- Email: [email protected]
Natural zeolite reactivity towards ozone: The role of compensating cations
Journal of Hazardous Materials 227–228 (2012) 34–40
Contents lists available at SciVerse ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Natural zeolite reactivity towards ozone: The role of compensating cations Héctor Valdés a,∗ , Serguei Alejandro b , Claudio A. Zaror b a b
Laboratorio de Tecnologías Limpias (F. Ingeniería), Universidad Católica de la Santísima Concepción, Alonso de Ribera 2850, Concepción, Chile Departamento de Ingeniería Química (F. Ingeniería), Universidad de Concepción, Concepción, Chile
h i g h l i g h t s
Chemical and thermal treatment enhances catalytic activity of natural zeolite. Modified natural zeolite exhibits high stability after thermal treatment. Reducing the compensating cation content leads to an increase on ozone abatement. Surface active atomic oxygen was detected using the DRIFT technique. The highest reactivity toward ozone was performed by NH4Z3 zeolite sample.
a r t i c l e
i n f o
Article history: Received 22 September 2011 Received in revised form 26 April 2012 Accepted 29 April 2012 Available online 5 May 2012 Keywords: Brønsted acid sites Compensating cations Lewis acid sites Ozone Zeolite
a b s t r a c t Among indoor pollutants, ozone is recognised to pose a threat to human health. Recently, low cost natural zeolites have been applied as alternative materials for ozone abatement. In this work, the effect of compensating cation content of natural zeolite on ozone removal is studied. A Chilean natural zeolite is used here as starting material. The amount of compensating cations in the zeolite framework was modified by ion exchange using an ammonium sulphate solution (0.1 mol L−1 ). Characterisation of natural and modified zeolites were performed by X-ray powder diffraction (XRD), nitrogen adsorption at 77 K, elemental analysis, X-ray fluorescence (XRF), thermogravimetric analysis coupled with mass spectroscopy (TGA-MS), and temperature-programmed desorption of ammonia (NH3 -TPD). Ozone adsorption and/or decomposition on natural and modified zeolites were studied by diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS). Results show that the zeolite compensating cation content affects ozone interaction with zeolite active sites. Ammonium ion-exchange treatments followed by thermal out-gassing at 823 K, reduces ozone diffusion resistance inside the zeolite framework, increasing ozone abatement on zeolite surface active sites. Weak and strong Lewis acid sites of zeolite surface are identified here as the main active sites responsible of ozone removal. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Environmental indoor contamination has been added to the list of problems that our modern society has to deal with. Among indoor pollutants, ozone is recognised to pose a threat to human health [1]. Ozone is normally generated in working environments due to the use of photocopiers, laser printers, fax machines and sterilisation apparatus [2]. In airplane cabins, ozone must be purged from air before it can be circulated [3]. Ozone can reduce lung functions, height lung sensitivity to allergens and irritants, and cause chronic damages to lung structure [4]. Microporous materials such as activated carbons have been used during the last decades for ozone
∗ Corresponding author. Tel.: +56 41 2345044; fax: +56 41 2345300. E-mail address: [email protected] (H. Valdés). 0304-3894/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jhazmat.2012.04.067
indoor elimination. Recently, low cost natural zeolites have been studied as alternative materials for this purpose. Zeolites and related microporous materials act as efficient heterogeneous catalysts, as adsorbents, and as molecular sieves in gas separation processes [5]. Zeolites may be used in catalysis or gas adsorption after chemical composition modification, mainly by ion exchange. Ion exchange is an intrinsic property of zeolites; and it is associated with the presence of cations that compensate the negative zeolite structure charge [6]. Various zeolite applications are based directly or indirectly on this property [6]. The influence of surface characteristics of Chilean natural zeolites in gaseous ozone abatement was reported in previous works [7,8]. Lewis acid surface sites were claimed as the main responsible for ozone gaseous elimination, using outgassed natural zeolite at 823 K. However, there are still some doubts related to the effect of zeolite compensating cation content on gaseous ozone removal. Unfortunately, such fundamental aspect has not been studied in detail, being addressed in this article.
H. Valdés et al. / Journal of Hazardous Materials 227–228 (2012) 34–40
2. Experimental methods 2.1. Materials Chilean natural zeolite (53% clinoptilolite, 40% mordenite and 7% quartz) was provided by Minera FormasTM . Zeolite sample was sieved to 0.3–0.425 mm, washed with ultra pure water, oven-dried at 398 K for 24 h, and finally stored in a desiccator until further use. Ozone was produced from instrumental dry air supplied by AGA, using an AZCOZON A-4 ozone generator (Vancouver, BC, Canada) rate at 4 g O3 h−1 . Ultra-pure water was obtained from an EASY pure® RF II system. 2.2. Modification of compensating cation content in natural zeolite Four de-cationised zeolite samples were obtained from natural zeolite (NZ) by applying ion-exchange treatments using an ammonium sulphate solution (0.1 mol L−1 ). Chemical modifications were carried out under shaking for 2 h at 363 K, using solution/solid ratios of 10:1, 20:1 and 30:1. These ammonium-exchanged zeolite samples were denominated NH4Z1, NH4Z2 and NH4Z3, respectively. After the ion-exchange step, samples were rinsed with ultra-pure water at 363 K for 4 h. Ultra-pure water was renovated after washing at 2 h and 3 h. Additionally, a second ion-exchange step was conducted to the NH4Z1 sample using a solution/solid ratio of 10:1. This sample was named 2NH4Z1. All samples were oven-dried at 398 K for 24 h and stored in a desiccator until further use. Natural and modified zeolite samples were thermally outgassed at 823 K, in a vertical tubular furnace. Zeolite samples (0.15 g) were placed in a quartz fixed-bed flow reactor, and heated at 10 K min−1 , under vacuum. When samples reached maximum temperature, isothermal conditions were kept for 2 h, before quenching to room temperature. Thus, zeolite samples were ready to ozone contact. 2.3. Physical–chemical characterisation of natural and modified zeolite samples X-ray powder diffraction (XRD) was applied to natural and modified zeolite samples in order to evaluate mineralogical and structural changes. XRD was performed with a Bruker AXS Model D4 ENDEAVOR diffractometer, equipped with a cupper X-ray tube and Ni filter. Powdered samples were mounted on quartz plates and stepped scanned over the angular range 3–70◦ (2), a step size of 0.02 and a time/step of 0.2 s. X-ray generator was fixed at 40 kV and 20 mA. Surface area of natural and modified zeolite samples were obtained by nitrogen adsorption isotherms, using a Micromeritics Gemini 3175 as described elsewhere [7,8]. Samples were previously degasified at 623 K for 12 h under vacuum. Surface areas (S) of natural and modified zeolite samples were calculated from nitrogen adsorption data, applying the Langmuir equation. Elemental analyses were performed on a LECO CHN 2000 apparatus. Samples of 0.1 g were put inside the oven at 1173 K and burned in the presence of excess oxygen to ensure complete combustion. Gaseous combustion products were evacuated using He (100 cm3 min−1 ) and the nitrogen content of modified zeolite samples was registered by a thermal conductivity detector, previously calibrated with ethylenediaminetetraacetic acid (EDTA). X-ray fluorescence (XRF) allowed determination of bulk chemical composition of natural and modified zeolites by using a RIGAKU Model 3072 spectrometer. This is a dispersive wave spectrometer equipped with a Rhodium X-ray generator, four crystal diffractors (LIF200-SC, LIF200-PC, PER, TAP100), and two detectors. 5 g of lithium tetraborate were thoroughly mixed with 0.5 g sample in
35
