One-pot synthesis of thiol- and amine-bifunctionalized mesoporous silica and applications in uptake and speciation of arsenic

Abstract

A series of thiol- and amine-bifunctionalized mesoporous silicas were synthesized via one-pot co-condensation of tetraethylorthosilicate, 3-mercaptopropyltrimethoxysilane and N-(2-aminoethyl)-3-aminopropyltriethoxysilane. The mesoporous materials were characterized by X-ray diffraction, scanning electron microscopy, transmission electron microscopy, nitrogen gas adsorption, infrared spectroscopy, thermogravimetric analysis and elemental analysis. As(V) and As(III) were effectively adsorbed by amine and thiol on the functionalized silica, respectively, through electrostatic interaction and chelation. Adsorption isotherms and kinetic uptake profiles of As(V) and As(III) onto these adsorbents were investigated by batch adsorption experiments. Moreover, this material was employed for the speciation analysis of arsenic using a home-made syringe-based solid phase extraction device. As(V) and As(III) were effectively separated and pre-concentrated in a single run through a sequential elution strategy, in which 0.1 M HNO₃ was first used to selectively elute As(V), and then 1 M HNO₃ with 0.01 M KIO₃ was used to elute As(III). The merits of easy preparation, low cost, high adsorption capacity and selective desorption make the bifunctional mesoporous silica an ideal solid material for the removal and speciation analysis of arsenic in environmental waters.

---

1. Introduction

In recent decades, heavy metal contamination of natural waters has been of great concern because of the toxicity of heavy metals in relatively low concentrations, and their tendency for bioaccumulation in plants, animals and human beings.^1,2 Thus It is necessary to control the harmful effects of heavy metal ions through daily monitoring and removal technologies. Solid materials such as adsorbent that are used for the adsorption of heavy metal ions has increasingly received considerable attention in recent years because they are simple, cost-effective and easy to be automated.³ Moreover, they can be easily incorporated into automated solid phase extraction (SPE) procedures for the determination of trace metal ions and their species.⁴ A variety of adsorbents, such as biomass,⁵ carbon,⁶ zeolites,⁷ and functionalized inorganic supports⁸ have been applied for the study of trace elements with satisfactory results. In contrast to some other solid materials, significant attention has been paid to mesoporous silica due to its large specific surface area, good dispersibility and controllable morphology.^9,10 Recently, Wang et al.¹¹ prepared an L-cysteine-functionalized SBA-15 for Hg(II) sorption, which shows a large adsorption capacity. Awual et al.¹² synthesized 6-((2-(2-hydroxy-1-naphthoyl)hydrazono)methyl)benzoic acid (HMBA) functionalized material as optical mesoporous adsorbent for the selective recognition and removal of ultra-trace Cu(II) and Pd(II) ions.

To date, a series of functionalized mesoporous silicas, generally synthesized by the post modification of pure mesoporous silica gel or direct co-condensation of organosilane reagents with tetra-alkoxysilanes (either tetraethoxysilane (TEOS) or tetramethoxysilane (TMOS)), are currently available to adsorb heavy metal ions. Three main types of functional groups (amine, thiol and carboxylic) have been demonstrated to be the specific ligands for most of the metal ions.^13–15 However, there are certain differences for some elements, e.g., arsenic, selenium, antimony and tellurium, because they exist as different species in environmental water, and the properties of each species are considerably different.¹⁶ For example, arsenic is an omnipresent toxic trace element, and it is mainly found in environmental waters in two oxidation states, As(III) and As(V). Because of the different properties of these two species, it is hard to simultaneously adsorb them on mono-functionalized mesoporous silica with thiol or amine functional groups.^8,17–19 In addition, it is worth noting that most of the adsorbents, including mesoporous silica, are facing a similar problem; thus, a pre-oxidation or pre-reduction operation is indispensable for the uptake of both arsenic species. For example, pre-oxidization of As(III) to As(V) is usually performed using oxidizing agents or photocatalytic oxidation to enhance As(III) removal.^20,21

The studies have shown that the toxic effects of arsenic in an environmental or biological system critically depend on its chemical form, and As(III) is the most toxic form of the water-soluble species, while As(V) is also considerably toxic. Thus, it is not only important to remove As(III) and As(V), but also necessary for the quantitative determination of each arsenic species in daily monitoring.^22–29 Conventionally, the speciation of inorganic arsenic without employing chromatographic separation can be achieved by the selective adsorption of one species onto adsorbents, and the other species calculated by subtraction from the total inorganic arsenic. To obtain the total inorganic arsenic content, an additional extraction procedure is usually needed after the pre-oxidization/pre-reduction of arsenic in the sample,^24–27 or adjustment of the sample pH,^28,29 which makes the sample pretreatment slightly time-consuming.

