K2xSn4–xS8–x (x = 0.65–1): a new metal sulfide for rapid and selective removal of Cs+, Sr2+ and UO22+ ions

The synthesis and crystal structure of K2xSn4–xS8–x (x = 0.65–1, KTS-3) a material which exhibits excellent Cs+, Sr2+ and UO22+ ion exchange properties in varying conditions are reported.


Introduction
The pursuit of efficient, cheap, sustainable and growing sources of energy, involves nuclear energy which has emerged as one of the prominent alternatives in many countries and accounted for 12.3% of the world's electricity production in 2012. 1 Over the last four decades the accumulation of radioactive spent nuclear fuel (nuclear waste) has reached a staggering volume of 71 780 metric tons and it is increasing by 2300 metric tons every year. 2 The rapidly increasing number of nuclear power plants will generate even larger amounts of nuclear waste. The prime source of nuclear fuel is various uranium salts, which are being used in different stages from mining, nuclear fabrication and processing. The uranium salts form a major component of the nuclear waste along with the ssion generated non-actinide isotopes. The estimated amount of uranium present in seawater is 4 Â 10 12 kg (at $3 ppb), so potentially it could supply nuclear fuel for thousands of years. 3 The primary issue with isolation of uranium from seawater in a cost effective manner is the presence of other ions (Na + , Cl À , Mg 2+ , SO 4 2À , Ca 2+ , and CO 3 2À ) in predominant amounts. 90 Sr and 137 Cs are the main hazardous ssion generated non-actinide isotopes present in nuclear waste, as they produce gamma and high energy beta particles. 4 90 Sr (with a half-time of t 1/2 $ 29 years) and 137 Cs (t 1/2 $ 30 years) pose a major long-term risk due to their long half-life. Recently, the tsunami-induced disaster at the Fukushima nuclear power plant in 2011 resulted in contamination of a wide region of the northern Kanto and Tohoku areas in Japan with radionuclides, 131 I, 134 Cs, 137 Cs, and 90 Sr. [5][6][7] Therefore, nuclear waste needs to be dealt with effectively, for safe storage and disposal due to its adverse health effects in humans and the environment.
The most commonly used technique for the separation of radioactive elements from industrially produced nuclear waste is solvent extraction using liquid phase organic compounds. [8][9][10] The use of ion exchange media is another alternative for the removal of radionuclides from the nuclear waste, [11][12][13][14][15][16][17][18][19][20][21][22] however, they are relatively less explored due to certain drawbacks: the organic ion exchange materials are efficient but costly, whereas the inorganic ion exchange materials are cheaper but they are less efficient because of low selectivity for the ions of interest. So, there is a growing need to develop efficient inorganic ion exchange materials for radioactive species.
Over the past decade or so metal suldes have emerged as a selective class of ion exchangers for capturing so metal ions such as Hg, Cd, Ag etc. 20,23,24 Chalcogenide open-framework compounds, such as K 6 Sn[Zn 4 Sn 4 S 17 ] 17 and (NH 4 ) 4 In 12 Se 20 (ref. 19) present unique advantages over their oxide analogues. The layered thiostannates are particularly interesting because they exhibit open accessible structures where ion-exchange chemistry can occur readily. [25][26][27][28][29][30] In previous work, we proposed that layered metal suldes K 2x M x Sn 3Àx S 6 (M ¼ Mn, KMS-1; M ¼ Mg, KMS-2) can be used for facile ion exchange of Sr 2+ , Cs + and UO 2 2+ . 18,19,31,32 A variety of synthetic parameters were explored in the search for new compounds based on a tin sulde layer structure to modulate the ion exchange properties. The advantage of the chalcogenide materials stems from the fact that they are based on soer chalcogen ligands (in the Lewis base sense) which can induce high selectivity for heavy metal ions, Cs + , Sr 2+ , UO 2 2+ against co-present hard ions such as Na + , Al 3+ and Ca 2+ . 17,19,[31][32][33] Herein, we report a new ternary layered compound, K 2x -Sn 4Àx S 8Àx (x ¼ 0.65-1, KTS-3) and its promising selectivity for removing Cs + , Sr 2+ and UO 2 2+ species via ion exchange processes. Specically, we nd that KTS-3 exhibits high distribution coefficients (K d ) for the capture of Cs + (5.5 Â 10 4 ), Sr 2+ (3.9 Â 10 5 ) and UO 2 2+ (2.7 Â 10 4 ) over a broad pH range (V/m $ 1000 mL g À1 ). We nd that KTS-3 remains highly effective for these ions even in presence of a large amount of Na + ions.
