ABSTRACT The use of zero-valent iron (ZVI) represents one of the latest innovative technologies for groundwater remediation. The reactivity of ZVI is enhanced when the particle size is in the nanometer range. However, nanoscale ZVI has limited field applications for in-situ groundwater remediation such as permeable reactive barriers due to its powdery form. Therefore, a method of adhering nanoparticles on a supporting material was suggested. In this paper, functionalized mesoporous silica beads were created using 3-mercaptopropyltrimethoxy silica, tetrabutyl orthosilicate, and cetyltrimethyl ammonium bromide, and their physical and chemical characteristics were measured. The highly active ZVI nanoparticles were adhered on these mesoporous silica beads. The reactivity of the resulting material was tested using nitrate solution. The reductive reaction of nitrate indicated that the degradation of nitrate appeared to be pseudofirst order with a high reaction rate constant of 0.1619 h-1. The reaction constant decreased to 0.0122 h-1 after 3 h of experiment due to mass transfer limitation. The higher dose of the supported nanosized ZVI increased the removal rate as well as the removal efficiency of nitrate
Journal of Water and Environment Technology, Vol. 6, No.1, 2008 - 35 - Reduction of nitrate by nanoscale zero-valent iron supported on mesoporous silica beads Heesu Park*, Kyeong-Ho Yeon*, Yong-Min Park*, Seong-Jae Lee**, Sang-Hyup Lee*, Yong-Su Choi*, Yunchul Chung* * Center for Environmental Technology Research, Korea Institute of Science and Technology, P.O. Box 131, Cheongryang, Seoul 130-650, South Korea ** Geoworks, Research Institute for Energy and Resources 135-410, Seoul National University, Gwanak-gu, Seoul 151-742, South Korea ABSTRACT The use of zero-valent iron (ZVI) represents one of the latest innovative technologies for groundwater remediation. The reactivity of ZVI is enhanced when the particle size is in the nanometer range. However, nanoscale ZVI has limited field applications for in-situ groundwater remediation such as permeable reactive barriers due to its powdery form. Therefore, a method of adhering nanoparticles on a supporting material was suggested. In this paper, functionalized mesoporous silica beads were created using 3-mercaptopropyltrimethoxy silica, tetrabutyl orthosilicate, and cetyltrimethyl ammonium bromide, and their physical and chemical characteristics were measured. The highly active ZVI nanoparticles were adhered on these mesoporous silica beads. The reactivity of the resulting material was tested using nitrate solution. The reductive reaction of nitrate indicated that the degradation of nitrate appeared to be pseudo- first order with a high reaction rate constant of 0.1619 h -1 . The reaction constant decreased to 0.0122 h -1 after 3 h of experiment due to mass transfer limitation. The higher dose of the supported nanosized ZVI increased the removal rate as well as the removal efficiency of nitrate. Keywords: mesoporous silica; nitrate; nanoscale zero-valent iron, permeable reactive barriers INTRODUCTION Permeable reactive barriers using zero-valent iron (ZVI) are an emerging technology for the remediation of contaminated groundwater. The reductive removal of chlorinated organic compounds, hexa-valent chromium, arsenate, and etc. using ZVI has been studied and successful results have been reported (Mathason and Tratnyet, 1994). Nitrate removal by ZVI has also received attention because nitrate contamination is ubiquitous and strongly regulated in many countries. Several reaction mechanisms of nitrate reduction by ZVI were suggested (Siantar, et al., 1996; Choe, et al., 2000; Huang and Zhang, 2002; Alowitz and Scherer, 2002): NO 3 - + 4Fe 0 + 10H + → 4Fe 2+ + NH 4 + + 3H 2 O (1) NO 3 - + 8Fe 0 + 10H + → 8Fe 3+ + NH 4 + + 3H 2 O (2) NO 3 - + 4Fe 0 + 7H 2 O → 4Fe 2+ + NH 4 + + 10OH - (3) 6NO 3 - + 10Fe 0 + 3H 2 O → 5Fe 2 O 3 + 3N 2 + 6OH - (4) 2NO 3 - + 5Fe 0 + 6H 2 O → 5Fe 2+ + N 2 + 12OH - (5) Address correspondence to Sang-Hyup Lee, Center for Environmental Technology Research, Korea Institute of Science and Technology, Email: yisanghyup@kist.re.kr Received Februar y 29, 2008, Accepted May 13, 2008. Journal of Water and Environment Technology, Vol. 6, No.1, 2008 - 36 - These reactions are known to be favorable only in acidic condition of solution (pH <4). This restriction was overcome by applying nanoscale ZVI (nZVI) which has enhanced reactivity originated from its higher specific surface area (Zhang, 2003). However, nanosized particles have limitations in in-situ groundwater remediation such as permeable reactive barriers, thus adhering nanoparticles on a suitable supporting material is important to prevent agglomeration of the nZVI and expand the area of its applications. In this study, mesoporous silica beads were created and nZVI was adhered to the beads. The resulting material was tested for nitrate removal. The main purpose of this paper was to evaluate whether the mesoporous silica is suitable for nZVI implementation and the resulting material can be used to treat nitrate in groundwater without ammonium release. MATERIALS AND METHODS Synthesis of mesoporous silica Mesoporous silica was selected as a supporting material for nZVI. The basic theory of the synthesis of mesoporous material is the formation of the framework with the templating assistance of long-chained amphiphilic molecules. Cetyltrimethyl ammonium bromide (CTAB) was used as a template and 3-mercaptopropyltrimethoxy silica (MPTS) and tetrabutyl orthosilicate (TBOS) were added as silica sources. MPTS was selected to provide the silica beads with functional groups for the removal of ammonium which is a possible byproduct of denitrification by ZVI. CTAB was dissolved in deionized water. The mixture of 3-mercaptopropyltrimethoxy silica (MPTS) and tetrabutyl orthosilicate (TBOS) were added to the solution containing CTAB. Then the final solution was mixed using a magnetic stirrer for 48 h. Mesoporous structures were formed through hydrolysis and condensation reactions (Nooney, et al., 2001). The transparent mesoporous silica materials were filtered out, rinsed, and dried in a dessicator at room temperature. The synthesis procedure is described Fig. 1. The effect of solution’s pH and agitation speed on the formation of mesoporous silica was evaluated. Initial pH was controlled by adding 2 M of NaOH solution and it ranged from 12.4 to 13.3for the experiments. Different agitation speed between 150 to 350 rpm was applied for solution mixing. After obtaining the optimal conditions for the bead formation, different percentage of MPTS ranging from 10 to 25 mole % was used to compare the characteristics of the resulting materials. BET surface area and compressibility were tested using the surface area and porosity analyzer ASAP 2010 (Micromeritics, Inc. USA) and the Universal Materials Testing System (model 4465, INSTRON, USA), respectively. The permeability was measured using a glass column with 30 mm of diameter and 150 mm of length. The pressure to maintain the certain height was observed and permeability was calculated using Darcy’s law. Cationic ion exchange capacity was also tested using 0.5 N HCl and 0.5 N NaOH solutions. The tested material was saturated with HCl and titrated by NaOH solution. Journal of Water and Environment Technology, Vol. 6, No.1, 2008 - 37 - Fig. 1 - Synthesis of mesoporous silica beads nZVI formation on the mesoporous silica beads 0.2 M of ferrous solution was made by dissolving ferrous chloride (FeCl 2 , Yakuri Pure Chemicals Co. Ltd., Japan) into deionized water (18. 2 MΩ, Milli-Q Water, MilliPore, USA) and poured into a tedlar bag (Supelco, USA). 