Applied Catalysis A: General 425–426 (2012) 145–152
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General journal homepage: www.elsevier.com/locate/apcata
Characterization of (Sn and Cu)/Pd catalysts for the nitrate reduction in natural water Cristina Franch a,1 , Enrique Rodríguez-Castellón b,∗ , Álvaro Reyes-Carmona b , Antonio E. Palomares a,1 a b
Instituto de Tecnología Química (UPV-CSIC) Universidad Politécnica de Valencia, Consejo Superior de Investigaciones Científicas, Avenida de los Naranjos s/n, 46022 Valencia, Spain Departamento de Química Inorgánica, Cristalografía y Mineralogía (Unidad Asociada al I.C.P.-C.S.I.C.) Facultad de Ciencias, Universidad de Málaga, 29071 Málaga, Spain
a r t i c l e
i n f o
Article history: Received 13 October 2011 Received in revised form 25 February 2012 Accepted 6 March 2012 Available online 15 March 2012 Keywords: Nitrate catalytic reduction Natural water Continuous process Sn–Pd Cu–Pd, XPS
a b s t r a c t The aim of this work is to characterize different (Cu and Sn)/Pd catalysts, supported on alumina, used for the catalytic removal of nitrates in natural water. The catalysts have been prepared with a Pd/(Cu or Sn) ratio of 2 but with different metal contents. Their activity and selectivity have been studied using a continuous stirred-tank reactor with nitrate polluted water from an aquifer. The catalysts have been characterized both before and after reaction by XPS, XRD, XRF, adsorption–desorption N2 isotherms at 196 ◦ C and TEM. XPS results show changes of the Pd/Sn surface atomic ratio upon catalyst activation and after reaction and the coexistence of different oxidation states for the active metals. The studied catalysts are active, being the catalyst with the best performance that with the highest metallic dispersion and with the lowest phases segregation. The characterization of the catalysts after reaction shows that catalyst deactivation could be related with the non-reversible oxidation of the Sn–Pd couple. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Nitrates groundwater pollution is an important problem in many rural and populated areas from the world. This pollution is related to the intensive use of fertilizers, the agricultural livestock production industry and by domestic and industrial effluents. In some of these areas, the natural water nitrate concentration is over the permitted limit for the human consumption, which in Europe is 50 mg L−1 . There are different commercial techniques for removing nitrates from water such as reverse osmosis and electrodialysis. They are effective but they generate a polluted waste that should be treated or disposed of [1]. Another possibility is the use of biological processes which are based on the nitrates reduction to nitrogen using microorganisms [2]. However, there are concerns regarding possible bacterial contamination. Another new technique, still under research is the catalytic reduction of nitrates to nitrogen, using hydrogen as reductant [2–10]. The catalysts used are based in a combination of a noble metal, such as Pd, Pt, Rh or Ir and another metal, such as Cu, Sn, Ag or In [5–8,11–15]. The best results are achieved with Cu–Pd or Sn–Pd catalysts supported on alumina, although recently new catalysts based in Pt have been proposed
∗ Corresponding author. Tel.: +34 952131873; fax: +34 952132000. E-mail addresses:
[email protected] (E. Rodríguez-Castellón),
[email protected] (A.E. Palomares). 1 Tel.: +34 963879632; fax: +34 963877639. 0926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.apcata.2012.03.015
with very interesting results [16,17]. The problem with this reaction is the formation of undesirable subproducts such as nitrites or ammonia and in order to obtain a commercial catalyst this problem must be solved. Although some advances have been made in recent years, the role of the different metallic active sites in the catalyst activity and selectivity is not clear. Some studies have appeared recently studying Cu–Pd [4,18–20], Cu–Pt [18] and Pd/SnO2 [21] catalysts by XPS and other techniques. In the case of Pd/SnO2 catalysts, it has been shown [21] that the catalytic activity is governed by the Pd reducibility and the degree of metal-support interaction between Pd and SnO2 , where Pd2+ is dissolved in the SnO2 support. Although these phenomena can be studied with the aid of spectroscopic X-ray techniques such as photoelectronic spectroscopy (XPS) and fluorescence spectroscopy (XRF), there are not many papers studying other Sn–Pd catalytic systems. This is the case of the Sn–Pd/Al2 O3 catalysts used for the catalytic reduction of nitrates in water. The catalyst characterization by these techniques is required for a better understanding of the catalytic role of the metallic active sites. On the other hand, the nitrate catalytic reduction has been studied with catalysts used in batch or semi-batch equipment and with distilled water containing nitrates whereas only a few studies have been reported using continuous reactors and with natural water [15,18,22]. In this work we have characterized by different techniques, as XPS, XRD, adsorption–desorption N2 isotherms at −196 ◦ C, XRF and TEM, different Sn–Pd catalysts supported on
146
C. Franch et al. / Applied Catalysis A: General 425–426 (2012) 145–152
