Behaviour of kaolinite intercalation compounds with selected ammonium salts in aqueous chromate and arsenate solutions

Accepted Manuscript Behaviour of kaolinite intercalation compounds with selected ammonium salts in aqueous chromate and arsenate solutions Jakub Matusik, Lucyna Matykowska PII: DOI: Reference:

S0022-2860(14)00430-X http://dx.doi.org/10.1016/j.molstruc.2014.04.063 MOLSTR 20570

To appear in:

Journal of Molecular Structure

Received Date: Revised Date: Accepted Date:

23 January 2014 18 April 2014 18 April 2014

Please cite this article as: J. Matusik, L. Matykowska, Behaviour of kaolinite intercalation compounds with selected ammonium salts in aqueous chromate and arsenate solutions, Journal of Molecular Structure (2014), doi: http:// dx.doi.org/10.1016/j.molstruc.2014.04.063

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Behaviour of kaolinite intercalation compounds with selected ammonium salts in aqueous chromate and arsenate solutions

Jakub Matusika,*, Lucyna Matykowskaa

a

Department of Mineralogy, Petrography and Geochemistry; Faculty of Geology,

Geophysics and Environmental Protection; AGH University of Science and Technology, al. Mickiewicza 30, 30-059 Krakow, Poland. Corresponding author. Tel. +48 126175233; Fax: +48 126334330. E-mail address: [email protected] (J. Matusik).

Abstract The removal of aqueous Cr(VI) and As(V) oxyanions from waters by different materials with sorption properties is of environmental importance. In this study, a methoxy-kaolinite derivative was intercalated with benzyltrimethylammonium (B1), tetramethylammonium (TMA), and benzyldimethylhexadecylammonium (B5) chlorides and the interaction of the obtained materials with oxyanions was examined. The PXRD (powder X-ray diffraction) and IR (Infrared spectroscopy) analyses indicated a monolayer arrangement of the B1 and TMA molecules in the interlayer space of the mineral, while a tilted arrangement was noticed in the case of B5. A complete or partial deintercalation of introduced molecules was observed in the reactions with aqueous solutions of Cr(VI) and As(V). In all studied systems a significant improvement of the oxyanions removal was observed as compared to the pure kaolinite. The highest uptake of oxyanions was noticed in the reaction with B5-intercalated material. This was due to precipitation of organic alkyl salts. The formation of alkylchromate was confirmed

1

using FTIR spectroscopy. The lower uptake of oxyanions by the B1- and TMAintercalated materials was due to lack of new solid precipitation and resulted from the ion-exchange of chlorides initially compensating the ammonia nitrogen charge. The experimental sorption isotherms for all the reactions were best represented by Langmuir equation. A gradual, two-step removal process of Cr(VI) and As(V) by B1- and TMAintercalated materials was observed. In turn, the precipitation of alkyl salts in reaction with B5-intercalated material resulted in a rapid immobilization of the oxyanions. The kinetic data modelled using pseudo-second order equation showed very good agreement with experimental results.

Keywords: ammonium salts, arsenate, chromate, intercalation, kaolinite.

1. Introduction The deposits of kaolin group minerals are widespread in the world [1]. Their physical and chemical properties make them useful materials with applications in industry and environmental protection [2]. The structure of kaolinite which is the main component of kaolin sedimentary rock is build from stacked 1:1 layers of Al2Si2O5(OH)4 composition held by hydrogen bonds. The layer is composed of tetrahedral (Si-O) and octahedral (Al-O) sheets bonded through common oxygens. The unmodified mineral does not have swelling properties, however the interlayer can be intercalated using selected organic molecules e.g. dimethyl sulfoxide (DMSO) [3], urea [4], or formamide [5] which could be easily exchanged. Moreover, kaolinite possesses an unique asymmetric interlayer environment with two chemically different surfaces: oxygens of tetrahedral sheet and inner surface hydroxyls of octahedral sheet. In

2

particular the latter surface is susceptible to grafting reactions and formation of new structures [6]. This opens the way to synthesize new kaolinite-based hybrid materials in reactions with organic molecules [7-9]. An appropriate selection of intercalated or grafted organic guest enables to design materials with improved sorption [10-12], luminescent [13], and catalytic properties [14, 15]. The obtained organo-functionalized materials may also be used as fillers in the production of polymer-clay nanocomposites due to their chemical compatibility with selected polymers [16, 17]. Chromate and arsenate oxyanions are highly mobile and their toxic and/or cancerogenic properties are known [18, 19]. The remediation of waters polluted with inorganic oxyanions are often based on sorption processes which include the use of different materials e.g. layered double hydroxides [20], activated carbon [21], oxides [22], modified smectite minerals [23], and organo-zeolites [24-26]. Recently new types of kaolinite intercalation compounds with selected benzylalkylammonium salts were synthesized and characterized [27, 28]. The introduced molecules in the form of chlorides may induce immobilization properties of the obtained materials towards oxyanions which are of environmental interest. Thus, the research objective was to investigate the efficiency of Cr(VI) and As(V) removal by intercalation compounds of well ordered kaolinite with selected ammonium salts. The sorption equilibrium and kinetics, pH effect as well as the immobilization mechanism were investigated in detail.

