Catalysts based on pillared clays for the oxidation of chlorobenzene

G Model

ARTICLE IN PRESS

CATTOD-9181; No. of Pages 13

Catalysis Today xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Catalysis Today journal homepage: www.elsevier.com/locate/cattod

Catalysts based on pillared clays for the oxidation of chlorobenzene A. Aznárez a , R. Delaigle b , P. Eloy b , E.M. Gaigneaux b , S.A. Korili a , A. Gil a,∗ a

Department of Applied Chemistry, Building Los Acebos, Public University of Navarra, Campus of Arrosadia, E-31006 Pamplona, Spain Université cathlolique de Louvain, Institute of Matter Condensed and Nanosciences (IMCN), MOlecules, Solids and reactiviTy (MOST), Croix du Sud, 2/L17.05.17, B-1348 Louvain la Neuve, Belgium b

a r t i c l e

i n f o

Article history: Received 21 April 2014 Received in revised form 30 June 2014 Accepted 15 July 2014 Available online xxx Keywords: Pillared clays Palladium supported catalysts Platinum supported catalysts Chlorobenzene oxidation

a b s t r a c t The aim of this work was to reveal the main factors which affect the oxidation of chlorobenzene (PhCl) over palladium and platinum supported on alumina pillared clays. The catalysts were prepared by wet impregnation of an alumina-pillared montmorillonite (Al-PILC) with palladium and platinum solutions and characterized by several physicochemical techniques before and after the catalytic tests. During oxidation of PhCl over the catalysts, the formation of carbon dioxide along with small quantities of carbon monoxide, PhClx and coke was found. The nature of the supported metal, the temperature, the metal loading, the support, and the time on stream, are factors affecting the combustion of PhCl. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Volatile Organic Compounds (VOCs), which often represent a serious environmental problem, can be produced from a variety of industrial and commercial processes, including printing, metal decorating, paint drying, metal degreasing, manufacturing of organic compounds and polymers and food processing [1]. Chlorinated VOCs (Cl-VOCs) are considered important environmental pollutants due to their toxicity and high stability. As a result of their widespread application in industry, their production and hence their emission into the environment are increasing rapidly. As ClVOCs have always been considered among the most hazardous organic compounds emitted into the environment, the decomposition and removal of these pollutants have always been subjects of great importance [2–4]. While the methods commonly employed in removing gaseous pollutants from industrial gas streams, wet scrubbing and adsorption, are relatively efficient in removing several gaseous pollutants, their efficiency towards Cl-VOCs is very limited [5]. A large number of methods have been applied in order to solve the problem of releasing Cl-VOCs into the atmosphere: e.g. thermal incineration, hydrodechlorination, biological processes, steam reforming and photocatalytic degradation [6]. Thermal incineration, usually above 1000 ◦ C, is the most extensively used method to remove VOCs

∗ Corresponding author. Tel.: +34 948 169602; fax: +34 948 169602. E-mail address: [email protected] (A. Gil).

from industrial air streams. Besides the high cost, due to the high temperatures necessary, it involves additional fuel, high operating costs and the use of temperature-resistant materials. Moreover, the incineration process is always associated with the formation of a wide range of toxic byproducts, such as NOx , and partial oxidation products, such as phosgene (COCl2 ), polychlorinated dibenzodioxins (PCDDs), and polychlorinated dibenzofurans (PCDFs) [2,3,7,8]. A well-managed incineration process is feasible but expensive, due to the advanced pollution control devices needed to prevent the emission of dioxins, for example [9]. Cost analysis has clearly shown that thermal incineration technologies should be dedicated to high flow rates or VOC contents [10]. Catalytic combustion is a promising air abatement technology for treating VOCs under moderate flow rates. The major advantage of catalytic combustion is that very dilute pollutants (<1%), which cannot be thermally combusted without additional fuel, can be treated efficiently. Thus, catalytic combustion offers environmental advantages, since it is an energy-efficient, low-cost process and operates at much lower temperatures, thus preventing the formation of NOx [8,11]. The low operating temperatures (<500 ◦ C) and high selectivity into harmless products, such as CO2 , H2 O and HCl/Cl2 , make it an attractive option [12–16]. Catalytic combustion has broader application for end-of-pipe pollution clean-up than thermal combustion [2]. For catalysts, there are two points to be considered. As the industrial catalytic reactors have to be operated at high space velocities, the use of highly active catalysts is needed [17]. It must be also taken into account that the treated pollutants are emitted from various industrial sectors, thus another important

http://dx.doi.org/10.1016/j.cattod.2014.07.024 0920-5861/© 2014 Elsevier B.V. All rights reserved.

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024

G Model CATTOD-9181; No. of Pages 13

ARTICLE IN PRESS A. Aznárez et al. / Catalysis Today xxx (2014) xxx–xxx

2

characteristic of these catalysts is the capability to remove all the contaminants present in the stream simultaneously [18]. Catalytic combustion has been widely applied to the destruction of Cl-VOCs [2,3,19–22]. However, the interaction of the catalysts with chlorine is the main problem in the design of catalytic systems for Cl-VOCs combustion. The optimal catalyst for this reaction should be active, stable and above all, highly selective towards CO2 , H2 O and HCl; limiting the formation of other environmentally hazardous organic compounds [23]. Catalysts based on noble metals (Pt, Pd) [8,9,12,19,22–24] and transition metals (Cu, Co, V, Mn, Fe, Cr) [8,25–27] are used in the combustion of VOCs. Noble metals are effective catalysts for combustion reactions, but can easily form inorganic chlorides that could also cause chlorination of organic compounds besides their oxidation [28]. Van den Brink et al. [19] and de Jong et al. [9] showed that the combustion of PhCl over 2% Pt/␥-Al2 O3 produced a significant amount of PhClx , in addition to the expected combustion products. Similarly, the combustion of PhCl carried out by Scirè et al. over 0.5% Pt supported on several zeolites also produced PhClx [23]. Becker and Förster [22] and Oliveira et al. [8] showed the formation of PhClx in the combustion of PhCl when using Pd supported catalysts. Giraudon et al. reported the formation of polychlorinated by-products with Pd supported on perovskites [24] and nanostructured TiO2 and ZrO2 [12]. These polychlorinated byproducts are more toxic and recalcitrant than the starting molecule. Despite these drawbacks, Pd and Pt catalysts have been investigated and commercially applied in the combustion of Cl-VOCs [1,29–31]. Noble metals deposited on conventional supports like Al2 O3 , SiO2 , and TiO2 have been studied as catalysts for the combustion of Cl-VOCs. Pillared InterLayered Clays (PILCs) are also good supports due to their high surface area, special surface acidity and thermal stability. In spite of being added-value materials from natural products and having very interesting properties, there is very little information in the literature about transition metal oxides and noble metal catalysts supported on PILCs for the catalytic combustion of chlorobenzene (PhCl). The advantage of using PILCs with respect to other supports is that, starting from natural, inexpensive clay minerals, it is possible to develop new materials with a microporous structure that can be controlled during the intercalation process, and create catalytic sites by incorporating metal cations. The possibility of catalyst design and the control of active site distribution are the two main reasons for choosing pillared clays as stable and selective catalysts in sustainable combustion reactions [32]. In this study the main factors affecting the combustion of PhCl related to the type of catalysts prepared are examined. The use of the catalytic series Pd/Al-PILC and Pt/Al-PILC in PhCl combustion, the comparison between the two series, the detailed study of the products obtained and the study of the stability of a selected catalyst are original in this study.