For the removal or speciation analysis of arsenic, the conversion of species from one form to another may not only cause operational complexity, but also introduce interferences. In this paper, we report the first attempt to synthesize a thiol and amine-bifunctionalized mesoporous silica by a one-step method for the simultaneous uptake of As(III) and As(V). The novel materials synthesized from different reactant proportions of organosilane reagents and TEOS were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), N₂ adsorption, Fourier-transform infrared spectroscopy (FT-IR), thermogravimetric analysis (TGA) and elemental analysis (EA). The batch adsorption experiments were performed to investigate their adsorption behaviors and capacities for As(III) and As(V). Then, using a facilely homemade syringe-based SPE device, a sequential elution strategy for arsenic speciation analysis was proposed, and the simultaneous separation and pre-concentration of As(III) and As(V) was achieved in a single run without changing any sample conditions.

2. Experimental

2.1 Reagents and materials

Cetyltrimethylammonium bromide (CTAB) was purchased from TCI (Tokyo, Japan). Tetraethylorthosilicate (TEOS) and 3-mercaptopropyltrimethoxysilane (MPTMS) were purchased from Alfa Aesar (Tianjing, China), and N-(2-aminoethyl)-3-aminopropyltriethoxysilane (AAPTES) was purchased from Ourchem (Shanghai, China). HNO₃ was of guaranteed reagent grade and obtained from Merck (Zurich, Switzerland). All other chemicals were at least of analytical grade, and used without further purification. Deionized water (DIW, 18.25 MΩ cm) obtained from a Milli-Q water system (Millipore, Bedford, MA, USA) was used throughout the experiment.

Stock standard solutions of As(III) and As(V), 1000.0 mg L⁻¹, were prepared by respectively dissolving appropriate amounts of Na₃AsO₃ and As₂O₅ (both of analytical grade, purchased from Johnson Matthey, UK) in DIW. Lower concentration standard solutions were prepared daily by appropriate dilutions from their stock solutions.

2.2 Synthesis of bifunctional mesoporous silica

Bifunctional mesoporous silica was synthesized via co-condensation of TEOS, MPTMS and AAPTES with surfactant (CTAB) as the template in alkaline solution (Scheme 1). Typically, 120 mL DIW and 3.5 mL NaOH (1 M) were added to 0.5 g CTAB and stirred at 80 °C for 30 min. Then, a certain amount of TEOS, MPTMS and AAPTES were mixed and added to the above clear basic surfactant solution. Following the injection, a white precipitate was observed within 4 min. The reaction mixture was allowed to stir at 80 °C for 2 h under the atmosphere of nitrogen, and then the resulting slurry was filtered and rinsed with excess of H₂O, followed by ethanol and naturally dried overnight. Subsequently, the dried solid was extracted with a mixed solution of 0.5 mL HCl and 150 mL ethanol by stirring at 50 °C for 3 h. The extraction was repeated twice to accomplish the complete removal of CTAB. Then, the solid was filtered and washed with copious amounts of water until neutral, and subsequently repeatedly rinsed with ethanol. The resulting mesoporous silica was finally dried under vacuum for 24 h.

Scheme 1 Preparation of the bifunctional mesoporous silica of MP-AAP-X.

The molar composition of the reaction mixture was (1 − 2X)TEOS : X MPTMS : X AAPTES :  0.11 CTAB :  0.28 NaOH :  532 H₂O, where the reaction molar proportion between MPTMS/AAPTES and total Si is represented by X. Four different molar proportions, X = 0, 0.025, 0.075 and 0.15 were taken, and the bifunctionalized mesoporous silica materials thus obtained are hereafter noted as MP-AAP-0%, MP-AAP-2.5%, MP-AAP-7.5% and MP-AAP-15%, respectively.