3

2. Experimental section 2.1. Synthesis of intercalation compounds A well ordered kaolinite with Hinckley index of 1.31 from Polish “Maria III” deposit (M sample) was chosen for the experiments. A detailed characterization of the mineral sample was reported earlier [29, 30]. The chemical compounds: dimethyl sulfoxide (DMSO) and methanol were obtained from POCH company (Polish Chemical Reagents). In turn, the following ammonium salts were purchased from Sigma-Aldrich: benzyltrimethylammonium chloride (B1), tetramethylammonium chloride (TMA) and benzyldimethylhexadecylammonium chloride (B5). The structures and dimensions of the molecules are given as supporting material (Fig. S1). The synthesis of intercalation compounds followed a three-step procedure described previously [27]. Briefly, the kaolinite pre-intercalated with DMSO was further grafted with methoxyl groups by washing with methanol at room temperature (22ºC). The formed methoxy-kaolinite derivative with hydrophobic character of the interlayer space was sufficient for the insertion of the ammonium salts. The organic compounds in the form of chlorides were dissolved in methanol and the solution concentration was set to 2 mol/L (B1 and TMA salts) and 1 mol/L (B5 salt). The appropriate solutions were mixed with methoxy-kaolinite precursor for 24 h at 22ºC in solid/solution ratio of 125 g/L. The obtained materials were abbreviated in analogy to used salts as MB1, MTMA, and MB5, respectively. Additionally, the MB1 and MTMA derivatives were washed in isopropanol to remove the excess of salts which could have crystallized outside the interlayer space. The washing was not possible in the case of MB5 sample as the isopropanol molecules could destroy the intercalation compound structure.

4

2.2. Immobilization experiments For the sorption equilibrium experiments the Cr(VI) and As(V) solutions prepared from potassium dichromate K2Cr2O7 and sodium arsenate dibasic heptahydrate Na2HAsO4 · 7 H2O, respectively, covering the 0.02-50.0 mmol/L range were used. The initial pH (pHin) was set in the 3-11 range using diluted solutions of HNO3 and/or KOH. The sorbent/solution suspensions of 20 g/L concentration were shaken at 22ºC for 24 h. For the kinetic experiments the 5.0 mmol/L starting oxyanion concentration was used and the pHin was set to 5.0. The suspension aliquots were collected within the time interval 0.5-20 minutes. All collected suspension samples were filtered through 0.22 µm PES filters. The concentration of Cr(VI) was measured colorimetrically by 1,5diphenylcarbazide method using UV-Vis Hitachi U-1800 spectrophotometer [31]. The As(V) concentration was determined by hydride generation method using GBC SavantAA atomic absorption spectrometer (HG-AAS). The spectrophotometrical measurement of the B1 and B5 salts concentration was carried out using 208 nm band attributed to π →π* transition.

2.3. Methods for structural characterization The powder X-ray diffraction (PXRD) patterns were recorded using a Philips APD PW 3020 X’Pert instrument with CuKα radiation and a graphite monochromator. The powdered samples were analyzed in the range of 1.5 to 16°2θ with a step of 0.05°2θ. FTIR spectra were collected by Nicolet 7600 spectrometer using DRIFT technique (3 wt.% sample/KBr) with 64 scans at 4 cm-1 resolution in the 4000-400 cm-1 mid-region. Elemental analysis was performed using VarioEL III Elementar CHNS analyzer. The

5

results were used to calculate the formulas of materials. In the case of methoxykaolinite, the carbon content was taken for the calculations, while the formulas of intercalation compounds were calculated on the basis of nitrogen content.

2.4. Equilibrium and kinetic theoretical models A detailed description of the equations used for modelling of the experimental equilibrium and kinetic data is given in table 1.

Table 1

3. Results and discussion 3.1. Structure of intercalation compounds The pure M kaolinite exhibits a 7.2 Å peak in the PXRD pattern which is characteristic for a 1:1 layered structure of the kaolin group minerals (Fig. 1). The IR spectrum shows four well resolved bands corresponding to different types of structural OH hydroxyls (3700-3600 cm-1 region) characteristic for a well ordered kaolinite (Fig. 2). The DMSO intercalation caused a swelling of the mineral as the initial basal spacing increased to 11.2 Å (MDS sample) (Fig. 1) [32]. Moreover, dramatic changes of the bands attributed to the inner surface OH groups (3700-3600 cm-1) are observed in the IR spectra (Fig. 2). This especially involves an intensity decrease of ~3695 cm-1 band as well as the appearance of 3540 and 3504 cm-1 bands (Fig. 2). The latter are due to formation of hydrogen bonds between DMSO and octahedral surface of kaolinite [33]. The appropriate C-H stretching vibrations of DMSO molecule are found in the 31002900 cm-1 region (Fig. 2). A repeated washing of the MDS with methanol leads to