2. Experimental 2.1. Catalyst preparation and characterization techniques The catalysts were prepared from an alumina-pillared montmorillonite which has been described in detail in a previous work [33]. The starting material was a montmorillonite from Tsukinuno, supplied by The Clay Science Society of Japan. The clay mineral was pillared with alumina, according to a conventional pillaring procedure, for use as a support in all the sample preparations. Supported metal catalysts were prepared by wet impregnation of the support with solutions of palladium (palladium(II) nitrate solution, 10 wt.% in 10 wt.% HNO3 , Sigma–Aldrich) and platinum salts

([Pt(NH3 )2 ](NO2 )2 , Strem Chemicals). The metal salt/clay slurries were evaporated to dryness under reduced pressure in a rotary evaporator and the resulting solids dried at 120 ◦ C for 16 h before being calcined in air at 500 ◦ C for 4 h to form the final supported catalysts. The catalysts had metal loadings of 0.1, 0.5, 1 and 2 wt.%, and are referred to hereinafter as wt.Met, where wt. indicates the metal content and Met the metallic phase (Pd or Pt). The raw clay is referred to as Na-Mont, and the alumina pillared clay as Al-PILC. The catalytic series are referred to as Pd/Al-PILC and Pt/AlPILC. The physicochemical characterization of the catalysts included chemical analysis, nitrogen physisorption at −196 ◦ C, carbon dioxide physisorption at 0 ◦ C, X-ray photoelectron spectroscopy (XPS), X-ray powder diffraction (XRD), CO and H2 chemisorption at 35 ◦ C, NH3 chemisorption at 70 ◦ C, and temperature-programmed reduction (TPR). Details of the experimental conditions are given elsewhere [33]. 2.2. Catalytic performance Chlorobenzene (PhCl) oxidation was carried out on an automated bench-scale catalytic unit (Microactivity Reference, PID Eng&Tech). The reactor was a tubular, fixed-bed, downflow type, with an internal diameter of 0.9 cm and a length of 30 cm. Catalyst samples were mixed with an inert material (glass spheres with diameters in the range 315–500 ␮m), at a weight ratio of 1:4 to dilute the catalyst bed and avoid hot spot formation. The catalytic bed was composed of 200 mg of catalyst powder selected within the 200–315 ␮m granulometric fraction and diluted in 800 mg of inactive glass spheres with diameters in the range 315–500 ␮m. The PhCl concentration in the feed was 0.01% (Praxair; PhCl-helium, 0.2–99.8%) and the oxygen-to-hydrocarbon molar ratio was 2000 (Praxair; 99.995%), with helium as diluting gas (Praxair; 99.996%), up to a total feed flow of 200 cm3 /min. Space velocities (GHSV), calculated at standard temperature and pressure and based on the volume of the catalytic bed, were about 22,000 h−1 . Before and after the reaction, stabilization stages of 1.5 h were carried out in by-pass mode at 100 and 25 ◦ C, respectively, to investigate the evolution of the concentration of the various reagents and products. After the first stabilization stage, the reaction was run from 100 to 400 ◦ C in 50 ◦ C steps. At each temperature considered, the catalyst was stabilized for 150 min in order to ensure steady-state conversion. The reactants and the reaction product streams were analyzed on-line using a Varian CP-3800 gas chromatograph system equipped with three detectors (one TCD and two FIDs). Analysis of the gases was performed by separating them into a four-column system (Hayesep Q, Hayesep T, Molsieve 13X and CP-Sil 8CB). The parameters of the GC system allowed an analysis approximately every 35 min (each injection lasts 28 min). All tubes and injection valves in the GC were heated at 190 ◦ C to prevent condensation of products. The identification and quantification of the PhClx was carried out by injecting in the GC a liquid PhCl mixture of known composition from Sigma–Aldrich (40621-U-DL 152/2006 Chlorobenzene mixture). The analytical standard, which contained the same concentration (100 ␮g/cm3 ) of each component in methanol, included: PhCl, 1,2-PhCl2 , 1,3PhCl2 , 1,4-PhCl2 , 1,2,4-PhCl3 , 1,2,4,5-PhCl4 , PhCl5 and PhCl6 . The retention times of all peaks, except the one ascribed to methanol, were compared and identified as specific congeners of PhClx . The denomination 1,2,x,5-PhCl4 represents both 1,2,3,5-PhCl4 and 1,2,4,5-PhCl4 isomers, because these compounds could not be separated on our GC. PhCl conversion was calculated from the disappeared reagent as well as from the products obtained. Only the data measured after 2–2.5 h of reaction time were taken into account.

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024

G Model CATTOD-9181; No. of Pages 13

ARTICLE IN PRESS A. Aznárez et al. / Catalysis Today xxx (2014) xxx–xxx

3. Results and discussion

3.2. Catalytic performance Conversion of PhCl over an inert material (glass spheres with diameters in the range 315–500 ␮m), Al-PILC, Pd/Al-PILC and Pt/AlPILC is shown in Fig. 1. To determine the extent of homogeneous reactions, a blank test was performed under exactly the same conditions but with inert material only. In this case, the conversion of PhCl at 400 ◦ C was 9%, and small amounts of PhCl2 were found, approximately the yield of PhCl2 was 1.5%. When the support alone was tested, conversion at 400 ◦ C was about 22%, which was significantly higher than the conversion reached with the inert material, and the yield of PhCl2 was about 1%. Conversion of PhCl with the catalysts occurred at lower temperatures than with the support. In both cases, the 0.1 wt.% metal-loaded catalyst improved the catalytic conversion of the support, with the improvement being much higher with 0.1Pt than with 0.1Pd. Neither the support nor 0.1Pd consumed 50% of the PhCl in the feed in the temperature range studied. Conversion of 90% of the PhCl in the feed was only reached with 2Pt, and complete conversion of PhCl was not reached at 400 ◦ C with any catalyst. The temperature needed for PhCl combustion with the Pt/Al-PILC catalysts was lower than with the Pd/Al-PILC ones. Also, the conversion of PhCl increased with increasing metal loading. T20 and T50 are defined as the temperatures at which 20 and 50% of the PhCl in the feed is consumed. For Al-PILC, Pd/Al-PILC and Pt/Al-PILC, T20 and T50 are shown in Table 2.

Conversion of chlorobenzene (%)

The results of the physicochemical characterization of the fresh catalysts have been presented and discussed in detail in a previous report [33]. The main structural and textural properties of the catalysts are summarized in Table 1. The basal spacing d(0 0 1) of Al-PILC was 1.64 nm, clearly higher than 1.20 nm, the basal spacing of Na-Mont. These XRD results, together with those of nitrogen adsorption, suggest that the intercalation-pillaring processes of Na-Mont were successful. Impregnation of the support with an aqueous solution of metal resulted in a loss of surface area and pore volume. The study of the surface and porous network of the samples provided useful information about the location of the impregnated cations. From the adsorption of nitrogen and carbon dioxide, it was deduced that pores with diameters of between 0.4 and 1.5 nm were not influenced by the presence of metal species, whereas those with diameter of between 1.5 and 2 nm were. Also, while the proportion of particles situated on the external surface is greater for the Pd/Al-PILC catalysts than for the Pt/Al-PILC ones, the proportion of particles situated on the micropore surface is greater for the Pt/Al-PILC catalysts than for the Pd/Al-PILC ones. In the case of 2Pt, the platinum particles block a significant portion of those micropores with diameters in the range 1.5–2 nm. Taking into account the results from H2 and CO chemisorption, for both catalytic series, metal dispersion decreases with increasing metal loading, especially in the case of the Pd/Al-PILC catalysts. Particle sizes in the ranges of 3.8–24.9 and 3.8–8.8 nm were obtained for the Pd/Al-PILC and Pt/Al-PILC catalysts. The XRD patterns of the samples revealed the presence of PdO and Pt. Our results for the Pt/Al-PILC catalysts show that, besides the metallic Pt detected by XRD, bulk PtO2 and small PtO2 particles, which partially interact with the support, can be detected by XPS and TPR. For the Pd/Al-PILC catalysts, although the XRD technique only showed diffraction peaks for bulk PdO, PdO2 or highly dispersed and deficiently coordinated Pd2+ in intimate contact with the alumina support to form palladium-alumina structures and Pd species could also be detected by XPS and TPR.