2.3 Characterization of bifunctional mesoporous silica

XRD, SEM, TEM, N₂ adsorption, FT-IR, TGA and EA were employed for the characterization of the synthesized bifunctional mesoporous silica. Powder XRD experiment was performed on the Philip X'Pert Pro using a Cu Kα radiation source. Low angle diffraction with a 2-theta range of 1 to 10° was used to investigate the long-range order of the materials. Particle morphology of these materials was obtained by SEM using a Hitachi S-3400N II with 10 kV accelerating voltage and 0.005 nA of beam current for imaging. The microstructure was characterized by high resolution transmission electron microscopy (HRTEM) on a JEM-200CX microscope operating at a 200 kV accelerating voltage. N₂ adsorption measurement was carried out on a Micromeritics ASAP 2020 BET surface analyzer system at liquid N₂ temperature (−196 °C). Before measurement, the sample was outgassed at 90 °C for 1 h and then at 150 °C for 4 h. The data were evaluated using the Brunauer–Emmett–Teller (BET) and Barrett–Joyner–Halenda (BJH) methods to calculate the surface area and pore volumes/pore size distributions, respectively. FT-IR spectra were recorded on a Nicolet-6700 spectrometer in the range of 4000–450 cm⁻¹. TGA of the materials was performed on a Perkin-Elmer Pyris 1 DSC from 303 to 973 K with a heating rate of 10 K min⁻¹. EA of the particles was carried out on an Elementar Vario EL III using oxygen as the combustion gas.

2.4 Batch adsorption experiments

The adsorption of As(III) and As(V) on the bifunctional mesoporous silica was measured in batch experiments by adding a fixed amount of adsorbent (0.010 g) into a definite volume (10 mL) of either As(III) or As(V) solution. The initial pH of the solutions was adjusted with diluted HNO₃ or NH₃·H₂O using a pH meter. The mixture was placed in a rotary mixer and shaken (100 rpm) at room temperature for 30 min. After centrifugation, the remaining arsenic in the supernatant liquid was quantified using an inductively coupled plasma atomic emission spectrometer (ICP-AES, Perkin-Elmer, Model Optima 5300DV) by measuring the signal intensity at the emission wavelength of As, 189.042 nm. The amount of arsenic adsorbed was calculated using the following equation:

Q_e = [(C₀ − C_e)V]/m

where C₀ and C_e are the original and equilibrium concentrations, respectively, in the solution of the arsenic species (mg L⁻¹), V is the solution volume (mL), and m is the adsorbent mass (mg).

2.5 Speciation analysis procedure

A syringe filter tip (0.45 μm) of 13 mm diameter was used to form a simple SPE device. Briefly, 200 mg mesoporous silica were dispersed into 10 mL DIW, the suspension was then sequentially and rapidly injected evenly into ten syringe filter tips using a 10 mL syringe cylinder. In this case, DIW was passed through the filter membrane while the silica particles were collected on the surface of the filter membrane. As a result, a layer of 20 mg silica particles was immobilized in each syringe filter tip.

In the extraction step, 10 mL As(III) or As(V) solution was allowed to flow through the filter tip at a flow rate of 1 mL min⁻¹ for the preconcentration of As(III) and As(V). The flow rate was precisely controlled by a pump (Baoding Longer, Model BT100) or roughly controlled by pushing the syringe by hand. The effluent was collected and ⁷⁵As was monitored by an inductively coupled plasma mass spectrometer (ICP-MS, Perkin-Elmer, Model Elan-9000) to calculate the adsorption percentage. In the elution step, 1.5 mL of 0.1 mol L⁻¹ HNO₃ was injected into the filter to firstly desorb As(V) from the silica and the eluent was collected. Then, 1.5 mL of 1 M HNO₃ with 0.01 M KIO₃ was injected into the filter to desorb As(III) and the eluent was also collected. Finally, two portions of the eluent were directly introduced into ICP-MS for the determination of As(V) and As(III), respectively.

3. Results and discussion

3.1 Structure, property and function of the bifunctional silica

3.1.1 Structural and morphological characterization of the bifunctional silica. The mesoporous structure and morphology of the bifunctionalized silica particles were studied using XRD, TEM, SEM and N₂ adsorption.