6

formation of a methoxy-kaolinite (MMD sample) which in a dry state has a 8.6 Å basal spacing (Fig. 1) [34, 35]. The IR spectrum of MMD sample reveals bands in the 30002800 cm-1 range due to attachment of the methoxyl CH3O- group to the octahedral sheet through Al-O-C bonds (Fig. 2) [34]. This is a result of a grafting which involves the reaction between inner surface OH groups of the mineral and OH groups of the methanol molecules [34]. In the reaction, which is often referred as esterification, water molecules are formed as a by-product [36]. The absence of 3021 cm-1 band (CH3-S bonding in DMSO) confirms the removal of DMSO during methanol washing (Fig. 2). The washing also caused changes in the OH region of the spectrum as a result of water formation in the grafting process leading to the appearance of 3563 and 3515 cm-1 bands. The intercalation of B1, TMA, and B5 molecules leads to a significant increase of the basal spacing to 14.6 Å, 12.6 Å, and 34.5 Å, respectively (Fig. 1). Additionally, a second order reflection at 17.0 Å is noticed for the MB5 material. A comparison of the molecules size (Fig. S1) with the d values for the MB1 and MTMA materials indicates that the B1 and TMA molecules are lying parallel to the methoxy-kaolinite layers forming a monolayer [27]. In turn, in theory the B5 molecules which adopt a position perpendicular to methoxy-kaolinite layers should give d value equal to ~39.2 Å. This is a sum of the B5 molecule length in trans conformation (30.6 Å) (Fig. S1) and the d value of MMD sample (8.6 Å). Analogically, the monolayer arrangement of the B5 molecules should result in d value close to ~16.0 Å. Therefore, a tilted arrangement of the B5 molecule to the methoxy-kaolinite layers is proposed. The IR spectra for all the materials show changes in the OH (3700-3500 cm-1) and C-H (3100-2800 cm-1) stretching regions (Fig. 2). The intercalation of the salts leads to a simultaneous incorporation of water molecules which are relatively highly ordered in the case of MB1

7

and MTMA materials. This is confirmed by relatively narrow water bands at ~3545 cm1

and is connected with hydration properties of the B1 and TMA molecules. On the

other hand a broad band is observed for the MB5 suggesting that water occupies random positions in the interlayer space. The most intense bands in the C-H region at 2921 and 2851 cm-1 for the MB5 material are due to vibrations of methylene and methyl groups which build the long aliphatic chain of the B5 molecule (Fig. 2). It is worth to underline that in the PXRD patterns after each modification step the 7.2 Å peak is hardly visible which indicates that almost all of the kaolinite layers underwent intercalation in the applied experimental conditions (Fig. 1). This is most often observed for kaolin group minerals of high structural order which show high reactivity towards organic molecules [5]. The calculations based on the CHNS analysis show that the MMD formula is the following: Al2Si2O5(OH)3.77(OCH3)0.23. This indicates that part of the inner surface OH groups were exchanged to CH3O- groups. The calculated molar amount of the B1, MTMA, and B5 salts per 1 mole of the MMD is equal to: 0.38, 0.54, and 0.59, respectively. The PXRD data which give information on the arrangement of the molecules and their dimensions enabled to calculate the following maximal molar amount which could be intercalated: B1: 0.29, TMA: 0.51, and B5: 0.46. The comparison of these values indicates that the surface of the obtained intercalates is not free of crystalline salts and an excess of the salts is observed.

Fig. 1 Fig. 2

8

3.2. Interaction of intercalates with aqueous Cr(VI) and As(V) 3.2.1. Equilibrium study and pH effect The sorption isotherms are best described using Langmuir equation in terms of linear correlation coefficient (R2) and modelled equilibrium sorption capacity (q) which should correspond to the experimental data (Fig. 3 and Table 2). Thus, the q values are used in the discussion. This is with exception of the MB5 material in reaction with Cr(VI) which is reported below in detail. The sorption capacity of the pure M kaolinite in reaction with Cr(VI) and As(V) is very low and does not exceed 2 and 5 mmol/kg, respectively (Fig. 3 and Table 2). The kaolin group minerals do not possess ionexchange properties due to insignificant isomorphous substitution effect within the 1:1 layer [37]. However, the sorption of oxyanions may take place on the deprotonated aluminol Al-OH and silanol Si-OH groups at the crystallites edges in convenient pH conditions [38, 39]. The Dubinin-Radushkevich equation enables to determine the adsorption energy (EDR) and its magnitude reflects the nature of the process [40]. If the EDR is lower than 8 kJ/mol than physical adsorption dominates, in turn if the EDR is in the 8-16 kJ/mol range than ion-exchange governs the process. Thus, for the M sample the EDR values indicate a physical adsorption both for the reaction with Cr(VI) and As(V) (Table 2). The PXRD data reveal that after reaction with aqueous oxyanions the introduced molecules were deintercalated as the initial basal spacing peaks disappeared, thus the salts were released to the solution (Fig. 1). Regardless of the oxyanion a similar increase of the sorption capacity was observed for the MB1 and MTMA materials. The Cr(VI) and As(V) removal for the MB1 reaches 81.3 and 76.3 mmol/kg, respectively. In turn,