Hydrothermal synthesis of K 2x Sn 4Àx S 8Àx (x ¼ 0.65-1, KTS-3) K 2 CO 3 (6 mmol, 0.830 g), elemental Sn (9 mmol, 1.068 g), S (30 mmol, 0.962 g) were taken in a 23 mL polytetrauoroethylene (PTFE) lined stainless steel autoclave and deionized water (0.5 mL) was added drop wise until the mixture acquired dough-like consistency. The autoclave was sealed properly and maintained in a preheated oven at 220 C for 15 h under autogenous pressure. Then, the autoclave was allowed to cool to room temperature. The product was found to contain yellow rod shaped crystals along with yellow polycrystalline powder (Fig. 1a). The product was isolated by ltration, washed several times with water, acetone and dried under vacuum. The yield was $2.0 g ($85%, based on Sn) and the product was air and moisture stable. Electron Dispersive Spectroscopy (EDS) analysis shows the presence of K, Sn and S and gave an average formula "K 1.34 Sn 3.26 S 7.32 ". given by the equation: HCl or NaOH solution to $6 ppm. The ion exchange experiment at different Na + concentration was done by dissolving the required amount of NaCl in 10 mL solution of A n+ ion ($6 ppm).
The exchange experiments were performed with V/m z 1000 mL g À1 , at room temperature and 15 h contact. Competitive ion exchange (Cs + and Sr 2+ ) experiments of KTS-3 were also carried out with a V/m ratio of 1000 mL g À1 , at room temperature with 15 h of contact time. The initial concentration was approximately $6 ppm for both the ions. The competitive ion exchange experiments were similar to those of the individual ion exchange experiments except they contained both Cs + and Sr 2+ ions in solution.
The kinetic studies of the adsorption of ions by KTS-3 were carried out as follows: ion-exchange experiments of various reaction times (5, 15, 30, 60, 120, 300 and 1200 min) were performed. For each experiment, 10 mg of KTS-3 was weighed into a 20 mL vial. A 10 mL sample of water solution containing $1 ppm of Cs + /Sr 2+ /UO 2 2+ was added to each vial, and the mixtures were kept under magnetic stirring (pH $ 7). The suspensions were ltered aer the designated reaction time and the ltrates were analyzed by inductively coupled plasma-mass spectroscopy (ICP-MS).
operating at 40 kV and 20 mA. The samples were prepared by grinding and spreading over a glass slide.

Single-crystal X-ray crystallography
A suitable single crystal was carefully selected under a polarizing microscope and glued to a thin glass ber. Single crystal data were collected on a STOE IPDS II diffractometer using Mo Ka radiation (l ¼ 0.71073Å) at room temperature. The generator was operated at 50 kV and 40 mA. The data were collected with a u scan width of 1 keeping the crystal to detector distance xed at 8.0 cm. Integration and numerical absorption corrections were performed using X-AREA, X-RED, and X-SHAPE. 34 The structure was solved using direct methods and rened by the SHELXTL program package 35 using a full-matrix least squares renement against the square of structure factors. Final structure renement included atomic positions and anisotropic thermal parameters for all Sn and S atoms. The thermal displacement parameters of the disordered K atoms was rened isotropically. Details of the structure solution and nal renements for the compound are given in Table 1.

Scanning electron microscopy and energy dispersive spectroscopy
The energy dispersive spectroscopy (EDS) was performed with a Hitachi S-3400N-II scanning electron microscope (SEM) equipped with an ESED II detector. An accelerating voltage of 20 kV and 60 seconds acquisition time were used for elemental analysis.

Thermogravimetric analysis
The thermogravimetric analysis (TG) was performed with a Shimadzu TGA-50 system under nitrogen atmosphere in an aluminum crucible. The analysis was performed with a heating rate of 10 C min À1 and a nitrogen ow rate of 40 mL min À1 from room temperature to 600 C.

