ARTICLE pubs.acs.org/est
Highly Selective PdCu/Amorphous SilicaAlumina (ASA) Catalysts for Groundwater Denitration Yongbing Xie,†,‡,§ Hongbin Cao,*,†,‡,§ Yuping Li,†,‡,§ Yi Zhang,†,‡,§ and John C. Crittenden^ †
National Engineering Laboratory for Hydrometallurgical Cleaner Production Technology, Beijing 100190, China Research Centre for Process Pollution Control and §Key Laboratory of Green Process and Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China ^ Brook Byers Institute for Sustainable Systems, Georgia Institute of Technology, Atlanta, Georgia 30332, United States ‡
bS Supporting Information ABSTRACT: Catalytic nitrate reduction is a promising technology in groundwater purification. In this study, PdCu bimetallic catalysts supported on an industrial amorphous silicaalumina (ASA) were synthesized and used to simulate catalytic removal of nitrate in groundwater. The catalysts exhibited very high activity and the highest catalytic selectivity toward N2O and N2 was 90.2%. The optimal Pd/Cu weight ratio was four. Relatively low reduction temperature was found benefit the catalytic stability and 300 C was the appropriate reduction temperature during catalyst preparation. With an average particle size 5.4 nm, the metal particles were very uniformly distributed on the catalyst surface prepared with the codeposition method. This kept the catalyst more stable than the PdCu/Al2O3 catalyst with larger metal particles. According to XRD, TEM, and XPS results, the metals maintained zero-valence but aggregated by about 2 nm during the denitration reaction, which caused gradual deactivation of the catalysts. Little leaching of Cu and Pd from the catalyst might also have a slightly negative impact to the stability of the catalysts. A simple treatment was found to redistribute the particles on the deactivated catalysts, and high catalytic activity was recovered after this process.
’ INTRODUCTION As one of the most widespread contaminants, nitrate constantly accumulates in soil and groundwater all over the world. Sources of nitrate include fertilizer abuse and animal feces disposal. Nitrate is potentially hazardous to human health when it enters drinking water, as it is converted to nitrite in the human body and causes blue baby syndrome and various cancers. Excess nitrate in rivers or lakes will also cause water eutrophication, which degrades ecological systems. Nitrate removal has become a compelling and important research issue in many countries.1,2 Many physical methods have been adopted to remove nitrate in water, such as reverse osmosis,3 ion exchange,4 and electrodialysis.5 Because these processes do not transform but simply transfer nitrate, additional treatment for high salt wastewater is thus required. Furthermore, the cost of nitrate removal processes based on physical methods is relatively high and operation is complicated. Similarly, biochemical methods seem to be very effective and have been applied in many industrial projects.6 But they also have serious drawbacks, such as the generation of a large amount of biological sludge and possible bacterial contamination of the purified water. A cheap and clean method is thus required to remove nitrate from groundwater efficiently. r 2011 American Chemical Society
Catalytic denitration was originally proposed by Vorlop and co-workers in 1989, in which nitrate was converted to harmless nitrogen over bimetallic PdCu catalysts under a mild reaction condition.7 This requires a very active and highly selective catalyst because the reaction is carried out at the temperature of groundwater and more harmful byproduct ammonium is possibly produced in this process. Extensive studies have been performed to enhance the catalyst performance, in terms of the bimetallic metal compositions, supports, and different preparation methods. PdCu was found to be very effective component in catalytic denitration.8,9 The catalytic performance of denitration catalyst was critically determined by the metal composition and the properties of the support. Meanwhile, the optimum ratio of noble metal and promoter varied on different kinds of supports.8,9 This indicated that the support plays a key role in determining the behavior of the catalyst. Much attention had been paid on different kinds of Received: December 5, 2010 Accepted: March 22, 2011 Revised: March 22, 2011 Published: April 07, 2011 4066
dx.doi.org/10.1021/es104050h | Environ. Sci. Technol. 2011, 45, 4066–4072
