New Ultrastable Mesoporous Adsorbent for the Removal of Mercury

Mar 26, 2010 - Fax: + 32 9 264 49 83. E-mail: ... To find a more stable adsorbent for the selective removal of mercury ions, a new mesoporous adsorben...
10 downloads 6 Views 2MB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

New Ultrastable Mesoporous Adsorbent for the Removal of Mercury Ions Els De Canck,† Linsey Lapeire,† Jeriffa De Clercq,‡ Francis Verpoort,† and Pascal Van Der Voort*,† †

Department of Inorganic and Physical Chemistry, Center for Ordered Materials, Organometallics and Catalysis (COMOC), Ghent University, Krijgslaan 281-S3, 9000 Ghent, Belgium, and ‡Faculty of Applied Engineering Sciences, University College Ghent, Schoonmeersstraat 52, 9000 Ghent, Belgium Received January 15, 2010. Revised Manuscript Received March 10, 2010

To find a more stable adsorbent for the selective removal of mercury ions, a new mesoporous adsorbent is developed and compared with a number of carefully selected mesoporous silica adsorbents described in literature. This new adsorbent is based on a pure trans-ethene bridged periodic mesoporous organosilica (PMO) which is subsequently modified to obtain a suitable adsorbent. The outcome is a new thiol-containing ethene bridged PMO which combines the adsorption efficiency of the thiol group toward mercury ions with the stability of ethene bridged PMOs. During the adsorption process, this material not only maintains its mesoporous structure and ordering, it also completely preserves the amount of organic functionalities, allowing recycling and reuse of the adsorbent. Additionally, this PMO is able to reduce the Hg2þ amount in aqueous solutions below 0.5 μg/L, and the adsorbent has a maximal adsorption capacity of 64 mg/g which means an apparent 1:1 ratio mercury(II) ion to thiol.

1. Introduction Periodic mesoporous organosilicas (PMOs)1,2 have drawn the attention of many research groups during the latest decade because of their unique properties.3-5 The materials are synthesized using bridged bis-silanes, for example, (EtO)3-Si-R-Si(OEt)3 and immediately incorporate the organic functionality R in their structure. The bis-silane polycondensates around a nonionic triblock copolymer, acting as a template. After formation of the PMO, the template is removed by extraction to reveal the pores. These mesoporous materials have well-defined pore sizes and pore shapes, high specific surface areas, and uniform distribution of the functionalities. As a result, they combine the strength of inorganic structures and the flexibility and chemical versatility of the organic moieties. These materials have a wide range of potential applications and are used as chromatographic packing materials,6-8 low-k devices,9 chemical sensors,3 catalysts,10 and host materials for biomolecules and drug delivery.3 In addition, promising results *To whom correspondence should be addressed. Telephone: þ32 964 44 42. Fax: þ 32 9 264 49 83. E-mail: [email protected].

(1) Asefa, T.; MacLachlan, M. J.; Coombos, N.; Ozin, G. A. Nature 1999, 402, 867. (2) Inagaki, S.; Guan, S.; Fukushima, Y.; Ohsuna, T.; Teresaki, O. J. Am. Chem. Soc. 1999, 121, 9611. (3) Van Der Voort, P.; Vercaemst, C.; Schaubroeck, D.; Verpoort, F. Phys. Chem. Chem. Phys. 2008, 10(3), 347. (4) Fryxell, G. E. Inorg. Chem. Commun. 2006, 9(11), 1141. (5) Vercaemst, C.; Ide, M.; Friedrich, H.; de Jong, K. P.; Verpoort, F.; Van Der Voort, P. J. Mater. Chem. 2009, 19(46), 8839. (6) Cho, E. B.; Kim, D.; Jaroniec, M. Langmuir 2007, 23(23), 11844. (7) Park, S. S.; Kim, S. J.; Seo, Y. K.; Park, D. H. 4th International Symposium on Nanoporous Materials, Niagara Falls, Canada; Sayari, A., Jaroniec, M., Eds.; Elsevier Science Bv: The Netherlands, 2005; Vol. 156, pp 183-190. (8) Rebbin, V.; Schmidt, R.; Fr€oba, M. Angew. Chem., Int. Ed. 2006, 45(31), 5210. (9) Goethals, F.; Meeus, B.; Verberckmoes, A.; Van Der Voort, P.; Van Driessche, I. J. Mater. Chem. 2010, 20, 1709. (10) Shylesh, S.; Samuel, P. P.; Sisodiya, S.; Singh, A. P. Catal. Surv. Asia 2008, 12(4), 266. (11) Zhong, Z. X.; Wei, Q.; Wang, F.; Li, Q. Y.; Nie, Z. R.; Zou, J. X. J. Inorg. Mater. 2008, 23, 408. (12) Wu, H. Y.; Liao, C. H.; Pan, Y. C.; Yeh, C. L.; Kao, H. M. Microporous Mesoporous Mater. 2009, 119, 109.