alumina used for the catalytic reduction of nitrates from a polluted aquifer in a continuous reactor. 2. Experimental 2.1. Materials The supports used for the catalysts were ␥-Al2 O3 supplied by Merck (99.5% purity, surface area 139 m2 g−1 ) and a high surface area alumina (specific surface area 355 m2 g−1 ), labeled as HSA, synthesized in our laboratory by a standard precipitating procedure using two solutions [9]. The first solution contained Al(NO3 )3 ·9H2 O 3.4 mol L−1 , the second solution contained 6.3 mol L−1 of NaOH and 1.4 mol L−1 of Na2 CO3 to obtain the total precipitation of the aluminium ions in the former solution and to fix the pH at a value of 13. Both solutions were added, while vigorously stirring, at a total flow-rate of 1 cm3 min−1 , at room temperature and atmospheric pressure. The gel was aged under atmospheric pressure conditions at 60 ◦ C for 12 h, then filtered and washed with distilled water until the pH was 7 and carbonates were not detected in the filtrate. The material was dried at 100 ◦ C during 12 h and calcined at 500 ◦ C for 5 h in order to obtain the high surface area alumina (HSA). The bimetallic catalysts were prepared by incipient wetness impregnation with the soluble salts of the desired metals. The impregnation was made in two steps, first the support was impregnated with a SnCl4 ·5H2 O or Cu(NO3 )2 ·3H2 O solution, after that, the material was dried at 60 ◦ C for 4 h and calcined at 500 ◦ C for 1 h. In a second step, the catalyst was impregnated with a Pd(NO3 )2 ·2H2 O solution, then the material was dried and calcined at 500 ◦ C for 1 h. Finally, the catalyst was activated in a hydrogen flow at 500 ◦ C for 4 h. This temperature was chosen in order to compare these results with those previously reported by the same authors [4,9,15]. The catalysts were pelletized to obtain a grain size of 0.1–0.4 mm. Previous experiments have shown that using this particle size there are not diffusional limitations. Samples have been labeled as x/y Sn–Pd z, where x is the amount of Sn (%wt.), y the amount of Pd (%wt.) and z the support used. The Cu–Pd catalyst was similarly labeled. The composition and the textural properties of the tested catalysts are shown in Table 1. As can be seen, the surface area of the catalysts decreases after the incorporation of the metallic sites, being this more pronounced for the highest surface area catalyst. Nevertheless, the surface area of this catalyst is more than two times higher than that of the other catalysts. 2.2. Catalytic tests The catalysts were tested in a continuous stirred-tank reactor. The start-up procedure consists in filling the reactor with 600 cm3 of natural water that is bubbled with hydrogen during 60 min (at 250 cm3 min−1 , P = 1 atm). After that, the catalyst was introduced into the reactor and a water flow rate of 5 cm3 min−1 was maintained during the experiment. In order to keep the catalyst inside the reactor, the water outflow pass through a 10 m membrane filter placed in the reactor outlet. The catalyst mass used for the test was 3 g and a mixture of CO2 and H2 (1:1) with a total flow of 500 cm3 min−1 was introduced into the reactor during the reaction. The experiments were carried out at standard conditions (25 ◦ C Table 1 Composition (wt%) and textural properties of the tested catalysts. Catalyst
Pd%
Sn%
Cu%
Surface area (m2 g−1 )
2.5/5 Sn–Pd Al2 O3 1.5/3 Sn–Pd Al2 O3 2.5/5 Cu–Pd Al2 O3 0.5/1 Sn–Pd HSA
5 3 5 1
2.5 1.5 – 0.5
– – 2.5 –
125 134 128 277
and 1 atm) and with a stirring velocity of 900 rpm. The reaction progress was followed by taking, at defined periods, small aliquots at the outlet of the reactor for the photometric determination of the nitrate, nitrite and ammonia concentration. Measurements were made in an UV/VIS spectrophotometer (Jasco UV/VIS spectrophotometer, model V-530) combined with reagent kits for the determination of nitrate (Spectroquant® nitrate test from Merck, measuring range 1–90 mg L−1 at 515 nm), nitrite (Spectroquant® nitrite test from Merck, measuring range 0.02–3 mg L−1 at 525 nm) and ammonia (Spectroquant® ammonia test from Merck, measuring range 0.01–3.5 mg L−1 at 690 nm). In order to study the possible leaching of the catalyst metals during the reaction, the concentration of Pd, Sn and Cu in the water outlet was directly analyzed ICP-OES on a Varian 715-ES ICP-Optical Emission Spectrometer. The limit of detection of this system was 1.5 g/L for Cu, 10 g/L for Sn and 5 g/L for Pd. 2.3. Samples characterization X-ray diffraction patterns were collected using a Philips X’Pert diffractometer equipped with a graphite monochromator, operating at 40 kV and 50 mA and employing nickel-filtered CuK␣ radiation ( = 0.1545 nm). Textural properties were determined from N2 adsorption isotherms at 77 K, which were obtained on a Micromeritics Tristar 3000 instrument. Prior to measurements the samples were outgassed at 400 ◦ C and vacuum overnight. The specific surface areas were obtained using the BET equation. Transmission electron microscopy (TEM) images were obtained on a Philips CM10 electron microscope operating at 100 kV. The samples were prepared by dispersing the powder products in pure dichloromethane and transferring them to carbon-coated copper grids. The average metal crystallite diameter and the metal dispersion in the reduced samples were estimated from CO chemisorptions with the double isotherm method using a Micromeritics ASAP 2010C equipment. Prior to adsorption, the samples were reduced in situ by flowing hydrogen at the same temperature applied for the catalyst activation (500 ◦ C for 4 h). After reduction the samples