9

for the MTMA the Cr(VI) and As(V) removal is equal to 22.4 and 25.3 mmol/kg, respectively (Table 2). The most significant increase of oxyanions immobilization is noticed for the MB5 material. The Cr(VI) removal reaches ~1000 mmol/kg as estimated from the sorption isotherm, while in the case of As(V) the q value is equal to 322.6 mmol/kg (Fig. 3). The observed sorption maxima are much higher that these earlier reported for an organo-kaolinite, 13 mmol/kg for Cr(VI) and 7 mmol/kg for As(V), which surface was modified by hexadecyltrimethylammonium bromide (HDTMA-Br) [41]. The MB5 material shows a sorption capacity comparable with other types of materials used for Cr(VI) and As(V) removal (Table 3).

Table 3 [42-51]

The pHin influences the reaction between Cr(VI) and the materials (Fig. 4). A decrease of sorption is observed with pH increase which is most visible for the MB5 material. The reaction between released B5 salt and the Cr(VI) leads to precipitation of a yellow color alkychromate. Its precipitation is favoured by low pH. The ratio of removed B5 to removed Cr(VI) is equal to 1.84 (pHin 3) and 1.97 (pHin 5) and corresponds to the molar proportions of Cr(VI) to organic cation in (B5)2CrO4 alkylchromate (Table 4). A similar product formed by reaction of Cr(VI) and HDTMABr was reported earlier [52]. The formation of alkylchromate caused a very high immobilization efficiency of Cr(VI) by the MB5 material. The presence of alkylchromate in the solid after reaction is confirmed by the IR spectra (Fig. 5). The incorporation of Cr(VI) oxyanion to the B5 salt structure and its exchange with Cl- is indicated by the appearance of 939 and 877 cm-1 bands attributed to the stretching

10

vibrations in the CrO4 tetrahedron [53]. The spectra of MB5 and MB5-Cr samples also contain bands connected with Si-O and Al-OH vibrations of kaolinite layer. Moreover, the PXRD pattern of the solid recorded after reaction shows a peak at 32.0 Å connected to the crystalline structure of alkylchromate (Fig. 5). Along with further pH increase the precipitation mechanism is hampered and the amount of removed Cr(VI) decreases. The concentration changes of Cr(VI), B1, and accompanying chlorides do not allow unambiguously to confirm the Cr(VI) -> Cl- exchange as a dominating mechanism (Table 4). This is due to the fact that the total Cl- concentration is not changed throughout the experiment as the chlorides are either compensating the B1 salt charge or the exchanged chlorides are present in the solution. In both situations the chlorides are in ionic form and the formation of B1-based alkylchromate solid was excluded. There is no macroscopic and/or spectroscopic evidence for its precipitation (data not shown). The data indicate that interaction between Cr(VI) and released B1 salt is not favourable. Even the large excess of B1 salt (~24 mmol/L) which is released from the MB1 material does not remove the Cr(VI) in significant amount. However the increase of sorption in comparison to the pure M kaolinite was significant and can be firstly explained by an increase of sorption centers at crystallites edges due to intercalation which doubled the interlayer distance. Secondly, the pure M sample has a CEC (cation exchange capacity) of 3.82 meq/100g (19.08 mmol/kg) and part of the B1 should occupy the ion-exchange positions within the 1:1 kaolinite layer forming a monolayer. The local formation of second salt layer through hydrophobic B1-B1 benzyl group interactions may induce anion-exchange properties in analogy to organokaolinites and organo-zeolites [25, 41]. The UV-Vis quantitative analysis of the TMA salt was not possible as the molecule does not absorb the radiation due to lack of π

11

electrons. However, due to similarities between B1 and TMA molecules the same mechanism as for the MB1 material is proposed. The lower sorption in the case of MTMA results probably from the lower affinity of TMA salt to react with Cr(VI) and As(V). The formation of a double layer for the TMA molecule which does not contain a benzyl group is less likely due to low interaction energy between methyl groups. The same observations apply to the reactions of MB1 and MTMA materials with As(V) where the sorption capacities are similar. The formation of alkylarsenate in the reaction of MB5 material with As(V) is not observed macroscopically. The removed As(V) to removed B5 salt ratio is close to 1.0 at low pH suggesting the formation of (B5)AsO4 salt. However, the IR spectra do not show bands which could be attributed to AsO4 vibrations probably due to low solid content (data not shown). The formation of precipitate is not as efficient as in the case of Cr(VI) thus, the immobilization of As(V) is lower. The calculated EDR shows that interaction energies in the system with As(V) are higher that in the corresponding Cr(VI) system (Table 2). This may be attributed to higher molar mass of the first oxyanion. It is worth to underline that precipitation of alkylchromate is accompanied with the lowest energy of 7.02 kJ/mol.