10076 DOI: 10.1021/la100204d

have been obtained in the field of environmental sciences, where the materials are applied as adsorbents.11,12 Extensive research13-36 has highlighted the importance of adsorbents in the area of removing harmful or regaining valuable components. In most cases, it is important that the adsorption occurs selectively. In this study, the focus is placed upon the removal of mercury(II) ion, known for its major toxicity.37,38 An excellent adsorbent19-22 for mercury ions is characterized by, among other properties, (1) a high loading of the functional adsorbing groups, (2) a uniform distribution of these groups, and (13) Zhang, L.; Zhang, W.; Shi, J.; Hua, Z.; Li, Y.; Yan, J. Chem. Commun. 2003, 210. (14) Lee, B.; Kim, Y.; Lee, H.; Yi, J. Microporous Mesoporous Mater. 2001, 50, 77. (15) Yang, H. J. Hazard. Mater. 2008, 152, 690. (16) Perez-Quintanilla, D.; del Hierro, I.; Fajardo, M.; Sierra, I. J. Mater. Chem. 2006, 16, 1757. (17) Perez-Quintanilla, D.; del Hierro, I.; Fajardo, M.; Sierra, I. J. Hazard. Mater. 2006, 134, 245. (18) Perez-Quintanilla, D.; del Hierro, I.; Fajardo, M.; Sierra, I. Mater. Res. Bull. 2007, 42, 1518. (19) Delac^ote, C.; Gaslain, F. O. M.; Lebeau, B.; Walcarius, A. Talanta 2009, 79, 877. (20) Walcarius, A.; Etienne, M.; Bessiere, J. Chem. Mater. 2002, 14, 2757. (21) Walcarius, A.; Etienne, M.; Lebeau, B. Chem. Mater. 2003, 15, 2161. (22) Walcarius, A.; Delac^ote, C. Chem. Mater. 2003, 15, 4181. (23) Brown, J.; Richer, R.; Mercier, L. Microporous Mesoporous Mater. 2000, 37, 41. (24) Puanngam, M.; Unob, F. J. Hazard. Mater. 2008, 154, 578. (25) Walcarius, A.; Delacote, C. Anal. Chim. Acta 2005, 547, 3. (26) Zhang, L. X.; Zhang, W. H.; Shi, J. L.; Hua, Z.; Li, Y. S.; Yan, J. Chem. Commun. 2003, 2, 210. (27) Brown, J.; Richer, R.; Mercier, L. Microporous Mesoporous Mater. 2000, 37, 41. (28) Bibby, A.; Mercier, L. Chem. Mater. 2002, 14, 1591. (29) Brown, J.; Mercier, L.; Pinnavaia, T. J. Chem. Commun. 1999, 1, 69. (30) Aguado, J.; Arsuaga, J. M.; Arencibia, A. Ind. Eng. Chem. Res. 2005, 44, 3665. (31) Aguado, J.; Arsuaga, J. M.; Arencibia, A. Microporous Mesoporous Mater. 2008, 109, 513. (32) Walcarius, A.; Delacote, C. Anal. Chim. Acta 2005, 547, 3. (33) Mattigod, S. V.; Feng, X. D.; Fryxell, G. E.; Liu, J.; Gong, M. L. Sep. Sci. Technol. 1999, 34, 2329. (34) Olkhovyk, O.; Jaroniec, M. J. Am. Chem. Soc. 2005, 127, 60. (35) Olkhovyk, O.; Pikus, P.; Jaroniec, M. J. Mater. Chem. 2005, 15, 1517. (36) Olhovyk, O.; Jaroniec, M. Ind. Eng. Chem. Res. 2007, 46, 1745. (37) Jarup., L. Br. Med. Bull. 2003, 68, 167. (38) Guzzi, G.; La Porta, C. A. M. Toxicology 2008, 244, 1.

Published on Web 03/26/2010

Langmuir 2010, 26(12), 10076–10083

De Canck et al.

