Pb2+, Cd2+ and Hg2+ removal by designed functionalized swelling high-charged micas

1  Pb2+, Cd2+ and Hg2+ removal by designed functionalized swelling high-1 charged micas. 2 F.J. Osunaa, E. Pavóna,b and M.D. Albaa,1 3 aInstituto Ciencia de los Materiales de Sevilla (CSIC-US). Avda. Americo Vespucio, 49. 4 41092 Sevilla (Spain). 5 bDepartamento de Física de la Materia Condensada. Universidad de Sevilla. Avda. Reina 6 Mercedes s/n. 41012 Sevilla (Spain). 7 8 Abstract 9 The increasing accumulation of toxic heavy metals in the environment has 10 generated the need of efficient removal systems, being the adsorption method the most 11 popular one applied in aqueous solutions. Of particular concern is the case of Pb2+, Cd2+ 12 and Hg2+ due to their high potential hazard. In this paper, we describe the feasibility of a 13 new family of nanomaterials, swelling high charge micas, in the removal of these cations 14 from aqueous solutions. Batch adsorption experiments were carried out in the as-made 15 micas, Na-Mn, and after functionalization with ethylammonium, EA-Mn, and 16 mercaptoethylammonium, MEA-Mn. The results have demonstrated that all of them are 17 efficient heavy metal adsorbents, being Na-M2 the best adsorbent for Pb2+ and Cd2+, and, 18 MEA-M2 for Hg2+. 19 20 21 Keywords. swelling mica, heavy metals, pollutant, adsorption, functionalized adsorbents, 22 thiol 23 24  1 Corresponding author. E-mail address: [email protected] 2  1. Introduction 25 Industrialization has led to the accumulation in the environment of toxic heavy 26 metals, such as Cr6+, Ni2+, Pb2+, Hg2+ and Cd2+, causing severe environmental problems 27 and damaging human health (Volesky, 1990; Yuan et al., 2020; Zhang et al., 2020). 28 Among them, Pb2+, Cd2+ and Hg2+ stand out due to their high potential hazard and their 29 presence in soils and close to aqueous effluents (Sun et al., 2020; Volesky, 1990). 30 Conventional methods such as soil washing (Sierra et al., 2010), cement-based 31 solidification/stabilization (Shen et al., 2019), containment (NJDEP, 2014) and 32 electrokinetic removal (Figueroa et al., 2016) have been used for the remediation of soils 33 with heavy metals contamination. However, great efforts are still demanding to remove 34 them from contaminated water and wastewater (Bailey et al., 1999; Farid et al., 2020; 35 Inglezakis et al., 2007; Shaikhiev et al., 2020). 36 Adsorption is highlighted as one of the preferred methods for the elimination of 37 toxic contaminants (heavy metals, among others) from aqueous systems, being an 38 effective and economic technique for their removal, recovery and recycling. Several 39 materials have been tested to maximise the adsorption capacity of heavy metal cations 40 (Tran et al., 1999). Conventional adsorbents such as granular or powdered activated 41 carbon are technically inefficient and economically unviable. Consequently, non-42 conventional materials (Ezzat et al., 2020; Shaikh, 2020; Shaikhiev et al., 2020), such as 43 synthetic and natural fly ash (Mathur and Rupainwar, 1988; Yang et al., 2020), lignite 44 (Balasubramanian and Ahamed, 2000), tree fern (Ho, 2003), peanut shell (Periasamy, K, 45 Namasivayam, 1994) or peat (Ho, Y S, Ng, J C Y, McKay, 2001) have been used for this 46 purpose. 47 Clay minerals are widely used as adsorbents as they are the major soil constituents 48 and act as natural scavengers for pollutants in water through both ion exchange and 49 3  adsorption mechanisms. Moreover, their compositional and structural characteristics are 50 crucial for their adsorption capacities. First, cation exchange capacity (CEC) is different 51 for each clay minerals and determines their adsorption capacity through this mechanism 52 (Grim, 1968). Second, the location of isomorphous substitution in the clay layer is an 53 important factor affecting both hydration and cation speciation in the interlayer space, 54 determining the strength of the Lewis base (Malferrari et al., 2007). 