Differential thermal analysis
The differential thermal analyses (DTA) were performed on a Shimadzu DTA-50 thermal analyzer. For a typical analysis, around 30 mg of sample was sealed in a quartz ampoule and sealed under vacuum, another sealed quartz ampoule with Al 2 O 3 was used as reference material. The analysis was performed with a heating rate of 2 C min À1 and a nitrogen ow rate of 30 mL min À1 from room temperature to 600 C.

Infrared (IR) and Raman spectroscopy
Infrared spectra of compounds were collected on a Bruker Tensor 37 FTIR (MID IR/ATR) using an attenuated total reectance attachment in the range 4000-600 cm À1 . The Raman spectra of the ground samples were collected on a DeltaNu Raman system that uses a 785 nm constant wavelength laser. The spectra were collected in the range of 100-2000 cm À1 with the sample inside a 0.5 mm capillary tube.

Band gap measurements
The UV-vis/near-IR diffuse reectance spectra of the ground samples were collected using a Shimadzu UV03010 PC double beam, double monochromator spectrophotometer in the wavelength range of 200-2500 nm. BaSO 4 powder was used as a reference and base material on which the powder sample was coated. Using the Kubelka-Munk 36 equation the reectance data were converted to absorption data and the band edge of the sample was calculated from the intercept of the line extrapolated from the high energy end to the baseline.

X-ray photoelectron spectroscopy (XPS) analysis
XPS of the KTS-3 and exchanged materials were performed on ground powders using a Thermo Scientic ESCALAB 250 Xi spectrometer equipped with a monochromatic Al Ka X-ray source (1486.6 eV) operating at 300 W. Samples were analyzed under vacuum (P < 10 À8 mbar) with a pass energy of 150 eV (survey scans) and 25 eV (high-resolution scans). A low-energy electron ood gun was employed for charge neutralization. Ion beam etching was performed to clean off some of the surface contamination. Prior to the XPS measurements, the crystalline powders were pressed on copper foil and mounted on stubs and successively put into the entry-load chamber to pump. All peaks were referenced to the signature C1s peak binding energy at 284.6 eV for adventitious carbon. Avantage soware was used to t the experimental peaks.

Inductively coupled plasma-mass spectroscopy
The Cs + , Sr 2+ , and UO 2 2+ ion exchange samples and the competitive ion exchange samples (Cs 2+ and Sr 2+ ) were analyzed with Inductively Coupled Plasma-Mass Spectroscopy (ICP-MS) using a computer-controlled ThermoFisher X Series II Inductively Coupled Plasma Mass Spectrometer with a quadruple setup equipped with Collision Cell Technology. Eleven standards were prepared in the range of 0.78-800 ppb by diluting commercial solutions (Sigma-Aldrich). The ion exchange samples were diluted to lower the concentrations below 800 ppb. All the samples and standards were prepared in a 5% (nitric acid + hydrochloric acid) solution with 1 ppb (Bi, Ho, In, Li, Tb, Y) internal standard in order to correct the instrumental dri and matrix effects during analysis.

Synthesis and characterization
The synthesis of K 2x Sn 4Àx S 8Àx (x ¼ 0.65-1, KTS-3) was accomplished by a hydrothermal method at 220 C. The product was found to contain a large amount of yellow powder along with few rod shaped yellow crystals. The powder X-ray diffraction of the samples of KTS-3 showed that the yellow powder and the crystals are the same material and conrmed the phase purity ( Fig. 1b) when compared against the calculated pattern obtained by the single crystal model. The product was also analyzed with semi-quantitative SEM-EDS ( Fig. 1c and d) S1a †). 37,38 The 247 cm À1 band may arise from collective 3.868 and À1.905 eÅ À3 4.342 and À2.783 eÅ À3 Weighting scheme lattice modes or from the vibrations associated with the potassium ions.
Thermogravimetric (TG) analysis of the KTS-3 compound was carried out in owing nitrogen gas (ow rate ¼ 20 mL min À1 ) in the temperature range 20-600 C (heating rate of 10 C min À1 ). The TG studies indicate that KTS-3 exhibits a singlestep weight loss of $10% up to 235 C which corresponds to the loss of adsorbed water molecules. The compound remains stable up to 525 C, aer which it starts to decompose (Fig. S1b †) into K 2 Sn 2 S 5 and SnS 2 as determined by powder XRD. Differential thermal analysis (DTA) of the samples shows no sign of melting up to 600 C (Fig. S2 †).