(3) a great stability during the adsorption and the regeneration in acidic media.23-26 The sulfur functionality has already performed excellently in the removal of mercury ions. Mercier and co-workers27,28 have described mesoporous silica materials synthesized by the cocondensation of tetraethoxyorthosilicate (TEOS) and 3-(mercaptopropyl)trimethoxysilane (MPTMS). The materials with different sulfur content exhibited specific surface areas between 763 and 1176 m2/g and possessed approximately 0.47-2.3 mmol of thiol groups per gram of material. Mercury(II) ion uptake between 0.90 and 2.3 mmol/g was observed. When these materials were tested in the presence of other metal ions such as Pb2þ, Cd2þ, Zn2þ, Ni2þ, and Cu2þ, the results showed that the thiol functionalized adsorbents have a more pronounced affinity for the mercury(II) ion than for the other metals. This was consistent with previously published findings.29-31 Aguado et al.30,31 also reported materials prepared by the co-condensation of TEOS and MPTMS, and these materials possess a remarkably high maximal adsorption capacity up to 4.1 mmol of mercury(II) ion per gram of mesoporous adsorbent. Walcarius and Delacote32 grafted thiolpropyl groups on MCM materials with SBET = 1000 m2/g and thiol amounts of 0.84-2.26 mmol/g. No influence of other metals (Agþ, Cu2þ, Ni2þ, Cd2þ, Zn2þ, Bi2þ) was observed when the mercury(II) adsorption was examined. In addition to the uptake of Hg2þ in aqueous solutions, the removal of strongly complexed mercury(II) was examined by Mattigod et al.33 Mesoporous self-assembled silica materials with surface areas higher than 800 m2/g have shown a mercury ion adsorption of 2.8 mmol/g from KI/I2 lixiviant streams which even contain Kþ, I-, I2, Fe2þ, SO42-, and Ca2þ. This mercaptopropyl unit can also be combined with a large heterocyclic bridging group such as isocyanurate.34 Jaroniec et al.35 investigated the mercury ion adsorption of these materials synthesized by the co-condensation of tris[3-(trimethoxysilyl)propyl]isocyanurate, MPTMS, and TEOS. Mesoporous materials with specific surface areas of approximately 500 m2/g and narrow pore size distributions were obtained. The materials exhibited around 0.89 mmol of SH groups per gram. The mixture of isocyanurate and thiol resulted in a very high mercury ion adsorption. These co-condensed materials could adsorb up to 5.64 mmol/g Hg2þ. This research group also introduced other PMO materials.36 The isocyanurate moiety was combined with bis(3-(triethoxysilylpropyl)tetrasulfide), N-(3-triethoxysilylpropyl)4,5-dihydroimidazole, or ureidopropyltrimethoxysilane. Materials were prepared with specific surface areas ranging from 668 to 707 m2/g. They could adsorb up to 1.37 mmol Hg2þ/g. To achieve not only a selective but also a stable adsorbent, this study develops a new periodic mesoporous organosilica starting from a pure trans-ethene PMO.39-41 A sulfur group is introduced by covalent bonding to the CdC. After characterization, this material is tested for its stability in a set of mercury(II) ion adsorption experiments and compared with other functionalized silica materials available in literature. Equally, two different postsynthetic pathways have been followed to graft a functionality on the surface of a mesoporous silica material, in this case SBA-15. Also, a one-pot synthesis has been used to form a silica material via a co-condensation process. Next to the adsorption kinetics and (39) Vercaemst, C.; Ide, M.; Wiper, P. V.; Jones, J. T. A.; Khimyak, Y. Z.; Verpoort, F.; Pascal Van Der Voort, P. Chem. Mater. 2009, 21(24), 5792. (40) Vercaemst, C.; de Jongh, P. E.; Meeldijk, J. D.; Goderis, B.; Verpoort, F.; Van Der Voort, P. Chem. Commun. 2009, 27, 4052. (41) Vercaemst, C; Jones, J. T. A.; Khimyak, Y. Z.; Martins, J. C.; Verpoort, F.; Van Der Voort, P. Phys. Chem. Chem. Phys. 2008, 10(35), 5349.

Langmuir 2010, 26(12), 10076–10083

Article

isotherm, special emphasis is on the long-term stability and recyclability in acidic medium of this PMO adsorbent.

