Exchange by Highly Charged Swelling Micas, Sodium Engelhard

ARTICLE pubs.acs.org/est

Mercury(II) Exchange by Highly Charged Swelling Micas, Sodium Engelhard Titanosilicate-4, and Sodium Titanosilicate Young Dong Noh and Sridhar Komarneni* Department of Crop and Soil Sciences and Materials Research Institute, The Pennsylvania State University, University Park, Pennsylvania 16802, United States

bS Supporting Information ABSTRACT: Selective Hg2+-exchange properties of highly charged sodium swelling micas (Na-2-, Na-3-, and Na-4-micas), sodium Engelhard titanosilicate-4 (Na-ETS4), and sodium titanosilicate were determined by use of distribution coefficients (Kd), ion-exchange isotherms, and Kielland plots for their potential use of Hg decontamination from groundwater and soils. X-ray diffraction (XRD) patterns after 2Na+ f Hg2+ exchange were collected to check for change in (001) spacings of differently charged sodium micas. The isotherms and Kielland plots suggested that Na-ETS-4 was highly selective for Hg2+. Also, the Kd value of Na-ETS-4 was the highest among the tested exchangers, supporting its high selectivity. Hg releases from Hg-exchanged Na-4-mica and Na-ETS-4 were found to be lower compared to other samples tested with simulated groundwater. The (001) spacings of sodium micas after Hg2+ exchange changed from ∼12 to ∼14 Å or/and 12 Å depending on their layer charge density and the uptake amount of Hg. Our results suggest that Na-ETS4 is a good candidate for mercury(II) decontamination from groundwater and soils.

’ INTRODUCTION Mercury is one of the major pollutants in the environment as it is released from fossil fuel burning, mining and smelting, and natural sources.1 It is not only easily transported in the form of vapor in the atmosphere but also transformed to organomercury, which is more toxic than nonorganomercury by 10 times.2 It has a severe impact on the environment because of its high toxicity and accumulation in organisms, even at trace levels. Therefore, in order to remediate mercury from contaminated sites and groundwater, various cleanup techniques have been developed including precipitation,3 electrokinetic treatment,4 ion exchange,5 bacteria accumulation,6 and adsorption on activated carbon.7 Among the several treatment technologies, ion exchange is one of the preferred processes due to its simplicity. Natural zeolites,8 clays,9 oxides,10 and synthetic ion exchangers11 were tested for Hg removal. The cation-exchange capacity (CEC) of ion exchangers and their selectivity for target ions are significant factors to consider in selecting ion exchangers. The synthetic ion exchangers described below have suitable structures, high cation-exchange capacities, selectivity for some pollutants, and compatibility with the environment. Synthetic highly charged sodium fluorophlogopites are expandable, unlike nonswelling natural micas, and have high negative charge in the 2:1 layers with Mg trioctahedral sheets.12 The high charge density of these micas is due to isomorphic substitution in only tetrahedral sheets, ideally accommodating 12 exchangeable sodium ions in the interlayer spaces per formula unit, which is based on 10 oxygen atoms.1214 They are informally called Na-2-, Na-3-, and Na-4-micas depending up r 2011 American Chemical Society

on either 1, 1.5, or 2 interlayer Na ions exist per formula unit, respectively. Na-ETS-4 (sodium Engelhard titanosilicate-4) is a synthetic microporous titanosilicate with a mixed octahedral and tetrahedral framework.15 Sodium titanosilicate (Na-TS) is also a synthetic ion exchanger built of tetrahedral Si and octahedral Ti units, which form a tunnel structure with exchangeable Na ions.16 These were reported to have high selectivity for radioactive species and heavy metal cations with high cation-exchange capacities (Table S1, Supporting Information).5,1720 Therefore, the objective of this work was to investigate the Hg2+ ion-exchange properties of the synthetic Na-2-, Na-3-, and Na-4-micas (referred to hereinafter as sodium micas), and NaETS-4 and sodium titanosilicate (designated from here on as titanosilicates) by use of distribution coefficients, equilibrium isotherms, Kielland plots, and Hg-release experiments.