X-ray photoelectron spectroscopy performed on KTS-3 ( Fig. 2) shows peaks at 292.5 and 295.6 eV which are characteristic for 2p 3/2 and 2p 1/2 of K + cations. 39 The peaks at 486.0 and 494.5 eV are consistent with the 3d 5/2 and 3d 3/2 levels observed for Sn 4+ cations. 39 The sulfur 2p orbital excitations appear as a broad peak in the range 158-165 eV. The deconvolution of the broad band gives two bands centered at 161.5 and 162.7 eV which are characteristic of 2p 3/2 and 2p 1/2 sulde anions, respectively. 39,40 Crystal structure The structure of K 1.92 Sn 3.04 S 7.04 is composed of innite layers of [Sn 3 S 7 ] stacked along the b-axis with K ions residing between the layers, Fig. 3a Potassium atoms are disordered and sandwiched between the [Sn 3 S 7 ] layers. An apparent C-centered orthorhombic cell with a ¼ 3.6831(2)Å, b ¼ 25.8877(19)Å, and c ¼ 16.8155(11)Å can index most of the reections but aer careful examination of the reciprocal lattice (Fig. 3b) we found the presence of additional broad and diffuse reections that could be indexed by doubling of the short a-axis with a transformed primitive monoclinic unit cell of a ¼ 13.092(3)Å, b ¼ 16.882(2)Å, c ¼ 7.375(1)Å and b ¼ 98.10(1) . The origin of the supercell is due to partial long range ordering of vacancies in the [SnS 4 ] slabs where every other [Sn 2 S 6 ] unit is missing along the c-axis, Fig. 3c.
The orthorhombic cell can be rened in Cmcm with a stoichiometry of 'K 2 Sn 4 S 8 ' but this is problematic because this composition does not charge balance assuming K + , Sn 4+ , and S 2À ions. Furthermore, the agreement factor for the 'K 2 Sn 4 S 8 ' renement was very high at $14.5% with large negative residual electron density around the Sn(2) and S(4) sites. Upon renement of the occupancy of Sn(2) and S(4) (50% disorderly occupied) but omitting the supercell reections, the agreement factor improved signicantly (7.5%, see Table 1) and the rened composition becomes K 1.92 Sn 3.04 S 7.04 which is charge balanced. By subsequently introducing the intensity of the supercell reections into the renement, an additional long range ordering of the vacancies in the Sn(2) and S(4) sites was found. The supercell of KTS-3 was solved using the monoclinic spacegroup P2 1 /c and twining was required for a successful renement. A rened twin fraction of 45.0(3)% was determined using a twin law of 180 degrees rotation along the c-axis, Table 1. The nal agreement factor is satisfactory given the very broad and diffuse nature of the supercell reections, Fig. 3b. The asymmetric unit of the KTS-3 supercell has 15 atoms. Four crystallographically independent Sn 4+ atoms (two sites are partially occupied), eight sulde atoms (two sites are partially occupied) and three K + ions (two sites are partially disordered). The Sn(1) and Sn (2)   occupancy values were used for the S atoms that edgeshare the [SnS 4 ] tetrahedra, i.e., the occupancy factor of S(7) and S(8) was constrained at 73.0(4) and 31.1(5)%, respectively. All K atoms have relatively large thermal factors which is characteristic for loosely bound intercalated atoms found in ionexchanged materials. 17,18,32 K(1A) and K(1B) are delocalized with an average disordered distance of 2.32(1)Å and fractional occupancy of 60.7(5) and 39.3(5)%, respectively where K(3) fully occupies its own site.