55 Alba et al. (Alba et al., 2006) have synthetized a new family of clay minerals that 56 has been demonstrated to be attractive adsorbents. These synthetic fluorophlogopites 57 have layer charge in the range of the brittle micas but unlike the natural ones, they exhibit 58 swelling properties and cation exchange capacity (Naranjo et al., 2014). Those as-made 59 micas, Na-Mn (n is the layer charge which derives from the tetrahedral Si+4/Al+3 60 substitutions and can be found between 2 and 4), have exhibited excellent capacity for 61 the removal of radioactive cations and organic pollutants from wastewater (Alba et al., 62 2006; García-Jiménez et al., 2016; Osuna et al., 2019b, 2018; Pazos et al., 2017) and are 63 promising materials for heavy metal removal, although their efficiency has not been 64 proven yet. 65 However, raw clay minerals have shown poor stabilization performance, as the 66 immobilized metals can be exchanged and released in the long term (Ugochukwu et al., 67 2013). Then, in order to improve their adsorption capability, their surfaces have been 68 functionalized by inserting functional groups and organic molecules such as 69 alkylammonium cations (Bhattacharyya and Sen Gupta, 2008; Osuna et al., 2019b; Pazos 70 et al., 2017), L-cystein (Mittal et al., 2016) and humic acid (Wu et al., 2011). In fact, 71 different adsorption sites are observed when the functionalization is produced by grafting 72 organic groups, mainly thiol or amine functional groups (Cruz-Guzman et al., 2006; Lin 73 and Juang, 2002; Malferrari et al., 2007; Mercier, L, Detellier, 1995; Shah et al., 2005). 74 4  Therefore, the aim of the present study was, on the one hand, to investigate the 75 feasibility of using a family of design swelling high-charge micas, Na-Mn, before and 76 after the functionalization with ammonium and thiol groups, for removal of Pb2+, Cd2+ 77 and Hg2+ from aqueous solution and, on the second hand, to analyse the different removal 78 mechanisms. 79 80 2. Materials and methods 81 2.1. Materials 82 As-made Na-Mn (n is the layer charge equals to 2 or 4) were synthetized by 83 mixing SiO2, Al2O3, MgF2, and NaCl in the molar ratio 8-n:n:6:2n. The starting materials 84 were SiO2 from Sigma (CAS no. 112945-52-5, 99.8% purity), Al(OH)3 from Sigma 85 Aldrich (CAS no. 21645-51-2, 99% purity), MgF2 from Aldrich (CAS no. 20831-0, 98% 86 purity), and NaCl from Panreac (CAS no. 131659, 99.5% purity). All reagents were 87 mixed and ground in an agate mortar and heated up to 900 °C for 15 h in a Pt crucible. 88 Finally, the solids were washed with deionized water and dried at room temperature (Alba 89 et al., 2006) 90 The chemical products used for the preparation of the organoclays were obtained 91 from Aldrich Chemical Co. 600 mg of mica was added to ethylammonium (EA: 92 C2H5NH3+) or mercaptoethylammonium (MEA: SHC2H4NH3+) solutions, assuring their 93 concentration, twice mica cation exchange capacity, in order to favour the cation 94 exchange reaction. They were left to react for 3h at 80 ºC. The solids were recovered by 95 centrifugation, washed with deionized water and ethanol and dried at room temperature. 96 The functionalized micas were named as EA-Mn or MEA-Mn (Osuna et al., 2019a). 97 98 2.2. Batch adsorption 99 5  Adsorption experiments were carried out at RT and prepared in a 50 ml centrifuge tube, 100 where 0.5 g of Y-Mn (Y= Na, EA or MEA and n=2 or 4) were dispersed with 50 ml of a 101 25 mM X(NO3)2 (X=Pb 2+, Cd2+ or Hg 2+) solution. Adsorption of Pb2+ and Cd2+ was 102 performed at their natural pH as showed in Fig. S1 (open stars). The pH of the initial 103 Hg(NO3)2 solution was adjusted to 1.5 to avoid the salt precipitation. The samples were 104 equilibrated for 4 weeks in an end-over-end shaker at 50 rpm. Afterwards, the phases 105 were separated by centrifugation at 10,000 rpm at 8 ºC for 25 min. The samples were 106 named as X-Y-Mn. 107 The supernatants were filtered by a 0.2 µm syringe filter. After measuring the pH and 108 Eh values, the solutions were preserved by adding HNO3 and cool stored for subsequent 109 a Na+ and heavy metal analysis of three aliquots by inductively coupled plasma-mass 110 spectrometry (ICP-MS); the difference for metals before and after the adsorption reveals 111 the amount of adsorbed metals (Cs, meq/kg): 112 𝐶𝐶𝑉 𝐶  𝑉  𝑚 113 and the desorbed Na+ (Cdes, meq/kg) is calculated as following: 114 𝐶 𝐶󰇛󰇜𝑉 𝐶 󰇛󰇜𝑉  𝑚 115 Vi (L) is the volume of the initial solution, Veq (L) is the volume of the supernatant, m is 116 mica weight (kg), Ci (meq/L) and Ceq (meq/L) are the concentration of the metals in the 117 initial and final solutions (Table S1), respectively. The errors on the concentration have 118 been calculated from the triplicate ICP-MS analysis. 119 Solids were dried at room temperature and characterized by X-ray diffraction (XRD), 120 ATD/TG and MAS-NMR spectroscopy. 