’ MATERIALS AND METHODS Preparation and Characterization of Sodium Micas and Sodium Titanosilicates. The five tested ion exchangers were

prepared by thermal or hydrothermal treatments of precursors, based on procedures previously reported in the literature16,17,21,22 (see Supporting Information). After syntheses, the reactant solids were washed with deionized water and ethanol several times and Received: March 2, 2011 Accepted: July 11, 2011 Revised: June 22, 2011 Published: July 11, 2011 6954

dx.doi.org/10.1021/es200712r | Environ. Sci. Technol. 2011, 45, 6954–6960

Environmental Science & Technology

ARTICLE

dried at 60 °C. X-ray diffraction (XRD) patterns were collected on a Scintag diffractometer with Cu KR radiation and a Ge solidstate detector at a scanning speed of 3° 2θ/min with a scan step of 0.02° 2θ. Distribution Coefficient (Kd) Determination. Twenty milligrams of each exchanger was equilibrated for 24 h with 25 mL of a 0.5 N NaNO3 solution containing 0.0001 N Hg(NO3)2 at room temperature (pH = 4.3). After equilibrium, all solutions were collected by centrifugation and were analyzed by Milestone’s DMA-80 direct mercury analyzer. Triplicates were used for all experiments. The mean variation of Kd values was less than 10%. Kd (milliliters per gram) for Hg2+ was calculated by the following equation: Kd ¼

ðCi Ce Þ V Ce M

ð1Þ

where Ci and Ce are the initial and equilibrium concentrations of Hg2+ and V/M is the solution volume to exchanger mass ratio. Ion-Exchange Experiments. Twenty-five milligrams of each sample was placed in a polypropylene tube and shaken with 25 mL of a Hg(NO3)2 and NaNO3 solution prepared with different equivalent ratios (Hg2+/Na+ = 0.1, 0.2, 0.3, 0.5, 0.75, and 1.0) at room temperature for 4 weeks. The total normalities of the solutions for each sample were kept constant at 0.002 47, 0.003 61, 0.004 68, 0.006 39, and 0.007 10 N, chosen from theoretical total CECs of Na-2-mica, Na-3-mica, Na-4-mica, Na-ETS-4, and sodium titanosilicate, respectively (Table S1, Supporting Information). In order to avoid precipitation of Hg2+ during the exchange process, the initial pH of the equilibration solutions was set to 2.5 by use of HNO3 solution. On the basis of the Visual Minteq program,23 [Hg(OH2)6]2+ was determined to be the predominant species with no precipitation in all Hg exchange solutions at pH 2.5. Separation of solid ion exchangers and solution was done by centrifugation, and then the concentrations of Hg2+ were analyzed by atomic absorption spectroscopy on Milestone’s DMA-80 direct mercury analyzer. All these batch experiments were conducted in duplicate. The average of a mean variation was below 5%. The initial and final pH values of the exchange reactions were measured with a pH meter (Orion Research Inc., Beverly, MA). Powder XRD analysis of the solid phases of sodium micas was conducted to check for changes in the (001) spacings after equilibrium. The equivalent fractions of ion in solution and solid phase were calculated by use of data obtained through the exchange experiment in order to prepare isotherms and Kielland plots. A corrected selectivity coefficient, KM Na, in a Kielland plot gives a good indication for ion selectivity (see Supporting Information for ion-exchange theory and description of Kielland plot). If log KM Na is greater than zero, ion exchangers prefer metal (M) ions over Na+ ions, whereas Na+ ions are more preferred over M ions when log KM Na is less than zero. When log KM Na = 0, no preference for either ion is indicated. Hg Release from Hg-Exchanged Ion Exchangers. After the above ion-exchange experiments, two Hg-exchanged samples, one with low and another with high Hg2+ exchange, were selected to test for fixation of Hg2+. These two samples were equilibrated with the solutions of 0.1 and 1.0 equivalent ratio (Hg2+/Na+ = 0.1 and 1.0) for each exchanger. For Hg release experiment, 20 mg of solids was shaken with 40 mL of groundwater simulant (Ca 100 ppm, Mg 6.3 ppm, Na 25 ppm, Cl 234 ppm)24 for 24 h. The solids and solutions were separated by

Table 1. Kd Values of Five Ion Exchangers for Hg2+ Exchange

Kd (mL/g)

Na-2-mica

Na-3-mica

Na-4-mica

Na-ETS-4

Na-TS

0

4

8

854

456

centrifugation and the solutions were analyzed for Hg2+ released by groundwater.

’ RESULTS AND DISCUSSION Characterization of Sodium Micas, Sodium Engelhard Titanosilicate-4, and Sodium Titanosilicate. The XRD pat-