The basic difference between the structure of KTS-3 and that of so-called KMS structures which are also layered (K 2x M x -Sn 3Àx S 6

Ion-exchange of KTS-3 with Cs + , Sr 2+ and UO 2 2+ ions
The interlayer potassium ions in the KTS-3 structure are disordered and move rapidly in an ion-exchange process. To check the feasibility of ion exchange of K + in KTS-3 we immersed it in a solution of A n+ (A n+ ¼ Cs + , Sr 2+ and UO 2 2+ ) ions for 15 h. These ion exchange processes are in fact very rapid and almost all the ions were exchanged within 5 min, but to ensure a complete ion exchange we used 15 h. The EDS analysis of the materials aer ion exchange showed the complete removal of the K + ions. The EDS of the exchanged materials showed a ratio of 1.5 : 3 for Cs : Sn, 0.7 : 3 for Sr : Sn and 0.51 : 3 U : Sn, which are comparable with the expected Cs to Sn ratio (1.3-2.0) and Sr, UO 2 to Sn ratio (0.65-1) (Fig. S3 †). The PXRD of the exchanged materials showed isotactic ion exchange with retention of the parent structure (Fig. 4). The ion exchange processes can be described by the following equations:    which is consistent with the ionic size of the ions. The TG analysis (Fig. S4 †) showed that the degree of hydration for the exchanged materials follows the order Sr 2+ > UO 2 2+ > Cs + > K + .
The band gap of the pristine KTS-3 material is 2.38 eV and the yellow color of the material changes marginally upon exchange with Cs + and Sr 2+ ions. The exchanged materials show a small increase in absorption and the measured band gaps were 2.54 eV (Cs + ) and 2.56 eV (Sr 2+ ). With UO 2 2+ exchange, the yellow color slowly changes to a darker orange color and the band gaps red shied to 2.30 eV and 2.40 eV (Fig. 5a). This can be attributed to partial dehydration of the UO 2 2+ ions and the presence of U/S interactions. The presence of two band gaps for the UO 2 2+ exchanged material was attributed to the differently hydrated UO 2 2+ ions.
The infra-red spectrum of the uranyl exchanged KTS-3 material shows a strong peak at $910 cm À1 , which is not found in pristine KTS-3 (Fig. 5b) 41 The XPS spectra of the Cs + exchanged samples show the characteristic 3d 5/2 and 3d 3/2 for Cs + at 724.7 and 738.7 eV (Fig. 6a). 39 The Sr 2+ exchanged samples show peaks at 133.9 and 135.7 eV characteristic for 3d 5/2 and 3d 3/2 of Sr 2+ cations (Fig. 6b). 39 The UO 2 2+ exchanged samples show two peaks at 379.6 and 390.6 eV characteristic for 3f 7/2 and 3f 5/2 of U 6+ centers (Fig. 6c). 31,39 All exchanged samples showed the characteristic peaks for tin and sulfur ions as observed for the pristine compound. The peaks for the potassium 2p 3/2 and 2p 1/2 could not be found in the exchanged samples (Fig. 6d), which conrms their complete exchange from the KTS-3 compound.
Ion exchange adsorption isotherm studies (Cs + , Sr 2+ and UO 2 2+ ) In order to understand the ion exchange capacity of KTS-3 a detailed adsorption study was carried out. The ion exchange equilibrium, kinetics, effect of salt concentration and pH on the Cs + , Sr 2+ and UO 2 2+ ion exchange were studied. The equilibrium data for the ions were modeled using the Langmuir and Langmuir-Freundlich adsorption isotherms. 42 Table 2 shows the equilibrium constants and different parameters obtained by the modeling of the equilibrium data. Langmuir isotherm where q (mg g À1 ) is the amount of cation adsorbed at equilibrium concentration, q m is the maximum cation adsorption capacity, b (L mg À1 ) is the Langmuir constant, C e (ppm) is the equilibrium concentration and n is a constant. The Langmuir isotherm describes adsorption on a homogenous surface and the maximum adsorption corresponds to a saturated monolayer. This model is based on the assumptions that (a) the adsorption sites are equivalent and each site can only accommodate one molecule, (b) the energy of adsorption is constant and independent of surface coverage, and (c) there is no transmigration of adsorbate from one site to another. [42][43][44] The Langmuir-Freundlich isotherm is an extension of the Langmuir model, which reduces to Freundlich isotherms at low surface coverage and to Langmuir isotherms at high surface coverage. 