5g of mesoporous silica beads was added to this solution and the solution was stirred for 24 h. The ferrous-saturated silica beads were taken out of the solution and rinsed with deionized water and drained. Then 1.5 L of 0.5 M sodium borohydride (Sigma , USA) solution was added incrementally to the silica beads stored in a column at 2 mL/min (Sushil, et al., 2005). Equation (6) describes the related reaction. Fe 2+ + 2BH 4 - + 6H 2 O → Fe 0 + 2B(OH) 3 + 7H 2 (6) After borohydride addition, the beads started to take black color. This suggested the ferrous ions attached to the support material were successfully reduced to zero-valent state. The iron content in the resulting material was measured by atomic adsorption spectroscopy (SOLAAR M, Thermo Electron Corporation, USA) after the acid digestion. In addition, the surface morphology of nZVI supported on the mesoporous silica beads was observed using a field-emission scanning electron microscope (FE- SEM, S-4700, Hitachi, Japan). Batch tests for kinetic study Denitrification rate by the resulting material was evaluated in batch tests. 5 g of the nZVI-loaded silica beads was added to 100 mL of nitrate solution. pH of the solution was not actively controlled during the experiments to mimic in-situ groundwater remediation system. The bottles were attached to a rotary shaker (099ARD4512, Gas- Journal of Water and Environment Technology, Vol. 6, No.1, 2008 - 38 - Col CO., USA) and shaken for 24 h at room temperature. At timed intervals, samples were taken by a needleless syringe, filtered through a 0.45 µm syringe filter, and tested for nitrate and nitrite content using an ion chromatography (DX-120, DIONEX, USA). To investigate the effect of dosage on the nitrate removal, 1 g to 30 g of the supported nZVI on the silica beads were added to 100 mL of nitrate solution and the same batch tests were conducted without active pH control. RESULTS AND DISCUSSION Characteristics of mesoporous silica and nZVI-loaded silica Optimal pH and agitation speed to prepare regular shaped mesoporous silica was determined from the experiments. One of the desirable characteristics of reactive media for permeable reactive barrier is to have a regular shape. The material with regular shape does not result in blocked intergranular spaces. Fig.2 presented the photos of the resulting materials synthesized at different solution’s pH and stirring speed. Beads failed to form at pH lower than 12.9 or higher than 13.0. The optimal pH for silica bead formation was found to be approximately pH 13. Silica materials with irregular shapes were synthesized at 150, 180, 250, and 300 rpm, but beads with more regular shapes were created at 200 rpm. pH 12.4 pH 12.7 pH 12.9 pH 13.0 pH 13.1 pH 13.2 pH 13.3 Fig. 2 - Effect of pH and agitation speed on the bead formation Table 1 presents the physical and chemical characteristics of mesoporous silica beads prepared using the solutions with different MPTS concentration. The synthesized beads were in the size range of 1~1.5 mm. The size is related with permeability and the permeability of reactive media should be considered for the successful operation of a permeable reactive barrier system. The material selected should minimize constraints on groundwater flow by not having excessively small particle size. The permeability of the 200 rpm 150, 180, 250 and 300 rpm Journal of Water and Environment Technology, Vol. 6, No.1, 2008 - 39 - silica beads was proportional to the size and its values were greater than that of groundwater-bearing layer (~ 10 -3 cm/sec). This suggests the synthesized silica beads can be used as a reactive medium for permeable reactive barriers. Cationic ion exchange capacity increased from 1.34 to 2.08 meq/g with the mole fraction of MPTS. But, the BET surface area showed the opposite relationship with MPTS contents. The maximum BET surface area of 1275 m 2 /g was measured from the silica beads with lowest mole fraction of MPTS. The high BET surface area suggested that the mesoporous silica beads can be an excellent supporting material for nZVI. Table 1 - Physical and chemical characteristics of mesoporous silica beads MPTS concentration in a preparation solution (mol %) size (mm) ion exchange capacity (meq/g) BET surface area (m 2 /g) compressibility (kg/cm 2 ) permeability (cm/sec) 10 1.0 1.34 1275 0.478 1.5×10 -2 15 1.2 1.54 938.1 0.518 1.7×10 -2 20 1.3 1.78 875.5 0.743 2.1×10 -2 25 1.5 2.08 858.4 1.012 2.7×10 -2 To evaluate the nitrate reduction, the silica beads prepared from 10% MPTS was selected for nZVI immobilization, since it has the maximum BET surface area and more nanoparticles can disperse on the larger surface. The SEM image at 80-fold showed the spherical shape of the silica bead. Clustered nanoparticles on the surface of silica were also observed at 2000-fold SEM picture (Fig. 3), and the content of nZVI was 9.3 mg/ g of the reaction material. 200μm 1 μ (a) (b) Fig. 3 - SEM images of the zero-valent iron adsorbded on mesoporous silica at (a) 80- fold and (b) 2000-fold Denitrification by nZVI on the mesoporous silica beads The kinetic study for chemical reduction of nitrate was conducted using the supported nZVI prepared from 10 mol % MPTS. The initial concentrations of nitrate was 35 mg- N/L and 5 g of supported nZVI beads were added to 100 mL of nitrate solution without Journal of Water and Environment Technology, Vol. 6, No.1, 2008 - 40 - pH control. Denitrification by ZVI is known to follow pseudo-first order reaction with respect to nitrate concentration (Choe, et al., 2000). [] [] − − = − = 3 3 NOk d t NOd r obs [] tk NO NO obs −= ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ − − 03 3 ][ (7) Where k obs is the observed first-order reaction rate constant, which is the slope of the regression lines by plotting a natural log graph with respect to nitrate concentration with reaction time. Fig. 4 showed the results plotted as ln([NO 3 - ]/ [NO 3 - ] 0 ) versus reaction time and conducting a linear regression of the first order kinetic equation. This graph is noticed by higher k obs of 0.1619 h -1 in the initial stage of the reaction followed by lower k obs of 0.0122 h -1 . The decrease in the value of k obs can be explained by mass transfer limitation. nZVI attached to pores with better accessibility become exhausted quickly as the reaction proceeds, and later nZVI clusters adhered to pores with mass transfer limitation started to participate in the reaction. The good fit of the linear model to data supports that the reaction is pseudo-first order with respect to nitrate concentration. During this batch test, nitrite was not detected most of the times and did not show any accumulation. This suggests that denitrification by the supported nZVI was fast and nitrite was transformed to ammonium after they produced. The ammonium concentration was not also detected, and this can be attributed to the adsorption to the mesoporous silica by ion-exchange. Fig. 4 - Nitrate reduction by nanoscale zero-valent iron supported on the mesoporous silica beads at unjusted pH 6.2 Denitrification of nitrate by ZVI involves reaction at the metal surface, therefore the quantity of available surface area is an important variable for nitrate removal. By adding different amount of the reaction material, the effect of the nZVI dose on nitrate removal was evaluated. The results of batch tests conducted for 48 h were presented in Fig. 5, Journal of Water and Environment Technology, Vol. 6, No.1, 2008 - 41 - and they showed that increasing the concentration of nZVI-silica beads in the solution removed nitrate in a faster rate. 1 g of the beads in 100 mL of solution with initial nitrate concentration of 30 mg-N/L removed 20 % of the nitrate after 8 h, and stayed at that point during the rest of the reaction time. The maximum dosage of 30 g achieved the complete removal of nitrate at 36 h. The k obs values were also obtained by plotting ln([NO 3 - ]/ [NO 3 - ] 0 ) versus reaction time and conducting a linear regression based