terns (see control patterns in Figures 24, vide infra) of Na-2-, Na-3-, and Na-4-micas match well with those previously reported.13,14,25 All the mica samples have strong 001 reflections of 12.112.2 Å,12 indicating the existence of Na+ and one layer of water molecules in the interlayers irrespective of their charge density differences. XRD patterns (Figure S2, Supporting Information) of Na-ETS-4 and sodium titanosilicate matched well with those reported in previous studies.16,26 The X-ray results showed that the five synthesized exchangers are highly crystalline and phase pure. pH Change after 2Na+ f Hg2+ Exchange Reaction. Initial and final pH values of solutions for the 2Na+ f Hg2+ exchange reaction are shown in Table S2 (Supporting Information). In all cases, the initial pH increased due to overall H+ consumption in solution phase by the sodium exchangers with Na+ f H+ exchange, although some H+ production occurred by hydrolysis of Hg2+ to Hg(OH)+ and Hg(OH)20 and H+ release from titanol groups of the sodium titanosilicates. H+ production from the latter processes is expected to be minor compared to the protons in a pH 2.5 solution. Also, the pH change from the initial pH was smaller as the Hg2+ concentration of exchange solutions increased, due to higher uptake of Hg2+. That is, the extent of pH increase is inversely related to Hg uptake: when more Hg ions occupied exchange sites, more protons were left in solution (Tables S2 and S3, Supporting Information). Therefore, unlike sodium titanosilicates, the large pH increase of sodium micas from the initial pH in Hg solution with the highest concentration suggests their low Hg uptake and selectivity. Distribution Coefficients for Mercury. Kd values for Hg ion exchange by sodium micas and sodium titanosilicates are shown in Table 1. The Na-ETS-4 sample showed the highest affinity for mercury in the presence of 0.5 N NaNO3 solution for a Na to Hg2+ equivalent ratio of 5000. However, the Kd values of sodium micas are very low compared to the two titanosilicates. Since the predominant species was Hg(OH)20 in the pH 4.3 solution (Table S4, Supporting Information), this significant difference in the Kd values between sodium micas and sodium titanosilicates suggests that sodium titanosilicates have a capacity to adsorb Hg(OH)20 with titanol groups, possibly via the formation of -TiO-Hg(OH)2 releasing protons to the solutions, similar to adsorption of Hg(OH)20 at silanol and aluminol groups of kaolinite.27 Previous work with titania (TiO2), which was shown to adsorb mercury as Hg(OH)20,28 also supports this mechanism. Other reports suggest that the titanol groups may be responsible for ion exchange in titanosilicates.19,29 2Na+ f Hg2+ Exchange with Sodium Micas. The isotherms for 2Na+ f Hg2+ exchange with sodium micas are given in Figure 1a. The isotherms show that Na-2-, Na-3-, and Na-4-micas seem not to take up mercury at X Hg > ∼0.19, 0.59, and 0.3, respectively. It means Na ions were exchanged with Hg ions up to 6955

dx.doi.org/10.1021/es200712r |Environ. Sci. Technol. 2011, 45, 6954–6960


ARTICLE

Figure 1. (a) Isotherms and (b) Kielland plots of sodium micas after 2Na+ f Hg2+ ion exchange.

Figure 2. XRD patterns of Na-2-mica (control) and after 2Na+ f Hg2+ ion-exchange reaction with (a) 0.25 mN Hg(NO3)2 + 2.22 mN NaNO3 (Hg:Na = 0.1:0.9), (b) 0.49 mN Hg(NO3)2 + 1.97 mN NaNO3 (Hg: Na = 0.2:0.8), (c) 0.74 mN Hg(NO3)2 + 1.73 mN NaNO3 (Hg:Na = 0.3:0.7), (d) 1.23 mN Hg(NO3)2 + 1.23 mN NaNO3 (Hg:Na = 0.5:0.5), (e) 1.85 mN Hg(NO3)2 + 0.62 mN NaNO3 (Hg:Na = 0.75:0.25), and (f) 2.47 mN Hg(NO3)2 (Hg:Na = 1:0).

∼19%, 59%, and 30% of theoretical CECs in Na-2-, Na-3-, and Na-4-micas, respectively. In Kielland plots (Figure 1b) for 2Na+ f Hg2+ exchange with sodium micas, all data points fell below the x-axis, indicating low selectivity of sodium micas for Hg (II). The low uptake and selectivity of sodium micas seems to be related to their low uptake capacity for Hg(OH)20, which is neutral and hence not involved in the exchange process, and this is the dominant species above pH 3.5 (Figure S1, Supporting Information). Hg speciation would be a key factor dictating Hg selectivity and uptake capacity of sodium micas. Interestingly, Na-3-mica showed higher selectivity and larger uptake for mercury than other micas. This might be due to its intermediate layer charge density affecting availability of exchange sites and interaction between exchange sites and mercury. That is, Na-4-mica is difficult to swell, which indicates that the accessibility of exchanging cations to interlayer space could be limited, while its high charge density attracts cations strongly.