an agate mortar. Then, it was placed in a Pt–Ru crucible, together with lithium iodine before spectrometric determination [9]. Thermogravimetric analyses (TGA) were performed on a NETZSCH thermobalance ST409PC, coupled with a quadrupole mass spectrometer (MS) detector. Natural and de-cationised zeolite samples (0.025 g), without thermal out-gassing treatment, were heated up to 950 K (heating rate of 10 K min−1 ) under He flow (100 cm3 min−1 ) and the change in sample weight in relation to change in temperature was registered (TG curve). A derivative weight loss curve was also obtained as function of temperature (DTG curve). Simultaneously, the temperature difference between Al2 O3 reference crucible and the sample was recorded and the differential temperature was plotted against temperature. Thus, a differential thermal analysis (DTA) provides data on the transformations that occur in zeolite samples (DTA curve). Moreover, during TGA analysis, evolved gases were monitored with a quadrupole mass spectrometer detector (QMS 403C), recording the change of ammonia and water concentration by following the mass-to-charge ratio (m/z) at 16 and 17, respectively. Acid site density of zeolite samples was determined by applying the ammonia temperature-programmed desorption method (NH3 -TPD) [10]. NH3 -TPD is one of the most conventional methods for global acid site characterisation of zeolites. Natural and modified zeolite samples (0.15 g) were previously outgassed at 823 K (10 K min−1 heating rate) for 2 h under Ar flow (50 cm3 min−1 ). Then, they were allowed to adsorb the saturated ammonia vapour from the ammonium hydroxide solution until reaching saturation at 373 K (in order to minimise physical adsorption of ammonia). Finally, ammonia saturated zeolites were placed into the TPD equipment coupled with a thermal conductivity detector (TCD). Samples were heated from 293 to 973 K at a rate of 10 K min−1 in Ar flow (100 cm3 min−1 ). The conductivity changes of the evolved gas were plotted against temperature and the acid site strength and density were estimated. The temperature range in which most ammonia was desorbed; is related to the acid strength distribution. The total amount of chemisorbed ammonia is proportional to the number of acid sites per unit mass of the adsorbents (acid site density) [11]. 2.4. Experimental system Ozone abatement experiments were carried out in a quartz fixed-bed flow reactor (4 mm ID) at 293 K and 101 kPa, loaded with 0.15 g zeolite. Untreated (NZ) and treated (NH4Z1, NH4Z2, NH4Z3, 2NH4Z1) zeolite samples were used. Ozone removal experiments were conducted after the thermal treatment at 823 K was completed and room temperature was quenched. Typically, the reactor was continuously fed with an O2 /O3 gas mixture at 740 cm3 min−1 , and 0.345 mol O3 dm−3 (CO3in ). These operating conditions were selected from preliminary experiments aimed at reducing mass transfer effects. Ozone concentrations in the reactor outlet stream (CO3out ) were monitored on-line by continuous absorbance measurement at 255 nm, using an UV spectrophotometer (Bausch & Lomb TU-1810 S). The exhaust gas stream was sent to an ozone trap before discharging to the ambient air. Data were registered as function of time and processed using UV-Win 5.0 software. Once the outlet ozone concentration equalled the inlet level, the O3 /O2 flow was stopped and argon gas was fed into the reactor system for 10 min, before removing the zeolite. 2.5. DRIFTS study of ozone–zeolite surface interactions Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was used in order to confirm ozone molecule adsorption and/or ozone decomposition by-products on natural and modified zeolite samples [12]. DRIFTS measurements were carried out using
36
H. Valdés et al. / Journal of Hazardous Materials 227–228 (2012) 34–40
Fig. 1. X-ray powder diffraction patterns of natural and modified zeolite samples: (A) after thermal out-gassing treatment at 398 K, (B) after thermal out-gassing treatment at 823 K. Clinoptilolite (C), mordenite (M) and quartz (Q).
a Nicolet 6700 FTIR spectrometer equipped with a Smart Collector diffuse reflectance mirror system and a high-temperature/highpressure DRIFTS cell (Thermo Scientific). Sample was mixed and ground with KBr in an agate mortar, to yield a mix with 1.0% (w/w) zeolite. Then, the DRIFTS cell cup was filled with the sample and activation was carried out by heating up to 823 K (heating rate 10 K min−1 ) in a nitrogen flow (50 cm3 min−1 ), at atmospheric pressure. The sample was kept in nitrogen flow for 2 h at 823 K, and then cooled to 293 K. The corresponding spectrum of KBr powder was used as a background for each sample spectrum. The spectrum of the adsorbed species was obtained as a difference of the recorded spectra in presence and in absence of the O2 /O3 gas mixture. DRIFTS spectra were usually obtained at a resolution of 2 cm−1 averaged over 300 interferometer scans. In DRIFTS experiments, ozone was produced in situ with an ELECT CAVA model OEC 480S ozone generator. Ozone concentration was registered using an UV spectrophotometer (SPECTRONIC GENESYS 5). Initially, the O2 /O3 gas mixture was sent to an ozone trap until experimental conditions were stabilised. After that, the O2 /O3 gas mixture was redirected into the DRIFTS cell, and measuring was initiated. All the experiments reported here were done at 293 K. Results were processed with OMNIC software. 3. Results and discussion 3.1. Physical–chemical characterisation of natural and modified zeolite samples XRD results, shown in Fig. 1A and B, indicate that zeolite samples are highly crystalline. Characteristic peaks of clinoptilolite (C), mordenite (M) and quartz (Q) structures can be observed in the X-ray patterns. Ion-exchange treatments of natural zeolite do not show any significant changes in the zeolite structures. Moreover, no changes were observed in XRD peak intensities in the outgassed
modified zeolites at 823 K (see Fig. 1B). However, a reduction in the intensity of the natural zeolite diffraction pattern after thermal outgassing treatment at 823 K would indicate a loss in cristallinity. Upon thermal treatment at 823 K, the clinoptilolite phase intensity of natural zeolite underwent a 20% decrease in the peak register at 22◦ , while the mordenite phase showed a 6% of reduction on intensity peak at 25.3◦ . Under these thermal treatment conditions, natural zeolite could lose bonded water molecules and compensating cations would be forced to move closer to the framework oxygen to form new bonds. This mechanism is known to induce local strain on the zeolite tetrahedral network and would eventually lead to bond breakage in the framework (particularly the T O T bridges), reducing the stability of the natural zeolite structure [13]. Table 1 summarises physical–chemical characterisation data of natural and de-cationised zeolite samples prior to the out-gassing step at 823 K. Results show that the de-cationisation treatment used here effectively reduced the amount of compensating cations (Na, K, Mg, Ca), without a significant change on the Si/Al ratio of the zeolite framework. The highest extent of decationisation for sodium ions was obtained for NH4Z3 sample and the lowest for NH4Z1 sample. Moreover, for sodium and potassium cations a higher reduction was attained after a second ion-exchange treatment for NH4Z1 sample. This is probably due to the second rinse applied here, which could eliminate possible ion exchanged by-products that might clog zeolite pores. Calcium content of ammonium-exchanged zeolites decreases from NH4Z1 to 2NH4Z1 and nitrogen content increases, as a result of the de-cationisation process. These experimental evidences are in agreement with other studies which report that sodium ions present in natural zeolites, like clinoptilolite, were weakly bonded to zeolite lattice and easy to remove [14]. Results for modified samples, indicate the following order of ammonium ion exchange preference toward compensating
Table 1 Physical–chemical characterisation of zeolite samples. Samples
S627 [m2 g−1 ]
SiO2 a
Al2 O3 a
Na2 Oa
CaOa
K2 Oa
MgOa
TiO2 a
Fe2 O3 a
MnOa
Si/Al ratio
N2 b
NZ NH4Z1 NH4Z2 NH4Z3 2NH4Z1
205 181 177 177 171
75.25 78.07 75.25 79.84 79.26
14.1 14.69 14.10 14.63 14.85
1.89 0.68 0.49 n.d. 0.26
4.57 2.36 1.96 1.97 1.82
0.74 0.67 0.6 0.46 0.39
0.66 0.46 0.46 n.d. 0.37
0.42 0.46 0.47 0.49 0.47
2.31 2.55 2.63 2.56 2.53
0.05 0.05 0.05 0.05 0.05
5.34 5.32 5.27 5.46 5.34
0.13 1.78 2.08 2.08 2.16
n.d. – Non-detected. a By XRF (% w/w). b By elemental analyses (% w/w).