121 122 2.3. Characterization 123 6  The electromotive force (Eh) and the pH values of the initial solutions and 124 supernatants were measured at room temperature using a Eutech Instruments PC 700 pH-125 meter. 126 ICP-MS (Inductively Coupled Plasma-Mass Spectrometry) was used to analyse, 127 on the one hand, the sulphur content on the X-MEA-Mn solid samples, and, on the other 128 hand, the metal content on the initial solutions and the supernatants obtained after the 129 equilibrium reaction. Measurements were carried out at the Microanalysis Laboratory 130 (CITIUS, University of Seville, Spain) using a HORIBA JOBIN YVON-ULTIMA 2 131 equipment. 132 Thermogravimetrial (TG/DTA) experiments were carried out using a TA (model 133 STD-Q600) instrument, in Characterization Service (CITIUS, University of Seville, 134 Spain), with alumina as reference. The samples were placed into a Pt crucible and 135 maintained at air throughout the heating period. The temperature was increased at a 136 constant rate of 10° C/min up to 1000 ºC. 137 Powder X-ray diffraction (XRD) was carried out to check the phase purity, to 138 monitor crystallinity and to determine the basal spacing of the micas. XRD patterns were 139 obtained at the X-ray laboratory (CITIUS, University of Seville, Spain) on a Bruker D8 140 Advance instrument equipped with a Cu K radiation source operating at 40 kV and 40 141 mA. Diffractograms were obtained in the 2 range of 3–70° with a step size of 0.015° 142 and a step time of 0.1 s. 143 Solid state nuclear magnetic resonance (MAS-NMR) was used to analyse the short 144 order range in the solid interlayer space. Single-pulse (SP) MAS-NMR experiments were 145 recorded on a Bruker AVANCE WB400 spectrometer equipped with a multinuclear 146 probe, at the Nuclear Magnetic Resonance Service of University of Córdoba (Córdoba, 147 Spain). Powdered samples were packed in 3.2 mm zirconia rotors and spun at 10 kHz. 148 7  23Na MAS NMR spectra were recorded at 105.84 MHz with a pulse width of 0.75 µs (π/2 149 pulse length = 4.5 µs) and a delay time of 0.1 s. 13C MAS NMR spectra with proton 150 decoupling were recorded at 104.26 MHz with a pulse width of 2.5 µs (π/2 pulse length 151 = 7.5 µs) and a delay time of 2 s. The chemical shift values were reported in ppm from 152 tetramethylsilane for 13C and from a 0.1 M NaCl solution for 23Na. 153 154 3. Results 155 The pH and Eh values for the initial and supernatant solutions were represented in 156 the Pourbaix diagrams (Fig. S1), showing the most stable species in the solutions. 157 In all the initial solutions, the divalent heavy metal cation (X2+) was the most 158 stable species, although in mercury solutions the pH and Eh values were close to the Hg22+ 159 stability region. 160 In general, supernatants had a less acidic pH and a lower Eh than the starting 161 solutions. Despite these changes, the most stable species in lead and cadmium 162 supernatants was still the X2+ cation. However, HgO was the most stable species in Hg-163 Na-Mn and Hg-EA-Mn supernatants. The decrease in Eh in Hg-MEA-Mn supernatants 164 was more evident than in Hg-Na-Mn supernatants, causing pH and Eh values to be close 165 to the HgO/ Hg22+ stability limit for Hg-MEA-M4 and between HgO and metallic Hg for 166 Hg-MEA-M2 (Fig. S1). 167 The concentration of sodium and heavy metal cations in the initial solutions and 168 supernatants was measured by ICP-MS (Table S1). The desorbed interlayer sodium (Cdes 169 (Na+)) and adsorbed heavy metal cations (Cs (X2+)) in each system were calculated from 170 these values (Fig. 1). Interlayer sodium was not completely desorbed in Na-Mn (Fig. 1, 171 upper). As an exception, Hg-Na-M2 desorbed a quantity of sodium cations greater than 172 its cation exchange capacity (CEC), probably due to the low pH value of the starting Hg2+ 173 8  solution that provoked the desorption of the non-exchanged sodium (Osuna et al., 2019b). 174 In both sodium micas, the values of adsorbed heavy metal cations were higher than 175 desorbed sodium. Thus, other adsorption mechanisms further than cation exchange may 176 be happening, such as the formation of precipitates and/or adsorption at specific sites. 177 Fig. 1, bottom, compares the adsorption data for each heavy metal cation on the 178 different adsorbents. EA-Mn adsorbed less Pb2+ and Hg2+ than Na-Mn. The adsorbed Cd2+ 179 was higher in EA-M4 and similar in EA-M2 in comparison with Na-Mn. The system that 180 adsorbs the least amount of Pb2+ and Cd2+ was MEA-Mn. In contrast, the amount of 181 adsorbed Hg2+ in MEA-Mn was higher than in the other two systems. 