Figure 3. XRD patterns of Na-3-mica (control) and after 2Na+ f Hg2+ ion-exchange reaction with (a) 0.36 mN Hg(NO3)2 + 3.24 mN NaNO3 (Hg:Na = 0.1:0.9), (b) 0.72 mN Hg(NO3)2 + 3.24 mN NaNO3 (Hg: Na = 0.2:0.8), (c) 1.08 mN Hg(NO3)2 + 2.52 mN NaNO3 (Hg:Na = 0.3:0.7), (d) 1.80 mN Hg(NO3)2 + 1.80 mN NaNO3 (Hg:Na = 0.5:0.5), (e) 2.70 mN Hg(NO3)2 + 0.90 mN NaNO3 (Hg:Na = 0.75:0.25), and (f) 3.61 mN Hg(NO3)2 (Hg:Na = 1:0).

In contrast, Na-2-mica easily swells, leading to high accessibility to exchange sites because of low layer charge density. Therefore, the intermediate layer charge of Na-3-mica might result in the highest uptake and selectivity for mercury among sodium micas. The XRD patterns of Na-2-mica after 2Na+ f Hg2+ exchange reaction with the equilibrium solutions containing different equivalent ratios of Hg to Na (Figure 2) suggest that some of the Na ions were replaced with Hg ions because an expanded phase with the (001) spacing of 14.3 Å appeared by the exchange of Na ions with larger, hydrated Hg ions in the interlayer space. In all XRD patterns of Figure 2, the (001) reflections with ∼14 Å are sharp and of high intensity, compared to expanded phases of Na-3- and Na-4-micas in Figures 3 and 4, while the intensities of ∼12 Å peaks of the original Na-2-mica are very low after Hg2+ exchange (Figure 2). This expansion is caused by its lowest layer charge density. A ∼1415 Å phase is regarded as a two-layer hydrate, while one-layer hydrate contributes to a ∼12 Å phase in 2:1 layer-type clays.30 Clay swelling from the one-layer to the 6956



Figure 4. XRD patterns of Na-4-mica (control) and after 2Na+ f Hg2+ ion-exchange reaction with (a) 0.47 mN Hg(NO3)2 + 4.22 mN NaNO3 (Hg:Na = 0.1:0.9), (b) 0.94 mN Hg(NO3)2 + 3.75 mN NaNO3 (Hg:Na = 0.2:0.8), (c) 1.41 mN Hg(NO3)2 + 3.28 mN NaNO3 (Hg:Na = 0.3:0.7), (d) 2.34 mN Hg(NO3)2 + 2.34 mN NaNO3 (Hg:Na = 0.5:0.5), (e) 3.51 mN Hg(NO3)2 + 1.17 mN NaNO3 (Hg:Na = 0.75:0.25), and (f) 4.68 mN Hg(NO3)2 (Hg:Na = 1:0).

two-layer hydrate was found to be due to transition of a partially dehydrated inner-sphere complex of an interlayer Na ion to a fully hydrated outer-sphere sodium ion in sodium montmorillonite.31 Therefore, the appearance of ∼12 Å phase in the (001) spacing indicated the existence of less hydrated Hg ions in the interlayer space, while the ∼14 Å phase resulted from more hydrated Hg ions. The existence of the two different hydration states of Hg2+ found here is supported by a previous study of muscovite (001) surface, where Hg adsorption occurred as an inner-sphere complex and outer-sphere complex.32 This difference in the degree of hydration of interlayer cations occurs from an energetic difference between an electrostatic attraction, which keeps pulling down a cation close to the surface, and ion hydration strength, which keeps a water sphere surrounding a cation.33 If the electrostatic attraction between a cation and a negatively charged layer is stronger than the ion hydration strength, it causes the partial dehydration of cations, consequently producing less hydrated cations. Figure 2 shows that the ∼12 Å phase increased as the uptake of Hg increased (Table S3, Supporting Information). This is because the electrostatic attraction can be affected by the amount of intercalated divalent cations. The strong electrostatic attraction by divalent cations can be explained by ionion correlations not occurring by monovalent cations.34 It has been used to explain the strong electrostatic attraction between like charged surfaces in the presence of divalent counterions.3437 Therefore, the electrostatic attraction between the negatively charged layer and Hg ions increases with the increasing adsorbed Hg2+ amount. This in turn caused an increase in the proportion of Hg ions with the reduced hydration state, which led to increased intensity of the ∼12 Å of Na-2-mica (Figure 2bf) as the exchanged Hg2+ increased (Table S3, Supporting Information). XRD patterns of Na-3-mica after 2Na+ f Hg2+ exchange reaction with the different equilibrium solutions (Figure 3) show that, like the case of Na-2-mica, there are two types of (001)