were degassed at 1333 Pa for 2 h at this temperature and then cooled down to 35 ◦ C. Then, pure CO was admitted and the first adsorption isotherm (i.e. the total CO uptake) was measured. After evacuation at 35 ◦ C, the second isotherm (i.e. the reversible CO uptake) was taken. The amount of chemisorbed CO was obtained by subtracting the two isotherms. The metal dispersion was calculated from the amount of irreversibly adsorbed CO assuming a Pd/CO = 1 atomic ratio stoichiometry, in line with previous reports [23–26]. The average metal crystallite diameter was determined from chemisorption data assuming spherical geometry for the metal particles according to the procedure described by Anderson [27]. X-ray fluorescence by dispersive energy (XRF) were acquired using a Horiba XGT-5000 spectrometer, working with a Rh source at 50 kV and 1 mA, focalized to work with a 10 m probe. Average of 10 points values with an acquisition time of 400 s were used to calculate elemental ratios. X-ray photoelectron spectra (XPS) were registered using a Physical Electronics PHI 5700 spectrometer with a non-monochromatic Mg K␣ radiation (300 W, 15 kV, h = 1253.6 eV) as excitation source. Spectra were recorded at a 45◦ take-off angle by a concentric hemispherical analyser operating in the constant pass energy mode at 25.9 eV, using a 720 m diameter analysis area. Under these conditions the Au 4f7/2 line was recorded with 1.16 eV FWHM at a binding energy of 84.0 eV. Core level C 1s, N 1s, O 1s, Al 2p, Cl 2p, Cu 2p, Pd 3d and Sn 3d signals were recorded. The spectrometer energy scale was calibrated using Cu 2p3/2 , Pd 3d5/2 and Sn 3d5/2 photoelectron lines at 932.7, 368.3 and 486.0 eV, respectively. Charge referencing was
C. Franch et al. / Applied Catalysis A: General 425–426 (2012) 145–152
8 1000 S/cm 150 ppm 79 ppm 200 ppm 101 ppm 50 ppm 40 ppm 28 ppm 0.16 ppm
done against the Al 2p signal at 74.3 eV. Samples were studied after preparation, after activation and after reaction. In order to avoid reoxidation, after reduction treatment, samples were immersed in cyclohexane before analysis. Powdered solids were mounted on a sample holder without adhesive tape and kept overnight in a high vacuum chamber before they were transferred inside the analysis chamber of the spectrometer. Each region was scanned with several sweeps until a good signal to noise ratio was observed. The pressure in the analysis chamber was maintained below 10−7 Pa. PHI ACCESS ESCA-V6.0 F and Multipak 8.2b software packages were used for acquisition and data analysis. A Shirley-type background was subtracted from the signals. Recorded spectra were always fitted using Gauss–Lorentz curves in order to determine the binding energy of the different element core levels more accurately. The accuracy of the binding energy (BE) values was within ±0.1 eV. Short acquisition times (<10 min) were used in order to avoid the photoreduction of the Cu species [28]. 3. Results and discussion 3.1. Catalysis The studies were made with water from a polluted aquifer located in the Autonomous Region of Valencia (Spain). The main characteristics of the water are described in Table 2, with the most significant parameters being the hardness of the water, the high nitrate, sulfate and chloride content and its high conductivity. The catalysts were tested during 8 h in a continuous stirred-tank reactor. In order to evaluate the possible leaching of the catalyst metals, the Pd, Cu and/or Sn contents of the water stream were analyzed by ICP-OES spectrometry. In all samples, the metal measured concentrations were under the detection limit, showing that the metal leaching in the catalyst is not produced during the reaction. In Fig. 1 the conversion and the selectivity obtained with a Sn–Pd catalyst supported on a commercial alumina are compared with those obtained with a Cu–Pd catalyst, both with 5 wt%. Pd and with a Pd/Metal ratio of 2. These materials have been described as active catalysts for this reaction when using distilled water containing nitrates [11,19,29]. As can be seen in Fig. 1 the conversion obtained with both catalysts, using natural water is quite similar, obtaining 85–95% conversion after 8 h of reaction. Nevertheless the selectivity towards ammonia is higher with the catalyst containing Cu, for this reason the catalyst containing Sn is preferred over that containing copper. The formation of nitrite is negligible for both catalysts. From an economical point of view, a reduction in the active phase metal content is very important, since palladium is costly. A catalyst with lower palladium content was prepared and its activity was compared with the activity of the catalyst with higher metal content. In the new catalyst, the Pd:Sn ratio was maintained because previous works have shown that a metal ratio of 2 is necessary in order to have an active catalyst [15,19,30]. As can be seen in Fig. 2, a 40% decrease in palladium content results in only a 20% decrease in conversion after 7 h of reaction, the selectivity towards ammonia being lower with the catalyst containing less Pd. These
NO3- conversion - NH4+ selecvity (%)
pH Conductivity [Ca2+ ] [Cl− ] [SO4 2− ] [NO3 − ] [Na+ ] [Fe2+ ] [Al3+ ] [F− ]
100
80
60
40
20
0 0
1
2
3
4
5
6
7
8
Time (h) Fig. 1. Nitrate conversion (dark symbols) and ammonium selectivity (open symbols) of Cu–Pd () and Sn–Pd () catalysts supported on alumina (5% Pd, Pd:(Cu or Sn) ratio = 2), using natural water from a polluted aquifer.