42 The equilibrium data for Cs + ion exchange (Fig. 7a) could be tted with both Langmuir, and Langmuir-Freundlich isotherm models with a good agreement (R 2 $ 0.97). The value of the Langmuir-Freundlich constant n ¼ 1.37 (23) was found to be closer to 1 which suggests that the adsorption behavior of Cs + ion exchange follows the Langmuir adsorption model. The agreement of the Langmuir adsorption isotherm with the Cs + ion exchange can be rationalized by taking into consideration the structural features of KTS-3. The [Sn 3 S 7 ] 2À layers of KTS-3 are separated by layers of disordered potassium ions, so the exchangeable Cs + ions form a layer between the [Sn 3 S 7 ] 2À layers that corresponds to the monolayer of Langmuir isotherms. The adsorption sites for the exchangeable ions are xed (S 2À ions) and chemically equivalent. Moreover, once the ions are exchanged it is not possible to migrate to other sites. The equilibrium data for Sr 2+ (Fig. 7b) and UO 2 2+ (Fig. 7c) were also tted with Langmuir (R 2 ¼ 0.92 and 0.95, for Sr 2+ , UO 2 2+ respectively) and Langmuir-Freundlich adsorption (R 2 ¼ 0.92 and 0.96, for Sr 2+ , UO 2 2+ respectively) isotherms in good agreement. The value of Langmuir-Freundlich constant [n ¼ 1.81 (54) for Sr 2+ and 1.52 (24) for UO 2 2+ ] shows that it deviates from the Langmuir isotherm model (n ¼ 1). The behavior of Cs + , Sr 2+ and UO 2 2+ vis-a-vis their isotherms can be rationalized by the fact that the number of ions exchanged in the case of Cs + is twice that of bivalent Sr 2+ and UO 2 2+ and hence it has higher surface coverage and tends to follow better the Langmuir model. The ion exchange of bivalent metal ions oen follow the Langmuir-Freundlich model rather than the Langmuir model. 45 The maximum ion exchange capacities, q m were found to be 280(11) mg g À1 (2.10 mmol g À1 ) for Cs + , 102(5) mg g À1 (1.16 mmol g À1 ) for Sr 2+ and 287(15) mg g À1 (1.20 mmol g À1 ) for UO 2 2+ from the Langmuir isotherm model. The theoretical capacities for K 2x Sn 4Àx S 8Àx (x ¼ 0.96) considering all the K + ion are exchanged are 2.90 mmol g À1 (385 mg g À1 ) for Cs + and 1.45 mmol g À1 for Sr 2+ (127 mg g À1 ), UO 2 2+ (347 mg g À1 ). The observed Cs + exchange is about 72%, Sr 2+ exchange $ 80% and UO 2 2+ exchange $ 83% of the theoretical capacity. All K + ions are exchanged aer the reaction and the observed exchange capacity is due to the fact that the polycrystalline sample (K 2x -Sn 4Àx S 8Àx , KTS-3) has a range of x values from 0.65-1. The observed ion exchange capacity of KTS-3 compares well with well-known Cs + and Sr 2+ sorbents (e.g., zeolites, sodium silicotitanates and zirconium titanium silicates; 1.86-4.1 mmol g À1 for Cs + and 1.0-2.0 mmol g À1 of Sr 2+ ). [46][47][48][49] The Langmuir constants b (L mg À1 ) for the Cs + , Sr 2+ and UO 2 2+ were found to be 0.09(2), 0.20 (8) and 0.23(6) L mg À1 , respectively. The value of b is an indicator for the affinity towards a particular ion. Higher b values for Sr 2+ and UO 2 2+ ions indicate that KTS-3 has larger affinity towards them compared  to Cs + . The affinity of a sorbent towards a particular ion can also be expressed in terms of distribution coefficient (K d ), where, V is volume of testing solution (mL), m is the mass of the exchanger (g), C 0 and C f are the initial and nal concentration of the ion. The K d values were found to be 5.5 Â 10 4 mL g À1 for Cs + , 3.9 Â 10 5 mL g À1 for Sr 2+ and 2.7 Â 10 4 mL g À1 for UO 2 2+ ($6-8 ppm, V/m ¼ 1000 mL g À1 and pH ¼ 7). K d values in the 10 4 or 10 5 ranges are considered to be very good for ion exchange processes. 12,13,[50][51][52] Ion exchange of Cs + and Sr 2+ Generally, nuclear waste contains a large amount of other nonradioactive ions (Na + , K + , Ca 2+ ), also the waste solutions can be very corrosive depending on the pH. 53 Ion exchange experiments with KTS-3 were performed over a range of pH and salt concentrations aimed to simulate the conditions likely to be found in nuclear waste treatment.