ARTICLE

spacings, one at ∼14 Å and another at ∼12 Å, indicating two different hydration states. The expanded phase of ∼14 Å results from the exchange of Na ions with Hg ions in the interlayer space, while the (001) spacing of ∼12 Å may be due to unexchanged Na ions or from some less hydrated Hg ions because of the strong electrostatic attraction between the negatively charged layer and Hg ions, which leads to a shedding of some water molecules surrounding Hg ions. It is found here that the peak intensity of ∼14 Å phase changes with the Hg occupancy in Na-3-mica (Figure 3 and Table S3, Supporting Information). Especially, when the Hg occupancy in the interlayer space is ∼52% (Figure 3e), the peak intensity of the expanded phase significantly decreases, compared to those of Figure 3ad. This ∼14 Å phase is almost completely diminished at ∼59% Hg occupancy (Figure 3f). The above result suggests a partial dehydration of large Hg ions occurred due to the increased electrostatic attraction caused by a high uptake of Hg ions (59% of Na ions were replaced by Hg ions) and this dehydration led to the decrease in (001) spacing. In addition to the effect of Hg ions, the layer charge density affects not only the expansibility of sodium micas but also the electrostatic attraction. The electrostatic force increases with increasing layer charge density. The high layer charge density, therefore, enhances the partial dehydration of adsorbed Hg2+ ions, leading to the decreased (001) spacing in sodium micas. The increasing order of the layer charge density in micas is as follows: Na-4-mica > Na-3-mica > Na-2-mica. Therefore, the electrostatic attraction of Na-3-mica is higher than that of Na-2mica when only layer charge is considered. This is why the intensity of ∼12 Å phase of Na-3-mica at ∼6.6% Hg occupancy is higher than that of Na-2-mica at ∼19% (Figures 2f and 3a and Table S3, Supporting Information), which indicates that the layer charge density may be more influential on the electrostatic force than the Hg occupancy. XRD patterns of Na-4-mica after 2Na+ f Hg2+ exchange reaction with various Hg concentrations are shown in Figure 4. The results showed that there is a strong (001) reflection of a ∼12 Å phase in all cases of Na-4-mica, but expanded phases of ∼14 Å are very weak or not present. In contrast to XRD patterns of Na-2-mica in Figure 2, the XRD patterns of Na-4-mica showed high-intensity reflections of ∼12 Å and low-intensity reflections of ∼14 Å, which is due to higher layer charge density of Na-4mica than those of Na-2- and Na-3-micas. This is because of high electrostatic attraction between the negatively charged layers and Hg ions in Na-4-mica, which leads to partial dehydration of intercalated Hg ions. As a result, the ∼12 Å phase is enhanced in Na-4-mica relative to Na-2- and Na-3-micas. 2Na+ f Hg2+ Exchange with Sodium Engelhard Titanosilicate-4 and Sodium Titanosilicate. In Figure 5a, the isotherm for 2Na+ f Hg2+ exchange with Na-ETS-4 increases steeply up to X Hg < ∼0.71, indicating Na-ETS-4 took up most of the mercury from solution at X Hg < ∼0.71, although it seems to take up no more mercury at X Hg > ∼0.82. All log K values in the Kielland plot (Figure 5b) are greater than 0, which suggests NaETS-4 shows a preference for Hg ions at X Hg< ∼0.82. The log K value overall increases until Hg ions occupy 50% of exchange sites of the structure of Na-ETS-4 (X Hg < ∼0.5). After that, it starts to decrease up to around 0 at X Hg = ∼0.82. This pattern indicates that steric hindrance begins to occur in the exchange sites when Hg ions occupy about 50% of the exchange sites. The isotherm for 2Na+ f Hg2+ exchange with sodium titanosilicate (Figure 5a) shows a steep increase at X Hg< ∼0.5 and then 6957



ARTICLE

Figure 5. (a) Isotherms and (b) Kielland plots of Na-ETS-4 and sodium titanosilicate after 2Na+ f Hg2+ ion exchange.

Table 2. Hg Release of Hg-Exchanged Ion Exchangers by Groundwater Simulant a

Hg2+ uptake

Hg2+ release

Hg release

b

(mequiv/g)

(%) c

sample

(mequiv/g)

Na-2-mica 0.1

0.07

0.008

11.0

Na-2-mica 1.0

0.47

0.135

28.7

Na-3-mica 0.1 Na-3-mica 1.0

0.24 2.13

0.019 0.166

7.8 7.8

Na-4-mica 0.1

0.4

0.014

3.5

Na-4-mica 1.0

1.29

0.030

2.3

Na-ETS-4 0.1

0.63

0.057

9.1

Na-ETS-4 1.0

5.21

0.215

4.1

Na-TS 0.1

0.68

0.190

28.0

Na-TS 1.0

5.14

0.449

8.7

a

Samples were equilibrated for 4 weeks with solutions of 0.1 and 1.0 equiv ratio (Hg2+/Na+ = 0.1 and 1.0) for each ion exchanger before the Hg release experiment. b Hg2+ uptake was calculated from ion-exchange experiments. c Hg release (%) = [Hg2+ release (mequiv/g)/Hg2+ uptake (mequiv/g)] 100.