Small-angle XRD patterns of the final materials with different initial reactant ratios are shown in Fig. 1. The well-resolved diffraction patterns characteristic of hexagonal MCM-41 silica, including (1 0 0), (1 1 0) and (2 0 0) peaks, were observed in MP-AAP-0%, MP-AAP-2.5% and MP-AAP-7.5%, while MP-AAP-15% appeared to be disordered. The peak intensity of the samples weakens as the organosilane dosage increases, implying that the organosilane in the synthesis mixture would perturb the self-assembly of surfactant micelles and the silica precursors. Moreover, a systematic shift of the diffraction peaks to larger 2θ angles with increasing organosilane contents suggests that slight lattice contractions occur as a result of the incorporation of the organosilane groups within the mesostructure synthesis (Table 1).

Fig. 1 XRD patterns of MP-AAP-0% (a), 2.5% (b), 7.5% (c) and 15% (d).

Table 1 Physicochemical characteristics of the bifunctional silicas

Adsorbent	Initial X	d spacing (nm)	SBET (m^2 g^−1)	Pore size (nm)	Pore volume (cm^3 g^−1)	EA (μmol g^−1)		Qm (μmol g^−1)
						Sulfur	Nitrogen	As(III)	As(V)
MP-AAP-0%	0	4.0	992	3.5	0.87	0	0	0	0
MP-AAP-2.5%	0.025	3.9	940	3.2	0.73	199	536	33	49
MP-AAP-7.5%	0.075	3.8	706	2.0	0.35	739	807	139	156
MP-AAP-15%	0.15	3.4	547	<2.0	0.34	1912	1500	192	417

---

The particle morphology and mesoporous structure of the resulting materials were analyzed by SEM and TEM. The SEM micrographs in Fig. 2A show a dramatic transformation of the particle morphology, with ellipsoidal particles being observed at the low concentrations of organosilane reagents (MP-AAP-0% and 2.5%), and elongated rods being observed at a higher concentration of organosilane reagents (MP-AAP-7.5%). The TEM micrographs of Fig. 2B revealed that the mesopores are uniformly aligned along the long axes of MP-AAP-0%, MP-AAP-2.5% and MP-AAP-7.5%. However, there were no ordered pores observed on the irregular particles for MP-AAP-15%, which is consistent with the result from XRD observation shown in Fig. 1.

Fig. 2 SEM (A) and TEM (B) images of MP-AAP-0%, 2.5%, 7.5% and 15%.

The N₂ adsorption analyses were performed on the synthesized materials. As shown in Fig. 3, MP-AAP-0%, 2.5% and 7.5% display a type IV isotherm, while the isotherm of MP-AAP-15% is different probably due to its out-of-order structure. The surface area, pore volume and pore size distribution of the bifunctional materials are listed in Table 1. It was found that the overall trend is a decrease in these parameters with the increase of the initial organosilanes (X).

Fig. 3 (A) N₂ adsorption–desorption isotherms and (B) pore size distribution of MP-AAP-0%, 2.5%, 7.5% and 15%.

3.1.2 Characterization of the functional groups on the bifunctional silica. FT-IR, TGA, and EA of N and S were carried out to confirm the organic functional groups on the bifunctional mesoporous materials. FT-IR spectra of the obtained four materials were acquired as shown in Fig. 4. The intensities of the characteristic absorptions of C–N bond around 1480 cm⁻¹, C–H stretching in the range of 2850–3000 cm⁻¹ and S–H bond at 2575 cm⁻¹ (inset) increase with increasing organic loading content, which demonstrates the successful incorporation of N and S containing functional groups into the matrix. N and S contents of the bifunctional silicas determined by EA are listed in Table 1. As can be seen, the larger X results in more N and S in the products.

Fig. 4 FT-IR spectra of MP-AAP-0% (a), 2.5% (b), 7.5% (c) and 15% (d).

TGA (Fig. 5) shows that the weight loss of MP-AAP-X materials in the temperature range between 473 and 923 K was 6.1%, 10.5%, 13.8% and 22.2% for MP-AAP-0%, MP-AAP-2.5%, MP-AAP-7.5% and MP-AAP-15%, respectively, which is attributed to the decomposition of the organic functional groups. The results were in accordance with the FT-IR and EA.

Fig. 5 TG curves of MP-AAP-0% (a), 2.5% (b), 7.5% (c) and 15% (d).