Fig. 3 Fig. 4 Fig. 5 Table 2 Table 3 Table 4

12

3.2.2. Kinetics of Cr(VI) and As(V) removal The kinetics of Cr(VI) and As(V) removal show similar trends (Fig. 6). For the MB1 and MTMA materials the immobilization takes place gradually. The WeberMorris plot (sorption vs t1/2) indicates a two-step removal process as two trend lines could be distinguished [54]. This may be connected to a relatively slower release rate of the B1 and TMA salts and their further interaction with oxyanions. In the case of MB5 material a rapid removal of both oxyanions is observed with equilibrium achieved already after 30 seconds. This is due to precipitation of alkyl salts as discussed above. In order to check the usability of the kinetic equation to model experimental data the following plots were prepared: log(qeq-qt) vs t (pseudo-first order) and t/qt vs t (pseudosecond order) [55, 56] (Fig. 6). The kinetics in all systems are best described using the pseudo-second order equation with R2 close to 1.0 (Fig. 6 and Table 5). The pseudo-first order equation is not sufficient to model the experimental data (R2 << 1.0), therefore the parameters of the equation are not given.

Fig. 6 Table 5

4. Conclusions The kaolinite intercalation compounds with selected ammonium salts were obtained using methoxy-kaolinite as a starting derivative. The PXRD and IR studies indicated a monolayer arrangement of the benzyltrimethylammonium (B1) and tetramethylammonium (TMA) chloride molecules in the mineral interlayer space, while

13

a tilted arrangement was noticed for the benzyldimethylhexadecylammonium chloride (B5). The interaction of aqueous solutions with intercalation compounds lead to a deintercalation of the introduced molecules and their further reaction with oxyanions as attested by the PXRD and FTIR studies. A significant immobilization efficiency was observed in all cases in comparison to very low sorption properties of pure kaolinite towards Cr(VI) and As(V). The most efficient uptake of oxyanions was noticed in the reaction with B5-intercalated material. This was connected with the precipitation of organic alkyl salts especially at low pH. In particular, the formation of alkylchromate was confirmed using FTIR spectroscopy. The relatively lower uptake of oxyanions by the B1- and TMA-intercalated materials resulted from the ion-exchange of chlorides, which initially compensated the positive nitrogen charge, by Cr(VI) and As(V). This could take place in the solution or on the kaolinite surface. The model studies showed that the sorption isotherms were best described using Langmuir equation. The kinetic studies revealed a two-step removal process in the reaction of Cr(VI) and As(V) with B1- and TMA-intercalated materials where a gradual uptake was noticed. In turn, the precipitation mechanism caused a rapid immobilization of oxyanions in reaction with B5-intercalated material. The kinetic experimental data were best modelled using pseudo-second order equation.

Acknowledgements We thank Danuta Mildner (Faculty of Pharmacy, Jagiellonian University, Krakow) for the CHNS analyses. This project was supported by the Polish National Science Centre under research project awarded by decision No. DEC2011/01/D/ST10/06814.