gently increases up to X Hg< ∼0.72, suggesting that most Na ions were replaced with mercury at X Hg < ∼0.5, and there seems to be no more uptake of mercury at X Hg > ∼0.72. The Kielland plot for 2Na+ f Hg2+ exchange with sodium titanosilicate (Figure 5b) shows that the log K values in the range of X Hg from 0.3 to 0.5 are greater than 0, indicating Hg is more selective than Na on sodium titanosilicate at that range. With an increase of X Hg, the data points continuously increase up to the break point (X Hg < ∼0.5). However, after the break point the log K value drastically decreases with an increase in the occupancy of Hg ions in the solid phase (X Hg > ∼0.5), indicating the ion-exchange reaction is retarded. On the basis of this data pattern of sodium titanosilicate in the Kielland plot, it is suggested that a severe steric limitation is developed in the exchange sites after Hg ions occupy about 50% of all the exchange sites in sodium titanosilicate. Similar results have been shown previously by Kodama et al.,38 who suggested a rapid decrease in the log K values from the steric hindrance for exchanging cations with another exchanger. Little or no change could be detected by XRD after Hg2+ exchange in both sodium titanosilicates. The high selectivity and uptake of sodium titanosilicates for mercury, unlike that of sodium micas, seems to result from their ability to retain Hg(OH)20 as well as Hg2+ and HgOH+,

as supported by the Kd value experiments and high Hg occupancy in the exchangers. Na-ETS-4 has external titanol groups at crystal surface as well as high internal surface due to its highly defective structure,29 and these TiOH functional groups may contribute to higher selectivity of Na-ETS-4 than sodium titanosilicate. The steric hindrance is related to several factors such as ion size, extent of hydration, and shape and topology of a framework of an ion exchanger.39 The severe spatial hindrance in sodium titanosilicate seems to be related to the structural features of its framework and high hydration enthalpy of Hg (1824 kJ/mol) compared to that of Na (409 kJ/mol).40 The high hydration energy of Hg aggravates the steric hindrance at exchanging sites, because it tends to hold the Hg hydration phere strongly. Sdium titanosilicate has two types of exchange sites in the tunnel and the framework.16,19 Ideally, half the sodium ions reside within the tunnel, so that the theoretical CEC of the tunnel sites is 3.55 mequiv/g,calculated from the ideal chemical formula (Table S1, Supporting Information), whereas half the remainder ions are located in the framework with about 3.55 mequiv/g of the theoretical CEC. The sodium ions in the tunnel sites are loosely held and can be released first during exchange. At the break point of the sodium titanosilicate in Figure 5b (X Hg = ∼0.5), Hg ion capacity is 3.52 mequiv/g, which is similar to the capacity of tunnel sites. Therefore, it is suggested that Hg ions dominantly occupy the sites in the tunnels at X Hg < ∼0.5 because the tunnel sodium ions can be released first, while at X Hg > ∼0.5 it is difficult for hydrated Hg(II) to occupy framework sites due to its high hydration energy holding the water sphere strongly. Therefore, a severe steric limitation occurred at X Hg > ∼0.5 and lowered selectivity. Hg Release from Hg-Exchanged Ion Exchangers. Table 2 shows Hg release from the Hg-containing ion exchangers after equilibration with groundwater simulant for 24 h. The effect of layer charge density of sodium micas on Hg release is apparent because their Hg release percentage decreases with increasing layer charge density, as expected from the XRD patterns after equilibrium (Figures 24). Na-ETS-4 is likely to have high affinity for Hg because its Hg release amount is not large despite its three-dimensional structure, while Hg release of sodium titanosilicate is high, probably due to its tunnel structure where cations are easily accessible. Among the ion exchangers tested, Na-4-mica shows the lowest percentage of Hg release, while the releases of Na-2-mica and sodium titanosilicate are higher compared to other exchangers. 6958



’ ASSOCIATED CONTENT

bS

Supporting Information. Additional text with information on ion-exchange theory and preparation; four tables listing theoretical CEC and ideal formula of ion exchangers, pH change after 2Na+ f Hg2+ reaction, Hg uptake and occupancy in ion exchangers, and Hg speciation for Kd value experiments; and two figures showing Hg speciation diagrams in ion-exchange solutions as a function of pH and X-ray patterns of sodium titanosilicates. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Phone: 1-814-865-1542; Fax: 1-814-865-2326; e-mail: Komarneni@ psu.edu.