3.2 Batch adsorption behaviors of the bifunctional mesoporous silica

3.2.1 Influence of pH on adsorption. The pH of the solution determines the distribution of arsenic species and surface properties of the bifunctional mesoporous silica, thus influencing the adsorption of arsenic onto the adsorbent. The effect of solution pH on the adsorption behaviors of As(III) and As(V) is presented in Fig. 6. It can be clearly seen that As(III) and As(V) adsorption by the bifunctional silica is sensitive to the solution pH. The affinity of the bifunctional silica to As(V) is mainly due to the strong electrostatic interaction of the protonated amine group towards anionic arsenate species, which is the dominant species of As(V) in the pH range of 3.0–7.0.²⁶ Whereas, the affinity to As(III) can be explained by the strong chelation of the mercapto group to As(III).^29,30 Therefore, the bifunctional material has the simultaneous or selective adsorption ability for As(V) and As(III) within different pH ranges. In addition, MP-AAP-0%, i.e., the pure SiO₂, shows no appreciable adsorption to any of the two arsenic species.

Fig. 6 Influence of solution pH on the adsorption of As(III) and As(V) onto MP-AAP-0% and 7.5%.

3.2.2 Adsorption equilibrium isotherm. To obtain adsorption equilibrium isotherm data, adsorption experiments were performed using a fixed adsorbent/liquid ratio and different concentrations of either As(III) or As(V) solution. The obtained adsorption isotherms of As(V) and As(III) are shown in Fig. 7A and B, respectively. According to Langmuir adsorption isotherm models, the adsorption capacities (Q_m) for As(III) and As(V) were calculated and listed in Table 1. From the results, for the four as-prepared bifunctional materials with the elevation of the X value from 0 to 15%, the uptake of As(III) increased from 0 to 192 μmol g⁻¹, and the uptake of As(V) increased from 0 to 417 μmol g⁻¹. Simultaneously, it can be noted that Q_m for As(III) slowly increased when X increased from 7.5% to 15%. This might be because more mercapto groups are embedded in the co-condensation and a smaller percentage of the active mercapto groups are exposed to As(III) species, which consequentially leads to a decrease in the As(III)/-SH molar ratio. In addition, the adsorption of As(III) on MP-AAP-7.5% at pH 1.0 revealed that the strong interaction between As(III) and the material cannot be destroyed by dilute acid.

Fig. 7 Adsorption isotherms of As(V) (A) and As(III) (B) onto MP-AAP-0%, 2.5%, 7.5% and 15%, solution pH = 4.0. Adsorption kinetics (C) of As(V) and As(III) onto MP-AAP-7.5% and 15%, solution pH = 4.0.

Fig. 7 displays the As(V) and As(III) uptake curves for the two bifunctional materials, MP-AAP-7.5% and MP-AAP-15%. It can be seen that two adsorbents exhibit similar performance in the equilibrium time. The adsorption for As(V) was completed in 30 s, while for As(III) in 5 min. As mentioned in the previous section of the characterization of the adsorbents, the difference between MP-AAP-7.5% and MP-AAP-15% is the degree of mesoporous structure order and surface area. However, despite the high surface areas and uniform mesoporosities of the adsorbents, the results do not demonstrate any remarkably rapid arsenic uptake by the mesostructure.

3.2.3 Adsorption capacity. A comparison of the adsorption capacity with the previous functionalized silica materials that are reported for arsenic was made, and the results are listed in Table 2. As could be seen, the adsorption capacity of the bifunctionalized mesoporous silica synthesized via one-pot co-condensation method is comparable with those obtained by other reported adsorbents.

Table 2 Comparison of the adsorption capacity of silica based functionalized adsorbents

Adsorbent	Adsorption capacity (mg g^−1)		Ref.
	As(V)	As(III)
MP-AAP-2.5%	3.7	2.5	This work
MP-AAP-7.5%	11.7	10.4
MP-AAP-15%	31.3	14.4
Mercapto-functionalized mesoporous silica	—	19.4	18
AAAPTS modified silica gel	13.9	—	19
AAPTS modified mesoporous silica	10.3	—	26
(NH2 + SH) modified silica gel	0.29	2.7	29

---

3.3 Speciation analysis of arsenic species

3.3.1 Characteristics of the homemade SPE tip. Conventionally, SPE adsorbents are usually filled into a polymer or glass tube plugged with a small portion of glass wool at both ends to construct a SPE microcolumn or cartridge. The fabrication of a microcolumn or cartridge is little tedious, and the reproduction between different batches is slightly difficult. More importantly, the back pressure of these devices is very high, in particular using the small size adsorbents at large flow rate. A syringe filter tip, which is ubiquitous in the laboratory, was employed for a simple SPE device in the present study. Fig. 8 is a schematic diagram of the device used for the separation of arsenic species. MP-AAP-7.5% was used as the adsorbent due to its uniform structure, large surface area and high adsorption capacity. As described in the experimental section, the fabrication procedure of the SPE tip is very simple. Furthermore, experimental results indicated that the SPE device possess various advantages, such as high permeability, low pressure and facile assembly.

