Ion exchange capacity of synthetic zeolite L: a promising way for cerium recovery

The increasing rare earth elements’ (REE) demand to meet the market request and the current political scenario show that it is essential to find good solutions to recover these elements from waste (both industrial and mining). Zeolites are microporous materials with high cation exchange capacity, up to now only little investigated for REE recycle. Here, we propose the use of NH4+-exchanged zeolite L for Ce recovery from a very diluted solution (0.002 M), mimicking the Ce3+ concentration of the liquors deriving from the leaching of spent catalysts. The aim of this work is twofold: (i) to investigate the exploitability of zeolite L as cation exchanger in the Ce recovery; and (ii) to determine the best working conditions. The investigated process consists of a coupled cation exchange: (1) in the first exchange the NH4+ cations — present in the zeolite porosities — are exchanged with the Ce3+ ions in the solution; and (2) in the second experiment, the Ce3+ trapped into the zeolite is recovered through a further exchange with NH4. The best working conditions for Ce3+ exchange of NH4-zeolite L are: batch system, liquid/solid ratio equal to 90 mL/g and 180 mL/g, 24 h of contact at 25 °C. The resulting Ce adsorption capacity (qt) is equal to ~25 mg/g and ~39 mg/g and the removal efficiency 100% and 77% for the two tested liquid/solid ratios, respectively. The kinetics was proved to be fast and consistent with industrial timing; no energy cost for temperature setting is required; and the acid pH (~4) of the solutions does not affect the zeolite structure stability and its exchange performance. It has been demonstrated that the zeolite framework is not affected by the exchange so that the same absorbent material can be employed many times.


Introduction
Nowadays, the demand for rare earth elements (REE) is increasing due to their extensive use in many technological fields. However, the current political scenario -where REE are considered as strategic resources (Mancheri et al. 2019, Charalampides et al. 2015, the monopoly of only a few countries, the so-called balance problem between demand and natural abundance (Binnemans et al. 2018), and the need to limit the environmental costs of their mining (Haque et al. 2014, Suli et al. 2017, stress the necessity of a recycling policy of these elements. In terms of volumes, one of the most exploited REE is Ce (Charalampides et al. 2015, Binnemans et al. 2018, which is exploited in a huge number of devices such as NiMH batteries (Porvali et al. 2020), fluid catalytic cracking (FCC) catalysts (Innocenzi et al. 2015), redox support for three-way automotive catalysts and other metal catalysts (Gorte 2010), glass polishing and glass films (Borra et al. 2018) and fluorescent lamps (Tunsu et al. 2016, Yurramendi et al. 2019. These materials contain a variable amount of REEs; for example, Ce dioxide reaches 40-99 wt% in glass polishing waste (Meshram and Abhilash, 2020) -where Ce is the most abundant REE, being at least 65 wt% of the whole REE content (Binnemans et al. 2013)and 5.3 wt% in automotive catalysts (Zhao et al. 2020), while in FCC Ce is 0.2-0.3 wt% (Meshram and Abhilash, 2020), only 10 wt% of the total REE amount (Binnemans et al. 2013). The liquors obtained by leaching these exhausted components can present Ce concentrations ranging from ~0.02 M for the leached solutions coming from the residual waste of exhausted automobile catalysts (Ilyas et al. 2021) to 0.002 M for FCC (Innocenzi et al. 2015). A strategy for the extraction of Ce from these acid solutions is mandatory for the efficient recycling of this metal. Different methods were already tested for REE recovery such as precipitation (double salt precipitation, oxalate precipitation, carbonate precipitation, hydroxide precipitation), solvent extraction (Meshram and Abhilash, 2020), adsorption (de Farias et al. 2021), or ion exchange (Royen & Fortkamp 2016). This last is usually applied using metal-organic frameworks (MOF) materials or resins; however, efficient recycling can be achieved using other porous materials such as zeolites.
In the present paper, we tested the efficiency of NH 4 + -exchanged zeolite L (NH 4 -L) in Ce recovery from acid solutions. Zeolite L was reported as a promising candidate, among various tested zeolites (FER and FAU), for REE recovery (Duplouy 2016). A recent study by our group investigated the Ce exchange capacity of three different cationic forms of zeolite L and proved the efficacy of this zeolite in exchanging Ce 3+ from highly concentrated solutions . The structural study gave new insight on the sites involved in the exchange and showed that after the Ce exchange, of the NH 4 + zeolite L, the Ce recovery from the porosities was complete.
For the above-mentioned reasons, in the present work a NH 4 + -exchanged form of zeolite L, was chosen to test the effectiveness in exchanging Ce from a very diluted solution ([Ce 3+ ] = 0.002 M), which mimics the concentration reported for residual liquor of FCC waste (Innocenzi et al. 2015). We experimented different liquid/solid (l/s) ratios while, to follow the requirements of a green process, the temperature was maintained at 25 °C. In fact, several tests performed at higher (but mild) temperature did not indicate improvements in the cation exchange process. In a second step, the extraction of Ce from the Ce-exchanged L zeolite is obtained through a reverse exchange in a NH 4 + solution. This final exchange is proposed to recover the Ce ions for further utilization by fully exploiting the reversibility of cation exchange in zeolites ( Fig. 1).
This work is part of a wider project aimed at exploiting zeolite cation exchange properties to recover REE from waste, to move toward a green circular economy model. In fact, although some conventional adsorbents display high adsorption capacity, they have several limitations, including expensive production and difficult regeneration. Using zeolites for REE recovery will offer the possibility to exploit a material that can be recycled many times (probably for infinite cycles) and that does not need to be regenerated, since once the REE is recovered, the powder can be immediately reused.
We started by testing synthetic zeolites, since the use of pure materials allows a clearer understanding of the process occurring during the exchange and an easier following of the exchange kinetics. The further step of the project, to pursue a greener process, will the exploitation of natural zeolites, this in fact will allow the use of cheap materials, without Fig. 1 Nutshell of the process the need of a dedicated synthesis, allowing further energy saving. In addition, the use of an NH 4 + solution for the first and final exchange will allow the use of the same solution for several exchange cycles, so to avoid waste of chemicals during the treatment stages.