H. Valdés et al. / Journal of Hazardous Materials 227–228 (2012) 34–40
37
Fig. 2. Thermogravimetric analysis coupled with mass spectroscopy (TGA-MS): (A) TG and DTG curves, (B) DTA curves, (C) ammonia evolution, and (D) H2 O evolution. Operation conditions: 0.025 g of zeolite, heating up to 950 K (rate 10 K min−1 ) in He (0.1 dm3 min−1 ), 101 kPa.
cations: Ca2+ > K+ > Mg2+ . Ion exchange capacity of clinoptilolite is known to be lower than some other natural zeolites, but generally exhibits a high affinity for NH4 + ion [15]. Surface area of natural zeolite sample decreases from 205 m2 g−1 to 170 m2 g−1 after thermal treatment at 823 K. This could be related to the interaction of compensating cations and water in the zeolite framework. Compensating cations are normally coordinated to a defined number of water molecules, affecting their mobility within the zeolite structure [16]. During the out-gassing process at 823 K, water molecules leave the zeolite, destabilising the zeolite charge structure [7]. In order to correct this, it has been suggested that compensating cations in the zeolite framework, form bonds with network oxygen atoms, affecting the zeolite structure channels [13,14]. However, in the ammonium-exchanged zeolites, surface areas are increased after thermal treatment at 823 K. The values of surface areas after the out-gassing step at 823 K, are as follow; NH4Z1 (222 m2 g−1 ), NH4Z2 (230 m2 g−1 ), NH4Z3 (251 m2 g−1 ) and 2NH4Z1 (261 m2 g−1 ). Such increase in the value of surface areas could be related to the elimination of ammonium due to thermal out-gassing at 823 K, thus reducing any blocking effect on microporous structures. Fig. 2 illustrates the results of thermogravimetric analyses coupled with mass spectroscopy (TGA-MS) conducted to natural and modified zeolite samples. Fig. 2A shows the change in sample weight (TG curve) and derivative weight loss (DTG curve) as the temperature increases. Fig. 2B shows DTA curve results, while Fig. 2C and D show, ammonia and H2 O evolutions as a function of temperature. The total weight loss registered during TGA analysis was as follows: 9.2% (NZ), 9.9% (NH4Z1), 9.4% (NH4Z2), 8.5% (NH4Z3), and
9.5% (2NH4Z1). DTG curves reveal two characteristic peaks. The first, around the temperature range of 407–431 K and the second around 723–738 K. The first peak could be related to the desorption of physisorbed gaseous molecules and the second could be mainly due to water evolution during Brønsted acid sites transformation into Lewis acid sites [17]. As it can be seen in Fig. 2B, exothermic peaks appear at 796, 866, 875, 928 and 875 K in the DTA curves for NZ, NH4Z1, NH4Z2, NH4Z3 and 2NH4Z1, respectively. These temperature peaks could indicate the beginning of thermal zeolite structure collapse. NZ shows the lowest thermal stability of all the zeolite samples used here, as indicated before by XRD analysis. The increases on thermal stability of ammonium-exchanged zeolites could be associated to a reduction on the compensating cation content as reported by other authors [16]. Fig. 2C shows the release of ammonia from the ammoniumexchanged zeolites as temperature rises. It can be seen that the amount of desorbed ammonia is higher in sample 2NH4Z1 than in the other samples. Zeolite compensating cations were exchanged with ammonium ions during chemical modification treatments. The results obtained here are in agreement with those obtained by elemental analyses, confirming the highest level of compensating cation exchange accomplished for this sample. In Fig. 2C, the peaks registered at different temperatures might be related to free re-adsorption of ammonia on vacant acid sites, as it has been reported by other authors [18]. Ammonium protons are quite mobile as temperature increases, evolving as ammonia; hence leading to Brønsted acid sites generation [17]. Fig. 2D presents the water evolution during the thermogravimetric analyses. Water signal intensity decreases when the
38
H. Valdés et al. / Journal of Hazardous Materials 227–228 (2012) 34–40
Fig. 3. NH3 -TPD analysis of zeolite samples. Operation conditions: 0.15 g of zeolite, heating up to 900 K (10 K min−1 heating rate) in Ar flow (100 cm3 min−1 ), 101 kPa.
temperature increases as a consequence of moisture elimination. Water signal intensity increases at 823 K and it could be associated to Brønsted acid sites transformation into Lewis acid sites [17]. Thermal out-gassing treatment at 823 K does not only eliminate adsorbed contaminants, but also contributes to new Lewis acid sites formation. Fig. 3 shows thermal desorption profiles of zeolite samples, previously saturated with NH3 . During NH3 desorption, two distinctive peaks appear at different temperature ranges, corresponding to sites of different acid strengths. The evolution of NH3 in the low (300–530 K) and high (530–810 K) temperature ranges could be attributed to the presence of weak and strong acid sites on the zeolite samples, respectively. It is known that the required energy to release ammonia from acid sites could be associated to the strength of chemical bonds (acid strength) [19]. Table 2 lists the maximum registered temperature peaks and the corresponding amount of desorbed NH3 . These parameters are used here as indicators of the zeolite acid site strengths and densities, respectively. As it can be seen, NH4Z3 sample shows a maximum desorption of NH3 in the high-temperature range, resulting in a zeolite sample with the highest acid strength. De-cationisation procedure applied in this study does not only lead to pore opening; but also allows Brønsted acid sites formation during ammonium protons transform into ammonia in the first part of the thermal out-gassing treatment. As the out-gassing temperature approaches to 823 K, Brønsted acid sites are transformed into Lewis acid sites, as confirmed by water evolution detected by MS (see Fig. 2D). 3.2. Effect of zeolite compensating cations content on gaseous ozone removal Fig. 4 illustrates the effect of compensating cation content on ozone abatement after out-gassing at 823 K. Natural zeolite shows
Fig. 4. Influence of zeolite compensating cation content on ozone removal. Operation conditions: 0.15 g of natural and modified zeolite samples thermally treated at 823 K, 0.345 mol dm−3 of inlet ozone concentration, 0.740 dm3 min−1 , 101 kPa, 293 K.
the lowest ozone removal. This zeolite sample has the highest content of compensating cations (M+ or M2+ ). Compensating cation characteristics such as size and charge could affect ozone interaction with Lewis acid sites within zeolite framework, reducing ozone abatement (see Scheme 1, section (A)). On the other hand, in the ammonium-exchanged zeolite samples, the zeolite framework could remain almost unchanged; and might decrease ozone diffusion resistance inside the zeolite structure, allowing higher ozone abatement on Lewis acid sites. Such chemical interactions are graphically illustrated in Scheme 1, sections (B) and (C), using the Lewis acid site schematic representation given by Karge [20]. Modified zeolite samples were less affected by the thermal out-gassing treatment, as confirmed by surface area determination. Ozone is known to act as a Lewis base and could be adsorbed as a molecule on weak Lewis acid sites and decomposes to form oxygen radicals (O• ) on strong Lewis acid sites [21]. Ozone interaction with acid surface sites could be due to its resonance structure, where the high electron density on one of oxygen atoms, might show high basicity resulting in strong affinity to Lewis acid surface sites [22]. Lewis acid sites on zeolite surface might play a key role enhancing the gaseous ozone elimination [7,8]. DRIFTS spectra of natural and modified zeolite samples are shown in Fig. 5 DRIFTS spectra were registered in the absence of ozone (t = 0 min) and after 10 and 20 min ozone exposure. Ozone interaction with zeolite samples increases the intensities of the bands at 1052–1054 cm−1 and at 1405–1416 cm−1 ; which could be associated to ozone molecular adsorption and to active atomic oxygen, respectively. In particular, IR band around 1037 cm−1 has been associated to ozone adsorption on weak Lewis acid sites [21,23]. Moreover, a peak at 1380 cm−1 has been assigned to a stable surface oxide species due to a strong bond where the atomic oxygen was
Scheme 1. Influence of compensating cations on ozone–zeolite surface interactions.