Fig. 8 Schematic diagram of the SPE procedure for speciation analysis. (A) Single SPE tip for inorganic arsenic. (B) Tandem SPE tips for total organic and inorganic arsenic.

3.3.2 Preconcentration and elution. According to the results of the batch experiments, As(V) and As(III) could be simultaneously concentrated on the bifunctional adsorbent at a certain pH. Considering the adsorption capacity, a pH of 4.0 was selected to guarantee a simultaneous quantitative adsorption of As(III) and As(V). The effect of sample flow rate on the adsorption was examined by ejecting 10 mL sample solution containing 100 μg L⁻¹ As(V) or As(III) through the prepared SPE tip with different sample flow rates. It was found that a quantitative adsorption for As(V) and As(III) could be obtained when the sample flow rate was below 4 mL min⁻¹ and 1 mL min⁻¹, respectively. Thus, a sample flow rate of 1 mL min⁻¹ controlled by pump or hands was adopted for speciation analysis. For sampling volume, a handheld syringe-based SPE was employed and sensitive ICP-MS was used for the detection of arsenic; thus, a volume of 10 mL of sample solution was taken for easy operation and appropriate detection levels. However, the sensitivity could be further increased by increasing the sampling volume.

As described in the previous section, As(III) can be adsorbed on the bifunctional silica at both the pH 1.0 and 4.0, suggesting 0.1 M HNO₃ (pH 1.0) does not affect the adsorption of As(III) on the adsorbent. In the case of As(V), the adsorption was considerably affected by pH, in particular the adsorption percentage, which decreased to 0% at pH 1.0. Therefore, a sequential elution strategy for As(III) and As(V) was proposed, in which 0.1 M HNO₃ was employed to first selectively elute As(V), and then 1 M HNO₃ with 0.01 M KIO₃ was employed to elute As(III). The probable elution mechanism can be explained as follows: As(V) is transformed from anion to uncharged species (H₃AsO₄) in 0.1 M HNO₃ and loses electrostatic interaction with the adsorbent, and then mercapto groups are oxidized by KIO₃ and the chelation between mercapto and As(III) is destroyed. Using individual As(V) or As(III) solution, it was found that As(V) can be quantitatively eluted by 1.5 mL 0.1 M HNO₃, and no As(III) was observed in this process, and then, As(III) can be quantitatively eluted by 1.5 mL 1 M HNO₃ with 0.01 M KIO₃. These results confirmed the feasibility of the sequential elution strategy.

3.3.3 Interference study. Commonly encountered co-existing ions in environmental waters, e.g., alkali, alkaline and transition metal ions may cause interferences to the preconcentration and determination of arsenic. The influences of these ions were investigated by 10 μg L⁻¹ As(V) and As(III). It was found that the main ions existing in water, such as Na⁺, K⁺, Ca²⁺, Mg²⁺ and Cl⁻, can be tolerated to at least 1000 mg L⁻¹. For the ions which may potentially compete with arsenic for thiol and amine sites, 5 mg L⁻¹ Al³⁺, Fe³⁺ and Zn²⁺, and 100 μg L⁻¹ Sb(III), Se(IV) and Hg²⁺, did not cause any obvious change to the results in the determination of As(V) and As(III). Therefore, the proposed method could be applied for the speciation analysis of inorganic arsenic in environmental waters.