182 It should be noted that the Y-M2 systems (Y = Na, EA and MEA) adsorbed an amount of 183 Hg2+ higher than the CEC of Na-M2. 184 In order to analyze possible changes in the structure of the adsorbents and identify 185 new crystalline phases, solid samples were analyzed by XRD. 186 Fig. 2 compares the XRD diagrams in the 2θ range 4-10º of Na-Mn before and after each 187 adsorption. The adsorption of Pb2+, Cd2+ and Hg2+ in Na-M4 did not produce changes in 188 the hydration state of the sheets. The XRD diagram of sample Pb-Na-M2 showed a unique 189 001 reflection to ca. 1.21 nm, same basal spacing than sodium mica whereas Cd-Na-M2 190 and Hg-Na-M2 XRD patterns showed a second reflection corresponding to a greater basal 191 spacing (1.45 nm) than Na-M2 (1.21 nm). This peak was compatible with the presence of 192 heavy metal cations or leached framework cations surrounded by a bilayer of water in the 193 interlayer space(Kodama and Komarneni, 1999). This layer swelling after the adsorption 194 of Cd2+ and Hg2+ did not occur in Na-M4. Na-M4 had a greater layer charge than Na-M2 195 and, therefore, a higher interaction between the interlayer cations and the surface of the 196 mica occurred, hindering the swelling capacity. 197 9  The XRD patterns of the EA-Mn before and after each adsorption were shown in 198 Fig. 3. EA-M4 XRD pattern (Fig. 3a, left) showed a single reflection due to the parallel 199 bilayer arrangement of the organic cation (1.60 nm) (Osuna et al., 2019a). The adsorption 200 of Pb2+ did not produce variation in the 001 reflection. In the case of EA-M4 after 201 adsorbing Cd2+, in addition to the reflection at 1.60 nm, a second less intense reflection 202 was observed at 1.22 nm. This basal spacing was compatible with the presence of organic 203 cations in a parallel monolayer or Cd2+ arrangement surrounded by a monolayer of water 204 (Osuna et al., 2019a). The Hg-EA-M4 XRD pattern showed a single reflection at a value 205 of 2Ɵ that corresponds to a basal spacing of 1.36 nm, lower than that of the adsorbent (ca. 206 1.60 nm). This basal distance may be related to the presence of hydrated mercury cations 207 in the interlayer or to organic cations arranged in parallel monolayer (Osuna et al., 2019a). 208 These changes in the hydration states indicated that the adsorption of cadmium and 209 mercury cations occurs by a partial exchange of organic cations. In the case of Pb2+, the 210 XRD patterns could indicate that a large amount of EA was not desorbed. 211 The XRD pattern of EA-M2 (Fig. 3, right) showed two reflections due to different 212 arrangements of the organic cation: in parallel bilayer (ca. 1.60 nm) and parallel 213 monolayer (ca. 1.26 nm) (Osuna et al., 2019a). After the adsorption of Pb2+, there were 214 no changes in the reflections, however, after the adsorption of Cd2+, a decrease in the 215 intensity of the reflection at ca. 1.60 nm was observed and the reflection at ca. 1.26 nm 216 disappeared. A new reflection could also be observed at ca. 1.45 nm that could be 217 attributed to Cd2+ surrounded by a bilayer of water (Kodama and Komarneni, 1999). This 218 could indicate that the adsorption of Cd2+ was produced by an exchange of the EA 219 molecules by the heavy metal cation. In the case of Hg2+, the XRD pattern showed two 220 reflections, one at ca. 1.21 nm that may correspond to a parallel monolayer arrangement 221 of EA or to Hg2+ surrounded by a monolayer of water. The second reflection at 1.36 nm, 222 16  1). The 23Na MAS NMR spectrum of Y-M2 after each adsorption (Fig. 5) showed a 371 decrease of the non-exchangeable and sodalite signals, according with heavy metal 372 cations adsorbed at specific sites and/or forming precipitates. 373 The functionalized micas were analyzed by 13C MAS NMR to obtain more 374 information on the changes produced in the organic cations when adsorbing the different 375 heavy metal cations. The 13C MAS NMR spectrum of EA-M4 (Fig. 6a, up left) was 376 characterized by two signals corresponding to the carbons of CH3 (13.8 ppm) and CH2 377 (33.1 ppm) groups (PerkinElmer, 2017). After each adsorption, the 13C MAS NMR 378 spectra of the X-EA-M4 (Fig. 6b-d, up left) showed the same set of signals as the 379 adsorbent, but the signal corresponding to the CH2 carbon resonated at lower frequencies. 