2. Experimental Section Vinyltriethoxysilane (VTES; 97%), (3-mercaptopropyl)triethoxysilane (MPTES; 98%), and (3-mercaptopropyl)trimethoxysilane (MPTMS; 98%) were purchased from ABCR. Tetraethoxyorthosilicate (TEOS; 98%), the pluronic PEO20PPO70PEO20 (P123), Grubbs’ first generation catalyst, 3-chloro1-propanethiol (98%), triethylamine (>99%), magnesium, iron(III) chloride (98%; anhydrous), hydrochloric acid (37%; p.a.), acetonitrile (99,5%; p.a.), acetone (>99,5%), ethanol (96%), and tetrahydrofuran (p.a.) were acquired from Sigma-Aldrich. Hg(NO3)2 in HNO3 (2 M) was purchased from VWR. Triethylamine, acetonitrile, and tetrahydrofuran were dried and degassed before use. 2.1. Synthesis of SBA-15. The mesoporous material is synthesized following the procedure first published by Zhao and co-workers.42,43 An amount of 4 g of P123 is dissolved into 120 mL of 2 M hydrochloric acid and 30 mL of distilled water. The mixture is stirred at room temperature until the surfactant completely dissolves. An amount of 9 mL of TEOS is added, and the temperature is increased to 45 °C for 5 h under stirring. A white precipitation is formed. Subsequently, the temperature is raised to 90 °C for 16 h under static conditions. Before filtering of the solids, the mixture is allowed to cool down slowly to room temperature. The solids are subsequently washed 3 times with 15 mL of water and 15 mL of acetone. Finally, the material (denoted as SBA-15) is calcined at 550 °C for 6 h. 2.2. Functionalization of SBA-15. 2.2.1. With NEt3. A total of 0.7 g of SBA-15 is mounted into a Schlenk flask under an inert atmosphere. Then 10 mL of acetonitrile and 2.7 mL of triethylamine, behaving as an activator,44 are added. The reaction mixture is stirred for 2 h at room temperature, and subsequently the mixture of triethylamine and acetonitrile is removed under inert atmosphere. Volumes of 10 mL of acetonitrile and 5.0 mL of MPTES are added to the material. The mixture is allowed to react for 2 h at 65 °C before the solid is filtered and washed three times with 15 mL of acetonitrile and 15 mL of acetone. The material (SBA-SH-NEt3) is dried at 90 °C under vacuum for 16 h. 2.2.2. With HCl. A total of 0.3 g of SBA-15 material is put in a flask. Then 10 mL of acetonitrile, 0.89 mL of hydrochloric acid (2 M), and 0.44 mL of MPTES are added. The reaction mixture is stirred for 96 h at 50 °C before the solids are filtered and washed three times with 15 mL of acetonitrile and 15 mL of acetone. The material (SBA-SH-HCl) is dried at 90 °C under vacuum for 16 h.

2.3. One-Step Synthesis of a Functionalized SBA-15-like Material. The synthesis is based on the recipes published by

Mercier et al.27 and Aguado et al.31 A total of 1 g of P123 is allowed to dissolve in 30 mL of 2 M hydrochloric acid for 45 min at room temperature. Following that, 2.10 mL of TEOS is added and stirred for 45 min at 40 °C. Then 0.20 mL of MPTMS is added and stirred for 20 h. A white precipitation is formed and is allowed to age at 100 °C for 24 h. After filtering and washing several times with water and acetone, the surfactant P123 is extracted using Soxhlet extraction with a mixture of ethanol/hydrochloric acid. The sample, listed below as Co-Con, is dried at 90 °C under vacuum for 16 h.

2.4. Synthesis of Ethene Bridged Periodic Mesoporous Organosilica (PMO). The ethene bridged PMO is synthesized

using the recipe previously published by our group.45 An amount (42) Zhao, D. Y.; Huo, Q. S.; Feng, J. L.; Chmelka, B. F.; Stucky, G. D. J. Am. Chem. Soc. 1998, 120, 6024. (43) Zhao, D. Y; Feng, J. L.; Huo, Q. S.; Melosh, N.; Fredrickson, G. H.; Chmelka, B. F.; Stucky, G. D. Science 1998, 279, 548. (44) Blitz, J. P.; Murthy, R. S. S.; Leyden, D. E. J. Colloid Interface Sci. 1988, 126, 387. (45) Vercaemst, C.; Ide, M.; Allaert, B.; Ledoux, N.; Verpoort, F.; Van Der Voort, P. Chem. Commun. 2007, 22, 2261.

DOI: 10.1021/la100204d

10077

Article of 1 g of P123 is dissolved in 47.8 mL of water, 3.42 mL of concentrated hydrochloric acid, and 2.45 mL of butanol. The mixture is vigorously stirred for 1.5 h. Then 1.86 mL of homemade 100% trans 1,2-bis(triethoxysilyl)ethene is added, and the temperature is raised to 45 °C. A white precipitation is formed. After 6 h of stirring, the mixture is heated to 90 °C under static conditions. After cooling down of the mixture, the solids are filtered and washed with water (three times) and acetone (three times). To remove the template, a Soxhlet extraction is performed with acetone for 5 h. Afterward, the PMO is dried. Subsequently, the pure trans-ethene PMO is functionalized by bromination. The material (E-ePMO) is dried at 90 °C under vacuum for 16 h before adding Br2 in gaseous phase.46 The physisorbed bromine is removed by drying the material at 90 °C for 24 h (Br-ePMO). The 100% trans 1,2-bis(triethoxysilyl)ethene is synthesized as follows: 0.0535 g of Grubbs’ first generation catalyst (PCy3)2Cl2RudCH-Ph39,47 was added to a Schlenk flask under argon which contains 42.95 mL of vinyltriethoxysilane (VTES). The mixture is stirred for 1 h at room temperature and refluxed for 1 h. Subsequently, the mixture is destilled to remove the remaining VTES.