NO3- conversion - NH4+ selectivity (%)
Table 2 Characteristics of the natural water used in the nitrates reduction reaction.
147
100
80
60
40
20
0 0
1
2
3
4 5 Time (h)
6
7
8
Fig. 2. Influence of the metal content in the nitrate conversion (dark symbols) and ammonia selectivity (open symbols) of the Sn–Pd Al2 O3 catalysts using natural water from a polluted aquifer. 2.5/5 Sn–Pd Al2 O3 (), 1.5/3 Sn–Pd Al2 O3 (), 0.5–1 Sn–Pd HSA (♦).
results led us to lower the quantity of palladium whilst increasing its dispersion on the catalyst surface using a high surface area alumina prepared in our laboratory as a support. A catalyst containing only 1% Pd and with the same Pd:Sn ratio was prepared in this way. The results obtained with this catalyst are also shown in Fig. 2. As it can be seen, when reaching the steady state, the conversion obtained with the catalyst containing 5 times less palladium is only 1.2 times lower than that obtained with the catalyst containing 5% Pd. In Table 3, the nitrate moles reacted during 7 h per mol of palladium and per second of the different catalysts are compared. It is shown that the catalyst activity increases with decreasing palladium content, the best results being obtained with the catalyst Table 3 Mol of nitrates converted per mol of palladium per second of the different Pd/Sn samples. Sample
mol NO3 − mol−1 Pd s−1
2.5/5 Sn–Pd Al2 O3 1.5/3 Sn–Pd Al2 O3 0.5/1 Sn–Pd HSA
8.13 × 10−5 1.35 × 10−4 4.39 × 10−4
C. Franch et al. / Applied Catalysis A: General 425–426 (2012) 145–152
NO3- conversion - NH4+ selectivity (%)
148
and a continuous reactor. It is not clear the influence of the different ions present in the water in the catalyst activity. Since TOC concentration is very low, dissolved organic compounds do not exhibit any appreciable influence on the reaction, as it has been also shown in other papers [33]. Thus the inhibitive effect is therefore due to the presence of various inorganic ionic species in the liquid-phase. This effect cannot be directly assigned to the additive effect of single ions, because the addition of the results obtained, when using distilled water with nitrates and with another ion, is quite different than those obtained when using natural water containing many different ions. It seems that there is a synergic effect of the different ions present in natural water that appreciably influences the reaction kinetics, decreasing the catalyst activity and increasing the selectivity towards ammonia.
100
80
60
40
20
0 0
1
2
3
4
5
6
7
8
Time (h)
3.2. XRD
Fig. 3. Nitrate conversion (dark symbols) and ammonium selectivity (open symbols) of 0.5/1 Sn–Pd (♦) catalyst supported on a high surface area alumina and 2.5/5 Sn–Pd (夽) catalyst supported on a high surface area alumina.