The stability of the KTS-3 phase over a range of pH values (2-12) was tested and was found to be impressive. The compound remains crystalline (3 # pH # 11) and retains the layered structure for days when suspended in solution. Even in highly acidic (pH # 2) or basic conditions (pH $ 12) it remains stable for hours; a small decomposition of the compound can be seen if kept for more than 24 h (Fig. S5 †). Fig. 8 represents the variation of K d values for individual and competitive Cs + and Sr 2+ ion exchange with pH. KTS-3 shows excellent Cs + ion exchange capacity over a pH range of 2-12. It absorbs over 97% of the ions from pH 4 to 10 and it absorbs around 53% of the ions even in a highly acidic environment (pH 2). The K Cs d values were found to be $3.4 Â 10 4 to 5.5 Â 10 4 in the pH range of 4-10. However, there is slight decrease in the K Cs d values at pH ¼ 2 (1.1 Â 10 3 ) and it falls sharply at pH ¼ 12 (253) (7.4 ppm, V/m ¼ 1000). The decrease in K Cs d values may be due to partial decomposition of KTS-3 in these regions of pH. The presence of Sr 2+ in solution does not affect the ion exchange of Cs + , as the K Cs d values found are comparable with those of the individual values (9.8 Â 10 2 to 6.7 Â 10 4 ). The small increase in K Cs d in the presence of Sr 2+ can be attributed to the overall increase in ionic charge of the solution.
KTS-3 exhibits remarkable Sr 2+ capture capacity with more than 98% of the ions absorbed between pH 4 to 10. It decreases slightly at pH ¼ 2 (81%) and 12 (88%) but is still much higher than Cs + . The K Sr d values for the Sr 2+ ion exchange over the pH range 2-12 were found to be 4.2 Â 10 3 to 3.9 Â 10 5 mL g À1 (6.9 ppm, V/m ¼ 1000) (Fig. 8). The presence of Cs + does not induce an appreciable change as the K Sr d value remains almost same 4.5 Â 10 3 to 3.9 Â 10 5 mL g À1 .
The K d value for Cs + in the presence of a huge excess of Na + ions decreases slightly from 5.5 Â 10 4 mL g À1 at 0 M concentration to 4.4 Â 10 3 mL g À1 at 0.1 M concentration (Fig. 9a). Further increase in the Na + concentration reduced the K d values and even at a Na + concentration of 1 M, KTS-3 exhibited a reasonable K d value of 644 mL g À1 . The K d values in the presence of both Sr 2+ and Na + ions vary from 5.5 Â 10 4 mL g À1 at 0 M Na + to 501 mL g À1 at 1 M Na + concentration. The K d value for Sr 2+ in both individual and competitive (Cs + ) ion-exchange reactions drops sharply in the presence of Na + . Namely, it Fig. 7 Equilibrium data for (a) Cs + , (b) Sr 2+ and (c) UO 2 2+ ion exchange, the solid data represents the fitted lines by various isotherm models. The V/m ratio was 1000 mL g À1 , pH $ 7 and the contact time was $15 h. Fig. 8 Variation of distribution coefficient K d of individual and competitive Cs + and Sr 2+ ion exchange with increasing pH. The initial concentrations were 7.4 (Cs + ) and 6.9 (Sr 2+ ) ppm (both individual and competitive) and V/m ratio was 1000 mL g À1 . decreases from 3.9 Â 10 5 mL g À1 at 0 M to 201 at 1 M Na + concentration (Fig. 9a).
The kinetics of Cs + adsorption for low concentration ($1.2 ppm) solutions showed that 94% of the ions were absorbed within 5 min. The competitive Cs + adsorption (in the presence of Sr 2+ ) showed $90% adsorption within 5 min, which remains unchanged thereaer (Fig. 9b). The Sr 2+ adsorption 92% (individual) and 92% (competitive in presence of Cs + ) occurred within 5 min and with a longer time it increases to 97% (individual and competitive). The small decrease in ion exchange with time for Cs and increase for Sr 2+ are due to the dynamic ion exchange process between K + and Cs + /Sr 2+ and the higher affinity of KTS-3 towards Sr 2+ . Upon contact with KTS-3, the Cs + replaces the K + ions immediately and only a small amount of K + ions gets reabsorbed in the interlayer spaces to release some of initially absorbed Cs + ions back to solution. However, in the case of Sr 2+ the higher affinity of KTS-3 shuts down this dynamic ion exchange.