380 The sum of the amount of sodium cations (ICP-MS) and organic cations (TG) desorbed 381 was lower than the amount of heavy metal adsorbed, which could indicate that part of the 382 organic cations was adsorbed as ion pair along with the ions NO3- from the solutions of 383 the heavy metals salts. These ions would shield the charge of the methylene group causing 384 the carbon signal to shift. 385 The 13C MAS NMR spectrum of EA-M2 (Fig. 6a, up right) showed two sets of 386 signals corresponding to the organic cation carbons in parallel bilayer (13.8 ppm and 33.9 387 ppm) and parallel monolayer arrangement (16.2 ppm and 37.0 ppm) (PerkinElmer, 2017). 388 After the adsorption with Pb2+ and Cd2+ (Fig. 6b-c, up right), a decrease in the relative 389 intensity of the signals associated with the organic cation arranged in parallel monolayer 390 was observed. This decrease indicated that the adsorption of heavy metals caused the 391 partial removal of the alkyl ammonium chains. In the case of Hg-EA-M2 (Fig. 6d, up 392 right), it was not possible to distinguish carbon signals in the 13C MAS NMR spectrum. 393 The low signal-to-noise ratio of the 13C MAS NMR spectrum indicated that the sample 394 contained a small amount of EA, allowing Hg2+ to form internal sphere complexes, as 395 17  inferred by TG. Therefore, the two 001 reflections observed by XRD (Fig. 3d, right) 396 corresponded to different hydration states of the heavy metal cation and not to different 397 configurations of the organic cation, as originally postulated. 398 The functionalization of MEA-M4 was not complete (Osuna et al., 2020) and the 399 presence of sodium in the interlayer produced different environments in the organic 400 cation. The 13C MAS NMR spectra of X-MEA-M4 showed the same set of signals as the 401 initial functionalized mica, but with different relative intensities due to the loss of the 402 organic cation during the adsorption (Fig. 6, down left). The change in intensity was 403 greater in Hg-MEA-M4 since it was the system that desorbed the highest amount of 404 organic cations during the adsorption as previously observed by ICP-MS (Table 4). 405 The 13C MAS NMR spectrum of MEA-M2 (Fig. 6a, down right) was characterized 406 by four signals that were assigned to two different cations (PerkinElmer, 2017): HS-CH2-407 CH2-NH3+ (40 ppm and 23 ppm) and +H2S-CH2 -CH2-NH3+ (43 ppm and 33 ppm). The 408 13C MAS NMR spectra after the adsorption of Pb2+ and Hg2+ in MEA-M2 (Figure 6b and 409 d, right) only showed the carbon signals of the CH2-NH3+ groups. This fact may indicate 410 that the heavy metal adsorption occurred in the thiol groups, causing a widening of the 411 signal and preventing it from being distinguished from noise. 412 413 4. Discussion 414 These as-made micas, Na-Mn, adsorbed more amount of heavy metals cations 415 than natural clay minerals, Table S2, (Uddin, 2017a). However, Na-M4 adsorbed a lower 416 amount of heavy metal cations than Na-M2 despite its higher CEC, because the 417 electrostatic interaction between the interlayer cations and the surface of the mica 418 increased, hindering their exchange (Pavón et al., 2014, 2013). Thus, the high attraction 419 between the cations and the surface of Na-M4 provoked that the interlayer space did not 420 18  change regardless the cation charge. However, a new reflection corresponding to a larger 421 basal spacing was observed in the Na-M2 XRD pattern after the adsorption of Cd2+ and 422 Hg2+. 423 In general, the order of the adsorbed amounts of each heavy metal was Hg2+> 424 Pb2+> Cd2+, as previously reported by Osuna et al. (Osuna et al., 2019b) for the cation 425 exchange reaction. Hg2+ had a greater tendency to form hydroxyl-complexes 426 (pkH(Hg2+)=3.4, pkH(Pb2+)=7.7 and pkH(Cd2+)=10.1) (Hubbard, 2002) which were 427 thermodynamically more stable and were better adsorbed than the hydrated metal cation 428 (James, R O, Healy, 1972). 429 The set of results indicated the presence of precipitates after the adsorption of 430 Pb2+, Cd2+ and Hg2+ as an additional adsorption mechanism to the cation exchange 431 process. 432 Although the long-range structure of the adsorbents (XRD) remained unchanged, 433 the hydration of the interlayer space depended on the type of the adsorbed metal cation. 