H. Valdés et al. / Journal of Hazardous Materials 227–228 (2012) 34–40
39
Table 2 Acidity of natural and modified zeolite samples, using NH3-TPD analysis. Samples
NZ NH4Z1 NH4Z2 NH4Z3 2NH4Z1
Maximum temperature peaks [K]
Acid site density [% desorbed NH3 ]
Weak acid sites (300–530 K)
Strong acid sites (530–810 K)
Weak acid sites (300–530 K)
Strong acid sites (530–810 K)
390 374 376 391 399
715 684 702 720 707
93 86 75 24 31
7 14 25 76 69
Fig. 5. DRIFTS spectra of natural and modified zeolite samples during adsorption of ozone. Spectra were taken after 0, 10, and 20 min. Operation conditions: natural and modified zeolite samples thermally treated at 823 K, 0.345 mol dm−3 of inlet ozone concentration, 101 kPa, 293 K.
40
H. Valdés et al. / Journal of Hazardous Materials 227–228 (2012) 34–40
attached to a Lewis acid site on synthetic zeolite surfaces [12,24]. This surface oxide specie has been identified as an oxidation product of ozone decomposition on zeolite surface [12,24].In the case of the natural zeolite sample, the IR bands indicated as molecular adsorbed ozone increase with ozonation time; whereas the frequency associated to the active atomic oxygen formation does not show major differences. However, in the case of the NH4Z3 sample, as ozone exposure increases, the peak area related to active atomic oxygen increases, without any arise on the peak frequency related to ozone molecular adsorption. These results could be due to the influence of the compensating cation content on the acid site strength of these samples, as confirmed by the NH3 -TPD analyses. For zeolite samples studied here, no spectral features were found that could be attributed to adsorbed molecular ozone on hydroxyl surface groups. However, an IR band was identified as ozone molecular adsorbed via coordinative bonding on weak Lewis acid sites. Moreover, ozone could transform immediately after adsorption on strong Lewis acid sites and decomposes into O2 and an active atom of oxygen, which could remain bound to a Lewis acid site. These active atoms of oxygen could then participate in catalytic removal of ozone [21]. 4. Conclusions The decreases of zeolite compensating cation content by applying ammonium ion-exchange treatments followed by thermal out-gassing at 823 K, reduces ozone diffusion resistance inside the zeolite framework, increasing ozone abatement on zeolite surface active sites. Ozone removal seems to be a combination of adsorption and surface-catalysed decomposition. DRIFTS results obtained here could be regarded as experimental evidence of the different reactivities of natural zeolite surface sites toward ozone. Weak and strong Lewis sites are identified here as zeolite active sites responsible of ozone reduction. Chemical modification of natural zeolite applied in this study could be an effective method for low cost zeolite production for ozone elimination in indoor environments. Acknowledgements Authors gratefully acknowledge FONDECYT (grant no. 1090182) for their financial support. S. Alejandro wishes to thank CONICYT for providing a doctoral scholarship and to the Carbon and Catalysis Group from University of Concepcion for their valuable collaboration. References [1] J.N. Cape, Surface ozone concentrations and ecosystem health: past trends and a guide to future projections, Sci. Total Environ. 400 (2008) 257–269.
[2] B. Dhandapani, S.T. Oyama, Gas phase ozone decomposition catalysts, Appl. Catal. B 11 (1997) 129–166. [3] S. Bhangar, S.C. Cowlin, B.C. Singer, R.G. Sextro, W.W. Nazaroff, Ozone levels in passenger cabins of commercial aircraft on North American and transoceanic routes, Environ. Sci. Technol. 42 (2008) 3938–3943. [4] M.S. O’Neill, D. Loomis, V.H. Borja-Aburto, Ozone, area social conditions, and mortality in Mexico City, Environ. Res. 94 (2004) 234–242. [5] G. Martra, R. Ocule, L. Marchese, G. Centi, S. Coluccia, Alkali and alkaline-earth exchanged faujasites: strength of Lewis base and acid centres and cation site occupancy in Na- and BaY and Na- and BaX zeolites, Catal. Today 73 (2002) 83–93. [6] L. Martins, R.T. Boldo, D. Cardoso, Ion exchange and catalytic properties of methylammonium FAU zeolite, Microporous Mesoporous Mater. 98 (2007) 166–173. [7] S. Alejandro, H. Valdés, C.A. Zaror, Natural zeolite reactivity towards ozone: the role of acid surface sites, J. Adv. Oxid. Technol. 14 (2011) 182–189. [8] H. Valdés, E. Padilla, C.A. Zaror, Influence of chemical surface characteristics of natural zeolite on catalytic ozone abatement, Ozone: Sci. Eng. 33 (2011) 279–284. [9] H. Valdés, V.J. Farfán, J.A. Manoli, C.A. Zaror, Catalytic ozone aqueous decomposition promoted by natural zeolite and volcanic sand, J. Hazard. Mater. 165 (2009) 915–922. [10] F. Benaliouche, Y. Boucheffa, P. Ayrault, S. Mignard, P. Magnoux, NH3-TPD and FTIR spectroscopy of pyridine adsorption studies for characterization of Ag- and Cu-exchanged X zeolites, Microporous Mesoporous Mater. 111 (2008) 80–88. [11] C.W. Kwong, C.Y.H. Chao, K.S. Hui, M.P. Wan, Removal of VOCs from indoor environment by ozonation over different porous materials, Atmos. Environ. 42 (2008) 2300–2311. [12] C.W. Kwong, C.Y.H. Chao, K.S. Hui, M.P. Wan, Catalytic ozonation of toluene using zeolite and MCM-41 materials, Environ. Sci. Technol. 42 (2008) 8504–8509. [13] J. Li, J. Qiu, Y. Sun, Y. Long, Studies on natural STI zeolite: modification, structure, adsorption and catalysis, Microporous Mesoporous Mater. 37 (2000) 365–378. [14] S.J. Allen, E. Ivanova, B. Koumanova, Adsorption of sulfur dioxide on chemically modified natural clinoptilolite. Acid modification, Chem. Eng. J. 152 (2009) 389–395. [15] Y.F. Wang, F. Lin, W.Q. Pang, Ammonium exchange in aqueous solution using Chinese natural clinoptilolite and modified zeolite, J. Hazard. Mater. 142 (2007) 160–164. [16] X.-w. Cheng, Y. Zhong, J. Wang, J. Guo, Q. Huang, Y.-c. Long, Studies on modification and structural ultra-stabilization of natural STI zeolite, Microporous Mesoporous Mater. 83 (2005) 233–243. [17] M. Stöcker, Gas phase catalysis by zeolites, Microporous Mesoporous Mater. 82 (2005) 257–292. [18] K. Suzuki, T. Noda, N. Katada, M. Niwa, IRMS-TPD of ammonia: direct and individual measurement of Brønsted acidity in zeolites and its relationship with the catalytic cracking activity, J. Catal. 250 (2007) 151–160. ˜ [19] E. Díaz, S. Ordónez, A. Vega, J. Coca, Catalytic combustion of hexane over transition metal modified zeolites NaX and CaA, Appl. Catal. B 56 (2005) 313–322. [20] H.G. Karge, Caracterization of zeolites, in: J. Weitkamp, L. Pupe (Eds.), Catalysis and Zeolites: Fundamentals and Application, Springer-Verlag, Berlin, 1999, pp. 215–231. [21] K.M. Bulanin, J.C. Lavalley, A.A. Tsyganenko, IR spectra of adsorbed ozone, Colloids Surf. A 101 (1995) 153–158. [22] S.T. Oyama, Catalytic properties of ozone, Catal. Rev. Sci. Eng. 42 (2000) 279–322. [23] K.M. Bulanin, A.V. Alexeev, D.S. Bystrov, J.C. Lavalley, A.A. Tsyganenko, IR study of ozone adsorption on SiO2 , J. Phys. Chem. A 98 (1994) 5100–5103. [24] C.Y.H. Chao, C.W. Kwong, K.S. Hui, Potential use of a combined ozone and zeolite system for gaseous toluene elimination, J. Hazard. Mater. 143 (2007) 118–127.