Inert Al-PILC 0.1Pd 0.5Pd 1Pd 2Pd

80

60

40

20

0 0

100

200 Temperature (ºC)

300

400

200

300

400

100 Conversion of chlorobenzene (%)

3.1. Physico-chemical characterization

100

3

Inert Al-PILC 0.1Pt 0.5Pt 1Pt 2Pt

80

60

40

20

0 0

100

Temperature (ºC) Fig. 1. Conversion of PhCl, calculated from products, over an inert material, the support, Pd/Al-PILC and Pt/Al-PILC.

During conversion of PhCl over the catalysts, the formation of carbon dioxide along with small quantities of carbon monoxide and PhClx (x = 2–4) was observed. HCl and Cl2 were not quantified. The yield of COx and PhClx , as the selectivity to COx and PhClx , for conversions of 20 and 50% are included in Table 2. For a given conversion, selectivity to PhClx is higher for the Pd/Al-PILC catalysts than for the Pt/Al-PILC ones, whereas selectivity to COx is higher for the Pt/Al-PILC catalysts than for their Pd/Al-PILC counterparts. Thus, the Pt/Al-PILC catalysts are more selective to COx and more active in PhCl combustion than their Pd/Al-PILC counterparts. Also, for a given conversion, the yield of COx decreases and that of PhClx increases with increasing metal loading. Thus, the yield of COx with respect to conversion, which is selectivity to COx , decreases for the Pd/Al-PILC and Pt/Al-PILC catalysts with increasing metal loading. Instead, selectivity to PhClx increases for the catalysts with increasing metal loading. From comparing data at PhCl conversions of 20 and 50%, it is concluded that for a given palladium and platinum loading, selectivity to COx decreases and selectivity to PhClx increases with increasing PhCl conversion. Related to CO formation, it has been seen that selectivity to CO decreases with increasing metal loading and conversion. Formation of CO was higher with the Pd/Al-PILC catalysts than with their Pt/AlPILC counterparts, but small over the entire range of conversions. For a conversion equal to 20%, selectivity to CO with the support was 22%. For the Pd/Al-PILC serie, it varied from 1.8% (0.1Pd) to 0.2% (2Pd), and for the Pt/Al-PILC one, from 0.7% (0.1Pt) to 0% (2Pt).

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024

G Model

ARTICLE IN PRESS

CATTOD-9181; No. of Pages 13

A. Aznárez et al. / Catalysis Today xxx (2014) xxx–xxx

4

Table 1 Some physicochemical properties of the materials indicated. Catalyst

SLang a (m2 /g)

Vpb (cm3 /g)

V(NH3 )c (mmol/g)

D(H2 )d (%)

dp(H2 )e (nm)

D(CO)d (%)

dp(CO)e (nm)

Metal contentf (wt.%)

Al-PILC 0.1Pd 0.5Pd 1Pd 2Pd 0.1Pt 0.5Pt 1Pt 2Pt

212 162 168 156 151 150 150 128 74

0.113 0.086 0.090 0.085 0.066 0.083 0.087 0.076 0.050

0.26 0.28 0.28 0.31 0.29 0.28 0.26 0.27 0.18

– – – 6.0 5.3 1.5 12.6 13.0 11.8

– – – 18.7 21.1 75.5 9.0 8.7 9.6

– 29.7 9.6 5.6 4.5 29.9 19.0 17.6 12.8

– 3.8 11.7 20.0 24.9 3.8 6.0 6.4 8.8

– 0.08 0.50 1.06 2.03 0.09 0.49 0.99 2.01

a b c d e f

Specific surface area calculated using the Langmuir equation. Specific total pore volume. Ammonia adsorption volume. Metal dispersion. Mean size of the metal particles. Metal content from chemical analyses.

Catalytic conversion was calculated from the PhCl disappearance as well as from the product appearance (CO, CO2 and PhClx ), and a comparison between them was made. There are some factors, coke formation and PhClx retention in the catalysts, which affect the carbon balance, and thus the comparison between conversions. In our study, in the stabilization stage carried out at 25 ◦ C in bypass mode after the catalytic test, some PhClx (PhCl3 and PhCl4 ) were detected with 0.5Pd, 1Pd, 2Pd, 1Pt and 2Pt. Van den Brink et al. [19], who studied combustion of PhCl over Pt/␥-Al2 O3 , observed that at temperatures higher than that at which PhCl was completely converted (ca. 420 ◦ C), considerable amounts of polychlorinated benzenes were still present in the gas stream. It was only near 600 ◦ C that no PhClx could be detected anymore. They also observed that higher chlorinated benzenes still left the reactor hours after the feed of PhCl was stopped. They thought that it might be the case that PhClx were already formed at lower temperatures but desorbed only at a higher temperature, concluding that some PhClx derivatives had strong interaction with the catalyst. In the case of the support, there is good agreement between both conversions; which means that a large quantity of carbonaceous products was not formed. However, in the case of the catalysts, there are differences throughout practically the entire range of temperatures/conversions. For 0.1Pd, the conversion calculated from product formation is smaller than that calculated from PhCl disappearance for the entire range of temperatures, with the differences observed due to the formation of carbonaceous residues (coke is counted when conversion is calculated from PhCl disappearance, but not when calculated from product formation). For the other Pd/Al-PILC catalysts, 0.5Pd, 1Pd and 2Pd, the same trend is observed up to 350 ◦ C, and the opposite is observed at 400 ◦ C. In these cases, the differences observed up to 350 ◦ C are due to the formation of carbonaceous residues as well as the retention of PhCl3 and PhCl4 compounds. At 400 ◦ C, and because of the desorption of the products retained at lower temperatures, the carbon balance for these catalysts increases above the 100% and the conversion calculated

from product formation is overestimated. In the case of the Pt/AlPILC catalysts, conversion from PhCl disappearance is higher than that calculated from product appearance because of the formation of carbonaceous residues. With 1Pt and 2Pt, although some PhCl3 and PhCl4 were retained at low temperatures, neither the carbon balance nor the conversion calculated from product formation was overestimated at 400 ◦ C; contrary to that noticed for the Pd/Al-PILC catalysts. This is because the amounts of PhCl3 and PhCl4 formed with 1Pt and 2Pt are very small in comparison with the amounts formed with their Pd/Al-PILC counterparts. The carbon balances show that for a given conversion of PhCl, the yield of carbonaceous residues is more important in the presence of the catalysts than in the presence of the support. Also, it is more important over the Pt/Al-PILC catalysts than over their Pd/Al-PILC counterparts. Finally, the yield of coke increases with increasing metal loading. During conversion of PhCl over the support, PhCl2 was the only PhClx congener formed. In the presence of the Pd/Al-PILC and Pt/AlPILC catalysts, PhCl2 , PhCl3 and PhCl4 were formed, whereas neither PhCl5 nor PhCl6 were detected. The yield of PhCl2 , PhCl3 , and PhCl4 increased with increasing reactor temperature. In both cases, the increasing order appearance of PhClx congeners as a function of the reactor temperature was: PhCl2 , PhCl3 , and PhCl4 . With 2Pd, PhClx congeners appeared at the following reactor temperatures: PhCl2 (250 ◦ C), PhCl3 (300 ◦ C) and PhCl4 (400 ◦ C), and with 2Pt, they appeared at a slightly higher temperature: PhCl2 (250 ◦ C), PhCl3 (350 ◦ C) and PhCl4 (400 ◦ C). The yield of PhClx congeners followed the order: PhCl2 > PhCl3 > PhCl4 for both catalytic series. The yield of PhCl2 , PhCl3 , and PhCl4 as well as the total yield of PhClx were considerably higher for 2Pd than for 2Pt. Palladium is known to be a more active chlorination catalyst than platinum [34]. The yield of PhClx (PhCl2 , PhCl3 , and PhCl4 ) over the support, Pd/Al-PILC and Pt/Al-PILC for various PhCl conversions (20, 50, 75 and 90%) is shown in Fig. 2. For a given conversion of PhCl, the following trends were observed: (1) the higher the metallic loading, the higher the yield of PhClx ; (2) the yield of PhClx congeners was

