and Li+ lowers the qe for Li+ ions. Therefore, the subsequent experiments were performed at pH 7. A linear relationship between Co and qe is illustrated in Fig.  11b. The qe values rose up with increasing Co, whereas the removal percentage decreased at higher concentrations since the adsorption sites were saturated with Li+ ions with no further available vacant sites (Table 4).

Table 4 Ce, qe and E% values of lithium adsorption.
Full size table
Fig. 11
figure 11

Plots for (a) qe versus various pH values for a constant Co (20 ppm Li+), (b) qe versus C0.

Full size image

Kinetic studies and isotherms of lithium adsorption

The adsorption mechanism can be illustrated by analyzing the relationship between Ce and qe. This relationship is established via adsorption isotherm models, which help describe the formation of adsorbate layers on the adsorption sites, the characteristics of these sites, and the nature of the interactions involved52. In this study, several isotherm models were applied to interpret adsorption behavior.

To determine the homogeneity and heterogeneity of the adsorption of lithium ions on a surface with uniform energy, the Langmuir and Freundlich isotherm models (Eqs. (5) and (6), respectively) were used. Linear plots for the Langmuir model and Freundlich isotherm were constructed (Fig. 12a, b, respectively). The constants qm and KL for the Langmuir isotherm and n and Kf for the Freundlich isotherm can be calculated from the graphs (Table 5). The adsorption intensity or heterogeneity factor (n) value (n = 2.84 > 1) indicates that favorable selective adsorption occurred primarily through a chemisorption mechanism53 and increased with increasing metal ion concentration but at a decreasing rate. The maximum adsorption qm was Determined to be 4.98 mg g−1. Heterogeneous surfaces of porous materials have been detailed in a review article54. Adsorption isotherms can shed light on the heterogeneity of the surfaces. For instance, based on Freundlich isotherm model, if 0 < 1/n < 1, the adsorption is considered favorable, and the surface is heterogeneous. Freundlich constant n of 2.84 thereby points towards the heterogeneity of the adsorbent surface55. A highly heterogeneous surface is indicated If 1/n approaches 0, whereas 1/n approaching 1 the surface becomes closer to homogeneous. A more heterogeneous surface with a smaller value of n signifies a wider distribution of adsorption site energies, where stronger binding occurs to more energetic sites initially. This is observed with initial faster followed by slow adsorption (vide infra Fig. 13a).

$$\:Langmuir\:isotherm\:model:\frac{{C}_{e}}{{q}_{e}}\:=\frac{1}{{q}_{m}}{C}_{e}+\:\frac{1}{{K}_{L}{q}_{m}}$$
(5)
$$\:{{Fruendlich\:isotherm\:model}:\text{log}q}_{e}=log{k}_{f}+\:\frac{1}{n}\:log{C}_{e}$$
(6)

For a uniform distribution of binding sites, the Temkin isotherm describes the interaction between the adsorbent and adsorbate56. A linear plot of qe versus ln (Ce) was generated (Fig. 12 c), where the binding constant (A) and heat of adsorption (B) were obtained from the respective intercept and slope, as defined by Eq. (7) (Table 5). The Temkin isotherm provided a good fit for lithium adsorption, indicating that its adsorption is influenced by surface coverage resulting from specific interactions between the adsorbate and the adsorbent.

$$\:Temkin\:isotherm\:model:\:{q}_{e}=\:\frac{RT}{b}\:lnA+\:\frac{RT}{b}\:ln\:{C}_{e}=BlnA+Bln\:{C}_{e}$$
(7)
Fig. 12
figure 12

(a) Langmuir, (b) Freundlich and (c) Temkin isothermal models for lithium adsorption using resin 11 at 23 °C.