Contents lists available at SciVerse ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Natural zeolite reactivity towards ozone: The role of compensating cations Héctor Valdés a,∗ , Serguei Alejandro b , Claudio A. Zaror b a b
Laboratorio de Tecnologías Limpias (F. Ingeniería), Universidad Católica de la Santísima Concepción, Alonso de Ribera 2850, Concepción, Chile Departamento de Ingeniería Química (F. Ingeniería), Universidad de Concepción, Concepción, Chile
h i g h l i g h t s
Chemical and thermal treatment enhances catalytic activity of natural zeolite. Modified natural zeolite exhibits high stability after thermal treatment. Reducing the compensating cation content leads to an increase on ozone abatement. Surface active atomic oxygen was detected using the DRIFT technique. The highest reactivity toward ozone was performed by NH4Z3 zeolite sample.
a r t i c l e
i n f o
Article history: Received 22 September 2011 Received in revised form 26 April 2012 Accepted 29 April 2012 Available online 5 May 2012 Keywords: Brønsted acid sites Compensating cations Lewis acid sites Ozone Zeolite
a b s t r a c t Among indoor pollutants, ozone is recognised to pose a threat to human health. Recently, low cost natural zeolites have been applied as alternative materials for ozone abatement. In this work, the effect of compensating cation content of natural zeolite on ozone removal is studied. A Chilean natural zeolite is used here as starting material. The amount of compensating cations in the zeolite framework was modified by ion exchange using an ammonium sulphate solution (0.1 mol L−1 ). Characterisation of natural and modified zeolites were performed by X-ray powder diffraction (XRD), nitrogen adsorption at 77 K, elemental analysis, X-ray fluorescence (XRF), thermogravimetric analysis coupled with mass spectroscopy (TGA-MS), and temperature-programmed desorption of ammonia (NH3 -TPD). Ozone adsorption and/or decomposition on natural and modified zeolites were studied by diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS). Results show that the zeolite compensating cation content affects ozone interaction with zeolite active sites. Ammonium ion-exchange treatments followed by thermal out-gassing at 823 K, reduces ozone diffusion resistance inside the zeolite framework, increasing ozone abatement on zeolite surface active sites. Weak and strong Lewis acid sites of zeolite surface are identified here as the main active sites responsible of ozone removal. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Environmental indoor contamination has been added to the list of problems that our modern society has to deal with. Among indoor pollutants, ozone is recognised to pose a threat to human health [1]. Ozone is normally generated in working environments due to the use of photocopiers, laser printers, fax machines and sterilisation apparatus [2]. In airplane cabins, ozone must be purged from air before it can be circulated [3]. Ozone can reduce lung functions, height lung sensitivity to allergens and irritants, and cause chronic damages to lung structure [4]. Microporous materials such as activated carbons have been used during the last decades for ozone
∗ Corresponding author. Tel.: +56 41 2345044; fax: +56 41 2345300. E-mail address: [email protected] (H. Valdés). 0304-3894/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jhazmat.2012.04.067
indoor elimination. Recently, low cost natural zeolites have been studied as alternative materials for this purpose. Zeolites and related microporous materials act as efficient heterogeneous catalysts, as adsorbents, and as molecular sieves in gas separation processes [5]. Zeolites may be used in catalysis or gas adsorption after chemical composition modification, mainly by ion exchange. Ion exchange is an intrinsic property of zeolites; and it is associated with the presence of cations that compensate the negative zeolite structure charge [6]. Various zeolite applications are based directly or indirectly on this property [6]. The influence of surface characteristics of Chilean natural zeolites in gaseous ozone abatement was reported in previous works [7,8]. Lewis acid surface sites were claimed as the main responsible for ozone gaseous elimination, using outgassed natural zeolite at 823 K. However, there are still some doubts related to the effect of zeolite compensating cation content on gaseous ozone removal. Unfortunately, such fundamental aspect has not been studied in detail, being addressed in this article.
H. Valdés et al. / Journal of Hazardous Materials 227–228 (2012) 34–40
2. Experimental methods 2.1. Materials Chilean natural zeolite (53% clinoptilolite, 40% mordenite and 7% quartz) was provided by Minera FormasTM . Zeolite sample was sieved to 0.3–0.425 mm, washed with ultra pure water, oven-dried at 398 K for 24 h, and finally stored in a desiccator until further use. Ozone was produced from instrumental dry air supplied by AGA, using an AZCOZON A-4 ozone generator (Vancouver, BC, Canada) rate at 4 g O3 h−1 . Ultra-pure water was obtained from an EASY pure® RF II system. 2.2. Modification of compensating cation content in natural zeolite Four de-cationised zeolite samples were obtained from natural zeolite (NZ) by applying ion-exchange treatments using an ammonium sulphate solution (0.1 mol L−1 ). Chemical modifications were carried out under shaking for 2 h at 363 K, using solution/solid ratios of 10:1, 20:1 and 30:1. These ammonium-exchanged zeolite samples were denominated NH4Z1, NH4Z2 and NH4Z3, respectively. After the ion-exchange step, samples were rinsed with ultra-pure water at 363 K for 4 h. Ultra-pure water was renovated after washing at 2 h and 3 h. Additionally, a second ion-exchange step was conducted to the NH4Z1 sample using a solution/solid ratio of 10:1. This sample was named 2NH4Z1. All samples were oven-dried at 398 K for 24 h and stored in a desiccator until further use. Natural and modified zeolite samples were thermally outgassed at 823 K, in a vertical tubular furnace. Zeolite samples (0.15 g) were placed in a quartz fixed-bed flow reactor, and heated at 10 K min−1 , under vacuum. When samples reached maximum temperature, isothermal conditions were kept for 2 h, before quenching to room temperature. Thus, zeolite samples were ready to ozone contact. 2.3. Physical–chemical characterisation of natural and modified zeolite samples X-ray powder diffraction (XRD) was applied to natural and modified zeolite samples in order to evaluate mineralogical and structural changes. XRD was performed with a Bruker AXS Model D4 ENDEAVOR diffractometer, equipped with a cupper X-ray tube and Ni filter. Powdered samples were mounted on quartz plates and stepped scanned over the angular range 3–70◦ (2), a step size of 0.02 and a time/step of 0.2 s. X-ray generator was fixed at 40 kV and 20 mA. Surface area of natural and modified zeolite samples were obtained by nitrogen adsorption isotherms, using a Micromeritics Gemini 3175 as described elsewhere [7,8]. Samples were previously degasified at 623 K for 12 h under vacuum. Surface areas (S) of natural and modified zeolite samples were calculated from nitrogen adsorption data, applying the Langmuir equation. Elemental analyses were performed on a LECO CHN 2000 apparatus. Samples of 0.1 g were put inside the oven at 1173 K and burned in the presence of excess oxygen to ensure complete combustion. Gaseous combustion products were evacuated using He (100 cm3 min−1 ) and the nitrogen content of modified zeolite samples was registered by a thermal conductivity detector, previously calibrated with ethylenediaminetetraacetic acid (EDTA). X-ray fluorescence (XRF) allowed determination of bulk chemical composition of natural and modified zeolites by using a RIGAKU Model 3072 spectrometer. This is a dispersive wave spectrometer equipped with a Rhodium X-ray generator, four crystal diffractors (LIF200-SC, LIF200-PC, PER, TAP100), and two detectors. 5 g of lithium tetraborate were thoroughly mixed with 0.5 g sample in
35