De Canck et al. Table 1. Characteristics of the Mesoporous Adsorbents Synthesized in This Study sample

SBET (m2/g)a

Vp (cm3/g)b

dp (nm)c

mmol SH/gd

no. SH/nm2

SBA-15 865 0.84 3.12 SBA-SH-HCl 286 0.28 2.74 10 7.0 587 0.51 2.74 1.6 1.1 SBA-SH-NEt3 Co-Con 704 0.78 2.74 1.1 1.6 E-ePMO 1182 1.08 4.05 Br-ePMO 558 0.55 3.55 SH-ePMO 636 0.55 3.55 0.44 0.22 a SBET: the specific surface area calculated via the BET equation. b Total pore volume. c Pore diameter calculated from the adsorption isotherm with the BJH method. d The amount of thiol groups is determined via thiol titration as described earlier.

2.5. Modification of Brominated Ethene Bridged PMO. A mixture of magnesium (0.74 g), iron(III) chloride (0.54 g), and tetrahydrofuran (30 mL) is prepared under an inert atmosphere. This viscous solution is allowed to stir for 30 min at a temperature of 50 °C to activate the magnesium. 3-Chloro-1-propanethiol (0.22 mL) is added, and the mixture is stirred for 2 h at room temperature. Immediately following, the mixture is added to a Schlenk flask containing Br-ePMO. After stirring for 5 h at 40 °C, filtration is executed where the material is washed three times with 15 mL of THF, 2 M HCl, water, and acetone. Finally, the solids (SH-ePMO) are collected and dried at 90 °C for 16 h. 2.6. Determination of Thiol Groups. The amount of reachable thiol groups on the synthesized mesoporous materials is determined by silver titration.48 The thiol groups on the materials react with a known concentration of silver nitrate, and the excess of silver is titrated with potassium thiocyanate, using an iron indicator (FeNH4(SO4)2 3 12H2O in 0.3 M HNO3). The number of thiol groups can be calculated.

Figure 1. DRIFT spectrum of (a) SBA-15, (b) SBA-SH-NEt3, (c) SBA-SH-HCl, and (d) Co-Con. Scheme 1. Pre-activation of the Silanol Group with Triethylamine before Adding the Organosilane

2.7. Mercury(II) Ion Adsorption Experiments in the Low Concentration Range. A 150 mg amount of the material is

measured out, and 50 mL of 10 μg/L mercury(II) solution is added. The mixture is stirred at room temperature for 15 min and filtered, and then the amount of mercury ion in the filtrate is quantified with cold vapor atomic fluorescence spectroscopy (CV-AFS). The regeneration of the material is performed as follows. A sample of the Hg2þ-loaded material is washed three times with 10 mL of 2 M HCl and three times 10 mL of water. The amount of Hg2þ leached out of the adsorbent was measured by CV-AFS. The materials are dried in a furnace at 90 °C for 1 h before reuse. Three regeneration cycles are performed on each sample. The structures of the adsorbents are evaluated afterward. Nitrogen adsorption/desorption measurements are performed each time, and the amount of thiol groups are specified.

2.8. Mercury(II) Ion Adsorption Experiments in the High Concentration Range. Adsorbent equilibrium experiments were performed with 150 mg of mesoporous adsorbent (SH-ePMO) and 50 mL of Hg2þ solution with different concentrations between 100 μg/L and 400 mg/L. The mixture is stirred at room temperature for 90 min and filtered, and then the amount of mercury ion in the filtrate is quantified with cold vapor atomic absorption spectroscopy (CV-AAS). The adsorption experiments were performed at the pH that resulted from the Hg2þ solution (46) Nakai, K.; Oumi, Y.; Horie, H.; Sano, T.; Yoshitake, H. Microporous Mesoporous Mater. 2007, 100, 328. (47) Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc. 1996, 118, 100. (48) Vogel, A. I. Quantitative Inorganic Analysis, 3rd ed.; Longmans: London, 1961; pp 790-791.

10078 DOI: 10.1021/la100204d

without further adjustment. The initial and final pH of these Hg2þ solutions were measured. For the mercury(II) ion adsorption kinetic experiments, 300 mg of mesoporous adsorbent (SH-ePMO) was contacted with 100 mL of Hg2þ solution with Hg2þ concentrations of 10 and 100 mg/L. At predetermined intervals of time, samples of the mixture were Langmuir 2010, 26(12), 10076–10083

De Canck et al. withdrawn at suitable time intervals, filtered, and analyzed by CV-AAS. 2.9. Characterization Techniques. The specific surface area, pore volume, and pore radius are determined using nitrogen adsorption/desorption. The isotherms are recorded on Belsorp Mini II equipment at -196 °C. The samples are pretreated at 120 °C while degassing. CHNS elemental analysis is executed by the Centre National de la Recherche Scientifique (CNRS, France). Diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy is performed on a hybrid IR-RAMAN spectrophotometer, Equinox 55S (FRA 106; Bruker) with a MCT-detector, using a Graseby Specac diffuse reflectance cell, operating in vacuo and at 120 °C. Scheme 2. Partial Self-Condensation of MPTES on the Silica Surface