Environmental Science & Technology support materials 815 since this method was proposed. Besides these widely used materials, several kinds of novel supports were developed in the very recent years, such as conducting polymers,16 carbon nanotubes,17 resin,18 and bacterial cellulose nanofibers.19 The acidity and texture,20 adsorption capability,2,13 and ion-exchange ability 15 of the supports were found to influence the bimetallic catalyst performance critically. Monometallic catalysts supported on TiO2 were even developed for catalytic and photocatalytic nitrate reduction.21,22 To promote the application of catalytic denitration process in real groundwater treatment, new supports can be further investigated to get efficient denitration catalyst. The aim of this work was to develop an amorphous silicaalumina (ASA) supported PdCu denitration catalyst with high selectivity, and to intensively investigate its stability, possible reasons for deactivation and method for catalyst regeneration. ASA was selected because it is a very cheap and important material in petrochemical industry because of its brønsted acidity 23 and the acidity is adjustable.24 This property is very helpful because the selectivity of PdCu denitration catalysts was considered to be affected by the acidity of support.20 Electrondeficient noble metal particles were possibly created on ASA supported catalyst,25 which will contribute to the adsorption of nitrate on metal particles in catalytic denitration. In this work, an industrial ASA was introduced to prepare a series of PdCu catalysts and their catalytic behavior was evaluated. The preparation method, metal ratio, and thermal reduction temperature were investigated to optimize the catalyst performance. The catalytic stability was also intensively investigated. After discovering the main reason for gradual deactivation of these catalysts, a regeneration method was developed. This is very helpful because the deactivation factually happened on most of the industrial catalysts.
’ EXPERIMENTAL SECTION Materials and Catalyst Preparation. All reagents used in this study were A.R. grade, purchased from China National Medicines Corporation Ltd. and used without further purification. The simulated nitrate rich solution was prepared by dissolving NaNO3 in distilled water. The catalysts were prepared by incipient wetness impregnation or by codeposition. The metal loading on the catalysts was 5 wt % Pd and a variable amount of copper. The support materials, ASA (BET surface area: 375 m2/g) and γ-Al2O3 (BET surface area: 318 m2/g) (Tianjin Chemical Research & Design Institute), were ground and sieved to obtain particles with sizes between 0.07 and 0.15 mm. The supports were calcined at 500 C for three hours before use. The catalysts were prepared with impregnation or codepositon methods, the details are shown in section S-1Supporting Information. Alumina supported PdCu catalysts were prepared in the same procedure for comparison. Catalysts supported on ASA were labeled as PdCuI and PdCuD; I and D indicated the impregnation and deposition methods, respectively. PdCuAl2O3I and PdCuAl2O3D were marked in the same way. Monometallic Pd catalyst prepared on ASA with a deposition method was labeled Pd/ASA. Catalytic Reduction Experiments. The catalytic performance of these catalysts was studied in a batch reactor under strong stirring. The simulated nitrate-rich solution was 500 mL solution of NaNO3 with a nitrate concentration of 100 ppm. The
ARTICLE
reactions were carried out at 25 C at atmospheric pressure. Catalyst (0.4 g) was used in each case. High-purity H2 was used as the reducing agent, and high-purity CO2 was introduced to adjust the pH of the solution to approximately 5.4. They were premixed and imported through a porous glass located at the bottom of the reactor. The flow rates of hydrogen and CO2 were 200 mL/min and 100 mL/min, respectively, if not indicated otherwise. This was much more than the stoichiometric amount and ensured a possible maximum concentration of hydrogen in the solution. The solution was bubbled with H2 to remove oxygen in the system before reaction. Typical kinetic runs lasted 90 min. The degradation of nitrate and generation of nitrite and ammonium were measured at 15 min intervals, when samples were taken from the reactor and filtered with a 0.45 μm membrane prior to analysis. The concentrations of nitrate and nitrite were detected with ion chromatography (IC, Dionex DX500), and the ammonium ion selective electrodes measured the ammonium concentration. The nitrate conversion (C) and reaction selectivity (S) in each kinetic run were calculated as follow: C ¼ ð1 C1i =C0 Þ 100%
ð1Þ
S ¼ ½1 ðC2i þ C3i Þ=ðC0 C1i Þ 100%
ð2Þ