containing 0.5% Sn and 1% Pd supported on a high surface area alumina. For a better comparison a catalyst containing 2.5% Sn and 5% Pd supported on a high surface area alumina was also prepared and its activity was compared with that of the 0.5/1 Sn–Pd catalyst supported on a high surface area alumina. The results obtained are shown in Fig. 3. It can be observed that the results obtained with the 2.5/5 Sn–Pd catalyst supported on a high surface area alumina are very similar to those obtained with the same catalyst supported on a commercial alumina, confirming the previously discussed results. On the other hand, as it can be seen in Fig. 3, the selectivity towards ammonia obtained with the catalyst containing 1% Pd is much lower, during all the reaction time, than that obtained with the catalyst containing 5% Pd. As it occurred with the previous catalysts, the formation of nitrites during the reaction with this catalyst was negligible. Although a direct comparison cannot be made with both catalysts because the conversions obtained are different, the results suggest that the catalyst containing less Pd and supported on a high surface area is not only of interest due to its lower cost but also owing to its high activity and selectivity. It has to be taken into account that a lower selectivity towards nitrogen is always obtained when using natural water than when using distilled water containing nitrates [31–33], and to the best of our knowledge these are the best results described for this reaction using natural water
The X-ray diffractograms of the samples used in the reaction are shown in Fig. 4. As can be observed, the diffractograms present the peaks corresponding to the ␥-alumina main diffractions at 2 = 37.5, 45.9, 66.9 and 85.1◦ in all cases. The high surface area alumina displays the most defined ␥-alumina peaks, indicating a high cristallinity of the oxide synthesized. Before activation and reaction the XRD also show peaks at 2 = 33.9, 42.2, 54.7, 60.5 and 71.5◦ , corresponding to PdO. As expected, the samples containing higher palladium content show more intense PdO peaks suggesting a worst dispersion of the noble metal. Peaks from copper or tin species are not observed, because the low content and high dispersion of the metals. After activation and reaction, the peaks assigned to PdO disappear, while the peaks corresponding to Pd0 at 2 = 39, 45.4, 66.2, 82.1 and 86.6◦ appearing instead. This indicates that most of the PdO sites have been reduced, although still a shoulder can still be observed in some samples at 2 = 33.9◦ suggesting the presence of PdO.
3.3. TEM The samples 0.5/1 Sn–Pd HSA and 2.5/5 Sn–Pd Al2 O3 were characterized by TEM. Approximately one hundred metallic nanoparticles were considered for the determination of the crystallite size distribution. In Fig. 5 the TEM image and the crystallite size distribution of the 2.5/5 Sn–Pd alumina sample are shown, which mainly comprises particles in the range of 2–9 nm. Fig. 6 shows the TEM image and the crystallite size distribution of the 0.5/1
Before reacon
Aer reacon
Pd Al2O3 PdO
(d) (c) (b) (a) 20
60
40
2 θ (º)
80
20
40
60
80
2 θ (º)
Fig. 4. X-ray diffractograms of the samples before and after reaction: (a) 2.5/5 Sn–Pd Al2 O3 ; (b) 1.5/3 Sn/Pd Al2 O3 ; (c) 0.5/1 Sn–Pd HSA and (d) 2.5/5 Cu–Pd Al2 O3 .
C. Franch et al. / Applied Catalysis A: General 425–426 (2012) 145–152
149
within this support can explain the high activity obtained with this catalyst. 3.5. X-ray photoelectron spectroscopy
Fig. 5. TEM image and crystallite size distribution of the 2.5/5 Sn–Pd Al2 O3 catalyst.
Fig. 6. TEM image and crystallite size distribution of the 0.5/1 Sn–Pd HSA catalyst.
Sn–Pd HSA catalyst, showing mainly monodispersed nanoparticles of 3–4 nm. 3.4. Metallic dispersion Metallic dispersion of Pd species was calculated using CO chemisorption isotherms at 35 ◦ C. These data are presented in Table 4, showing an increase of the metallic dispersion while decreasing the metal content. The small average crystal size and the highest metallic dispersion of the 0.5/1 Sn–Pd HSA sample should be pointed out. The high dispersion of the Pd centers achieved Table 4 Palladium metallic parameters. Catalyst
Active metal area (m2 g−1 )
Metal dispersion (%)
Crystal size (nm)
2.5/5 Sn–Pd Al2 O3 1.5/3 Sn–Pd Al2 O3 0.5/1 Sn–Pd HSA
0.4 0.9 1.7
1.9 6.4 38.9
57.6 17.6 2.9