14

References

[1] H.H. Murray, Appl. Clay Sci., 17 (2000) 207-221. [2] H.H. Murray, Clay Miner., 34 (1999) 39-49. [3] S. Olejnik, L.A.G. Aylmore, A.M. Posner, J.P. Quirk, J. Phys. Chem., 72 (1968) 241-249. [4] É. Makó, J. Kristóf, E. Horváth, V. Vágvölgyi, J. Colloid Interface Sci., 330 (2009) 367-373. [5] F. Franco, M.D. Ruiz Cruz, Clay Miner., 39 (2004) 193-205. [6] D. Hirsemann, T.K.J. Köster, J. Wack, L. van Wüllen, J. Breu, J.r. Senker, Chem. Mater., 23 (2011) 3152-3158. [7] J.E.F.C. Gardolinski, G. Lagaly, Clay Miner., 40 (2005) 537-546. [8] B. Caglar, Ç. Çırak, A. Tabak, B. Afsin, E. Eren, J. Mol. Struct., 1032 (2013) 12-22. [9] E. Moretti, L. Storaro, G. Chessa, A. Talon, E. Callone, K.J. Mueller, F. Enrichi, M. Lenarda, J. Colloid Interface Sci., (2012). [10] J. Matusik, T. Bajda, J. Colloid Interface Sci., 398 (2013) 74-81. [11] G.K. Dedzo, C. Detellier, The Analyst, 138 (2013) 767-770. [12] G.K. Dedzo, S. Letaief, C. Detellier, J. Mater. Chem., 22 (2012) 20593-20601. [13] E.H. de Faria, E.J. Nassar, K.J. Ciuffi, M.A. Vicente, R. Trujillano, V. Rives, P.S. Calefi, ACS Appl. Mater. Interfaces, 3 (2011) 1311-1318. [14] F.R. Araújo, J.G. Baptista, L. Marçal, K.J. Ciuffi, E.J. Nassar, P.S. Calefi, M.A. Vicente, R. Trujillano, V. Rives, A. Gil, S. Korili, E.H. de Faria, Catal. Today, (2013). [15] E.H. de Faria, G.P. Ricci, L. Marçal, E.J. Nassar, M.A. Vicente, R. Trujillano, A. Gil, S.A. Korili, K.J. Ciuffi, P.S. Calefi, Catal. Today, 187 (2012) 135-149.

15

[16] J. Matusik, E. Stodolak, K. Bahranowski, Appl. Clay Sci., 51 (2011) 102-109. [17] M.H.A. Rehim, A.M. Youssef, H.A. Essawy, Mater. Chem. Phys., 119 (2010) 546552. [18] R. Saha, R. Nandi, B. Saha, J. Coord. Chem., 64 (2011) 1782-1806. [19] B.J. Alloway, Heavy Metals in Soils: Trace Metals and Metalloids in Soils and their Bioavailability, Springer, 2013. [20] Y. Kiso, Y.-J. Jung, T. Yamada, M. Nagai, K.S. Min, Water Sci. Technol., 5 (2005) 75-81. [21] D. Mohan, C.U. Pittman Jr, J. Hazard. Mater., 137 (2006) 762-811. [22] S. Jessen, F. Larsen, C.B. Koch, E. Arvin, Environ. Sci. Technol., 39 (2005) 80458051. [23] W. Mozgawa, M. Król, T. Bajda, J. Mol. Struct., 993 (2011) 109-114. [24] S. Shevade, R.G. Ford, Water Res., 38 (2004) 3197-3204. [25] G.M. Haggerty, R.S. Bowman, Environ. Sci. Technol., 28 (1994) 452-458. [26] M. Majdan, S. Pikus, Z. Rzączyńska, M. Iwan, O. Maryuk, R. Kwiatkowski, H. Skrzypek, J. Mol. Struct., 791 (2006) 53-60. [27] J. Matusik, Z. Kłapyta, Appl. Clay Sci., 83-84 (2013) 433-440. [28] J. Matusik, Z. Kłapyta, Z. Olejniczak, Appl. Clay Sci., 83-84 (2013) 426-432. [29] J. Matusik, A. Gaweł, E. Bielańska, W. Osuch, K. Bahranowski, Clays Clay Miner., 57 (2009) 452-464. [30] J. Matusik, E. Wisła-Walsh, A. Gaweł, E. Bielańska, K. Bahranowski, Clays Clay Miner., 59 (2011) 116-135. [31] APHA, Standard methods for the examination of water and wastewater, 18th ed., American Public Health Association, Washington D.C., 1992.

16

[32] A. Michalková, D. Tunega, L. Turi Nagy, J. Mol. Struct. Theochem, 581 (2002) 37-49. [33] E. Scholtzová, L. Smrčok, Clays Clay Miner., 57 (2009) 54-71. [34] J. Matusik, E. Scholtzová, D. Tunega, Clays Clay Miner., 60 (2012) 227-239. [35] Y. Kuroda, K. Ito, K. Itabashi, K. Kuroda, Langmuir, 27 (2011) 2028-2035. [36] J.J. Tunney, C. Detellier, J. Mater. Chem., 6 (1996) 1679-1685. [37] P.V. Brady, R.T. Cygan, K.L. Nagy, J. Colloid Interface Sci., 183 (1996) 356-364. [38] B.K. Schroth, G. Sposito, Clays Clay Miner., 45 (1997) 85-91. [39] Z. Zhou, W.D. Gunter, Clays Clay Miner., 40 (1992) 365-368. [40] S. Debnath, U.C. Ghosh, J. Chem. Thermodyn., 40 (2008) 67-77. [41] Z. Li, R.S. Bowman, Water Res., 35 (2001) 3771-3776. [42] S. Deng, R. Bai, Water Res., 38 (2004) 2424-2432. [43] F. Gode, E. Pehlivan, J. Hazard. Mater., 119 (2005) 175-182. [44] L. Khezami, R. Capart, J. Hazard. Mater., 123 (2005) 223-231. [45] Z. Hu, L. Lei, Y. Li, Y. Ni, Sep. Purif. Technol., 31 (2003) 13-18. [46] N.K. Lazaridis, D.D. Asouhidou, Water Res., 37 (2003) 2875-2882. [47] C.P. Huang, P. Fu, Journal of the Water Pollution Control Federation, 56 (1984) 232-242. [48] H. Genç-Fuhrman, J.C. Tjell, D. McConchie, J. Colloid Interface Sci., 271 (2004) 313-320. [49] P.K. Dutta, A.K. Ray, V.K. Sharma, F.J. Millero, J. Colloid Interface Sci., 278 (2004) 270-275. [50] B.J. Lafferty, R.H. Loeppert, Environ. Sci. Technol., 39 (2005) 2120-2127.