Table 2 Percentage of PhCl feed converted into COx and PhClx , and selectivity to COx and PhClx for conversions of 20 and 50%. Catalyst

T20 a (◦ C)

COx (%PhClin )

SCOx (%)

PhClx (%PhClin )

SPhClx (%)

T50 a (◦ C)

COx (%PhClin )

SCOx (%)

PhClx (% PhCl in )

SPhClx (%)

Al-PILC 0.1Pd 0.5Pd 1Pd 2Pd 0.1Pt 0.5Pt 1Pt 2Pt

389 353 338 347 320 302 278 289 250

19 16.5 12.5 7 6.5 17.5 16 15 14

95 82.5 62.5 35 32.5 87.5 80 75 70

0.8 2 5 9 12 0 0.4 0.5 0.8

4 10 25 45 60 0 2 2.5 4

– – 397 384 367 343 336 338 321

– – 24 16.5 15 42 37 34 28

– – 48 33 30 84 74 68 56

– – 27 33 46 0 3 5.5 6

– – 54 66 92 0 6 11 12

a

T20 and T50 , temperatures at which the conversion is 20 and 50%.

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024

G Model CATTOD-9181; No. of Pages 13

ARTICLE IN PRESS A. Aznárez et al. / Catalysis Today xxx (2014) xxx–xxx

5

Fig. 2. Yield of PhClx congeners (x = 2–4) in the presence of the support, Pd/Al-PILC and Pt/Al-PILC for various PhCl conversions: (A) 20%, (B) 50%, (C) 75% and (D) 90%.

as follows: PhCl2 > PhCl3 > PhCl4 ; (3) the yield of PhClx was much higher for the Pd/Al-PILC catalysts than for the Pt/Al-PILC ones; (4) with increasing conversion and thus temperature, the order of appearance of PhClx congeners is: PhCl2 , PhCl3 and PhCl4 and (5) for a given catalyst, the yield of PhClx increases with increasing PhCl conversion. The distribution of PhCl2 in the presence of the support, Pd/Al-PILC and Pt/Al-PILC for various PhCl conversions (20, 50, 75 and 90%) is presented in Fig. 3. The percentage of each PhCl2 isomer (o-PhCl2 , m-PhCl2 , and p-PhCl2 ) was calculated as the

ratio between the selectivity to these and the selectivity to total PhCl2 . The distribution of PhCl2 isomers in the blank test at 400 ◦ C was: p-PhCl2 (54%) > o-PhCl2 (46%). For the support, and for a conversion of 20%, it follows the order: m-PhCl2 (40%) > p-PhCl2 (35%) > o-PhCl2 (25%). For the catalysts, the distribution is affected by the conversion and by the metal loading. Formation of PhCl2 isomers becomes easier with increasing conversion and metal loading. For example, in the presence of 0.5Pt, 1Pt and 2Pt, for a conversion of 20% only p-PhCl2 and m-PhCl2 were formed, whereas for higher conversions o-PhCl2 was also formed. Also, for

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024

G Model CATTOD-9181; No. of Pages 13 6

ARTICLE IN PRESS A. Aznárez et al. / Catalysis Today xxx (2014) xxx–xxx

Fig. 3. Isomer selectivity of PhCl2 in the presence of the support, Pd/Al-PILC and Pt/Al-PILC for various PhCl conversions: (A) 20%, (B) 50%, (C) 75% and (D) 90%.

a conversion of 20% and in the presence of 0.1Pt no PhCl2 isomer was formed, whereas in the presence of a catalyst with a higher metal loading, p-PhCl2 and m-PhCl2 were formed. However, once the three PhCl2 isomers have been formed, the distribution does not depend either on the conversion or on the metal loading, and the trend was: p-PhCl2 > m-PhCl2 > o-PhCl2 . Nevertheless, the distribution of PhCl2 for the Pd/Al-PILC and Pt/Al-PILC catalysts was somewhat different. For Pd/Al-PILC was: p-PhCl2 (45%) > m-PhCl2 (33%) > o-PhCl2 (22%); and for Pt/Al-PILC: p-PhCl2 (60%) > m-PhCl2 (26%) > o-PhCl2 (14%). It is clear from the blank test that, when

PhCl2 is formed through homogeneous reactions, the chlorination of PhCl occurs by means of a classical aromatic electrophilic substitution mechanism, with the ortho and para isomers being preferentially formed. Distributions of PhCl2 for the support and catalysts, and their comparison with that obtained for the blank test, rule out an aromatic electrophilic substitution mechanism, and thus the homogeneous reactions, as the main mechanism of PhCl chlorination. In the case of the support, it is conceivable that chlorination of PhCl occurs through a radical mechanism over the support (m-PhCl2 > p-PhCl2 > o-PhCl2 ). However, to explain the

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024

G Model CATTOD-9181; No. of Pages 13

ARTICLE IN PRESS A. Aznárez et al. / Catalysis Today xxx (2014) xxx–xxx

7

Fig. 4. Congener selectivity of PhClx for various PhCl conversions (20%, 50%, 75% and 90%) in the presence of: (A) 0.5Pt and B) 2Pt.

particular selectivity of the Pd/Al-PILC and Pt/Al-PILC catalysts (p-PhCl2 > m-PhCl2 > o-PhCl2 ), several authors [20,35] have reported that secondary reactions on the support (e.g. isomerization of PhClx formed on the Pt) could cause variations in congener and isomer patterns, although this has not been demonstrated experimentally. As an example, Taralunga et al. [35] studied the catalytic combustion of PhCl over Pt supported on a HFAU zeolite. They explained the distribution obtained (p-PhCl2 > m-PhCl2 > oPhCl2 ), by a monomolecular isomerization of the m-PhCl2 isomer, formed by radical mechanism over Pt species, on the strong Brönsted acid sites of the zeolite. They also suggested that the electronic state and/or the location of Pt could be an important parameter for the formation of p-PhCl2 , as it had been previously reported that platinum could be directly involved in the formation of PhClx compounds. Derouane et al. [36] reported that the types of diffusion constraints imposed by coke deposition, blockage of the pore mouth and pore mouth restriction, could affect the molecular shape properties of the zeolites. These coking modes would progressively reduce the diffusivity of reactant and product molecules as a result of a reduction in effective pore opening, enhancing the para-aromatic selectivity of the coked catalysts. In our case, the deposition of coke could explain the PhCl2 selectivity of the catalysts. In the presence of the Pt/Al-PILC catalysts, the formation of coke as well as the selectivity to the p-PhCl2 isomer, are more important than in the presence of their Pd/Al-PILC counterparts. Van den Brink et al. [20] studied the catalytic combustion of PhCl over Pt/SiO2 , Pt/SiO2 –Al2 O3 , Pt/␥-Al2 O3 and Pt/ZrO2 , as well as on the supports. They observed that the congener and isomer distribution of PhClx varied with the support. In contrast to the ␥-Al2 O3 support, with which no PhClx was observed at any temperature, some PhCl2 was observed on SiO2 , SiO2 –Al2 O3 and ZrO2 . The levels of PhClx , especially PhCl4 , PhCl5 , and PhCl6 , were quite substantial on the zirconium dioxide supported catalyst. On Pt/SiO2 , only small amounts of PhCl2 were formed. Pt/SiO2 –Al2 O3 showed considerable proportions of PhCl2 and PhCl3 , while higher chlorinated products were almost absent. Finally Pt/␥-Al2 O3 gave a more even distribution of lower and higher chlorinated