’ ACKNOWLEDGMENT We acknowledge financial support provided by the Clay Mineral Society through a student research grant. ’ REFERENCES (1) Selin, N. E. Global biogeochemical cycling of mercury: A review. Annu. Rev. Environ. Resour. 2009, 34, 43–63. (2) Boening, D. W. Ecological effects, transport, and fate of mercury: a general review. Chemosphere 2000, 40, 1335–1351. (3) Matlock, M. M.; Howerton, B. S.; Atwood, D. A. Irreversible precipitation of mercury and lead. J. Hazard. Mater. 2001, 84, 73–82. (4) Reddy, K. R.; Chaparro, C.; Saichek, R. E. Iodide-enhanced electrokinetic remediation of mercury-contaminated soils. J. Environ. Eng. 2003, 129, 1137–1148. (5) Lopes, C. B.; Otero, M.; Coimbra, J.; Pereira, E.; Rocha, J.; Lin, Z.; Duarte, A. Removal of low concentration Hg2+ from natural waters by microporous and layered titanosilicates. Microporous Mesoporous Mater. 2007, 103, 325–332. (6) Chen, S. L.; Wilson, D. B. Genetic engineering of bacteria and their potential for Hg2+ bioremediation. Biodegradation 1997, 8, 97–103. (7) Krishnan, S. V.; Gullett, B. K.; Jozewicz, W. Sorption of elemental mercury by activated carbons. Environ. Sci. Technol. 1994, 28, 1506–1512. (8) Chojnacki, A.; Chojnacka, K.; Hoffmann, J.; Gorecki, H. The application of natural zeolites for mercury removal: from laboratory tests to industrial scale. Miner. Eng. 2004, 17, 933–937. (9) Brigatti, M. F.; Colonna, S.; Malferrari, D.; Medici, L.; Poppi, L. Mercury adsorption by montmorillonite and vermiculite: a combined XRD, TG-MS, and EXAFS study. Appl. Clay Sci. 2005, 28, 1–8. (10) Collins, C. R.; Sherman, D. M.; Ragnarsdottir, K. V. Surface complexation of Hg2+ on goethite: Mechanism from EXAFS spectroscopy and density functional calculations. J. Colloid Interface Sci. 1999, 219, 345–350. (11) Monteagudo, J. M.; Ortiz, W. J. Removal of inorganic mercury from mine waste water by ion exchange. J. Chem. Technol. Biotechnol. 2000, 75, 767–772. (12) Gregorkiewitz, J.; Rausell-Colom, J. A. Characterization and properties of a new synthetic silicate with highly charged mica-type layers. Am. Mineral. 1987, 72, 515–527. (13) Shimizu, K.; Hasegawa, K.; Nakamuro, Y.; Kodama, T.; Komarneni, S. Alkaline earth cation exchange with novel Na-3-mica: kinetics and thermodynamic selectivities. J. Mater. Chem. 2004, 14, 1031–1035. (14) Stout, S. A.; Komarneni, S. Synthesis of Na-2-mica from talc and kaolinite: characterization and Sr2+ uptake. J. Mater. Chem. 2003, 13, 377–381.