an agate mortar. Then, it was placed in a Pt–Ru crucible, together with lithium iodine before spectrometric determination [9]. Thermogravimetric analyses (TGA) were performed on a NETZSCH thermobalance ST409PC, coupled with a quadrupole mass spectrometer (MS) detector. Natural and de-cationised zeolite samples (0.025 g), without thermal out-gassing treatment, were heated up to 950 K (heating rate of 10 K min−1 ) under He flow (100 cm3 min−1 ) and the change in sample weight in relation to change in temperature was registered (TG curve). A derivative weight loss curve was also obtained as function of temperature (DTG curve). Simultaneously, the temperature difference between Al2 O3 reference crucible and the sample was recorded and the differential temperature was plotted against temperature. Thus, a differential thermal analysis (DTA) provides data on the transformations that occur in zeolite samples (DTA curve). Moreover, during TGA analysis, evolved gases were monitored with a quadrupole mass spectrometer detector (QMS 403C), recording the change of ammonia and water concentration by following the mass-to-charge ratio (m/z) at 16 and 17, respectively. Acid site density of zeolite samples was determined by applying the ammonia temperature-programmed desorption method (NH3 -TPD) [10]. NH3 -TPD is one of the most conventional methods for global acid site characterisation of zeolites. Natural and modified zeolite samples (0.15 g) were previously outgassed at 823 K (10 K min−1 heating rate) for 2 h under Ar flow (50 cm3 min−1 ). Then, they were allowed to adsorb the saturated ammonia vapour from the ammonium hydroxide solution until reaching saturation at 373 K (in order to minimise physical adsorption of ammonia). Finally, ammonia saturated zeolites were placed into the TPD equipment coupled with a thermal conductivity detector (TCD). Samples were heated from 293 to 973 K at a rate of 10 K min−1 in Ar flow (100 cm3 min−1 ). The conductivity changes of the evolved gas were plotted against temperature and the acid site strength and density were estimated. The temperature range in which most ammonia was desorbed; is related to the acid strength distribution. The total amount of chemisorbed ammonia is proportional to the number of acid sites per unit mass of the adsorbents (acid site density) [11]. 2.4. Experimental system Ozone abatement experiments were carried out in a quartz fixed-bed flow reactor (4 mm ID) at 293 K and 101 kPa, loaded with 0.15 g zeolite. Untreated (NZ) and treated (NH4Z1, NH4Z2, NH4Z3, 2NH4Z1) zeolite samples were used. Ozone removal experiments were conducted after the thermal treatment at 823 K was completed and room temperature was quenched. Typically, the reactor was continuously fed with an O2 /O3 gas mixture at 740 cm3 min−1 , and 0.345 mol O3 dm−3 (CO3in ). These operating conditions were selected from preliminary experiments aimed at reducing mass transfer effects. Ozone concentrations in the reactor outlet stream (CO3out ) were monitored on-line by continuous absorbance measurement at 255 nm, using an UV spectrophotometer (Bausch & Lomb TU-1810 S). The exhaust gas stream was sent to an ozone trap before discharging to the ambient air. Data were registered as function of time and processed using UV-Win 5.0 software. Once the outlet ozone concentration equalled the inlet level, the O3 /O2 flow was stopped and argon gas was fed into the reactor system for 10 min, before removing the zeolite. 2.5. DRIFTS study of ozone–zeolite surface interactions Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was used in order to confirm ozone molecule adsorption and/or ozone decomposition by-products on natural and modified zeolite samples [12]. DRIFTS measurements were carried out using
36
H. Valdés et al. / Journal of Hazardous Materials 227–228 (2012) 34–40
Fig. 1. X-ray powder diffraction patterns of natural and modified zeolite samples: (A) after thermal out-gassing treatment at 398 K, (B) after thermal out-gassing treatment at 823 K. Clinoptilolite (C), mordenite (M) and quartz (Q).
a Nicolet 6700 FTIR spectrometer equipped with a Smart Collector diffuse reflectance mirror system and a high-temperature/highpressure DRIFTS cell (Thermo Scientific). Sample was mixed and ground with KBr in an agate mortar, to yield a mix with 1.0% (w/w) zeolite. Then, the DRIFTS cell cup was filled with the sample and activation was carried out by heating up to 823 K (heating rate 10 K min−1 ) in a nitrogen flow (50 cm3 min−1 ), at atmospheric pressure. The sample was kept in nitrogen flow for 2 h at 823 K, and then cooled to 293 K. The corresponding spectrum of KBr powder was used as a background for each sample spectrum. The spectrum of the adsorbed species was obtained as a difference of the recorded spectra in presence and in absence of the O2 /O3 gas mixture. DRIFTS spectra were usually obtained at a resolution of 2 cm−1 averaged over 300 interferometer scans. In DRIFTS experiments, ozone was produced in situ with an ELECT CAVA model OEC 480S ozone generator. Ozone concentration was registered using an UV spectrophotometer (SPECTRONIC GENESYS 5). Initially, the O2 /O3 gas mixture was sent to an ozone trap until experimental conditions were stabilised. After that, the O2 /O3 gas mixture was redirected into the DRIFTS cell, and measuring was initiated. All the experiments reported here were done at 293 K. Results were processed with OMNIC software. 3. Results and discussion 3.1. Physical–chemical characterisation of natural and modified zeolite samples XRD results, shown in Fig. 1A and B, indicate that zeolite samples are highly crystalline. Characteristic peaks of clinoptilolite (C), mordenite (M) and quartz (Q) structures can be observed in the X-ray patterns. Ion-exchange treatments of natural zeolite do not show any significant changes in the zeolite structures. Moreover, no changes were observed in XRD peak intensities in the outgassed
modified zeolites at 823 K (see Fig. 1B). However, a reduction in the intensity of the natural zeolite diffraction pattern after thermal outgassing treatment at 823 K would indicate a loss in cristallinity. Upon thermal treatment at 823 K, the clinoptilolite phase intensity of natural zeolite underwent a 20% decrease in the peak register at 22◦ , while the mordenite phase showed a 6% of reduction on intensity peak at 25.3◦ . Under these thermal treatment conditions, natural zeolite could lose bonded water molecules and compensating cations would be forced to move closer to the framework oxygen to form new bonds. This mechanism is known to induce local strain on the zeolite tetrahedral network and would eventually lead to bond breakage in the framework (particularly the T O T bridges), reducing the stability of the natural zeolite structure [13]. Table 1 summarises physical–chemical characterisation data of natural and de-cationised zeolite samples prior to the out-gassing step at 823 K. Results show that the de-cationisation treatment used here effectively reduced the amount of compensating cations (Na, K, Mg, Ca), without a significant change on the Si/Al ratio of the zeolite framework. The highest extent of decationisation for sodium ions was obtained for NH4Z3 sample and the lowest for NH4Z1 sample. Moreover, for sodium and potassium cations a higher reduction was attained after a second ion-exchange treatment for NH4Z1 sample. This is probably due to the second rinse applied here, which could eliminate possible ion exchanged by-products that might clog zeolite pores. Calcium content of ammonium-exchanged zeolites decreases from NH4Z1 to 2NH4Z1 and nitrogen content increases, as a result of the de-cationisation process. These experimental evidences are in agreement with other studies which report that sodium ions present in natural zeolites, like clinoptilolite, were weakly bonded to zeolite lattice and easy to remove [14]. Results for modified samples, indicate the following order of ammonium ion exchange preference toward compensating
Table 1 Physical–chemical characterisation of zeolite samples. Samples
S627 [m2 g−1 ]
SiO2 a
Al2 O3 a
Na2 Oa
CaOa
K2 Oa
MgOa
TiO2 a
Fe2 O3 a
MnOa
Si/Al ratio
N2 b
NZ NH4Z1 NH4Z2 NH4Z3 2NH4Z1
205 181 177 177 171
75.25 78.07 75.25 79.84 79.26
14.1 14.69 14.10 14.63 14.85
1.89 0.68 0.49 n.d. 0.26
4.57 2.36 1.96 1.97 1.82
0.74 0.67 0.6 0.46 0.39
0.66 0.46 0.46 n.d. 0.37
0.42 0.46 0.47 0.49 0.47
2.31 2.55 2.63 2.56 2.53
0.05 0.05 0.05 0.05 0.05
5.34 5.32 5.27 5.46 5.34
0.13 1.78 2.08 2.08 2.16
n.d. – Non-detected. a By XRF (% w/w). b By elemental analyses (% w/w).