benzenes. The distribution of PhCl2 at 320 ◦ C was about the same for Pt/␥-Al2 O3 , Pt/ZrO2 and Pt/SiO2 : p-PhCl2 (45%) > m-PhCl2 (40%) > oPhCl2 (15%); but it was somewhat different for Pt/SiO2 -Al2 O3 : p-PhCl2 (55%) > m-PhCl2 (25%) > o-PhCl2 (20%). In spite of having found that the difference in Pt crystallite size was not important in those variations, they did not find which factors determined it. The distribution of PhCl3 and PhCl4 isomers is also affected by the conversion and by the metal loading. For a given metal loading, the formation of PhCl3 and PhCl4 isomers becomes easier with increasing conversion. In the same way, for a given conversion, the formation of PhCl3 and PhCl4 isomers becomes easier with increasing metal loading. But when the three PhCl3 isomers are formed, the distribution does not depend on the mentioned factors. In this case, the PhCl3 distribution for the Pd/Al-PILC catalysts was almost equal and as follows: 1,2,4-PhCl3 (75%) > 1,2,3-PhCl3 (15%) > 1,3,5PhCl3 (10%); while for the Pt/Al-PILC ones, it was: 1,2,4-PhCl3 (84%) > 1,2,3-PhCl3 (9%) > 1,3,5-PhCl3 (7%). Thus, in both cases the formation of the 1,2,4-PhCl3 isomer was considerably higher than that of the other two isomers. In the case of PhCl4 , the distribution for the Pd/Al-PILC catalysts was: 1,2,x,5-PhCl4 (65%) > 1,2,3,4-PhCl4 (35%); while for the Pt/Al-PILC ones only the isomers referred to as 1,2,x,5-PhCl4 were formed. Van den Brink et al. [19] studied the formation of polychlorinated benzenes during the catalytic combustion of PhCl over Pt/␥-Al2 O3 . At 400 ◦ C, the amount of PhCl2 isomers followed the order: p-PhCl2 > m-PhCl2 > o-PhCl2 . For PhCl3 , 1,2,4-PhCl3 was the major compound formed, followed by 1,3,5PhCl3 , with a small amount of 1,2,3-PhCl3 . PhClx distributions in the presence of 0.5Pt and 2Pt at various PhCl conversions are depicted in Fig. 4, and in the presence of 0.5Pd and 2Pd in Fig. 5. The percentage of each PhClx congener has been calculated with respect to the total amount of PhClx formed, as the ratio between the selectivity to these and the selectivity to total PhClx . The distribution of PhClx congeners is also affected by the conversion and by the metal loading. For a given metal loading, the formation of higher PhClx isomers becomes easier with increasing conversion, and thus reaction temperature. The formation of PhCl2 noticeably decreases in favor of higher PhClx . As an example, in the case of 2Pt, the percentage of p-PhCl2 decreased from

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024

G Model CATTOD-9181; No. of Pages 13 8

ARTICLE IN PRESS A. Aznárez et al. / Catalysis Today xxx (2014) xxx–xxx

Fig. 5. Congener selectivity of PhClx for various PhCl conversions (20%, 50%, 75%, and 90%) in the presence of: (A) 0.5Pd and (B) 2Pd.

68.2% (conv. = 20%) to 36.7% (conv. = 90%), while the percentage of 1,2,4-PhCl3 increased from 0% (conv. = 20%) to 26.4% (conv. = 90%). In spite of this, the p-PhCl2 isomer was predominant over all PhClx for all PhCl conversions. In the case of 2Pd, the percentage of p-PhCl2 decreased from 38.5% (conv. = 20%) to 24.2% (conv. = 75%), while the percentage of 1,2,4-PhCl3 increased from 12.6% (conv. = 20%) to 27.7% (conv. = 75%). Thus, at a conversion of 75%, the 1,2,4-PhCl3 isomer was the most abundant of the PhClx . This clearly indicates that at higher temperatures further chlorination of the organic compound becomes easier. In the same way, for a given conversion, the formation of higher PhClx congeners becomes easier with increasing metal loading. For example, for a conversion of 75% and in the presence of 0.5Pt only the o-PhCl2 , m-PhCl2 , p-PhCl2 and 1,2,4PhCl3 byproducts were formed, whereas in the presence of 2Pt, 1,2,3-PhCl3 , 1,3,5-PhCl3 and 1,2,x,5-PhCl4 were also formed. Finally, the stability of the Pt/Al-PILC catalysts in the reaction was tested by reusing 0.5Pt in consecutive combustion cycles, and by performing long catalytic runs at a constant combustion temperature. When 0.5Pt was submitted to successive combustion cycles it exhibited good maintenance of the catalytic performance after the three combustion cycles, and practically no beneficial-detrimental effect in activity over the combustion cycles was observed. No pretreatment was applied between the successive cycles. In contrast to what was observed on submitting the catalyst to successive combustion cycles, conversion of PhCl over 0.5Pt decreased by approximately 7% after 30 h reaction time. The carbon balance also decreased during the 30 h test. From these results, it is concluded that the long time-on-stream at a high temperature plays a key role in coke deposition. Increasing time and temperature of the reaction enhances the deposition of coke on the catalyst and, therefore, its deactivation [37]. It is generally accepted that coke affects the activity of porous catalysts in two different ways: poisoning (or coverage) of the active sites or/and pore blockage [36]. By comparing the reutilization of 0.5Pt in consecutive combustion cycles with the long catalytic run at high temperature, it is concluded that the more demanding operating conditions used in the long catalytic run (30 h at 400 ◦ C) promote coke deposition and reduce catalytic performance. The distribution of PhCl2 isomers is to the same as that