434 The adsorption of Cd2+ or Hg2+ produced an increase in the basal spacing of Na-Mn. This 435 was due to the fact that both cations had greater hydration enthalpy, in absolute value, 436 than Pb2+ (ΔH hyd(Pb2+)=-1481 kJ/mol, ΔHhyd(Cd2+)=-1807 kJ/mol and ΔHhyd(Hg2+) =-437 1824 kJ/mol) (Smith, 1977), and, therefore, they had a greater tendency to surround 438 themselves with water molecules (Etci et al., 2010). 439 The XRD patterns of the EA-Mn after the adsorption of Cd2+ and Hg2+ showed 440 new reflections attributed to the hydrated heavy metal cations in the interlayer space. As 441 for MEA-Mn, the adsorption of these two heavy metals did not affect the hydration of 442 MEA-M4 (Fig. 4, left), however, a change in the relative intensities of the 001 reflections 443 in MEA-M2 was observed (Fig. 4, right). The Pb2+ adsorption did not produce changes 444 19  in the basal space of the adsorbents. In general, the amount of adsorbed Pb2+, Cd2+ and 445 Hg2+ depended on the hydrolysis constants of each metal. 446 The effect of the ammonium group on the adsorption of heavy metals was 447 analyzed thorough the comparison of the adsorption by Na-Mn and EA-Mn. 448 Functionalized micas, EA-Mn, adsorbed a smaller amount of heavy metal than Na-Mn. 449 Some authors observed that in natural clays, the adsorption capacity of heavy metals 450 decreased after the functionalization with organic cations (Lee et al., 2002; Oyanedel-451 Craver and Smith, 2006; Uddin, 2017b). In absence of functional groups capable of 452 adsorbing heavy metal cations, such as ammonium, the adsorption occurred by cationic 453 exchange as shown by XRD, TG and MAS NMR. The exchange between heavy metal 454 cations and organic cations was disadvantaged in relation to sodium because organic 455 cations were more strongly adsorbed on mica than Na+ (Lee et al., 2002). In addition, the 456 organophilic character of EA-Mn hindered the adsorption of heavy metal cations. 457 However, the adsorption of Cd2+ in both Y-M2 was practically the same and in EA-M4 458 was higher than in Na-M4, probably due to the formation of amorphous precipitates that 459 were not observed by XRD. 460 Finally, the effect of thiol groups on the adsorption was evaluated from the 461 adsorption results obtained by EA-Mn and MEA-Mn. 462 ICP-MS data showed that the presence of thiol groups in the organic cation 463 decreased the adsorption of Pb2+ and Cd2+, but improved the adsorption of Hg2+. In MEA-464 Mn, the adsorption of Pb2+ and Cd2+ occurred mainly by cation exchange mechanism 465 while the Hg 2+ adsorption was also at specific sites (thiol groups) (Table 4). The 466 adsorption of mercury cations in the thiol group was favored with respect to the other two 467 heavy metal cations, since Hg2+ Gibbs free energy was more negative than that of Pb2+ 468 and Cd2+ (Brown et al., 1999): 469 20  470 Several authors, (do Nascimento et al., 2016; Guerra et al., 2009; Tran et al., 471 2015a, 2015b) had previously observed that natural clays adsorbed a higher amount of 472 Hg2+ after functionalizing them with thiol groups. After the functionalization of 473 motmorillonites and sepiolites with MEA and MPTMS, respectively, their Hg2+ 474 adsorption capacity increased and their Pb2+ adsorption capacity decreased (Celis et al., 475 2000). 476 The data obtained from the XRD, ICP-MS and 13C MAS NMR spectroscopy 477 indicated that the loss of organic cations during the adsorption was lower in MEA-Mn 478 than in EA-Mn, so thiol groups increased the adsorption strength of the organic cation 479 and hindered cation exchange reaction. Thus, the presence of thiol groups in the organic 480 chain provided specific sites for the Hg2+ adsorption, but hindered the adsorption of Pb2+ 481 and Cd2+. 482 483 5. Conclusions 484 The efficient adsorption of heavy metal cations in water from both as-made and 485 functionalized micas has been demonstrated, being the efficiency higher in Y-M2 than in 486 Y-M4. In all the adsorbents, a higher amount of Hg2+ is adsorbed, followed by Pb2+, and, 487 finally Cd2+. 488 In general, the functionalization of micas decreases the amount of adsorbed heavy 489 metal, except for Hg-MEA-Mn, since the presence of thiol groups generates specific 490 adsorption sites that improves their adsorption capacity. 