Where C0 is the initial NO3N concentration and C1i, C2i, and C3i represent real time concentrations of NO3N, NO2N, and NH3N, respectively. The very little amount of gas phase products was not detected in this work. According to the literature,9 harmless N2O gas was possibly produced in nitrate reduction along with nitrogen. Herein, S meant the selectivity for hydrogenation of nitrate to N2O and N2. The stability of the PdCu catalysts on ASA was investigated in the reaction system in this work. The procedure was very similar to that reported in literature.26 Successive pulses of nitrate were introduced into the batch reactor after every 90 min. The concentration change of nitrate, nitrite, and ammonium was recorded in each cycle. Adsorption Experiments. The adsorption capacity of the two supports to nitrate and nitrite was studied. The experiments were carried out in the same batch reactor at 25 C in which the catalytic denitration was carried out. The adsorbent concentration in solution was 0.8 g/L. The initial concentrations of nitrate or nitrite were 100 ppm. High-purity CO2 at a flow rate of 100 mL/min was continuously bubbled into the suspension under vigorous stirring to maintain the same acidic environment as that when catalytic denitration was performed. The adsorption of the two supports to ammonium was also studied. An amount of each support, 0.4 g, was dispersed in 500 mL solution of (NH4)2CO3, and the initial concentration of NH4þ was 50 ppm. 200 mL/min H2 and 100 mL/min CO2 were bubbled into the suspension under constant stirring. After 90 min adsorption, the concentration of residual NH4þ in the solution was detected. Catalysts Characterization. The catalysts were characterized by transmission electron microscopy (TEM), X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS). The details of the characterization methods can be found in the section S-2 of the Supporting Information. 4067
dx.doi.org/10.1021/es104050h |Environ. Sci. Technol. 2011, 45, 4066–4072
Environmental Science & Technology
ARTICLE
Figure 2. Nitrate removal over PdCu/ASA catalysts with different metal ratios.
Figure 1. Conversion of nitrate (A) and selectivity (B) of catalytic denitration over PdCu bimetallic catalysts.
’ RESULTS AND DISCUSSION Adsorption Removal of Anions. The ability of the two supports to adsorb nitrate or nitrite was investigated first time in this work. The concentration of the anions in the solution was detected every 15 min. It was found that the concentration of nitrate decreased slightly in 90 min. The final concentrations of nitrate decreased only by 0.9 and 2.6 ppm in the case of ASA and alumina adsorption, respectively. This meant the adsorption removal of the supports to nitrate was very little in the presence of CO2 bubbling. The adsorption removal of ASA and alumina to nitrite was also studied. After 90 min adsorption, the final concentrations of nitrite were 95.7 and 94.1 ppm, respectively. Thus the adsorption removal of the supports to nitrate and nitrite was negligible and was not considered in the following study of catalytic removal of nitrate. The adsorption of the two supports to ammonium was also studied. The concentration of ammonium in the solution did not decrease after 90 min adsorption in both cases. This meant the adsorption of the two supports to ammonium was also very weak under acidic environment at pH 5.4. Catalytic Removal of Nitrate. The catalytic performance of the four catalysts was characterized over three reaction cycles; the results are shown in Figure 1. The catalysts used in these cases were reduced by high-purity hydrogen gas at 300 C. Part A of Figure 1 shows that the fresh catalysts were very active in the first cycle. More than half of the nitrate in the solution was removed in the first 15 min over all four catalysts. The average disappearance rates of nitrate of ASA-supported catalysts were 0.079 and
0.087 mmol/(gcata.min) for PdCuD and PdCuI catalysts in this period, respectively. As the concentration of nitrate in the suspension decreased, the disappearance rates of nitrate slowed. Nitrate was almost completely eliminated within 60 min, and the average disappearance rate of nitrate was 0.033 mmol/(gcata.min) both for PdCuD and PdCuI catalysts. Similar rates were found on alumina-supported fresh catalysts. In the second cycle, PdCuAl2O3I lost its activity very fast, but the other three catalysts were relatively stable. ASA-supported catalysts were more stable and active than alumina-supported ones, which was further confirmed in the third cycle of reaction. The catalytic degradation of nitrate was considered a stepwise reaction. Nitrate was reduced initially to nitrite over copper, and then nitrite was reduced to N2, N2O, or ammonium over palladium or palladiumcopper bimetallic particles. The concentration of nitrite produced over the four catalysts was also detected and shown in Figures S-1 and S-2 of the Supporting Information. A large amount of nitrite was produced over the PdCuAl2O3I catalyst in the first cycle, and then the amount greatly decreased in the second and third cycles. The