Article X-ray diffraction is performed with an ARL X0 tra X-ray diffractometer of Thermo Scientific equipped with a Cu KR1 tube and a Peltier cooled lithium drifted silicon solid stage detector. Cold vapor atomic fluorescence spectroscopy (CV-AFS) is performed on a Mercur spectrophotometer (Analytik Jena) with a UV source of 253.7 nm. Experimental data are processed with WinAAS version 4.2.0. Cold vapor atomic absorption spectroscopy (CV-AAS) is performed on a GBC-933 instrument combined with GBCHG3000 (GBC Scientific Equipment).

3. Results and Discussion 3.1. Adsorbent Characterization. The physical characteristics and the -SH loading of the functionalized ethene PMO together with a variety of mesoporous reference materials are presented in Table 1. The standard SBA-15 is functionalized by a grafting procedure with MPTES. Two different pathways are used. The first consists of the preactivation of the silanol groups on the surface by triethylamine44 (SBA-SH-NEt3; Scheme 1). When using the amine, the silanol group becomes more nucleophilic and an SN2 reaction with the organosilane occurs. This synthetic pathway results in a mesoporous material with 1.1 thiol groups/ nm2. The DRIFT spectrum (Figure 1b) shows the absence of the sharp silanol peak at 3740 cm-1, indicating the complete reaction of all the silanol groups available on the surface with MPTES. Characteristic are the peaks for the propyl unit at 2980, 2935, and 2896 cm-1. The second pathway uses hydrochloric acid as an activator to speed up the hydrolysis process (SBA-SH-HCl; Scheme 2). The amount of thiol groups/nm2 is significantly higher compared to the case of amine activation process. The high loading of 7.0 thiol groups/nm2 indicates more organosilanes attached to the surface than original silanol groups available. This is caused by a partial self-condensation of MPTES on the silica surface. Several MPTES units are attached to the surface via one silanol group. A network of condensed organosilanes is formed in the pores of the materials, lowering the specific surface area by

Scheme 3. (a) Schematic Presentation of the Ethene Bridged PMO with its Hexagonal Structure. (b) Functionalization of the Ethene Bond: (i) Bromination with Br2(g); (ii) Substitution of the Bromine by 3-Chloro-1-propanethiol

Langmuir 2010, 26(12), 10076–10083

DOI: 10.1021/la100204d

10079

Article

De Canck et al.

Figure 2. N2 adsorption/desorption isotherms of the samples. Every time before (9) and after three (4) mercury(II) ion adsorption cycles. (a) SBA-SH-HCl, (b) SBA-SH-NEt3, (c) Co-Con, and (d) SH-ePMO.

almost 600 m2/g. The DRIFT spectrum (Figure 1c) shows that silanol groups are still available on the surface: a small peak at 3733 cm-1. In addition to these two grafting procedures, another method can be used to synthesize mesoporous materials. Co-Con has been made by the co-condensation of TEOS and MPTMS, and its characteristics and DRIFT spectrum are also shown in Table 1 and Figure 1d. This material has the major advantage that inclusion of the organosilane immediately occurs during synthesis of the material, without further functionalization steps. The amount of thiol groups is comparable with the amount of SBASH-NEt3 and a high specific surface area is obtained. The final material examined in this study was a functionalized ethene PMO, reported for the first time. The bromination of the pure trans-ethene PMO39 has already been described by our group and yields almost 2 bromine groups/nm2. The SN2 substitution with a Grignard reagent (Scheme 3) toward mercaptopropyl 10080 DOI: 10.1021/la100204d

groups is not 100% efficient; it yields 0.44 available surface SH functions per nm2 as determined by silver adsorption. 3.2. Adsorption Experiments. For the development of a good quality adsorbent, the regeneration ability is a key feature. Only materials which show high stability and do not collapse when used are good candidates. Therefore the stability of the adsorbent after a mercury ion adsorption/desorption cycle needs to be studied. In addition to the preservation of the structure, the materials should keep their functionalities during the adsorption of mercury ion and the regeneration process. Hence, the number of functional groups is evaluated before and after the regeneration. Preliminary mercury ion adsorption experiments were performed with the four adsorbents to determine the time necessary for each material to reach the equilibrium in an experiment with a low mercury(II) ion concentration. The material was stirred for a certain amount of time in a mercury(II) solution. After removal of Langmuir 2010, 26(12), 10076–10083

De Canck et al.