3.3.4 Analytical performance and its validation. With a sampling volume of 10 mL and desorbing volume of 1.5 mL, the obtained limits of detection (LODs, defined as 3-fold signal-to-noise ratio) were 0.015 μg L⁻¹ for As(V) and 0.025 μg L⁻¹ for As(III), which were considerably lower than the allowed limit of arsenic in environmental waters. The calibration curves were established in the range of 0.1–100 μg L⁻¹ As(V) and As(III) with linear equations of y = 12 039.4x + 507.1 (R² = 0.999) and y = 13 841.8x + 1494.0 (R² = 0.998), respectively. The precisions (relative standard deviations, RSDs) for six replicate determinations of 10 μg L⁻¹ As(V) and As(III) were 5.6% and 4.5%, respectively. Adsorbents (MP-AAP-7.5%) prepared within batch (n = 4) and between different batches (n = 3) were examined by measuring the recoveries of the same solution containing 10 μg L⁻¹ As(V) and As(III) under the optimized conditions. It was found that the RSDs of eluted As(V) and As(III) were 4.0% and 6.9% for intra-batch, and they were 4.3% and 4.7% for inter-batch, respectively.

The accuracy of the proposed separation scheme was evaluated by analyzing standard solutions and two certified reference materials, namely GSBZ 50004-88 (standard environmental water sample) and GSB 080230 (standard seawater sample). The results of these analyses are summarized in Table 3. As can be seen, the concentrations of As(V), As(III) and As(Total) were in good agreement with the standard or certified values. These results indicated that As(III), As(V) and As(Total) in these water samples can be successfully determined based on the simultaneous retention of As(Total) on the SPE tip, and sequential elution with appropriate eluent.

Table 3 Analytical results of As(V) and As(III) in standard solutions (ST) and certified materials

Sample	Certified (μg L^−1)			Found (μg L^−1)
	As(V)	As(III)	As(Total)	As(V)	As(III)	As(Total)
ST1	10	10	20	10.8 ± 0.7	11.0 ± 0.9	21.8 ± 1.6
ST2	100	100	200	98.6 ± 7.8	107.5 ± 8.6	216.1 ± 16.4
50004-88	—	—	124 ± 8	2.1 ± 0.8	128.5 ± 9.6	130.6 ± 10.5
080230	—	—	1000 ± 40	70.4 ± 16.6	945.3 ± 73.4	1015.7 ± 80.0

---

In addition, the levels of monomethylarsenic acid (MMA) and dimethylarsenic acid (DMA) were also considered, a tandem SPE method that combined the MP-AAP-7.5% with strong cation-exchange resin (SCX 732, Sinopharm Chemical Reagent Co., Ltd) was investigated (Fig. 8B). The first SPE unit was filled with SCX, and the second SPE unit was filled with MP-AAP-7.5%. After loading the sample solution, the tandem syringe filter tips were separated. The first one was eluted by 1 M HNO₃ for the retained DMA, and the second one was orderly eluted by 50 mM acetic acid (HAc) for MMA, 0.1 M HNO₃ for As(V) and 1 M HNO₃ with 0.01 M KIO₃ for As(III). The recoveries in ranges of 97–104%, 83–92%, 92–101% and 96–108% were obtained for 10 mL of 10 μg L⁻¹ DMA, MMA, As(V) and As(III), respectively.

4. Conclusions

Thiol- and amine-bifunctionalized mesoporous silica was synthesized for the first time by one-pot co-condensation of TEOS, MPTMS and AAPTES under basic CTAB solution. The amounts of MPTMS and AAPTES in the initial synthetic mixture were varied, and their effects on the structural and chemical properties of these porous materials were systematically investigated. Due to the different interaction mechanisms between arsenic with bifunctional groups, As(V) and As(III) can be simultaneously adsorbed by the adsorbents in a wide pH range, and their separation can be realized by selective and sequential elution of As(V) with diluted 0.1 M HNO₃ and then As(III) with 1 M HNO₃ with 0.01 M KIO₃. The homemade syringe-based SPE device possessing the characteristics such as portability and simplicity provides a promising application for field sampling and pretreatment. These advantages make the bifunctional mesoporous silica an attractive and desirable adsorbent not only for arsenic removal from contaminated water, but also for arsenic speciation analysis.

Further investigations on the adsorption of other heavy metal ions on the bifunctionalized mesoporous silica have been conducted to fully evaluate the properties of this adsorbent, which would be reported elsewhere.

Acknowledgements

This work was supported by National Natural Science Foundation of China (21275069, 90913012), National Basic Research Program of China (973 program, 2009CB421601, 2011CB911003), National Science Funds for Creative Research Groups (21121091), and Analysis & Test Fund of Nanjing University.