In order to obtain information about the evolution of the oxidation state of the active phases, X-ray photoelectron spectroscopy studies of fresh, reduced (activation step) and used samples were performed, centering our interest in the dispersion of the active phases and the chemical state. Core level C 1s signals were decomposed in two contributions at 284.8 and 288.0 eV, assigned to adventitious carbon and surface carbonates, respectively. The core level N 1s signal was only detected in used samples as a weak peak at 399.4 eV and associated to ammonia adsorbed over the surface. All samples show a band with the main peak centered in the Cl 2p region at 198.3 eV, assigned to chloride ions from the Sn salt precursor used. O 1s core level signals cannot be clearly decomposed because of the coexistence of several oxygen containing species (Al2 O3 , SnO2 and PdO) that give rise to a photoemission between 530 and 531.3 eV. Table 5 summarizes the surface chemical concentration (in atomic percentages) for all the samples. Fresh 2.5/5 Sn–Pd Al2 O3 sample exhibits a low surface atomic concentration of Al and a high one of Pd and Sn due to the poor dispersion of the active phases. After reduction, the atomic ratio Pd/Sn decreases, indicating an increase of the metallic dispersion. After reaction, the observed atomic ratio Pd/Sn is similar to that observed after reduction. That is, the metallic dispersion is maintained. At the same time the surface concentration of Al increases to values similar to those observed in the other studied catalysts. In the case of 0.5/1 Sn–Pd HSA sample, a decrease of the surface atomic concentration of the active metals are also observed upon reduction and after the catalytic test. Interestingly, the found Pd/Sn atomic ratio (1.89) for the used sample is very close to the theoretical one (2.2). The ratios are far from this value for the fresh and reduced sample. It should be pointed out that as no leaching was observed in the catalyst metals, the changes observed by XPS in the surface metal content after reaction cannot be related with a metal lixiviation. A study of the oxidation state evolution of the active phases is very important because it has been described [5] that mainly, reduced Pd–(Cu or Sn) sites promote the first nitrate reduction step to nitrite. These sites maintain reduced by spill-over hydrogen from Pd centers. Over metallic Pd sites, reduction of nitrite to NH3 or N2 also occurs. XPS can provide relevant information on this subject [11,34–36]. In this way the photoelectronic profiles of Pd 3d core level of the 2.5/5 Sn–Pd Al2 O3 and 0.5/1 Sn–Pd HSA samples are shown in Fig. 7. The core level Pd 3d5/2 signal shows two contributions at 334.6 eV for Pd0 and 336.9 eV for PdO. A reduction of Pd2+ to Pd0 after activation treatment and a partial re-oxidation to Pd2+ after reaction were observed [37], as shown in Table 6. The existence of a small contribution of Pd4+ species is noted in the 2.5/5 SnPd–Al2 O3 fresh sample, with a Pd 3d5/2 signal with maximum at 338.0 eV. The photoelectronic profiles of the Sn 3d core level are shown in Fig. 8. Reduction of the Sn species was also observed in the catalysts after activation and in the used catalysts. The Sn 3d5/2 core level signal can be decomposed in two contributions at 484.7 and 486.7 eV for Sn0 and SnOX (Sn4+ and Sn2+ ) species, respectively. It is difficult to distinguish between Sn2+ and Sn4+ due to the proximity of the B.E. values for these oxidation states. The obtained values are summarized in Table 6 and they give evidence of the presence of Sn species with different oxidation states after reduction. After reaction, the observed Sn 3d binding energy values are similar to those observed after reduction. These values and their similarity after reaction are indicative of a strong interaction Sn–Pd, where Sn is
150
C. Franch et al. / Applied Catalysis A: General 425–426 (2012) 145–152
Table 5 Surface atomic concentration (%) of studied samples. Sample 2.5/5 Sn–Pd Al2 O3 Fresh Reduced Used 1.5/3 Sn–Pd Al2 O3 Fresh Reduced Used 0.5/1 Sn–Pd HSA Fresh Reduced Used 2.5/5 Cu–Pd Al2 O3 Fresh Reduced Used
N%
O%
Al%
Cl%
Cu%
Pd%
Sn%
– – 0.42
63.68 61.31 60.95
9.44 22.80 33.84
2.85 0.71 0.17
– – –
15.56 9.68 2.66
8.46 5.48 1.95
– – 0.31
60.72 60.39 62.15
34.56 35.16 34.64
0.79 0.48 0.16
– – –
0.81 3.45 1.44
3.11 3.45 1.29
– – 0.33
63.40 61.96 63.53
33.35 35.37 34.47
0.36 0.50 0.14
– – –
2.56 1.57 0.99
6.82 1.75 0.53
– – 0.38
59.87 59.50 60.52
36.17 37.48 36.56
– – –
0.86 0.46 0.72
3.10 2.55 1.81
– – –
Fig. 7. Photoelectronic profiles of Pd 3d core level of the samples (a) 2.5/5 Sn–Pd Al2 O3 and (b) 0.5/1 Sn–Pd HSA.
(a)
2.5/5 Sn-Pd Al2O3
Sn 3d
(b)
x+
0.5/1Sn-Pd HSA
Sn
Sn
0
Sn 3d
Reduced
Used Intensity (a.u.)
Intensity (a.u.)
Used
Reduced
Fresh
500 498 496 494 492 490 488 486 484 482 480
Binding Energy (eV)
Fresh
500 498 496 494 492 490 488 486 484 482 480
Binding Energy (eV)
Fig. 8. Photoelectronic profiles of Sn 3d core level of the samples (a) 2.5/5 Sn–Pd Al2 O3 and (b) 0.5/1 Sn–Pd HSA.