Zeolite L
The synthetic zeolite L has formula (K 6 Na 3 (H 2 O) 21 Al 9 Si 27 O 72 , a = 18.4 Å c=7.52 Å, γ = 120 °) and LTL framework type (Baerlocher and McCusker, n.d.). This framework is composed by small cages and two mono-dimensional channels both running along the c axis of the unit cell: a channel built by connection of rings of 12 tetrahedra (12MR) and a smaller elliptical one delimited by rings of 8 tetrahedra (8MR). In the as-synthesized K form, cations occupy different crystallographic positions: two of them participate to the cation exchange -being hosted in the channels -whereas those hosted in the cages cannot be exchanged. The NH 4 + exchanged form of synthetic K-L zeolite (NH 4 -L from now on), with chemical formula K 2.25 (NH 4 ) 6.86 Si 26.91 Al 9.08 O 72 ·17.07(H 2 O), was fully characterized in Confalonieri et al. (2021). In the exchanged form, NH 4 + ions substitute ~75% of the original K and are hosted just in the two channels (Fig. 2). NH 4 + -exchanged form was chosen for this study since, once the Ce exchange will be obtained, the zeolite will be further NH 4 + exchanged to recover Ce from the pores. In this way, the presence of NH 4 + as unique exchangeable cation in the starting material will allow to obtain a final recovery solution containing only NH 4 + (easily removable and recyclable by evaporation) and Ce ions.

Ce exchange
A 0.002 M Ce 3+ solution was prepared dissolving Ce(NO 3 ) 3 ·6H 2 O (Cerium(III) nitrate hexahydrate, REac-ton®, 99.5% (REO) -Alfa Aesar) into double-distilled water. NH 4 -L zeolite was put in contact for 72 h with the solution at 25 °C under stirring. The pH was initially set to 4.3 using HNO 3 to avoid Ce compound precipitation and to mimic the acid conditions of a real industrial waste liquor. The stability of the zeolite framework at this pH condition was tested in order to exclude the possibility of a dealumination process or structural modifications. The pH of the solution was constantly monitored showing a maximum variation from 4.3 to 4.9 after 72 h of contact with the zeolite. No buffer solution was used to avoid competition in the zeolite ion exchange. Various experiments were performed using different liquid/solid ratios (30, 60, 90, 180, 290, 750 mL/g) and the same contact time of 72 h. Several tests were performed at higher temperatures (up to 80 °C, data not reported), but, since they did not indicate improvements in the cation exchange process, we decided to work at room temperature. The exchanged samples were labelled Ce-X, where X is the liquid/solid ratio in mL/g. At the end of the process, all the exchanged zeolites were rinsed three times with double-distilled water using an ALC multi-speed centrifuge PK 121, operated at 5000 rpm for 10 min. The samples were finally dried at 60 °C for 24 h.