17

[51] O.M. Vatutsina, V.S. Soldatov, V.I. Sokolova, J. Johann, M. Bissen, A. Weissenbacher, React. Funct. Polym., 67 (2007) 184-201. [52] M. Majdan, O. Maryuk, S. Pikus, E. Olszewska, R. Kwiatkowski, H. Skrzypek, J. Mol. Struct., 740 (2005) 203-211. [53] H. Stammreich, D. Bassi, O. Sala, H. Siebert, Spectrochim. Acta, 13 (1958) 192196. [54] K.V. Kumar, V. Ramamurthi, S. Sivanesan, J. Colloid Interface Sci., 284 (2005) 14-21. [55] Y.S. Ho, G. McKay, Process Biochem., 34 (1999) 451-465. [56] S. Azizian, J. Colloid Interface Sci., 276 (2004) 47-52.

Figure captions

Fig. 1. PXRD pattern of: M - pure kaolinite, MDS - M sample intercalated with DMSO, MMD - methoxy-kaolinite, MB1, MTMA, and MB5 - intercalation compounds with B1, TMA, and B5 salts, respectively, MB1-Cr, MTMA-Cr, and MB5-Cr - intercalates after reaction with Cr(VI).

Fig. 2. IR spectra of: M - pure kaolinite, MDS - M sample intercalated with DMSO, MMD - methoxy-kaolinite, MB1, MTMA, and MB5 - intercalation compounds with B1, TMA, and B5 salts, respectively.

18

Fig. 3. Sorption isotherms of: (a) M, MB1, and MTMA samples in reaction with Cr(VI), (b) M, MB1, and MTMA samples in reaction with As(V), (c) MB5 sample in reaction with Cr(VI) and As(V). The dashed lines show calculated Langmuir isotherms.

Fig. 4. The effect of pH on the sorption capacity of MB1, MTMA, and MB5 materials in reaction with Cr(VI) (lower graph) and As(V) (upper graph).

Fig. 5. IR spectra of: B5-Cr - alkylchromate salt formed by reaction of Cr(VI) with B5, MB5 - B5 intercalation compound, and MB5-Cr - MB5 after reaction with Cr(VI).

Fig. 6. Kinetics of MB1, MTMA, and MB5 samples reaction with (a) Cr(VI) and (b) As(V). Lower graph - Webber-Morris plot (sorption vs t1/2). Upper graph: pseudo-first order (log(qeq-qt) vs t)) and pseudo-second order (t/qt vs t) plots.

Fig. S1. The dimensions of B1, TMA, and B5 molecules.

Table captions

Table 1. Sorption equilibrium and kinetic theoretical models used for calculations.

Table 2. Model coefficients calculated from Freundlich, Langmuir and DubininRadushkevich equations.

Table 3. Comparison of sorption capacity for different types of sorbents.

19

Table 4. Summary of Cr(VI), As(V), B1, B5, and Cl- concentration changes versus initial pH.

Table 5. Pseudo-first and pseudo-second order kinetic parameters.

20

21

22

23

24

25

26

Graphical abstract

27

Highlights The kaolinite intercalation compounds with selected ammonium salts were obtained.

The XRD and IR indicated a monolayer or tilted arrangement in the interlayer space.

The reaction with aqueous Cr(VI) and As(V) lead to deintercalation of molecules.

A significant improvement of oxyanions immobilization was observed in all cases.

The removal was due to precipitation of oxyanions alkyl salts and/or ion-exchange.