Ion exchange of UO 2 2+
KTS-3 exhibits the best UO 2 2+ adsorption near neutral pH, the K UO2 d value at pH 7 is 2.7 Â 10 4 mL g À1 . The UO 2 2+ adsorption capacity remains more or less the same between pH 4 to 8, however, it decreases at low (9.0 Â 10 2 mL g À1 at pH 2) and high (2.6 Â 10 3 mL g À1 at pH 12) pH values (Fig. 10a). The effect of Na + on K UO2 d is negligible; it decreases only slightly from 2.7 Â 10 4 mL g À1 at 0 M concentration to 4.8 Â 10 3 mL g À1 at 0.1 M concentration. Even at 1 M Na + concentration the K UO2 d is as high as 3.6 Â 10 3 mL g À1 (Fig. 10b). The kinetics of UO 2 2+ adsorption ($0.95 ppm) solution showed that the ion exchange is rapid and 80% of the ions were adsorbed within 5 min, which increases to 90% with time (Fig. 10c) (3.9 Â 10 5 ) and UO 2 2+ (2.7 Â 10 4 ) ion exchange (7.4 ppm, 6.9, 5.7 ppm Cs + , Sr 2+ and UO 2 2+ , respectively; V/m $ 1000 mL g À1 ). The ion exchange capacity of the material remains mostly unaffected between pH 4-10 and decreases slightly in higher acidic or basic environment. The kinetics of the ion exchange showed Fig. 9 (a) Variation of distribution coefficient K d of individual and competitive Cs + and Sr 2+ with increasing molar concentration of Na + (initial concentrations of the ions were 7.4 (Cs + ) and 6.9 (Sr 2+ ) ppm, V/m ratio was 1000 mL g À1 and pH $ 7), and (b) kinetics of individual Cs + , Sr 2+ and competitive ion-exchange vs. time t (min). The initial concentration was $1.2 ppm (for both Cs + and Sr 2+ ) and V/m ratio was 1000 mL g À1 and pH $ 7. Fig. 10 Variation of distribution coefficient K UO2 d (a) with pH, (b) increasing molar concentration of Na + . The initial concentration was 6 ppm, the V/m ratio was 1000 mL g À1 and (c) kinetics of individual UO 2 2+ ion-exchange vs. time (1 ppm, V/m ratio was 1000 mL g À1 and pH $ 7). that the process is very facile and it absorbs most of the ions within minutes.
The ion exchange capacity of K 2x Sn 4Àx S 8Àx (x ¼ 0.65-1, KTS-3) is excellent and compares well with K 2x M x Sn 3Àx S 6 (M ¼ Mn, KMS-1; M ¼ Mg, KMS-2). The Cs and UO 2 2+ ion exchange capacity of KTS-3 (q m ¼ 226 mg g À1 for Cs + and 382 mg g À1 for UO 2 2+ ) is comparable with KMS-1 (q m ¼ 280 mg g À1 for Cs + and 287 mg g À1 for UO 2 2+ ) and the Cs + ion exchange capacity is much lower than KMS-2 (q m ¼ 532 mg g À1 for Cs + ). However, KTS-3 (q m ¼ 102 mg g À1 for Sr 2+ ) outperforms both KMS-1 (q m ¼ 77 mg g À1 for Sr 2+ ) and KMS-2 (q m ¼ 87 mg g À1 for Sr 2+ ) in terms of Sr 2+ ion exchange capacity. Moreover, the relative ease and inexpensive synthesis of K 2x Sn 4Àx S 8Àx make it a promising material for future studies. Our work shows that the metal chalcogenide family can provide promising ion exchange materials for the selective removal of radionuclide from nuclear waste. Further work is to assess the utility of KTS-3 in remediation applications of nuclear wastes is justied.

Author contributions
DS and MGK designed and conducted the research. The structure was solved by CDM. KSS and MSI helped characterizing the exchanged compounds by TGA, IR, Raman, UV-Vis and XPS. The manuscript was written by DS, CDM and MGK. All authors have approved the nal version of the manuscript.