H. Valdés et al. / Journal of Hazardous Materials 227–228 (2012) 34–40
37
Fig. 2. Thermogravimetric analysis coupled with mass spectroscopy (TGA-MS): (A) TG and DTG curves, (B) DTA curves, (C) ammonia evolution, and (D) H2 O evolution. Operation conditions: 0.025 g of zeolite, heating up to 950 K (rate 10 K min−1 ) in He (0.1 dm3 min−1 ), 101 kPa.
cations: Ca2+ > K+ > Mg2+ . Ion exchange capacity of clinoptilolite is known to be lower than some other natural zeolites, but generally exhibits a high affinity for NH4 + ion [15]. Surface area of natural zeolite sample decreases from 205 m2 g−1 to 170 m2 g−1 after thermal treatment at 823 K. This could be related to the interaction of compensating cations and water in the zeolite framework. Compensating cations are normally coordinated to a defined number of water molecules, affecting their mobility within the zeolite structure [16]. During the out-gassing process at 823 K, water molecules leave the zeolite, destabilising the zeolite charge structure [7]. In order to correct this, it has been suggested that compensating cations in the zeolite framework, form bonds with network oxygen atoms, affecting the zeolite structure channels [13,14]. However, in the ammonium-exchanged zeolites, surface areas are increased after thermal treatment at 823 K. The values of surface areas after the out-gassing step at 823 K, are as follow; NH4Z1 (222 m2 g−1 ), NH4Z2 (230 m2 g−1 ), NH4Z3 (251 m2 g−1 ) and 2NH4Z1 (261 m2 g−1 ). Such increase in the value of surface areas could be related to the elimination of ammonium due to thermal out-gassing at 823 K, thus reducing any blocking effect on microporous structures. Fig. 2 illustrates the results of thermogravimetric analyses coupled with mass spectroscopy (TGA-MS) conducted to natural and modified zeolite samples. Fig. 2A shows the change in sample weight (TG curve) and derivative weight loss (DTG curve) as the temperature increases. Fig. 2B shows DTA curve results, while Fig. 2C and D show, ammonia and H2 O evolutions as a function of temperature. The total weight loss registered during TGA analysis was as follows: 9.2% (NZ), 9.9% (NH4Z1), 9.4% (NH4Z2), 8.5% (NH4Z3), and
9.5% (2NH4Z1). DTG curves reveal two characteristic peaks. The first, around the temperature range of 407–431 K and the second around 723–738 K. The first peak could be related to the desorption of physisorbed gaseous molecules and the second could be mainly due to water evolution during Brønsted acid sites transformation into Lewis acid sites [17]. As it can be seen in Fig. 2B, exothermic peaks appear at 796, 866, 875, 928 and 875 K in the DTA curves for NZ, NH4Z1, NH4Z2, NH4Z3 and 2NH4Z1, respectively. These temperature peaks could indicate the beginning of thermal zeolite structure collapse. NZ shows the lowest thermal stability of all the zeolite samples used here, as indicated before by XRD analysis. The increases on thermal stability of ammonium-exchanged zeolites could be associated to a reduction on the compensating cation content as reported by other authors [16]. Fig. 2C shows the release of ammonia from the ammoniumexchanged zeolites as temperature rises. It can be seen that the amount of desorbed ammonia is higher in sample 2NH4Z1 than in the other samples. Zeolite compensating cations were exchanged with ammonium ions during chemical modification treatments. The results obtained here are in agreement with those obtained by elemental analyses, confirming the highest level of compensating cation exchange accomplished for this sample. In Fig. 2C, the peaks registered at different temperatures might be related to free re-adsorption of ammonia on vacant acid sites, as it has been reported by other authors [18]. Ammonium protons are quite mobile as temperature increases, evolving as ammonia; hence leading to Brønsted acid sites generation [17]. Fig. 2D presents the water evolution during the thermogravimetric analyses. Water signal intensity decreases when the
38
H. Valdés et al. / Journal of Hazardous Materials 227–228 (2012) 34–40
Fig. 3. NH3 -TPD analysis of zeolite samples. Operation conditions: 0.15 g of zeolite, heating up to 900 K (10 K min−1 heating rate) in Ar flow (100 cm3 min−1 ), 101 kPa.
temperature increases as a consequence of moisture elimination. Water signal intensity increases at 823 K and it could be associated to Brønsted acid sites transformation into Lewis acid sites [17]. Thermal out-gassing treatment at 823 K does not only eliminate adsorbed contaminants, but also contributes to new Lewis acid sites formation. Fig. 3 shows thermal desorption profiles of zeolite samples, previously saturated with NH3 . During NH3 desorption, two distinctive peaks appear at different temperature ranges, corresponding to sites of different acid strengths. The evolution of NH3 in the low (300–530 K) and high (530–810 K) temperature ranges could be attributed to the presence of weak and strong acid sites on the zeolite samples, respectively. It is known that the required energy to release ammonia from acid sites could be associated to the strength of chemical bonds (acid strength) [19]. Table 2 lists the maximum registered temperature peaks and the corresponding amount of desorbed NH3 . These parameters are used here as indicators of the zeolite acid site strengths and densities, respectively. As it can be seen, NH4Z3 sample shows a maximum desorption of NH3 in the high-temperature range, resulting in a zeolite sample with the highest acid strength. De-cationisation procedure applied in this study does not only lead to pore opening; but also allows Brønsted acid sites formation during ammonium protons transform into ammonia in the first part of the thermal out-gassing treatment. As the out-gassing temperature approaches to 823 K, Brønsted acid sites are transformed into Lewis acid sites, as confirmed by water evolution detected by MS (see Fig. 2D). 3.2. Effect of zeolite compensating cations content on gaseous ozone removal Fig. 4 illustrates the effect of compensating cation content on ozone abatement after out-gassing at 823 K. Natural zeolite shows
Fig. 4. Influence of zeolite compensating cation content on ozone removal. Operation conditions: 0.15 g of natural and modified zeolite samples thermally treated at 823 K, 0.345 mol dm−3 of inlet ozone concentration, 0.740 dm3 min−1 , 101 kPa, 293 K.
the lowest ozone removal. This zeolite sample has the highest content of compensating cations (M+ or M2+ ). Compensating cation characteristics such as size and charge could affect ozone interaction with Lewis acid sites within zeolite framework, reducing ozone abatement (see Scheme 1, section (A)). On the other hand, in the ammonium-exchanged zeolite samples, the zeolite framework could remain almost unchanged; and might decrease ozone diffusion resistance inside the zeolite structure, allowing higher ozone abatement on Lewis acid sites. Such chemical interactions are graphically illustrated in Scheme 1, sections (B) and (C), using the Lewis acid site schematic representation given by Karge [20]. Modified zeolite samples were less affected by the thermal out-gassing treatment, as confirmed by surface area determination. Ozone is known to act as a Lewis base and could be adsorbed as a molecule on weak Lewis acid sites and decomposes to form oxygen radicals (O• ) on strong Lewis acid sites [21]. Ozone interaction with acid surface sites could be due to its resonance structure, where the high electron density on one of oxygen atoms, might show high basicity resulting in strong affinity to Lewis acid surface sites [22]. Lewis acid sites on zeolite surface might play a key role enhancing the gaseous ozone elimination [7,8]. DRIFTS spectra of natural and modified zeolite samples are shown in Fig. 5 DRIFTS spectra were registered in the absence of ozone (t = 0 min) and after 10 and 20 min ozone exposure. Ozone interaction with zeolite samples increases the intensities of the bands at 1052–1054 cm−1 and at 1405–1416 cm−1 ; which could be associated to ozone molecular adsorption and to active atomic oxygen, respectively. In particular, IR band around 1037 cm−1 has been associated to ozone adsorption on weak Lewis acid sites [21,23]. Moreover, a peak at 1380 cm−1 has been assigned to a stable surface oxide species due to a strong bond where the atomic oxygen was
Scheme 1. Influence of compensating cations on ozone–zeolite surface interactions.