Ce recovery
All the Ce-exchanged zeolites, obtained as described in the previous section, were contacted for 24 h under stirring, at room temperature, with a solution [NH 4 + ] = 0.8 M, prepared dissolving NH 4 NO 3 (ammonium nitrate, 99%, Carlo Erba) in double-distilled water. The liquid/solid ratio was 60 mL/g; pH was not set. Zeolite samples so obtained are identified with the prefix Rev. At the end of the NH 4 + -exchange, the powders were washed, dried following the procedure reported above, and analysed to check the Ce recovery.

Inductively coupled plasma (ICP) analysis
The kinetic of Ce 3+ exchange was monitored daily analysing the exchange solutions with ICP-OES technique. A small fraction of supernatants was withdrawn from the solutions, filtered by cellulose nitrate membrane (0.45 μm) and properly diluted. ICP analysis was performed using a Perkin Elmer Optima 4200 DV to measure the Ce content. Three   analyses were performed for each sample, and the initial solution with [Ce 3+ ] = 0.002 M was also analysed and used as reference.

Thermogravimetric analysis (TG)
Seiko SSC/5200 thermal analyser was employed to perform thermogravimetric analysis. The temperature range was room temperature (r.t.) −1050 °C with a ramp of 10 °C/min under air flow.

Elemental analysis
Nitrogen content of zeolite samples was analysed through Flash 2000 CHNS/MAS200R instrument. "FLASH" dynamic combustion (modified Dumas method) allows to treat the sample at 1800 °C. Released elementary gases are separated through a chromatographic column and detected by a highly sensitive thermal conductivity (TCD) detector.

SEM-EDS analysis
Chemical analyses of the exchanged zeolites were performed by a scanning electron microscope (SEM) Nova NanoSEM 450 equipped with X-EDS Bruker QUANTAX-200 detector. Samples were prepared compressing the zeolite powders into thin self-supporting discs, then coated with gold. The instrument operated at 15 kV; five analyses for each sample were performed scanning the beam to avoid dehydration and cation migration in the zeolite. The obtained intensities were corrected using the phi-rho-Z method.