previously found for the Pt/Al-PILC catalysts, and does not change either with the combustion cycles or with time-on-stream at high temperature: p-PhCl2 (60%) > m-PhCl2 (26%) > o-PhCl2 (14%). 3.3. Chlorinating agent According to Van den Brink et al. [19], a chlorinating agent and a benzene ring are required to explain the formation of PhClx in the catalytic combustion of PhCl. As no PhClx were observed in the catalytic combustion of C2 Cl4 , they concluded that the aromatic ring in PhClx stemmed from PhCl. Van den Brink et al. [19] proposed that, in the formation of PhClx , PhCl could be chlorinated by the molecular chlorine formed as a result of the reverse Deacon reaction. The Deacon reaction is exothermic (28.5 kJ/mol) and very slow in the uncatalyzed gas phase, even at temperatures as high as 625 ◦ C [38], with a catalyst being required to obtain equilibrium. There are only a few metals able of catalyzing the Deacon reaction. As no data involving noble metals had been reported, Van den Brink et al. [19] considered speculative if Cl2 was formed on Pt/␥-Al2 O3 . If, however, they assumed that the Deacon equilibrium was obtained under their conditions (30–600 ◦ C), they had to assume that, if the PhCl species could be chlorinated by gaseous Cl2 , they could also be oxidized by free molecular oxygen. Sommeling et al. [39] performed competitive experiments between chlorination and oxidation of PhCl in the gas phase, and noted that PhCl2 production was very low when the oxygen concentration was higher than chlorine. Therefore, with an O2 :Cl2 ratio similar to 500:1, PhCl chlorination by gas phase Cl2 , if any, was expected to be very minor. Later, Van den Brink et al. [20] reported Cl2 formation in the combustion of PhCl over platinum-based catalysts. Recently, Finocchio et al. [40] also reported the formation of Cl2 in the combustion of PhCl over Pt/Al2 O3 . Giraudon et al. [12,24] reported the formation of Cl2 during the catalytic combustion of PhCl over Pd/perovskites, Pd/ZrO2 and Pd/TiO2 , and attributed it to the catalytically activated Deacon reaction by Pd. Chlorination of physically adsorbed PhCl was a more probable option. The C–Cl bond in PhCl breaks at room temperature on the surface of metals [41]. In PhCl, the C–Cl bond (390 kJ/mol) is weaker

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024

G Model

ARTICLE IN PRESS

CATTOD-9181; No. of Pages 13

A. Aznárez et al. / Catalysis Today xxx (2014) xxx–xxx

9

Table 3 Binding energies (BEs) and atomic ratios obtained from XPS experiments performed on the fresh and used Pd/Al-PILC and Pt/Al-PILC. Catalyst (fresh)

Al-PILC 0.1Pd 0.5Pd 1Pd 2Pd 0.1Pt 0.5Pt 1Pt 2Pt

BEs (EV)

Atomic ratios

O1s

Si 2p

Mg 2s

Al 2p

Na 1s

Ca 2p

532.0 532.0 532.1 532.0 532.0 531.9 531.8 532.0 532.0

102.9 102.9 102.9 102.8 102.8 102.8 102.6 102.9 102.9

89.5 89.4 89.4 89.4 89.3 89.4 89.2 89.5 89.5

74.9 74.9 74.9 74.8 74.8 74.8 74.7 75.0 74.9

– – – – – – – – –

– 347.1 – – – – – – –

Pd 3d5/2 338.1 338.1 338.1 338.1

Pt 4f7/2 – 336.5 336.5 336.5 336.5 – – – –

Catalyst (used)

BEs (EV) O1s

Si 2p

Mg 2s

Al 2p

Na 1s

Ca 2p

Cl 2p3/2

Cl 2p1/2

Al-PILC U 0.1Pd U 0.5Pd U 2Pd U 0.1Pt U 0.5Pt U 2Pt U

532.1 532.1 532.0 532.1 532.0 532.1 532.0

103.0 103.0 102.9 102.9 102.9 103.0 102.9

89.6 89.6 89.4 89.3 89.5 89.5 89.5

75.0 75.0 74.9 74.9 74.9 75.0 75.0

1071.9 1071.9 1071.8 1072.1 1072.0 1072.0 1071.7

347.5 347.8 – 347.0 347.9 347.0 347.7

198.9 199.1 198.9 199.2 199.1 198.7 198.8

200.9 200.7 200.8 201.2 200.8 200.3 200.6

– – – – –

335.1 335.1 335.1 335.1 73.8 73.5 73.9 73.6

71.7 71.3 71.6 71.5

O/Si

Al/Si

C/Si

Metal/Si

3.26 3.08 3.25 3.15 3.33 3.13 3.16 3.20 3.12

0.62 0.62 0.61 0.61 0.60 0.61 0.58 0.59 0.59

0.52 0.62 0.57 0.62 0.63 0.54 0.50 0.50 0.50

– 0.0021 0.0325 0.0326 0.071 0.0022 0.0039 0.0049 0.0073

Atomic ratios

than the C–H (470 kJ/mol) [42]. Also, the Cl–Pt bond is stronger than the H–Pt [34]. Hence, Van den Brink et al. [19] expected that dissociative adsorption of PhCl on the platinum surface occurred in their study temperature range, forming phenyl groups and chlorine on the platinum surface. Therefore, the chlorination of adsorbed PhCl entities by chlorine attached to the catalyst surface was more likely. Surface chlorine could originate from the adsorption of HCl or remain on the surface after dissociation of the C–Cl bond of chlorinated benzenes. The chlorinating species could be in the form of Pt-chlorides, as suggested by Miyake et al. [34], or Pt-oxychlorides as observed by Lieske et al. [43,44] at temperatures up to 600 ◦ C. When a Pt/Al2 O3 catalyst was treated with HCl and heated under oxygen, a PtIV OHx Cly -surface complex was initially formed. At a higher temperature of 400 ◦ C, it converted into a PtIV Ox Cly -surface complex. Only above 700 ◦ C, did this complex decompose, forming crystalline Pt. No Pt-chlorides were observed at any temperature. The suggestion that platinum oxychlorides were the active species in the formation of polychlorinated benzenes was supported by Van den Brink et al. [19], due to PhClx formation increasing with oxygen pressure: high oxygen concentrations were likely to promote the formation of oxychlorides. Later, many authors [8,9,12,23,24,35] reported that the chlorination of PhCl was carried on PdOx Cly and PtOx Cly related phases. Taralunga et al. [35] reported that the electronic state of platinum was an important parameter for the formation of polychlorinated by-products. When a large quantity of Pt species was present on the catalyst, as in catalysts with a platinum loading higher than 0.6% or in pre-reduced ones, the PhClx concentration increased. The initial presence of Pt species or the large formation of these species during the reaction would be largely responsible for the PhClx formation during PhCl combustion. Reduced platinum species enhanced the formation of PhCl2 by-products through the chlorination of Pt particles. Reduced platinum particles were more easily chlorinated than oxidized platinum ones, leading to platinum chlorides (e.g. PtCl4 ). In the presence of low-loaded catalysts, of Pt ≤ 0.6%, where PtO2 particles predominate, their chlorination leads to the formation of platinum oxychlorides (e.g. PtOCl2 ). In addition to the main combustion reaction, they also reported the existence of minor reactions. Benzene was considered to be an intermediate formed via the platinum species and to be rapidly oxidized into CO2 when the oxygen concentration was sufficient. R.1 and R.2 show oxychlorination/chlorination of PtO2 /Pt by PhCl

Pt 4f7/2

O/Si

Al/Si

C/Si

Metal/Si

Cl/Si

– – – –

3.11 3.28 3.33 3.26 3.20 3.25 3.21

0.62 0.57 0.59 0.63 0.59 0.58 0.59

0.73 .077 0.72 0.79 0.71 0.65 0.65

– 0.0024 0.0197 0.0761 0.0018 0.0031 0.0072

0.006 0.030 0.021 0.021 0.029 0.010 0.020

73.6 73.6 73.6

71.3 71.4 71.3

molecules to give benzene and PtCl2 O/PtCl4 . Water is that released from the main reaction of combustion. PtO2

2 C6 H5 Cl + H2 O −→ 2 C6 H6 + O2 + PtCl2 O

(R.1)

For the reduced platinum particles, (R.1) becomes: Pt 0

4 C6 H5 Cl + 2 H2 O −→ 4 C6 H6 + O2 + PtCl4

(R.2)