Article

Figure 3. Powder X-ray diffraction patterns of (a) SBA-SH-NEt3, (b) SBA-SH-HCl, (c) Co-Con, and (d) SH-ePMO before and after three regeneration cycles. The XRD pattern of ePMO is also displayed.

the solids, the concentration of the mercury(II) was determined with CV-AFS. When using a typical concentration of 10 μg/ L, complete equilibrium was observed before 15 min for all materials. The recyclability of these materials is of key importance in the total evaluation of an adsorbent. In the recycling experiments, a low concentration Hg2þ solution (10 μg/L) is added to the adsorbent and stirred for 15 min. After separation of the solids, the solids are washed several times with a 2 M hydrochloric acid solution to recover the mercury ion. These adsorption cycles with the low mercury ion solution of 10 μg/L are repeated three times, and every mesoporous material could remove >99.99% of the mercury(II) ions. Thus, their adsorption capacity is comparable after three cycles at 10 μg/L mercury(II) ion concentration. However, remarkable differences can be seen when their stability is examined. We will discuss separately the structural stability and the chemical stability of the materials. The overall mesoporous structure of the material is examined with nitrogen adsorption/desorption and X-ray diffraction measurements before and after adsorption. Significant differences can be observed (Figure 2). Starting with SBA-SH-HCl, the isotherm shows a change in structure of the material, where the specific surface area has increased with approximately 220 m2/g. This demonstrates the unblocking of the pores by removal of the functionalities. The SBA-SH-NEt3 material displays a similar Langmuir 2010, 26(12), 10076–10083

behavior. The specific surface area also increases, yet not as much as the SBA-SH-HCl. The isotherm maintains its typical type IV hysteresis, and thus, the material still exhibits its ordered structure. When examining the isotherm of sample Co-Con, a structural collapse can be observed. Whereas the material shows a good ordering before the adsorption experiments, it loses its structure upon adsorption and regeneration. For the PMO material, the shape of the isotherm is similar, albeit slightly shifted to higher volumes of adsorbed nitrogen gas. Consequently, the specific surface area increases only 30 m2/g. This slight increase might be caused by the opening of blocked micropores by using hydrochloric acid during the rinsing step of the adsorption experiment. It is known that some micropores can be blocked with the surfactant P123 since the surfactant extraction45 only removes about 95% of the surfactant. The XRD patterns of the adsorbents before and after three regeneration cycles are displayed in Figure 3. The (100), (110), and (200) reflections, characteristic for a P6mm space group, can be observed in the XRD patterns of SBA-SH-NEt3 and SBA-SHHCl. Co-Con possesses a sharp (100) reflection and less pronounced (110) and (200) reflections. e-PMO and SH-ePMO both possess these characteristic reflections even after performing the bromination and substitution with chloro-propylthiol. The pattern of all the adsorbents slightly broadens when mercury ion adsorption and regeneration of the materials are performed. The (110) and (200) reflections of SBA-SH-NEt3 and DOI: 10.1021/la100204d

10081

Article

De Canck et al.

Figure 4. (left) N2 adsorption/desorption isotherm of SH-ePMO before (9) and after (4) 96 h of stirring in 2 M HCl. (right) Powder X-ray diffraction patterns of SH-ePMO before and after 96 h of stirring in 2 M HCl. Table 2. Characteristics of the Adsorbents after the Adsorption of Mercury(II) Iona sample

SBET (nm2/g)b

Vp (cm3/g)c

dp (nm)d

mmol SH/ge

leaching (%)

SBA-SH-HCl 510 0.61 3.12 1.00 90 614 0.57 3.12 0.76 53 SBA-SH-NEt3 Co-Con 416 0.44 2.13 0.85 23 SH-ePMO 666 0.61 3.55 0.44 0 a Every time, three cycles were performed before examining the material by nitrogen adsorption measurements and thiol titration. b SBET: the specific surface area calculated via the BET equation. c Total pore volume. d Pore diameter calculated from the adsorption isotherm with the BJH method. e The amount of thiol groups is determined with the thiol titration as described earlier.

SBA-SH-HCl become smaller and especially the (110) reflection is less pronounced in the XRD pattern after the three regeneration cycles. This is consistent with the nitrogen adsorption desorption measurement. The material made by the one-pot synthesis (CoCon) does not completely maintain its ordered structure. The (100) reflection broadens after regeneration, and the (110) and (200) reflections disappear completely. The SH-ePMO adsorbent on the other hand preserves its ordered hexagonal structure; only a slight broadening of the pattern can be observed. To examine the stability of the PMO adsorbent in strong acidic solution, the material was stirred for 96 h in 2 M hydrochloric acid. After filtering, washing, and drying the adsorbent, the material was characterized again to evaluate the stability of its structure. Figure 4 shows the nitrogen adsorption/desorption isotherms. Only a small difference can be seen in the isotherms, the result of further unblocking of the micropores. This effect is heavily pronounced in the isotherm as the removal of surfactant also makes the sample lighter, and these isotherms are always expressed in milliliters of N2 adsorbed per gram of sample. This proves that the materials can resist acidic media without loss of structure. The results of the thiol titration are presented in Table 2. When a comparison is made between the SBA-materials, the cocondensation and the PMO, the differences in the stabilities of the functional groups are remarkable. Whereas the periodic 10082 DOI: 10.1021/la100204d