on the equation (7). The results are presented in Fig. 6. The k obs values from the initial phase ranged from 0.037 h -1 to 0.218 h -1 and increased with the dose. Under a low stoichiometric excess of nZVI in nitrate solution, the k obs increased because the reactive Fe site increased proportionally with the Fe/solution ratio. The sluggish growth of k obs against the dose of nZVI in the initial phase was explained by Choe, et al. (2000). They used the enzyme reaction type of equation for the saturation effect shown in high dose of nZVI in nitrate solution. The k obs in later phase was from 0.002 h -1 to 0.084 h -1 and also increased with the dose of the reaction material. Elapsed time (hr) 0 102030405060 Relative concentration of residual nitrate-nitrogen (C/C o ) 0.0 0.2 0.4 0.6 0.8 1.0 dosage 1g dosage 3g dosage 5g dosage 7g dosage 20g dosage 30g Fig. 5 - Effect of different dosage of nZVI-silica on nitrate removal Fig. 6 - The dependency of the rate constants on the nZVI-silica dose Journal of Water and Environment Technology, Vol. 6, No.1, 2008 - 42 - CONCLUSIONS Functionalized mesoporous silica beads were successfully created using TBOS and MPTS as silica sources and CTAB as a surfactant and their characteristics were evaluated. The BET surface area increased as mole percent of MPTS decreased. The maximum surface area of 1275 m 2 /g was obtained with 10 mol % of MPTS. However, the highest ion exchange capacity of 2.08 meq/g was measured from the silica beads with 25 mol % of MPTS. nZVI was successfully incorporated into mesoporous silica through borohydride reduction of ferrous ions and the reduction kinetics of nitrate removal by these materials was evaluated using batch tests. The results indicated that the degradation of nitrate appeared to be a pseudo first-order reaction with a high reaction constant of 0.1619 h -1 without pH control. The reaction constant decreased to 0.0122 h -1 after 3 h of experiment due to mass transfer limitation. The enhanced reactivity and stability of these materials should make them good candidates for use in permeable reactive barriers. ACKNOWLEDGEMENT This work was supported by a grant from the Ministry of Environment of Korea – ECOPIA program and the Korea Research Foundation Grant funded by the Korean Government (KRF-2007-355-D00013). REFERENCES Alowitz, M. J., Scherer, M. M. (2002) Kinetics of nitrate, nitrite, and Cr(VI) reduction by iron metal. Environ. Sci. Technol., 36(3), 299-306. Choe, S., Chang, Y.-Y., Hwang, K.Y., and Khim, J. (2000) Kinetics of reductive denitrification by nanoscale zero-valent iron. Chemosphere, 41, 1307-1311. Huang, Y. H., Zhang, T. C. (2002) Kinetics of nitrate reduction by iron at near neutral pH. J. Environ. Eng.-ASCE, 128(7), 604-611. Mathason, L. J. and Tratnyek, P.G (1994) Reductive Dehalogenation of Chlorinated Methanes by Iron Metal. Environ. Sci. Tehchnol., 29(12), 2045-2053. Nooney, R. I. , Kalyanaraman, M. , Kennedy, G. , and Maginn, E. J. (2001) Heavy Metal Remediation Using Functionalized Mesoporous Silicas with Controlled Macrostructure. Langmuir, 17(2), 528-533. Siantar, D. P., Schreier, C. G., Chou, C. S., and Reinhard, M. (1996) Treatment of 1,2- dibromo-3-chloropropane and nitrate-contaminated water with zero-valent iron or hydrogen/palladium catalysts. Water Res., 30(10), 2315-2322. Sushil R. K., Bruce M., Laurent C., and Heechul C. (2005) Removal of Arsenic(Ⅲ) from Groundwater by Nanoscale Zero-Valent Iron. Environ. Sci. Technol., 39(5), 1291-1298. Zhang, W. X. (2003) Nanoscale iron particles for environmental remediation: An overview. J. Nanopart. Res., 5(3-4), 323-332. . 7H 2 O → 4Fe 2+ + NH 4 + + 10OH - (3) 6NO 3 - + 10Fe 0 + 3H 2 O → 5Fe 2 O 3 + 3N 2 + 6OH - (4) 2NO 3 - + 5Fe 0 + 6H 2 O → 5Fe 2+ + N 2 + 12OH - (5) Address. (Sushil, et al., 2005). Equation (6) describes the related reaction. Fe 2+ + 2BH 4 - + 6H 2 O → Fe 0 + 2B(OH) 3 + 7H 2 (6) After borohydride addition, the