(R.3) is the combustion of benzene formed in (R.1) and (R.2). C6 H6 +

PtO2 /Pt 0 15 O2 −→ 6 CO2 + 3 H2 O 2

(R.3)

As mentioned above, PtCl2 O/PtCl4 species are able to catalyze PhClx formation. (R.4) and (R.5) show the chlorination of PhCl into PhCl2 via PtCl2 O/PtCl4 with the regeneration of PtO2 /Pt: 2 C6 H5 Cl

PtCl2 O, O2 , H2 O

−→

2 C6 H4 Cl2 + PtO2

PtCl4 , O2

4 C6 H5 Cl −→ 4 C6 H4 Cl2 + 2 H2 O + Pt 0

(R.4) (R.5)

In our case, selectivity to PhClx increases with increasing metal loading. Thus, PhClx formation is more affected by the metal loading than by the particle size. Thus, the high-loaded catalysts, with a large amount of reduced and less chlorinated species, are more active in PhCl chlorination than the low-loaded ones, where the oxidized and more chlorinated species predominate. This finding is in agreement with that reported by Taralunga et al. [35]. Also, palladium is more active in PhClx formation than platinum. 3.4. Characterization of the catalyts used in combustion of chlorobenzene The carbon balances, the XPS and TG analyses, and the dark color of the catalysts after the reaction confirmed the deposition of coke on their surface. The state of the Pd/Al-PILC and Pt/Al-PILC catalysts after the reaction was studied by means of XRD, XPS, TGA and TPR. The binding energies (BEs) and atomic ratios of the used catalysts were determined from XPS experiments and presented in Table 3. Na 1s, Ca 2p and Cl 2p lines, which were not present in the surface of the fresh catalysts, were detected in the used ones. The Na+ and Ca2+ cations of the support structure, which were not exchanged by alumina polycations during the pillaring process, migrated from the interlayer space to the surface of the sample during the reaction. As confirmed by the metal/Si atomic ratio,

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024

G Model CATTOD-9181; No. of Pages 13 10

ARTICLE IN PRESS A. Aznárez et al. / Catalysis Today xxx (2014) xxx–xxx

Fig. 6. XPS spectra. Retention of inorganic and organic chlorine entities in the presence of: (A) 0.1Pd and (B) 0.1Pt.

there were not significant changes in metallic dispersion in the used catalysts, except for 0.5Pd. By comparing the atomic ratio O/Si of the fresh support and catalysts, it is concluded that the oxygen in the catalysts is mainly due to the support, which means that the ratio O/Si is not an optimal indicator of the oxidation state of the catalytic metals. The atomic ratio O/Si of the used support decreases with respect to its fresh counterpart, while the opposite trend is observed for the catalysts, except for 2Pd. Larrubia et al. [45] reported that chloroaromatics adsorb on ionic oxides such as alumina (Lewis acid sites). A rapid and irreversible reaction (nucleophilic substitution) occurs at the surface, already at room temperature, resulting in a dissociative adsorption via Cl abstraction by a Lewis acid site, and the formation of surface phenolates. Khaleel et al. [4], who studied the oxidative decomposition of PhCl over Fe2 O3 /TiO2 , reported similar results and indicated the involvement of the lattice oxygen in the combustion process, as some oxidation into CO2 had been observed in reactions in the absence of oxygen. In the case of Al-PILC, it is strongly accepted that Lewis acidity is due mainly to the Al2 O3 pillars. Hence, the involvement of the alumina oxygen in the PhCl combustion process would explain the decrease in the O/Si atomic ratio. Instead, the increase of the atomic ratio O/Si in the used catalysts could be related to the oxygenated coke composition. Tsou et al. [46], who studied the combustion of o-xylene over Pt/HBEA catalysts, found that coke was formed by two main families: the aromatic hydrocarbons of formula Cn H2n−14 , and the aromatic oxygenated compounds with the formula Cn H2n−16 O, in the proportion 35/65. These coke compositions had been already detected during the combustion of benzene [47] over doped zeolites. The atomic ratio C/Si increased in all the used samples, support and catalysts, which means that carbonaceous residues were formed with all the samples. The atomic ratio Cl/Si is similar for both series of used catalysts, and considerably higher than that of the used support. Thus, the retention of chlorine in the used catalysts is mainly due to the metallic phase. The binding energy values of Cl 2p3/2 in the range 198.7–199.2 eV are consistent with inorganic chloride entities [12,48]. By comparison with those in the literature [49], the following species could be proposed: NaCl (198.7 eV), CaCl2 (199.0 eV), FeCl3 (199.0 eV), K2 PdCl6 (198.7 eV), K2 PdCl4 (198.8 eV), PdCl2 (198.9 eV), K2 PtCl6 (198.7 eV) and K2 PtCl4 (198.9 eV). Hence, it seems that the Na+ and Ca2+ cations in the support structure could be related to the retention of inorganic chlorine in the catalysts. Several authors have proposed that the chlorinating agent of PhCl could be a metal-chloride or metal-oxychloride species, see Section 4.1. The binding energies of Cl 2p1/2 in the range 200.0–201.0 eV correspond to chlorine in organic structures, PhCl, PhClx and probably coke deposits [49]. The distribution of

inorganic and organic chlorine in the surface of all the used samples, support and catalysts, was about the same and as follows: organic chlorine (33%) and inorganic chlorine (67%). Thus, the percentage of inorganic chlorine in the surface of the used samples is higher than that of organic chlorine. If the Cl 2p3/2 /metal atomic ratio is calculated from the Cl 2p3/2 /Si and metal/Si atomic ratios, the following values are obtained for the Pd/Al-PILC catalysts: 0.1Pd (8.3) > 0.5Pd (0.7) > 2Pd (0.2), and for the Pt/Al-PILC catalysts, it is: 0.1Pt (10.5) > 0.5Pt (2.2) > 2Pt (1.8). The atomic ratio Cl 2p3/2 /metal is particularly important in the case of the 0.1 wt.%-loaded catalysts, meaning that the more dispersed metallic phase favors the formation of inorganic chlorine entities in the catalyst. Also, platinum favors the formation of inorganic chlorine entities more than palladium. Thus, the metallic phase is chlorinated to a higher extent in the low-loaded catalysts than in the high-loaded ones, and in the Pt/Al-PILC catalyts more than in their Pd/Al-PILC counterparts. Retention of inorganic and organic chlorine entities in the presence of the 0.1 wt.%-loaded catalysts is presented in Fig. 6. The study of the palladium phases after the reaction showed that, in addition to the three phases of palladium present in the fresh catalysts and detected at 335.1, 336.5 and 338.1 eV, a new palladium phase at 335.7 eV was formed. Carbonaceous species induced a shift of the surface core levels to higher binding energy of 335.7 eV compared to the Pd signal (335.1 eV). Therefore this component was assigned to a Pd state modified by C, as for example palladium surface covered by carbon containing surface species or a Pd-C phase. Thus, the formation of a carbon phase in Pd alters the electronic structure of the palladium surface [50]. The palladium surface composition of the used Pd/Al-PILC catalysts is shown in Table 4. The XRD patterns of 2Pd and 2Pt, for the fresh and used catalysts in PhCl combustion, are shown in Fig. 7. The fresh and used catalysts have the same diffraction peaks, corresponding to PdO and Pt phases, respectively. The diffraction peaks of the support and metallic phases in the used catalysts were slightly less intense than those of the fresh catalysts, which could be related to coke formation in the solids. No graphite-like crystalline structure was identified for any catalyst. The TPR patterns of the used and fresh catalysts in the combustion of PhCl are presented in Fig. 8. The results confirm the differences between the fresh and used catalysts. A considerably smaller negative peak at 70 ◦ C was observed for the used palladium catalyst than for the fresh one. It could be that the formation of a Pd state modified by carbon would not allow hydrogen to adsorbe on the surface of Pd. Thus, compared to the fresh catalysts, the lower consumption of H2 in the case of the used ones could be attributed to the deposition of coke on their surfaces.