production of nitrite over the PdCuI catalyst is not large during the first cycle, but the maximum concentration greatly raised in the second and third reaction cycles. The amount of intermediate nitrite produced over the catalysts prepared by the deposition method was much lower than that over impregnated catalysts, except for that over PdCu/Al2O3I catalyst in the third cycle. At the end of each reaction cycle, the residual concentrations of ammonium in the solution were determined. The amount of ammonium remaining in the solution over ASA-supported fresh catalysts was only one-fifth of that over Al2O3-supported fresh catalysts. The amounts of ammonium produced over aluminasupported catalysts decreased in later reaction cycles, but they were still more than those over PdCuD and PdCuI catalysts. The overall reaction selectivity in each cycle was calculated and is shown in part B of Figure 1. The selectivity to N2O and N2 over the PdCuD and PdCuI catalysts were 90.2% and 87.9%, respectively. Alumina supported catalysts showed much lower reaction selectivity because of the production of a large amount of nitrite and/or ammonium. The corresponding selectivities were 61.1% and 62.1% over PdCuAl2O3D and PdCuAl2O3I, respectively. In all three cycles of the reaction, the selectivities 4068
dx.doi.org/10.1021/es104050h |Environ. Sci. Technol. 2011, 45, 4066–4072
Environmental Science & Technology
ARTICLE
Table 1. Denitration Performance of PdCuD Catalysts Reduced at Different Temperatures a catalysts PdCuD 300
PdCuD 300b PdCuD 400
PdCuD 500
nitrate (ppm)
nitrite (ppm)
ammonium (ppm)
selectivity (%)
rate (mmol min1 gcata1)
0
0.8
2.4
90.2
0.045
7.8
2
1.9
89.6
0.041
37.7
2.6
1.6
85.0
0.028
1.1
0.6
3.8
81.5
0.044
20.9
2.1
2.6
84.4
0.035
0
1.1
4.8
81.1
0.045
15.6
1.2
2.6
86.8
0.038
39.2 4.2
3.2 1.1
1.7 4
82.7 83.2
0.027 0.043
42.2
1.4
1.2
89.2
0.026
49
1.7
1.8
82.6
0.023
a
All catalysts were tested in the solution with successive pulse of nitrate, the concentrations of anions were detected after 90 min of reaction. b Flow rate of CO2: 50 mL/min.
over the PdCuD and PdCuI catalysts were higher than those over the PdCuAl2O3D and PdCuAl2O3I catalysts. From the data above, it was concluded that ASA-supported catalysts are more advantageous for nitrate hydrogenation than aluminasupported catalysts in terms of activity, selectivity and stability. The PdCuD catalyst exhibited the best performance among the four catalysts. In the following work, we optimized this PdCuD catalyst and found possible reasons for its deactivation as well as a regeneration method. It has been widely reported that catalytic performance is highly dependent on the metal ratio of the catalyst.9,14,15 In this work, different ratios of Pd/Cu on PdCuD catalysts were studied; Figure 2 shows the results. The flow rate of H2 was 100 mL/min ad 0.2 g of PdCuD catalyst was used in this case. After 90 min of reaction, the conversion over the catalyst with a Pd/Cu weight ratio of three was only 5.8%. This was greatly enhanced to 81.9% when the ratio of Pd/Cu was raised to four, but it dropped to 16% when the ratio was further raised to five. This indicated that a Pd/Cu weight ratio of four was optimal for the ASA-supported bimetallic catalysts, and the corresponding mole ratio of Pd/Cu was 2.39. Catalytic performance was highly related to the sizes and shapes of metal particles, which early treatment greatly influenced. A series of PdCuD catalysts were prepared at different reduction temperature and their catalytic performance was investigated. The catalysts were repeatedly used in three cycles of reaction to test their stability preliminary and Table 1 summarizes the results. PdCuD 300 clearly exhibits the highest activity and PdCuD 500 had the lowest activity in the three cycles. It indicated that a relatively lower reduction temperature helped the catalyst to get a higher activity. This is because higher thermal reduction temperature will lead to the sintering of metal particles on the surface, which reduces the amount of catalytically active sites on the catalyst surface. All catalysts lost their activity to different extents in the second and third cycles, and higher reducing temperature led to lower catalytic stability. The stability of the catalysts was according to the following trend: reduced at 300 C > reduced at 400 C > reduced at 500 C. Generally, the production of nitrite increased in the second and third cycles whereas the amount of ammonium in the solution decreased. As a result, the overall reaction selectivity did not drop dramatically. When the CO2 flow rate was halved to 50 mL/min, the selectivity and the removal rate of nitrate both decreased to different extent. This was because a lower pH is preferred for catalytic dinitration.