Full size image
Table 5 Isotherm constants for lithium adsorption on resin 11 at 23 °C.
Full size table

The kinetic behavior of the adsorptions was studied via pseudo first- and second-order linear equations (Eqs. 8  and 9)57.

$$\:pseudo\:first\:order\:kinetic:\:{\text{l}\text{n}(q}_{e}-{q}_{t})=ln{\:q}_{e}-{k}_{1}t$$
(8)
$$\:pseudo\:second\:order\:kinetic:\:\frac{t}{{q}_{t}}=\frac{1}{{k}_{2}{q}_{e}^{2}}+\frac{t}{{q}_{e}}$$
(9)

where the rate constants k1 (min−1) and k2 (g·mg−1·min−1) correspond to the respective first- and second-order kinetic models. The qt and Ct values were determined at various times. The uptake of Li+ ions by 11 reached 50% within 5 min (Fig. 13a). The pseudo-second-order model indicated a better fit (Fig. 13b vs. Figure 13c); k2 was found to be 0.351 g mg−1 min−1 (Table 6).

Fig. 13
figure 13

Kinetic behavior of lithium adsorption by resin 11 at 23 °C: (a) plot of time versus removal percentage (Co = 20 ppm), (b) pseudo-first order and (c) pseudo-second order kinetic graphs.

Full size image
Table 6 Constant values of second-order kinetics for lithium adsorption at 23 °C on resin 11.
Full size table

Selectivity

Li+ selectivity was evaluated against Na+, K+ and Mg2+ contents. Table 7  shows the results of the adsorption of various ions by CLR 11. The respective adsorption capacities (qe) of CLR 11 for Li+, Na+ and K+ and Mg2+ (20 mg L−1 for each Mn+, pH = 7.0) reached 1.38–1.54, 0, 0 and 1.02 mg g−1, respectively (entries 1–3, Table 7). The higher qe for Li+ on CLR 11 than for the other ions is because of the cavity size of the 12-crown-4 ether motif. The cavity radius of ~ 75 pm for 12-crown-4 matches the ionic radius of Li+ (76 pm), whereas the sizes of Na+, K+ and Mg2+ are 102, 138 and 72 pm, respectively58,59. The ionic size of Mg2+ is nearly the same as Li+ ions. The selectivity coefficients \(\:k\) of CLR 11 for Li+ with respect to the competing ions Na+, K+, and Mg2 are > > 1, >> 1, and 1.53, respectively. Note the selectivity coefficient \(\:k=\:\frac{{K}_{d\left({Li}^{+}\right)}}{{K}_{d\left({M}^{n+}\right)}}\) from a solution containing 20 ppm each of Li+ and Mg2+ is 1.533, whereas it became 0.533 from a solution containing 20 ppm Li+ and 40 ppm Mg2+ (entry 4, Table 7). The adsorbent is more selective to Li+ than Mg2+ from a solution containing equal concentrations (i.e. 20 ppm each), whereas k changes in favor of Mg2+ from a solution containing its higher concentration.

Table 7 Distribution and selectivity coefficient for the adsorption of several ion pairs.
Full size table

Regeneration

Li+ from the adsorbent 11-Li+ complex was Desorbed using 0.5 M HCl. After 5 cycles, qe retained 89% of its original value, thereby suggesting preservation of the structural integrity and binding sites for Li⁺.

Conclusion

Homopolymer 7, copolymer 8, and CLR 10 were synthesized via the cyclopolymerization of monomer 6. The new adsorbent 11 was synthesized via the polymerization of a diallyl amine salt monomer bearing 12-crown-4 ether motif, SO2, and cross-linker 9. The lithium removal study was conducted at ppm level. Resin 11 adsorbs lithium ions rapidly with excellent efficiency following second-order kinetics and fitting the Temkin and Langmuir adsorption isotherms. The extraction efficiency of resin 11 for the selective adsorption of Li+ versus Na+, K+, and Mg2+ was evaluated, and the selectivity coefficient \(\:k\) was found to be > > 1, >> 1, and 1.54, respectively. These results show that CLR 11 selectively recognizes the target ion Li+.