mesoporous organosilica material retains 100% of the thiol groups, the materials prepared via the postsynthetic route and the one-pot synthesis both show significant loss of their functionalities. As shown in Table 2, sample SBA-SH-HCl even leaches its thiol groups for 90%. This can be explained by the sensitivity of the siloxane bridges that are responsible for the attachment of the functionalities. During the grafting procedure, these bridges are very sensitive to hydrolysis, especially when the material is treated with the acidified solution used for regeneration. Samples SBASH-NEt3 and Co-Con also suffer from functionality loss, although the loss is smaller. In the PMO, the propylthiol group is attached to the surface by means of a C-C group. The strength of this bond results in a material which does not leach any functionalities. The XRD, nitrogen adsorption/desorption measurements and thiol determination results clearly show the sustainability of the thiol functionalized PMO in comparison with the silica materials where the functionality is anchored via Si-O-Si bonds. The stability of PMOs in general is also described by Burleigh et al. and recently by our research group.49,50 Furthermore, additional mercury(II) ion adsorption experiments were performed to provide a better insight into the adsorption behavior of the ultrastable SH-ePMO. The effects of the initial mercury(II) ion concentration on the adsorption and adsorptivity (% of the Hg2þ adsorbed) are shown in the Figure 5. The adsorptivity decreases with increasing Hg2þ concentrations, whereas the adsorption capacity increases. The maximal adsorption capacity of the mesoporous adsorbent was approximately 64 mg/g. This implies a 1:1 stoichiometry of the mercury(II) ion toward the thiol group of the mesoporous material and is consistent with previously reported literature (Scheme 4).20,51,52 After adsorption, the pH of the solution (49) Burleigh, M. C.; Markowitz, M. A.; Jayasundera, S.; Spector, M. S.; Thomas, C. W.; Gaber, B. P. J. Phys. Chem. B 2003, 107, 12628. (50) Goethals, F.; Vercaemst, C.; Cloet, V.; Hoste, S.; Van Der Voort, P.; Van Driessche, I. Microporous Mesoporous Mater. 2010, 131, 68. (51) Mercier, L.; Pinnavaia, T. J. Environ. Sci. Technol. 1998, 32, 2749. (52) Wu, H. Y.; Liao, C. H.; Pan, Y. C.; Yeh, C. L.; Kao, H. M. Microporous Mesoporous Mater. 2009, 119, 109.

Langmuir 2010, 26(12), 10076–10083

De Canck et al.

Article

Figure 5. Effect of the initial Hg2þ concentration on the equilibrium Hg2þ adsorption onto SH-ePMO (150 mg) in Hg(NO3)2 solution (50 mL) at 20 °C.

The adsorption capacity as a function of time is shown in Figure 6. The kinetics of the adsorption of mercury(II) ion are best described by a pseudo-second-order rate.

Figure 6. Effect of adsorption time on Hg2þ adsorption onto SHePMO at initial Hg2þ concentration of 10 (b) and 100 (0) ppm.

Scheme 4. Proposed Adsorption Mechanism of Mercury(II) Ion on the Mesoporous SH-ePMO Adsorbent

decreased; that is protons were released during the adsorption, confirming the adsorption mechanism, earlier suggested in literature.

Langmuir 2010, 26(12), 10076–10083

4. Concluding Remarks The continuing search for new metal adsorbents has resulted in the development of a new functionalized periodic mesoporous organosilica. The PMO material was compared with adsorbents prepared via a postsynthetic route starting from SBA-15 and via one-pot synthesis. The outcome of modifying an ethene bridged PMO with propylthiol is an ultrastable adsorbent for the adsorption of mercury(II) ion. The material keeps its structure after multiple regeneration cycles and maintains its amount of thiol functionalities. The hydrochloric acid solution, necessary for the regeneration, does not affect the material on any level. The material was compared with the three other mesoporous silica materials synthesized in this study and was proven to be by far the most stable material. Even when treated for a longer period in hydrochloric acid, the material sustains its mesoporous structure. All other materials (functionalized SBA-15, co-condensed materials) quickly lose their structure and/or functionalities during adsorption or regeneration. The thiol functionalized PMO showed a maximal adsorption capacity of 64 mg/g and a 1:1 ratio of Hg2þ/SH. Acknowledgment. The authors acknowledge the Ghent University for financial support. We thank Cindy Claeys and Danny Vandeput from our Department of Inorganic and Physical Chemistry (Ghent University) for performing the nitrogen adsorption/desorption measurements and Karen Leus for the XRD measurements.

DOI: 10.1021/la100204d

10083