Figure 3. Catalytic performance of regenerated PdCu/ASA catalysts.
Many kinds of denitration catalysts have encountered the problem of deactivation, and the catalysts in this research are no exception. To recover their activity, deactivated PdCuD 300, PdCuD 400, and PdCuD 500 catalysts were collected from solution and dried at 110 C. They were further calcined at 500 C for one hour and reduced at 300 C under high-purity hydrogen flow for two hours. The catalytic performance of these regenerated catalysts (0.25 g per experiment) was then studied in the same reaction system; Figure 3 shows the results. Comparing the data before (Table 1) and after regeneration (Figure 3), recalcination, and reduction treatment was very effective in reactivating the catalysts. In the third cycle of the denitration reaction, the average removal rates of nitrate were 0.028, 0.027, and 0.023 mmol/(gcata 3 min) for the PdCuD 300, PdCuD 400, and PdCuD 500 catalysts, respectively. After regeneration, the corresponding removal rates of nitrate were enhanced to 0.034, 0.040, and 0.045 mmol/(gcatal 3 min) over 90 min. The disappearance rate of nitrate was enhanced by 21.5% for PdCuD 300 catalyst and doubled for PdCuD 500 catalyst. This indicated that the combination of recalcination and reduction was very effective in regenerating the gradually deactivated catalysts. Catalyst Characterization. The particle size and distribution are visible via TEM, as shown in Figure 4. Part A of Figure 4 is a TEM image of fresh PdCuD catalyst, the particles are very 4069
dx.doi.org/10.1021/es104050h |Environ. Sci. Technol. 2011, 45, 4066–4072
Environmental Science & Technology
ARTICLE
Figure 4. TEM images of PdCuD catalysts. Fresh catalyst (A), aged catalyst (B), and regenerated catalyst (C).
uniform on the surface, and have a narrow size distribution. Most of the particles were in the 47 nm range, and the average diameter was 5.4 nm. After three cycles of reaction, the particles grew larger by approximately 2 nm. They became more inhomogeneous on the surface, as shown in part B of Figure 4. The statistic average particle size was 7.5 nm, which was 2.1 nm larger than fresh PdCu/ASA catalyst. After regeneration, the particles became much more scattered, as shown in part C of Figure 4. The average particle size decreased to 6.5 nm. On the basis of the images and reaction data above, particle size and distribution significantly influence catalytic removal rate of nitrate. Bimetallic particles aggregated during catalytic denitration in solution, which lead to catalytic activity loss. The recalcination and reduction process enhanced the particle distribution, thus the nitrate removal rate over the aged catalysts increased. TEM images revealed that particles on fresh PdCuAl2O3D catalysts were distributed in a wider range, mostly 511 nm, and were much larger than those in part A of Figure 4. On the basis of the TEM images, ASA-supported catalysts have great advantages in terms of particle distribution when compared to the alumina-supported catalysts in this work. These highly distributed particles may have a higher resistance to aggregation, which induced a lower activity loss rate. This is consistent with the results in Table 1. PdCuD Catalyst reduced at 500 C with larger particles lost its activity faster than those reduced at lower temperatures. Aggregation of bimetallic particles was also found in other studies.18,23 PtIn, PdIn, and PdCu metallic particles also grew larger after reactions in solution, which led to lower catalytic activity.18,23 The XRD characterization of bimetallic catalysts on ASA was carried out, and the results were shown in Figure S-8 of the Supporting Information. The peaks at 40.6and 47are characteristic of Pd (111) and Pd (200). Cu peaks were not observed due to the relatively small amount of copper used and lack of isolated copper particles on the surface. After three cycles of rection, the characteristic peaks of Pd(0) remained and no peaks for PdOx (x = 1, 2) and CuOx (x = 0.5, 1) were found in XRD spectra. This meant the metals on the catalysts were not oxidized. According to Scherrer’s equation, wider full width at half-maximum (fwhm) of a peak indicates smaller particle size. PdCuDR and PdCuIR showed narrower peaks at 40.6, which meant the particle grew bigger during reaction. The same trend was also found on alumina-supported catalysts. Particles on both PdCuAl2O3D and PdCuAl2O3I catalysts grew bigger after reaction. This is consistent with TEM results. Catalysts prepared with the deposition method had wider fwhm, indicating smaller particle size and better particle distribution, which were favorable to catalytic activity. XPS was used to detect the valence change of metals on the catalysts surface and the results were shown in Figure 5. The 5%
Figure 5. XPS spectrum for Pd/ASA and PdCuD catalysts before and after the reaction.