C. Franch et al. / Applied Catalysis A: General 425–426 (2012) 145–152
151
Table 6 Binding energies (in eV) of studied samples (fresh, reduced, used). Sample 2.5/5 Sn–Pd Al2 O3 Fresh Reduced
Sn 3d5/2 (eV/%)
Pd 3d5/2 (eV/%)
Cu 2p3/2 (eV/%)
486.1
336.0 (82) 338.2 (18) 334.6 (92) 336.9 (8) 334.6 (86) 336.9 (14)
–
484.7 (39) 486.6 (61) 484.6 (26) 486.3 (76)
Used 1.5/3 Sn–PdAl2 O3 Fresh Reduced
– –
486.2 484.0 (40) 486.2 (60) 484.0 (24) 486.6 (76)
336.0 335.1
– –
334.7 (80) 336.9 (20)
–
486.7 484.0 (38) 486.7 (62) 484.0 (39) 486.6 (61)
335.7 334.0 (73) 335.3 (27) 334.7 (80) 336.5 (20)
– –
–
335.8
Reduced
–
Used
–
334.0 (78) 335.8 (22) 334.5 (73) 335.9 (27)
933.5 Isat /Ipp = 0.55 931.9 Isat /Ipp = 0 932.2 (47) 934.6 (53) Isat /Ipp = 0.33
Used 0.5/1 Sn–Pd HSA Fresh Reduced Used 2.5/5 Cu–PdAl2 O3 Fresh
partially oxidized and associated with Pd. Some authors [13,18,20] have proposed that the redox cycles of the Pd and Sn (or Cu) species explain this reaction, that is, the active sites are reoxidized and reduced repeatedly, for this reason a non-reversible oxidation of some Pd–Sn species can be the reason for the slow deactivation observed for these catalysts after some hours of reaction. In this way the 0.5/1 Sn–Pd HSA catalyst, presents the lowest deactivation ratio (Fig. 2) and is the sample with less oxidized species after reaction (Table 6). Nevertheless an additional deactivation produced by the adsorption or deposition of the other ions present in water cannot be discarded for all the catalysts tested. The Cu–Pd system has also been studied and compared with the Sn–Pd system using the same amount of active phases (2.5% Cu and 5% Pd). While the Pd 3d core level signal shows a similar behavior to that observed for Sn–Pd, a total reduction of Cu was achieved after activation. These results, shown in Table 6 and Fig. 9, are concluded from the shift to lower B.E. values and the practically disappearing
Cu 2p 2.5/5.0Cu-Pd Al2O3
Shake up Satellites
Intensity (a.u.)
Used
Reduced
Fresh
970
965
960
955
950
945
940
935
930
925
Binding Energy (eV) Fig. 9. Photoelectronic profiles of Cu 2p core level of the sample 2.5/5 Cu–Pd Al2 O3 .
–
Table 7 Fresh and used XRF and XPS weight ratios from samples. Sample
2.5/5 Sn–Pd Al2 O3 Fresh Used 1.5/3 Sn–Pd Al2 O3 Fresh Used 0.5/1 Sn–Pd HSA Fresh Used 2.5/5 Cu–Pd Al2 O3 Fresh Used
XPS
XRF
Sn/Pd
Cu/Pd
Sn/Pd
Cu/Pd
0.606 0.818
– –
0.845 0.848
– –
4.282 1.000
– –
0.900 0.815
– –
2.972 0.596
– –
0.463 0.503
– –
– –
0.167 0.237
– –
0.605 0.605
of the shake up satellite (from 933.5 eV and a Isat /Ipp = 0.55, typical values for CuO, to 931.9 eV and Isat /Ipp = 0, assigned to metallic Cu). After reaction, the Cu species were partially oxidized, as observed with the Sn species for the Pd–Sn catalytic systems. Dispersion of Sn and Pd species were studied converting the atomic concentration values of Table 5 to weight ratios and they are compared with the results obtained by XRF. All these values are shown in Table 7. The ratios obtained by both techniques show similar values mainly in the Sn–Pd used samples, which indicate a homogeneous dispersion of active phases after reaction. In this table also can be observed that the catalyst with higher activity and with a better selectivity to nitrogen (the 0.5/1 Sn–Pd HSA catalyst) is the only one where the Pd/metal ratio is kept in a value close to 2 after reaction. This ratio has also been described as the most adequate for an active catalyst [15] in this reaction and this could be one of the reasons for the good results obtained with this catalyst. 4. Conclusions A family of alumina supported Sn–Pd catalysts with a Pd/Sn ratio of 2 has been prepared. These catalysts have been characterized by
152
C. Franch et al. / Applied Catalysis A: General 425–426 (2012) 145–152
different techniques and their activity for the reduction of nitrates from polluted natural water in a continuous reactor has been tested. It has been observed that the most active and selective catalyst (the 0.5/1 Sn–Pd HSA catalyst) is the one that presents the highest metallic dispersion and the lowest phase segregation. XPS results show changes of the surface Pd/Sn atomic ratio upon reduction and after reaction and the coexistence of different oxidation states for the active metals. Deactivation of the catalyst can be explained by the non-reversible oxidation of the Sn–Pd couple, showing the lowest deactivation rate, the catalyst with lowest oxidized species after reaction. Nevertheless, an additional deactivation by the deposition of the other ions present in the natural water cannot be discarded. Acknowledgments We gratefully acknowledge the support from the Ministry of Science and Innovation, Spain (MICINN, Spain) through the project MAT2009-10481 and FEDER funds. Authors also thank the Spanish Government (projects MAT2009-14528-C02-01 and CONSOLIDER INGENIO 2010) and the European Union (European Community’s Seventh Framework Programme FP7/2007-2013 under Grant Agreement No. 226347 Project) for financial support. References [1] [2] [3] [4] [5] [6] [7]
A. Kapoor, T. Viraraghavan, J. Environ. Eng. 123 (1997) 371–380. S. Hörold, K.D. Vorlop, T. Tacke, M. Sell, Catal. Today 17 (1993) 21–30. J. Batista, A. Pintar, M. Ceh, Catal. Lett. 43 (1997) 79–84. A.E. Palomares, J.G. Prato, F. Márquez, A. Corma, Appl. Catal. B 41 (2003) 3–13. U. Prüsse, M. Hälhein, J. Daum, K.D. Vorlop, Catal. Today 55 (2000) 79–90. R. Gavagnin, L. Biasetto, F. Pinna, G. Strukul, Appl. Catal. B 38 (2002) 91–99. H. Berndt, I. Mönnich, B. Lücke, M. Menzel, Appl. Catal. B 30 (2001) 111–122.