H. Valdés et al. / Journal of Hazardous Materials 227–228 (2012) 34–40
39
Table 2 Acidity of natural and modified zeolite samples, using NH3-TPD analysis. Samples
NZ NH4Z1 NH4Z2 NH4Z3 2NH4Z1
Maximum temperature peaks [K]
Acid site density [% desorbed NH3 ]
Weak acid sites (300–530 K)
Strong acid sites (530–810 K)
Weak acid sites (300–530 K)
Strong acid sites (530–810 K)
390 374 376 391 399
715 684 702 720 707
93 86 75 24 31
7 14 25 76 69
Fig. 5. DRIFTS spectra of natural and modified zeolite samples during adsorption of ozone. Spectra were taken after 0, 10, and 20 min. Operation conditions: natural and modified zeolite samples thermally treated at 823 K, 0.345 mol dm−3 of inlet ozone concentration, 101 kPa, 293 K.
40
H. Valdés et al. / Journal of Hazardous Materials 227–228 (2012) 34–40
attached to a Lewis acid site on synthetic zeolite surfaces [12,24]. This surface oxide specie has been identified as an oxidation product of ozone decomposition on zeolite surface [12,24].In the case of the natural zeolite sample, the IR bands indicated as molecular adsorbed ozone increase with ozonation time; whereas the frequency associated to the active atomic oxygen formation does not show major differences. However, in the case of the NH4Z3 sample, as ozone exposure increases, the peak area related to active atomic oxygen increases, without any arise on the peak frequency related to ozone molecular adsorption. These results could be due to the influence of the compensating cation content on the acid site strength of these samples, as confirmed by the NH3 -TPD analyses. For zeolite samples studied here, no spectral features were found that could be attributed to adsorbed molecular ozone on hydroxyl surface groups. However, an IR band was identified as ozone molecular adsorbed via coordinative bonding on weak Lewis acid sites. Moreover, ozone could transform immediately after adsorption on strong Lewis acid sites and decomposes into O2 and an active atom of oxygen, which could remain bound to a Lewis acid site. These active atoms of oxygen could then participate in catalytic removal of ozone [21]. 4. Conclusions The decreases of zeolite compensating cation content by applying ammonium ion-exchange treatments followed by thermal out-gassing at 823 K, reduces ozone diffusion resistance inside the zeolite framework, increasing ozone abatement on zeolite surface active sites. Ozone removal seems to be a combination of adsorption and surface-catalysed decomposition. DRIFTS results obtained here could be regarded as experimental evidence of the different reactivities of natural zeolite surface sites toward ozone. Weak and strong Lewis sites are identified here as zeolite active sites responsible of ozone reduction. Chemical modification of natural zeolite applied in this study could be an effective method for low cost zeolite production for ozone elimination in indoor environments. Acknowledgements Authors gratefully acknowledge FONDECYT (grant no. 1090182) for their financial support. S. Alejandro wishes to thank CONICYT for providing a doctoral scholarship and to the Carbon and Catalysis Group from University of Concepcion for their valuable collaboration. References [1] J.N. Cape, Surface ozone concentrations and ecosystem health: past trends and a guide to future projections, Sci. Total Environ. 400 (2008) 257–269.
[2] B. Dhandapani, S.T. Oyama, Gas phase ozone decomposition catalysts, Appl. Catal. B 11 (1997) 129–166. [3] S. Bhangar, S.C. Cowlin, B.C. Singer, R.G. Sextro, W.W. Nazaroff, Ozone levels in passenger cabins of commercial aircraft on North American and transoceanic routes, Environ. Sci. Technol. 42 (2008) 3938–3943. [4] M.S. O’Neill, D. Loomis, V.H. Borja-Aburto, Ozone, area social conditions, and mortality in Mexico City, Environ. Res. 94 (2004) 234–242. [5] G. Martra, R. Ocule, L. Marchese, G. Centi, S. Coluccia, Alkali and alkaline-earth exchanged faujasites: strength of Lewis base and acid centres and cation site occupancy in Na- and BaY and Na- and BaX zeolites, Catal. Today 73 (2002) 83–93. [6] L. Martins, R.T. Boldo, D. Cardoso, Ion exchange and catalytic properties of methylammonium FAU zeolite, Microporous Mesoporous Mater. 98 (2007) 166–173. [7] S. Alejandro, H. Valdés, C.A. Zaror, Natural zeolite reactivity towards ozone: the role of acid surface sites, J. Adv. Oxid. Technol. 14 (2011) 182–189. [8] H. Valdés, E. Padilla, C.A. Zaror, Influence of chemical surface characteristics of natural zeolite on catalytic ozone abatement, Ozone: Sci. Eng. 33 (2011) 279–284. [9] H. Valdés, V.J. Farfán, J.A. Manoli, C.A. Zaror, Catalytic ozone aqueous decomposition promoted by natural zeolite and volcanic sand, J. Hazard. Mater. 165 (2009) 915–922. [10] F. Benaliouche, Y. Boucheffa, P. Ayrault, S. Mignard, P. Magnoux, NH3-TPD and FTIR spectroscopy of pyridine adsorption studies for characterization of Ag- and Cu-exchanged X zeolites, Microporous Mesoporous Mater. 111 (2008) 80–88. [11] C.W. Kwong, C.Y.H. Chao, K.S. Hui, M.P. Wan, Removal of VOCs from indoor environment by ozonation over different porous materials, Atmos. Environ. 42 (2008) 2300–2311. [12] C.W. Kwong, C.Y.H. Chao, K.S. Hui, M.P. Wan, Catalytic ozonation of toluene using zeolite and MCM-41 materials, Environ. Sci. Technol. 42 (2008) 8504–8509. [13] J. Li, J. Qiu, Y. Sun, Y. Long, Studies on natural STI zeolite: modification, structure, adsorption and catalysis, Microporous Mesoporous Mater. 37 (2000) 365–378. [14] S.J. Allen, E. Ivanova, B. Koumanova, Adsorption of sulfur dioxide on chemically modified natural clinoptilolite. Acid modification, Chem. Eng. J. 152 (2009) 389–395. [15] Y.F. Wang, F. Lin, W.Q. Pang, Ammonium exchange in aqueous solution using Chinese natural clinoptilolite and modified zeolite, J. Hazard. Mater. 142 (2007) 160–164. [16] X.-w. Cheng, Y. Zhong, J. Wang, J. Guo, Q. Huang, Y.-c. Long, Studies on modification and structural ultra-stabilization of natural STI zeolite, Microporous Mesoporous Mater. 83 (2005) 233–243. [17] M. Stöcker, Gas phase catalysis by zeolites, Microporous Mesoporous Mater. 82 (2005) 257–292. [18] K. Suzuki, T. Noda, N. Katada, M. Niwa, IRMS-TPD of ammonia: direct and individual measurement of Brønsted acidity in zeolites and its relationship with the catalytic cracking activity, J. Catal. 250 (2007) 151–160. ˜ [19] E. Díaz, S. Ordónez, A. Vega, J. Coca, Catalytic combustion of hexane over transition metal modified zeolites NaX and CaA, Appl. Catal. B 56 (2005) 313–322. [20] H.G. Karge, Caracterization of zeolites, in: J. Weitkamp, L. Pupe (Eds.), Catalysis and Zeolites: Fundamentals and Application, Springer-Verlag, Berlin, 1999, pp. 215–231. [21] K.M. Bulanin, J.C. Lavalley, A.A. Tsyganenko, IR spectra of adsorbed ozone, Colloids Surf. A 101 (1995) 153–158. [22] S.T. Oyama, Catalytic properties of ozone, Catal. Rev. Sci. Eng. 42 (2000) 279–322. [23] K.M. Bulanin, A.V. Alexeev, D.S. Bystrov, J.C. Lavalley, A.A. Tsyganenko, IR study of ozone adsorption on SiO2 , J. Phys. Chem. A 98 (1994) 5100–5103. [24] C.Y.H. Chao, C.W. Kwong, K.S. Hui, Potential use of a combined ozone and zeolite system for gaseous toluene elimination, J. Hazard. Mater. 143 (2007) 118–127.