Pd/ASA catalyst exhibited intensive peaks at 335.7 and 340.9 eV. According to the literature,27 this pair of peaks is characteristic of Pd (0). Fresh PdCuD catalysts reduced at 300 C exhibited peaks at the same positions. After three cycles of reaction, these peaks of the PdCuD catalyst were the same, indicating that the bimetallic catalyst was not oxidized during reaction in water. This was consistent with XRD results and confirmed the catalytic activity loss could not be due to a valence change in the bimetallic particles on catalyst surface. The XPS results also indicated the atomic ratio of Pd to Si and Al on the surface of the catalysts. The loading of palladium for all catalysts was 5% by weight. After adding 1.25% Cu, the ratio of Pd/Si on the surface increased from 0.16 to 0.20. Promoter copper enhanced the distribution of palladium particles on the catalyst, which was advantageous to the catalytic activity. After three reaction cycles, the ratio decreased to 0.15. This was clear evidence that the metal particles aggregated after denitration reactions in solution. A similar trend in the Pd/Al ratio confirmed this conclusion, which was consistent with TEM and XRD results as well. The dissolution of Pd and Cu from the PdCuD catalyst during nitrate reduction in solution was also investigated. The concentrations of Pd2þ and Cu2þ in the solution were detected by ICP-AES after each cycle of reaction. The concentration of PdCuD catalyst in solution was 400 mg/L, so the initial concentrations of Pd and Cu in the suspension were 20 and 5 ppm, respectively. For comparison, the same amount of catalyst was put in the batch reactor with distilled water under strong stirring without H2 and CO2, and the leached amount of Pd2þ and Cu2þ were detected. Little Pd2þ was detected in the blank solution, whereas Cu2þ was not found. Apparently, very little 4070
dx.doi.org/10.1021/es104050h |Environ. Sci. Technol. 2011, 45, 4066–4072
Environmental Science & Technology precursor Pd(NO3)2 remained on the catalyst, and Cu(NO3)2 was completely deposited during catalyst preparation. Subtracting the Pd2þ from the precursors, the concentrations of Pd2þ leached during the reactions were 0.003, 0.003, and 0.032 ppm, respectively. Thus, 0.16% of the Pd on the PdCuD catalyst leached after three reaction cycles. The pH of the solution was steady at approximately 5.4. In the acidic solution, more Cu leached from the catalyst and the leached amount accounted for 1.26% of the total copper on the catalyst after three reaction cycles. The little leaching of Pd and Cu from the catalyst would also have a slightly negative impact on the catalyst stability. Environmental Significance. Active PdCu catalysts were developed on an industrial ASA material in this work and the selectivity over PdCu/ASA toward N2 and N2O in nitrate reduction was no less than 80%. The highest selectivity is 90.2%, which is higher than that over PdCu/Al2O3 catalysts in this work and in other literatures.10,15,17,28 ASA is very cheap and the catalyst preparation method is simple. Bimetallic particles were uniformly distributed on the catalyst surface prepared by codeposition method. Particles aggregation on the catalysts caused gradual deactivation, but the activity is easily recovered through particle redistribution on the surface. The regeneration of denitration catalyst can help to reduce the consumption of noble metals. The deactivation factually happened on industrial catalysts but the regeneration of denitration catalysts was seldom reported in literatures, except for the work of Charplin,29 in which a simple method was proposed to reactivate the catalysts poisoned by sulfide fouling. Further improvement of PdCu/ASA is still needed, catalysts with higher selectivity and stability can also be found in literature, such as PdCu/mordenite.9 To further contribute to the catalytic denitration for groundwater treatment, better catalysts preparation methods and usage of ASA with different surface properties can be tried.