[8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37]
F. Gauthard, F. Epron, J. Barbier, J. Catal. 220 (2003) 182–191. A.E. Palomares, J.G. Prato, A. Corma, J. Catal. 221 (2004) 62–66. N. Barrabes, J. Sa, Appl. Catal. B 104 (1–2) (2011) 1–5. O.S.G.P. Soares, J.J.M. Órfão, M.F.R. Pereira, Appl. Catal. B91 (2009) 441–448. G. Strukul, R. Gavagnin, F. Pinna, E. Modaferri, S. Perathoner, G. Centi, M. Marella, M. Tomaselli, Catal. Today 55 (2000) 139–149. F. Epron, F. Gauthard, C. Pinéda, J. Barbier, J. Catal. 198 (2001) 309–318. F.A. Marchesini, S. Irusta, C. Querini, E. Miró, Appl. Catal. A 348 (2008) 60–70. A.E. Palomares, C. Franch, A. Corma, Catal. Today 149 (2010) 348–351. N. Barrabés, A. Dafinov, F. Medina, J.E. Sueiras, Catal. Today 149 (2010) 341–347. N. Barrabés, M.A. Garrido, A. Frare, A. Monzón, D. Tichit, Catal. Today 175 (2011) 328–337. N. Barrabés, J. Just, A. Dafinov, F. Medina, J.L.G. Fierro, J.E. Sueiras, P. Salagre, Y. Cesteros, Appl. Catal. B62 (2006) 77–85. W. Gao, N. Guan, J. Chen, X. Guan, R. Jin, H. Zeng, Z. Liu, F. Zhang, Appl. Catal. B46 (2003) 341–351. A. Edelman, W. Schieber, H. Vinek, A. Jentys, Catal. Lett. 69 (2000) 11–16. G.C. Boskovic, M.K. Kovacevic, S.N. Ratkovic, E.E. Kiss, J. Radnic, Acta Period. Technol. 39 (2008) 93–99. A. Pintar, J. Batista, J. Levec, Catal. Today 66 (2001) 503–510. A. Mastalir, B. Rác, Z. Király, A. Molnár, J. Mol. Catal. A 264 (2007) 170–178. A. Cabiac, T. Cacciaguerra, P. Trens, R. Durand, G. Delahay, A. Medevielle, D. Plée, B. Coq, Appl. Catal. A 340 (2008) 229–235. F. Yin, S. Ji, P. Wu, F. Zhao, C. Li, J. Catal. 257 (2008) 108–116. R. Strobel, F. Krumeich, W.J. Stark, S.E. Pratsinis, A. Baiker, J. Catal. 222 (2004) 307–314. J.R. Anderson, Structure of Metallic Catalysts, Academic Press, New York, 1975. S. Poulston, P.M. Parlett, P. Stone, M. Bowker, Surf. Interface Anal. 24 (1996) 811–820. M.J. Chollier-Brym, R. Gavagnin, G. Strukul, M. Marella, M. Tomaselli, P. Ruiz, Catal. Today 75 (2002) 49–55. A. Pintar, J. Batista, I. Musevic, Appl. Catal. B52 (2004) 49–60. Y. Wang, J. Qu, H. Liu, J. Mol. Catal. A: Chem. 272 (2007) 31–37. A. Pintar, J. Batista, Catal. Today 53 (1999) 35–50. A. Pintar, M. Setinc, J. Levec, J. Catal. 174 (1998) 72–87. J. Kim, T. Momma, T. Osaka, J. Power Sources 189 (2009) 909–915. I.G. Casella, M. Contursi, J. Electroanal. Chem. 588 (2006) 147–154. A. Garron, K. Lázár, F. Epron, App. Catal. B 59 (2005) 57–69. Y. Wang, J. Qu, R. Wu, P. Lei, Water Res. 40 (2006) 1224–1232.