References

1. S. Babel and T. A. Kurniawan, J. Hazard. Mater., 2003, 97, 219–243.
2. F. L. Fu and Q. Wang, J. Environ. Manage., 2011, 92, 407–418.
3. T. M. Mututuvari and C. D. Tran, J. Hazard. Mater., 2014, 264, 449–459.
4. B. Hu, F. Zheng, M. He and N. Zhang, Anal. Chim. Acta, 2009, 650, 23–32.
5. X. W. Chen, A. M. Zou, M. L. Chen, J. H. Wang and P. K. Dasgupta, Anal. Chem., 2009, 81, 1291–1296.
6. M. L. Chen, Y. M. Lin, C. B. Gu and J. H. Wang, Talanta, 2013, 104, 53–57.
7. W. Qiu and Y. Zheng, J. Hazard. Mater., 2007, 148, 721–726.
8. S. Mandal, T. Padhi and R. K. Patel, J. Hazard. Mater., 2011, 192, 899–908.
9. J. Aguado, J. M. Arsuaga, A. Arencibia, M. Lindo and V. Gascón, J. Hazard. Mater., 2009, 163, 213–221.
10. B. G. Trewyn, I. I. Slowing, S. Giri, H. T. Chen and V. S. Lin, Acc. Chem. Res., 2007, 40, 846–853.
11. Q. Li, Z. Wang, D. M. Fang, H. Y. Qu, Y. Zhu, H. J. Zou, Y. R. Chen, Y. P. Du and H. L. Hu, New J. Chem., 2014, 38, 248–254.
12. M. R. Awual, I. M. Rahman, T. Yaita, M. A. Khaleque and M. Ferdows, Chem. Eng. J., 2014, 236, 100–109.
13. D. Rahmi, Y. Takasaki, Y. Zhu, H. Kobayashi, S. Konagaya, H. Haraguchi and T. Umemura, Talanta, 2010, 81, 1438–1445.
14. L. Zhang, W. Zhang, J. Shi, Z. Hua, Y. Li and J. Yan, Chem. Commun., 2003, 210–211.
15. Y. L. Wang, L. J. Song, L. Zhu, B. L. Guo, S. W. Chen and W. S. Wu, Dalton Trans., 2014, 43, 3739–3749.
16. Z. Mester, R. Sturgeon and J. Pawliszyn, Spectrochim. Acta, Part B, 2001, 56, 233–260.
17. D. Mohan and C. U. Pittman Jr, J. Hazard. Mater., 2007, 142, 1–53.
18. E. McKimmy, J. Dulebohn, J. Shah and T. J. Pinnavaia, Chem. Commun., 2005, 3697–3699.
19. H. T. Fan, T. Sun, H. B. Xu, Y. J. Yang, Q. Tang and Y. Sun, Desalination, 2011, 278, 238–243.
20. T. Nakajima, Y. H. Xu, Y. Mori, M. Kishita, H. Takanashi, S. Maeda and A. Ohki, J. Hazard. Mater., 2005, 120, 75–80.
21. A. H. Fostier, M. d. S. S. Pereira, S. Rath and J. R. Guimarães, Chemosphere, 2008, 72, 319–324.
22. X. C. Le, X. Lu and X. F. Li, Anal. Chem., 2004, 76, 26 A–33 A.
23. T. Llorente-Mirandes, J. Calderón, F. Centrich, R. Rubio and J. F. López-Sánchez, Food Chem., 2014, 147, 377–385.
24. X. P. Yan, X. B. Yin, X. W. He and Y. Jiang, Anal. Chem., 2002, 74, 2162–2166.
25. X. P. Yan, R. Kerrich and M. J. Hendry, Anal. Chem., 1998, 70, 4736–4742.
26. D. H. Chen, C. Z. Huang, M. He and B. Hu, J. Hazard. Mater., 2009, 164, 1146–1151.
27. M. L. Chen, C. B. Gu, T. Yang, Y. Sun and J. H. Wang, Talanta, 2013, 116, 688–694.
28. C. Z. Huang, B. Hu and Z. C. Jiang, Spectrochim. Acta, Part B, 2007, 62, 454–460.
29. E. Boyacı, A. Çağır, T. Shahwan and A. E. Eroğlu, Talanta, 2011, 85, 1517–1525.
30. A. Howard, M. Volkan and D. Y. Ataman, Analyst, 1987, 112, 159–162.

---