491 H2S + M2+  MS + 2H+ ΔG° (Hg2+) = -181.6 kJ/mol ΔG° (Cd2+) = -45.3 kJ/mol ΔG° (Pb2+) = -40.9 kJ/mol 21  In summary, the best adsorbent for Pb2+ and Cd2+ is Na-M2, and, in the case of 492 Hg2+, the best adsorbent is the MEA-M2. The adsorption amount of Pb2+ and Cd2+ by Na-493 M2 is 3373.3 meq/kg and 2263.0 meq/kg, respectively, and, 4918.6 meq Hg2+ are 494 adsorbed per 1 kg of MEA-M2, being the adsoption capacity of these micas ten time 495 higher than the capacity of the reported natural clay minerals. 496 497 Acknowledgements 498 The authors would like to thank the Junta de Andalucía (Spain) and FEDER 499 (Proyecto de Excelencia de la Junta de Andalucía, project P12-FQM-567), to the Spanish 500 State Program R+D+I oriented societal challenges and FEDER (Project MAT2015-501 63929-R) for financial support. F.J. Osuna thanks his grant to the training researcher 502 program associated to the excellence project of the Junta de Andalucía (P12-FQM-567). 503 Dr. Pavón thanks the University of Seville for the financial support of her current contract 504 from VI PPIT-US program. 505 506 References 507 Alba, M.D., Castro, M.A., Naranjo, M., Pavón, E., 2006. Hydrothermal reactivity of Na-508 n-micas (n = 2, 3, 4). Chem. Mater. 18, 2867–2872. 509 Bailey, S.E., Olin, T.J., Bricka, R.M., Adrian, D.D., 1999. A review of potentially low-510 cost sorbents for heavy metals. Water Res. 33, 2469–2479. 511 Balasubramanian, N., Ahamed, A.J., 2000. Adsorption dynamics - temperature effect on 512 adsorption of lead(II) species onto lignite surface. J. Indian Chem. Soc. 77, 11–13. 513 Bhattacharyya, K.G., Sen Gupta, S., 2008. Adsorption of a few heavy metals on natural 514 and modified kaolinite and montmorillonite: A review. Adv. Colloid Interface Sci. 515 140, 114–131. 516 22  Brown, J., Mercier, L., Pinnavaia, T.J., 1999. Selective adsorption of Hg2+ by thiol-517 functionalized nanoporous silica. Chem. Commun. 69–70. 518 Celis, R., Hermosin, M.C., Cornejo, J., 2000. Heavy metal adsorption by functionalized 519 clays. Environ. Sci. Technol. 34, 4593–4599. 520 Cruz-Guzman, M., Celis, R., Hermosin, M.C., Koskinen, W.C., Nater, E.A., Cornejo, J., 521 2006. Heavy metal adsorption by montmorillonites modified with natural organic 522 cations. Soil Sci. Soc. Am. J. 70, 215–221. 523 do Nascimento, F.H., de Souza Costa, D.M., Masini, J.C., 2016. Evaluation of thiol-524 modified vermiculite for removal of Hg(II) from aqueous solutions. Appl. Clay Sci. 525 124, 227–235. 526 Etci, O., Bektas, N., Oncel, M.S., 2010. Single and binary adsorption of lead and cadmium 527 ions from aqueous solution using the clay mineral beidellite. Environ. Earth Sci. 61, 528 231–240. 529 Ezzat, H., Menazea, A.A., Omara, W., Basyouni, O.H., Helmy, S.A., Mohamed, A.A., 530 Tawfik, W., Ibrahim, M., 2020. DFT: B3LYP/LANL2DZ Study for the Removal of 531 Fe, Ni, Cu, As, Cd and Pb with Chitosan. Biointerface Res. Appl. Chem. 10, 7002–532 7010. 533 Fahim, R.B., Kolta, G.A., 1970. Thermal decomposition of hydrated Cadmium oxide. J. 534 Phys. Chem. 74, 2501–2507. 535 Farid, I.M., Abbas, M.H.H., Bassouny, M.A., Gameel, A., Abbas, H.H., 2020. Indirect 536 Impacts of Irrigation with Low Quality Water on The Environmental Safety. 537 Egyotian J. Soil Sci. 60, 1–15. 538 Figueroa, A., Cameselle, C., Gouveia, S., Hansen, H.K., 2016. Electrokinetic treatment 539 of an agricultural soil contaminated with heavy metals. J. Environ. Sci. Heal. Part 540 A-Toxic/Hazardous Subst. Environ. Eng. 51, 691–700. 541 23  García-Jiménez, M.J., Cota, A., Osuna, F.J., Pavón, E., Alba, M.D., 2016. Influence of 542 temperature and time on the Eu3+ reaction with synthetic Na-Mica-n (n=2 and 4). 543 Chem. Eng. J. 284, 1174–1183. 544 Grim, R.E., 1968. Clay Mineralogy. New York. 545 Guerra, D.L., Santos, M.R.M.C., Airoldi, C., 2009. Mercury Adsorption on Natural and 546 Organofunctionalized Smectites - Thermodynamics of Cation Removal. J. Braz. 547 Chem. Soc. 20, 594–603. 548 Ho, Y S, Ng, J C Y, McKay, G., 2001. Removal of lead(II) from effluents by sorption on 549 peat using second-order kinetics. Sep. Sci. Technol. 36, 241–261. 550 Ho, Y.S., 2003. Removal of copper ions from aqueous solution by tree fern. Water Res. 551 37, 2323–2330. 552 Hu, G, Liu, G, Wu, D , Fu, B., 2018. Geochemical behavior of hazardous volatile 553 elements in coals with different geological origin during combustion. Fuel 233, 361–554 376. 555 Hubbard, A.T., 2002. Encyclopedia of surface and colloid science. Vol. 1. CRC Press. 