28

Table 1. Sorption equilibrium and kinetic theoretical models used for calculations. Sorption equilibrium equations Freundlich

qeq = KFCeq1/n

Langmuir

qeq = (KLqmCeq)/(1+KLCeq)

DubininRadushkevich

qeq = qDR · exp(-KDRε2)

qeq - sorption capacity at equilibrium (mmol/kg), Ceq - equilibrium concentration (mmol/L) KF - Freundlich adsorption capacity (mmol/kg), n - Freundlich dimensionless constant KL - Langmuir adsorption constant (L/mmol), qm - maximum adsorption capacity (mmol/kg) qDR - Dubinin-Radushkevich adsorption capacity (mmol/kg), KDR - DubininRadushkevich adsorption constant (mmol2/J2), ε - Polanyi potential: ε = RTln(1+1/Ceq) Energy of adsorption: EDR = (-2KDR)-1/2

Sorption kinetic equations Pseudo-first order

log(qeq-qt) = logqeq (k1t)/2.303

Pseudo-second order

t/qt = 1/(k2qeq2) + t/qeq

qeq and qt - sorption capacity at equilibrium and at time t (mmol/kg), respectively, k1 (min-1) and k2 (kg/mmol·min) - first order and second order rate constants, respectively.

29

Table 2. Model coefficients calculated from Freundlich, Langmuir and Dubinin-Radushkevich equations. Freundlich Anion Sample

Langmuir

q K K R2 R2 (mmol/kg) (L/mmol) (mmol/kg) M 0.453 1.98 0.95 1.46 1.66 0.99 MB1 0.702 12.39 0.99 81.30 0.27 0.79 Cr(VI) MTMA 0.672 4.31 0.96 22.37 0.40 0.93 MB5 0.50 0.31 M 0.579 1.31 0.92 4.22 0.91 0.98 MB1 0.375 15.10 0.79 76.34 1.13 0.99 As(V) MTMA 0.216 18.02 0.97 25.32 5.81 0.99 MB5 0.544 98.92 0.86 322.58 1.11 0.97 Explanation of parameters is given in Table 1 1/n

DubininRadushkevich EDR R2 (kJ/mol) 2.44 0.94 8.14 0.99 8.08 0.98 7.02 0.99 2.23 0.86 10.98 0.84 10.85 0.99 10.00 0.98

30

Table 3. Comparison of sorption capacity for different types of sorbents. Anion

Cr(VI)

As(V)

Sorbent Polyacrylonitrile fibers (APANFs) Lewatit M610 Anion exchange resin Activated carbon Acticarbone CXV Activated carbon GA-3 Hydrotalcite Mg-Al-CO3 Activated carbon Draco Activated Bauxsol TiO2 Hombikat UV 1000 Ferrihydrite Hybrid polymer/ inorganic fibrous sorbent (FIBAN-As)

Sorption capacity (mmol/kg)

Ref.

307.7

[42]

425.0

[43]

623.1

[44]

1950.0

[45]

2307.7

[46]

50.0

[47]

101.9

[48]

299.6

[49]

916.7

[50]

1088.8

[51]

31

Table 4. Summary of Cr(VI), As(V), B1, B5, and Cl- concentration changes versus initial pH. Anion

Cr(VI)

As(V)

Anion Salt Salt Ceq Salt Cin1 Cl- Cin1 removed removed (mmol/L) (mmol/L) (mmol/L) (mmol/L) (mmol/L) 3 0.89 24.07 24.13 23.12 5 0.58 24.07 23.55 23.12 7 0.63 24.07 23.65 23.12 MB1 9 0.55 24.07 24.13 23.12 11 0.59 24.07 24.92 23.12 3 5.71 12.54 2.01 10.53 12.50 5 5.28 12.54 2.15 10.39 12.50 MB5 7 5.37 12.54 4.10 8.44 12.50 9 5.21 12.54 5.99 6.55 12.50 11 5.22 12.54 7.58 4.96 12.50 3 0.58 24.07 25.23 23.12 5 0.40 24.07 24.75 23.12 7 0.38 24.07 23.81 23.12 MB1 9 0.47 24.07 24.44 23.12 11 0.19 24.07 23.81 23.12 3 1.43 12.54 11.32 1.22 12.50 5 1.98 12.54 9.45 3.09 12.50 MB5 7 2.53 12.54 8.44 4.10 12.50 9 2.48 12.54 9.45 3.09 12.50 11 2.25 12.54 10.32 2.22 12.50 pHin - initial pH, Cin and Ceq - initial and equilibrium concentrations 1 amount of salt or Cl- released from the material during water washing

Material

pHin

Cl- Ceq (mmol/L) 22.22 22.12 20.14 22.62 21.62 11.92 11.99 12.11 12.62 12.49 21.13 21.13 22.62 22.62 22.12 12.14 12.09 11.98 11.52 12.46

32

Table 5. Pseudo-first and pseudo-second order kinetic parameters. Pseudo-first Pseudo-second order order R2 k2 qeq R2 MB1 0.84 0.042 13.48 0.97 Cr(VI) MTMA 0.49 0.108 13.64 0.99 MB5 0.15 0.267 250.00 0.99 MB1 0.01 0.311 27.03 0.97 As(V) MTMA 0.05 0.036 32.47 0.98 MB5 0.32 0.044 104.17 0.99 Explanation of parameters is given in Table 1 Anion

Material

33