ARTICLE

(15) Nair, S.; Jeong, H.; Chandrasekaran, A.; Braunbarth, C. M.; Tsapatsis, M.; Kuznicki, S. M. Synthesis and structure determination of ETS-4 single crystals. Chem. Mater. 2001, 13, 4247–4254. (16) Poojary, D. M.; Cahill, R. A.; Clearfield, A. Synthesis, crystal structures, and ion-exchange properties of a novel porous titanosilicate. Chem. Mater. 1994, 6, 2364–2368. (17) Ravella, R.; Komarneni, S.; Martinez, C. E. Highly charged swelling mica-type clays for selective Cu exchange. Environ. Sci. Technol. 2008, 42, 113–118. (18) Komarneni, S.; Kozai, N.; Paulus, W. J. Superselective clay for radium uptake. Nature 2001, 410, 771–771. (19) Clearfield, A.; Bortun, L. N.; Bortun, A. I. Alkali metal ion exchange by the framework titanium silicate M2Ti2O3SiO4 3 nH2O (M = H, Na). React. Funct. Polym. 2000, 43, 85–95. (20) Ferreira, T. R.; Lopes, C. B.; Lito, P. F.; Otero, M.; Lin, Z.; Rocha, J.; Pereira, E.; Silva, C. M.; Duarte, A. Cadmium(II) removal from aqueous solution using microporous titanosilicate ETS-4. Chem. Eng. J. 2009, 147, 173–179. (21) Braunbarth, C. M.; Boudreau, L. C.; Tsapatsis, M. Synthesis of ETS-4/TiO2 composite membranes and their pervaporation performance. J. Membr. Sci. 2000, 174, 31–42. (22) Du, H. B.; Chen, J. S.; Pang, W. Q. Synthesis and characterization of a novel layered titanium silicate JDF-L1. J. Mater. Chem. 1996, 6, 1827–1830. (23) Gustafsson, J. P. Visual MINTEQ, ver. 2.61; Department of Land and Water Resources Engineering, KTH, Stockholm, Sweden, 2009. (24) Solbra, S.; Allison, N.; Waite, S.; Mikhalovsky, S.; Bortun, A.; Bortun, L.; Clearfield, A. Cesium and strontium ion exchange on the framework titanium silicate M2Ti2O3SiO4 3 nH2O (M = H, Na). Environ. Sci. Technol. 2001, 35, 626–629. (25) Park, M.; Lee, D. H.; Choi, C. L.; Kim, S. S.; Kim, K. S.; Choi, J. Pure Na-4-mica: Synthesis and characterization. Chem. Mater. 2002, 14, 2582–2589. (26) Cruciani, G.; De Luca, P.; Nastro, A.; Pattison, P. Rietveld refinement of the zorite structure of ETS-4 molecular sieves. Micro. Meso. Mater. 1998, 21, 143–153. (27) Sarkar, D.; Essington, M.; Misra, K. Adsorption of mercury(II) by kaolinite. Soil. Sci. Soc. Am. J. 2000, 64, 1968–1975. (28) Aguado, M. A.; Cerveramamarch, S.; Gimenez, J. Continuous photocatalytic treatment of mercury(II) on titania powders: Kinetics and catalyst activity. Chem. Eng. Sci. 1995, 50, 1561–1569. (29) Usseglio, S.; Calza, P.; Damin, A.; Minero, C.; Bordiga, S.; Lamberti, C.; Pelizzetti, E.; Zecchina, A. Tailoring the selectivity of Ti-based photocatalysts (TiO2 and microporous ETS-10 and ETS-4) by playing with surface morphology and electronic structure. Chem. Mater. 2006, 18, 3412–3424. (30) Sato, T.; Watanabe, T.; Otsuka, R. Effects of layer charge, charge location, and energy change on expansion properties of dioctahedral smectites. Clays Clay Miner. 1992, 40, 103–113. (31) Hensen, E. J. M.; Smit, B. Why clays swell. J. Phys. Chem. B. 2002, 106, 12664–12667. (32) .Lee, S. S.; Nagy, K. L.; Park, C.; Fenter, P. Enhanced uptake and modified distribution of mercury(II) by fulvic acid on the muscovite (001) surface. Environ. Sci. Technol. 2009, 43, 5295–5300. (33) Park, C.; Fenter, P. A.; Nagy, K. L.; Sturchio, N. C. Hydration and distribution of ions at the mica-water interface. Phys. Rev. Lett. 2006, 97, 016101. (34) Kjellander, R.; Marcelja, S.; Pashley, R. M.; Quirk, J. P. Doublelayer ion correlation forces restrict calcium clay swelling. J. Phys. Chem. 1988, 92, 6489–6492. (35) Kjellander, R.; Marcelja, S.; Pashley, R. M.; Quirk, J. P. A theoretical and experimental study of forces between charged mica surfaces in aqueous CaCl2. J. Chem. Phys. 1990, 92, 4399–4407. (36) Segad, M.; Jonsson, B.; Akesson, T.; Cabane, B. Ca/Na montmorillonite: Structure, forces and swelling properties. Langmuir 2010, 26, 5782–5790. 6959



ARTICLE

(37) Guldbrand, L.; Jonsson, B.; Wennerstrom, H.; Linse, P. Electrical double layer forces. A Monte Carlo study. J. Chem. Phys. 1984, 80, 2221–2228. (38) Kodama, T.; Ueda, M.; Nakamuro, Y.; Shimizu, K.; Komarneni, S. Ultrafine Na-4-mica: Uptake of alkali and alkaline earth metal cations by ion exchange. Langmuir 2004, 20, 4920–4925. (39) Tanaka, Y.; Tsuji, M.; Tarnaura, Y. ESCA and thermodynamic studies of alkali metal ion exchange reactions on an R-MnO2 phase with the tunnel structure. Phys. Chem. Chem. Phys. 2000, 2, 1473–1479. (40) Smith, D. W. Ionic hydration enthalpies. J. Chem. Educ. 1977, 54, 540–542.

6960


Exchange by Highly Charged Swelling Micas, Sodium Engelhard

Recommend Documents