556 Inglezakis, V.J., Stylianou, M.A., Gkantzou, D., Loizidou, M.D., 2007. Removal of Pb(II) 557 from aqueous solutions by using clinoptilolite and bentonite as adsorbents. 558 Desalination 210, 248–256. 559 James, R O, Healy, T.W., 1972. Adsorption of hydrolyzable metal-ions at oxide-water 560 interface .3. Thermodynamic model of adsorption. J. Colloid Interface Sci. 40, 65-. 561 Kodama, T., Komarneni, S., 1999. Na-4-mica: Cd2+, Ni2+, Co2+, Mn2+ and Zn2+ ion 562 exchange. J. Mater. Chem. 9, 533–539. 563 Lee, J.J., Choi, J., Park, J.W., 2002. Simultaneous sorption of lead and chlorobenzene by 564 organobentonite. Chemosphere 49, 1309–1315. 565 Lin, S.H., Juang, R.S., 2002. Heavy metal removal from water by sorption using 566 24  surfactant-modified montmorillonite. J. Hazard. Mater. 92, 315–326. 567 Malferrari, D., Brigatti, M.F., Laurora, A., Medici, L., Pini, S., 2006. Thermal behavior 568 of Cu(II)-, Cd(II)-, and Hg(II)-exchanged montmorillonite complexedwith cysteine. 569 J. Therm. Anal. Calorim. 86, 365–370. 570 Malferrari, D., Brigatti, M.F., Laurora, A., Pini, S., Medici, L., 2007. Sorption kinetics 571 and chemical forms of Cd(II) sorbed by thiol-functionalized 2 : 1 clay minerals. J. 572 Hazard. Mater. 143, 73–81. 573 Mathur, A., Rupainwar, D.C., 1988. Removal of lead from polluted waters by adsorption 574 on fly ash. Asian Environ. 10, 19–25. 575 Mercier, L, Detellier, C., 1995. Preparation, characterization and applications as heavy-576 metals sorbent of covalently grafted thiol functionalities on the interlamellar surface 577 of montmorillonite. Environ. Sci. Technol. 29, 1318–1323. 578 Mittal, A., Ahmad, R., Hasan, I., 2016. Biosorption of Pb2+, Ni2+ and Cu2+ ions from 579 aqueous solutions by L-cystein-modified montmorillonite-immobilized alginate 580 nanocomposite. Desalin. Water Treat. 57, 17790–17807. 581 Naranjo, M., Castro, M.A., Cota, A., Osuna, F.J., Pavón, E., Alba, M.D., 2015. Synthesis 582 temperature effect on Na-Mica-4 crystallinity and heteroatom distribution. 583 Microporous Mesoporous Mater. 204, 282–288. 584 Naranjo, M., Castro, M.A., Cota, A., Pavón, E., Pazos, M.C., Alba, M.D., 2014. A new 585 route of synthesis of Na-Mica-4 from sodalite. Microporous Mesoporous Mater. 586 186, 176–180. 587 NJDEP, 2014. Technical Guidance on the Capping of Sites Undergoing Remediation. 588 Osuna, F.J., Cota, A., Fernandez, M.A., Pavon, E., Torres Sanchez, R.M., Alba, M.D., 589 2019a. Influence of framework and interlayer on the colloidal stability of design 590 swelling high-charged micas. Colloids Surfaces A-Physicochemical Eng. Asp. 561, 591 25  32–38. 592 Osuna, F.J., Cota, A., Pavón, E., Carolina Pazos, M., Alba, M.D., 2018. Cesium 593 adsorption isotherm on swelling high-charged micas from aqueous solutions: Effect 594 of temperature. Am. Mineral. 103. 595 Osuna, F.J., Pavon, E., Alba, M.D., 2020. An insight on the design of mercapto 596 functionalized swelling brittle micas. J. Colloid Interface Sci. 561, 533–541. 597 Osuna, F.J., Pavon, E., Alba, M.D., 2019b. Design swelling micas: Insights on heavy 598 metals cation exchange reaction. Appl. Clay Sci. 182, 105298. 599 Oyanedel-Craver, V.A., Smith, J.A., 2006. Effect of quaternary ammonium cation 600 loading and pH on heavy metal sorption to Ca bentonite and two organobentonites. 601 J. Hazard. Mater. 137, 1102–1114. 602 Pavón, E., Castro, M.A., Cota, A., Osuna, F.J., Pazos, M.C., Alba, M.D., 2014. Interaction 603 of hydrated cations with mica-n (n = 2, 3 and 4) surface. J. Phys. Chem. C 118, 604 2115–2123. 605 Pavón, E., Castro, M.A., Naranjo, M., Orta, M.M., Pazos, M.C., Alba, M.D., 2013. 606 Hydration properties of synthetic high-charge micas saturated with different cations: 607 An experimental approach. Am. Mineral. 98, 394–400. 608 Pazos, M.C., Castro, M.A., Cota, A., Osuna, F.J., Pavón, E., Alba, M.D., 2017. New 609 insights into surface-functionalized swelling high charged micas: Their adsorption 610 performance for non-ionic organic pollutants. J. Ind. Eng. Chem. 52, 179–186. 611 Periasamy, K, Namasivayam, C., 1994. Process-development for removal and recovery 612 of cadmium from waste-water by a low-cost adsorbent - Adsorption rates and 613 equilibrium studies. Ind. Eng. Chem. Res. 33, 317–320. 614 PerkinElmer, 2017. ChemDraw Proffesional 16.0.1.4/77. 615 Praus, P., Motakova, M., Ritz, M., 2012. Montmorillonite ion exchanged by mercury (II). 616 32  Fig. 2 703 704 33  Fig. 3 705 706 707 34  Fig. 4 708 709 35  Fig. 5 710 711 712 36  Fig. 6 713 714 715 716