’ ASSOCIATED CONTENT
bS
Supporting Information. Additional information of the preparation methods and characterization of the catalysts (sections S-1 and S-2), the concentration change of nitrite and ammonium during the reactions over PdCuD, PdCuI, PdCuAl2O3D and PdCuAl2O3I catalysts (Figures S-1 and S-2), size distribution of the four catalysts (Figures S-3, S-4, S-5, and S-7), TEM image of PdCu/Al2O3 catalyst (Figure S-6), XRD spectra of PdCu/ASA catalysts (Figure S-8). This material is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION Corresponding Author
*Phone: þ86-010-82544845, fax: þ86-010-82544845, e-mail:
[email protected].
’ ACKNOWLEDGMENT The authors greatly appreciate the financial support from China Postdoctoral Science Foundation (Grant No. 20090460527), National Natural Science Foundation of China (Grant No. 20607023), the Water Pollution Control and Management Project (Grant No.2009ZX07529-004-3), Chinese Academy of Sciences Visiting Professorships for Senior International Scientists (Grant No. 2009G2-28) and the Brook Byers
ARTICLE
Institute for Sustainable Systems. We also thank to Tianjin Chemical Research & Design Institute for providing the supports.
’ REFERENCES (1) Pintar, A. Catalytic processes for the purification of drinking water and industrial effluents. Catal. Today 2003, 77, 451–465. (2) Wang, Y.; Qu, J. H.; Liu, H. J.; Hu, C. Z. Adsorption and reduction of nitrate in water on hydrotalcite-supported Pd-Cu catalyst. Catal. Today 2007, 126, 476–482. (3) Schoeman, J. J.; Steyn, A. Nitrate removal with reverse osmosis in a rural area in South Africa. Desalination 2003, 155, 15–26. (4) Primo, O.; Rivero, M. J.; Urtiaga, A. M.; Ortiz, I. Nitrate removal from electro-oxidized landfill leachate by ion exchange. J. Hazard. Mater. 2009, 164, 389–393. (5) Sahli, M. A. M.; Annouar, S.; Mountadar, M.; Soufiane, A.; Elmidaoui, A. Nitrate removal of brackish underground water by chemical adsorption and by electrodialysis. Desalination 2008, 227, 327–333. (6) Ghafari, S.; Hasan, M.; Aroua, M. K. Bio-electrochemical removal of nitrate from water and wastewater - A review. Bioresour. Technol. 2008, 99, 3965–3974. (7) Vorlop, K. D.; Tacke, T. Chem. Ing. Technol. 1989, 61, 836–837. (8) Yoshinaga, Y.; Akita, T.; Mikami, I.; Okuhara, T. Hydrogenation of nitrate in water to nitrogen over Pd-Cu supported on active carbon. J. Catal. 2002, 207, 37–45. (9) Nakamura, K.; Yoshida, Y.; Mikami, I.; Okuhara, T. Selective hydrogenation of nitrate in water over Cu-Pd/mordenite. Appl. Catal., B 2006, 65, 31–36. (10) Chaplin, B. P.; Roundy, E.; Guy, K. A.; Shapley, J. R.; Werth, C. J. Effects of natural water ions and humic acid on catalytic nitrate reduction kinetics using an alumina supported Pd-Cu catalyst. Environ. Sci. Technol. 