Ce exchange of NH 4 -L
The results of elemental (Table S1) and SEM-EDS (Table S2) analyses of the exchanged zeolite are reported in the Supporting Information, while thermogravimetric curves and chemical formulas are shown in Fig. 3 and Table 1, respectively.
As evident in Table 1, the H 2 O content increases with the Ce amount. This is because: (i) Ce 3+ has a larger solvation sphere in comparison to NH 4 + (~9.1 (Lutz et al. 2012) versus ~5.8 (Chang & Dang 2003) H 2 O molecules in the first hydration shell, respectively); and (ii) due to the reduced number of cations (trivalent Ce vs monovalent cations), more space is available in the channel . The weight loss imputable to H 2 O corresponds to the first peak of the derivative thermogravimetric analysis -DTG -curves. From Fig. 3, it is possible to observe a   Gottardi and Galli (1985), using the following formula: Error% = shift at higher temperature of this peaks as a function of the liquid/solid ratio used in the exchange. This is related to the increase of exchanged Ce with increasing liquid/solid ratio used. In fact, being the bonds between the H 2 O molecules and Ce stronger with respect to those with ammonium ions, the H 2 O molecules are released at higher temperature . The weight losses above 300 °C are related to the NH 4 + loss. TG results suggest a slightly higher NH 4 + content for samples with low liquid/solid ratio, corresponding to a lower Ce exchange, in agreement with elemental analyses (Table S1).
The kinetics of the Ce 3+ exchange was monitored every 24 h ( Fig. 4; Table S3).
From ICP results ( Fig. 4; Table S3), it appears that the Ce 3+ concentration into the solution decreases for all the liquid/solid ratios tested, so confirming that NH 4 -L zeolite extracts Ce 3+ from the solutions. The exchange occurs mainly in the first 24 h (Fig. 4, bottom panel) and affects mainly the NH 4 + ions (Table 1). A slight decrease in the K content with respect to starting NH 4 -L sample is observed (Table 1), indicating that some residual K can be further exchanged by Ce (see discussion below). The data after 72 h of contact is stable enough to be considered as equilibrium values (Table S3), and the experiments with different residual Ce 3+ amounts in solution allow to establish an experimental Ce 3+ -NH 4 + exchange isotherm, as shown in Fig. 5. The experimental points are expressed as Az vs As, where Az = 3 Ce 3+ / Al (molar) in the zeolite and As = C (Ce 3+ ) / initial C (NH 4 + ) (equivalents) in solution. The isotherm can be fitted by a constant separation factor model Az = (K As Az max ) / ( 1 + K As ), in which K = α B A − 1, where α B A = (Az / As ) / (Bz / Bs) is the separation factor, in which Bz = NH 4 + / Al (molar) in the zeolite and Bs = C (NH 4 + ) / initial C (NH 4 + ) (equivalents) in the solution (LeVan et al. 1997) (Breck 1984). The value of the separation factor of the exchange of Ce for ammonium is 67, indicating a much higher affinity of NH 4 -L for Ce than for ammonium. Such a value is not uncommon in the exchange of multivalent-monovalent cations in zeolites (Breck 1984). The isotherm reaches an asymptotic value of Az max at 0.31. This corresponds to a clear limit of Ce content at ~1 Ce atom p.u.c. This value is very close to the 1.17 Ce p.u.c. reported for the NH 4 -L using a highly concentrated Ce solution, which represents the maximum Ce adsorption in zeolite L and involves only sites positioned in the main channel .
In conclusion, Ce cannot saturate all the anion sites of zeolite L; however, the aim of this work is not to obtain the complete cation exchange of the zeolite, but rather to define the best conditions to recover Ce from a very diluted liquor. In particular, we would like to: (1) obtain an efficient uptake  Table S3.

Fig. 5
Ce-NH 4 exchange isotherm in NH 4 -L. The experimental points are expressed as Az vs As, where Az = 3 Ce 3+ / Al (molar) in the zeolite and As = C (Ce 3+ ) / initial C (NH 4+ ) (equivalents) in solution of Ce, reducing at the minimum the Ce left in the solution and, in the meantime (2) achieve a high concentration of Ce in the zeolite to reduce the number of uptake-recovery cycles; and (3) prove that the same zeolite powder can be used many times to recover Ce from solutions. Among the tested systems, two demonstrated to be the most suitable for these purposes: Ce-90 and Ce-180, showing an uptake of 100% and 77% of the Ce in the solution the first 24 h ( Fig. 4; Table S3)

Ce recovery
The Ce recovery from the Ce exchanged zeolites was fulfilled through a further exchange with a NH 4 + 0.8 M solution.
As reported in Table 2 (and in Tables S4 and S5), an almost complete release of Ce is obtained for all the samples.
The effectiveness of the final exchange is also shown by the nitrogen content, which recovers the original value of NH 4 -L. To test the impact of these treatments on the zeolite structure and evaluate the possible reuse of the same zeolite powder in further cycles of Ce recovery, X ray powder diffraction patterns before and after the whole process were collected for the two most promising samples. In Fig. 6, the diffraction pattern of the original NH 4 -L ) is compared with that of samples Ce-90 and Ce-180 before and after the Ce recovery. The two samples containing Ce clearly show differences in the peak intensity ratios related to the incorporation of the REE in comparison to the original NH 4 -L (Fig. 6, left panel. These changes disappear completely after the second exchange with NH 4 + solution (Fig. 6, right panel), demonstrating a complete recovery of Ce. Neither structure modifications nor loss of crystallinity was detected.

Comparison with other adsorbent materials
The adsorption capacity of the two most promising samples was calculated as in Zhao et al. ( 2019): where C 0 and C t are the initial concentration and the residual concentration (mg/L) of Ce in the solution; V is the volume of Ce 3+ solution (L), and m is the weight of the zeolite (g).