2006, 40, 3075–3081. (11) Garron, A.; Lazar, K.; Epron, F. Effect of the support on tin distribution in Pd-Sn/Al2O3 and Pd-Sn/SiO2 catalysts for application in water denitration. Appl. Catal., B 2005, 59, 57–69. (12) Zhang, F. X.; Miao, S.; Yang, Y. L.; Zhang, X.; Chen, J. X.; Guan, N. J. Size-dependent hydrogenation selectivity of nitrate on Pd-Cu/ TiO2 catalysts. J. Phys. Chem. C 2008, 112, 7665–7671. (13) Wan, D. J.; Liu, H. J.; Zhao, X.; Qu, J. H.; Xiao, S. H.; Hou, Y. N. Role of the Mg/Al atomic ratio in hydrotalcite-supported Pd/Sn catalysts for nitrate adsorption and hydrogenation reduction. J. Colloid Interface Sci. 2009, 332, 151–157. (14) Xu, Z. Y.; Chen, L. Q.; Shao, Y.; Yin, D. Q.; Zheng, S. R. Catalytic hydrogenation of aqueous nitrate over Pd-Cu/ZrO2. Ind. Eng. Chem. Res. 2009, 48, 8356–8363. (15) Rao, G. R.; Mishra, B. G. Al-pillared clay supported CuPd catalysts for nitrate reduction. J. Porous Mater. 2007, 14, 205–212. (16) Dodouche, I.; Barbosa, D. P.; Rangel, M. C.; Epron, F. Palladium-tin catalysts on conducting polymers for nitrate removal. Appl. Catal., B 2009, 93, 50–55. ao, J. J. M.; Pereira, M. F. R. P. Pd-Cu and (17) Soares, O. S. G. P.; Orf~ Pt-Cu catalysts supported on carbon nanotubes for nitrate reduction in water. Ind. Eng. Chem. Res. 2010, 49, 7183–7192. (18) Neyertz, C.; Marchesini, F. A.; Boix, A.; Miro, E.; Querini, C. A. Catalytic reduction of nitrate in water: Promoted palladium catalysts supported in resin. Appl. Catal., A 2010, 372, 40–47. (19) Sun, D. P.; Yang, J. Z.; Li, J.; Yu, J. W.; Xu, X. F.; Yang, X. J. Novel Pd-Cu/bacterial cellulose nanofibers: Preparation and excellent performance in catalytic denitrification. Appl. Surf. Sci. 2010, 256, 2241–2244. (20) Pintar, A.; Batista, J.; Musevic, I. Palladium-copper and palladium-tin catalysts in the liquid phase nitrate hydrogenation in a batchrecycle reactor. Appl. Catal., B 2004, 52, 49–60. (21) Sa, J.; Aguera, C. A.; Gross, S.; Anderson, J. A. Photocatalytic nitrate reduction over metal modified TiO2. Appl. Catal., B 2009, 85, 192–200. 4071
dx.doi.org/10.1021/es104050h |Environ. Sci. Technol. 2011, 45, 4066–4072
Environmental Science & Technology
ARTICLE
(22) Sa, J.; Berger, T.; F€ottinger, K.; Riss, A.; Anderson, J. A.; Vinek, H. Can TiO2 promote the reduction of nitrates in water?. J. Catal. 2005, 234, 282–291. (23) Poduval, D. G.; Veen, J. A. R.; Rigutto, M. S.; Hensen, E. J. M. Brønsted acid sites of zeolitic strength in amorphous silica-alumina. Chem. Commun. 2010, 46, 3466–3468. (24) Hensen, E. J. M.; Poduval, D. G.; Magusin, P. C. M. M.; Coumans, A. E.; Veen, J. A. R. Formation of acid sites in amorphous silica-alumina. J. Catal. 2010, 269, 201–218. (25) Niquille-R€othlisberger, A.; Prins, R. Hydrodesulfurization of 4, 6-dimethyldibenzothiophene over Pt, Pd, and Pt-Pd catalysts supported on amorphous silicaalumina. Catal. Today 2007, 123, 198–207. (26) Marchesini, F. A.; Irusta, S.; Querini, C.; Miro, E. Spectroscopic and catalytic characterization of Pd-In and Pt-In supported on Al2O3 and SiO2, active catalysts for nitrate hydrogenation. Appl. Catal., A 2008, 348, 60–70. (27) Xie, Y. B.; Liu, C. J. Stability of ionic liquids under the influence of glow discharge plasmas. Plasma Process Polym. 2008, 3, 239–245. (28) Sa, J.; Vinek, H. Catalytic hydrogenation of nitrate in water over a bimetallic catalyst. Appl. Catal., B 2005, 57, 247–256. (29) Chaplin, B. P.; Shapley, J. R.; Werth, C. J. Regeneration of sulfur-fouled bimetallic Pd-based catalysts. Environ. Sci. Technol. 2007, 41, 5491–5497.
4072
dx.doi.org/10.1021/es104050h |Environ. Sci. Technol. 2011, 45, 4066–4072