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024

G Model

ARTICLE IN PRESS

CATTOD-9181; No. of Pages 13

A. Aznárez et al. / Catalysis Today xxx (2014) xxx–xxx

11

Table 4 Palladium surface composition (%) of the fresh and used Pd/Al-PILC obtained from XPS experiments. Catalyst (fresh)

BEs (EV)—Pd 3d5/2

0.1Pd 0.5Pd 1Pd 2Pd

338.1 (Pdn+ ) (n ≥ 2)

336.5 (Bulk PdO)

335.1 (Pd)

35.0 18.8 8.7 5.7

60.0 73.7 71.1 69.9

5 7.5 20.2 24.4

Catalyst (used)

5 8 20 24

336.5 (Bulk PdO)

335.1 (Pd) + 335.7 (Pd state modified by C)

30.0 13.2 3.7

40.0 58.3 54.7

30.0 28.5 41.6

2Pd_Used in chlorobenzene combustion

PdO (103) PdO (200)

55

60

TCD signal (a.u.)

PdO (112)

2Pd_Fresh

2Pd_Fresh

Al-PILC_Fresh

40

45

50

formation with the platinum [52,53]. In spite of this, they found that the addition of alkanes almost eliminated the formation of PhClx , with PhCl destruction being firmly accelerated. They concluded that alkanes, which are known to react rapidly with elementary chlorine, caused the removal of Cl from the Pt surface. Without alkanes, Cl was removed only at high temperatures, thus blocking the surface for PhCl and oxygen adsorption. Scirè et al. [23] suggested that the structure of H-ZSM5 zeolite, a medium pore size zeolite possessing two types of channels, with openings 0.53 mm × 0.56 nm and 0.51 mm × 0.55 nm, induced a product shape selectivity effect, which hindered the chlorination of PhCl into PhClx and affected PhClx distribution, favoring the formation of isomers with lower steric hindrance.

PdO (103) PdO (200)

PdO (112)

2Pd_Used in chlorobenzene combustion

PdO (110)

PdO (002) PdO (101)

PdO (002) PdO (101)

PdO (110)

To carry out a safe and effective destruction of Cl-VOCs, PhClx formation has to be avoided. Several solutions have been proposed for PhCl combustion performed over noble metals: Van den Brink et al. [51] and de Jong et al. [9] proposed co-feeding a non-chlorinated VOC and PhCl, as contaminants in industrial fuel gases comprise chlorinated compounds and hydrocarbons (alkynes, alkenes, alkanes and oxygenates). PhCl adsorbed better on the catalyst than the alkanes, because of the ␲-complex

Intensity (a.u.)

95 92 80 76

338.1 (Pdn+ ) (n ≥ 2)

3.5. Posible solutions to minimize PhClx formation in combustion of chlorobenzene

35

Reduced palladium

BEs (EV)—Pd 3d5/2

0.1Pd U 0.5Pd U 2Pd U

30

Oxidized palladium

65

70

2 Theta (º)

0

100

200

300

400

500

600

700

800

600

700

800

Ptº (220)

2Pt_Used in chlorobenzene combustion 2Pt_Fresh

TCD signal (a.u.)

Ptº (200) Ptº (200)

2Pt_Used in chlorobenzene combustion

Ptº (220)

Ptº (111) Ptº (111)

Intensity (a.u.)

Temperature (ºC)

2Pt_Fresh

Al-PILC_Fresh

30

35

40

45

50

55

60

65

70

2 Theta (º) Fig. 7. XRD patterns of the support, 2Pd and 2Pt: fresh and used in PhCl combustion.

0

100

200

300

400

500

Temperature (ºC) Fig. 8. TPR patterns of 2Pd and 2Pd: fresh and used in PhCl combustion.

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024

G Model CATTOD-9181; No. of Pages 13 12

ARTICLE IN PRESS A. Aznárez et al. / Catalysis Today xxx (2014) xxx–xxx

Van den Brink et al. [20], who studied the addition of low amounts of water (1–1.75%) to the feed stream, found that the output of PhClx was greatly reduced on all platinum catalysts. The total organic chlorine levels were 3–5 times lower, while the formation of higher PhClx congeners was especially suppressed. Conversion was also improved in the presence of water, and T50 was reduced with 10–30 ◦ C. However, in contrast to platinum catalysts, water retarded the conversion on palladium catalysts. While water removed Cl from the platinum surface [54,55], Pd–OH groups, which decreased the number of active PdO sites for C–H activation, were formed on palladium catalysts. Van den Brink et al. [19,20] and De Jong et al. [9], who studied the influence of oxygen concentration in PhCl combustion over platinum catalysts, reported that at 15% oxygen, the production of PhClx was the highest, but that it decreased rapidly with decreasing oxygen pressure. Also, under oxygen-rich conditions (above 4% O2 ) the catalysts are less active. They thought that the higher O2 concentrations caused an increased coverage by oxy- and/or oxychloride species, diminishing the number of sites available for the reaction.

4. Summary and conclusions During oxidation of PhCl over the palladium and platinum catalysts, the formation of carbon dioxide along with very small quantities of carbon monoxide, PhClx and coke were observed. It has been shown that some PhClx derivatives (x = 3, 4) had strong interaction with the catalysts, as they were formed at 300 and 400 ◦ C and also desorbed at a higher temperature and during the stabilization stage after the catalytic reaction. Selectivity to PhClx was considerably higher over the Pd/Al-PILC catalysts than over their Pt/Al-PILC counterparts or the support. Palladium is known to be a more active chlorination catalyst than platinum. The reduced metal species, chlorinated to a lesser extent than the oxidized ones, and present in larger quantities in catalyst with a high metal loading, would be responsible for a large part of the PhClx formation. Thus, PhClx selectivity increases with increasing metal loading; and also with increasing reaction temperature. Through the homogeneous reaction, detected by means of the blank test at 400 ◦ C, the chlorination of PhCl occurred through a classical aromatic electrophilic substitution (p-PhCl2 > o-PhCl2 ). Distributions of PhCl2 for the support (m-PhCl2 > p-PhCl2 > o-PhCl2 ) and catalysts (p-PhCl2 > m-PhCl2 > o-PhCl2 ), and their comparison with that obtained for the blank test, ruled out an aromatic electrophilic substitution mechanism and thus, the homogeneous reaction as the main mechanism of PhCl chlorination. In the case of the support, it is conceivable that PhCl chlorination occurs through a radical mechanism over the support. In the case of the catalysts, the selectivity of PhCl2 isomers could be explained due to secondary reactions on the support, isomerization of the products formed on the metallic phase, or to coke deposition. In fact, in the presence of the Pt/Al-PILC catalysts, the formation of coke and the selectivity to p-PhCl2 were more important than in the presence of their Pd/Al-PILC counterparts. PhClx distribution is also affected by the metallic phase, the metal loading and the reaction temperature. PhClx distribution is different for the support, Pd/Al-PILC and Pt/Al-PILC. Selectivity to higher PhClx congeners increases with increasing metal loading and temperature. The study of the palladium phases after the reaction showed the formation of a new palladium phase at 335.7 eV, assigned to a Pd state modified by C. Retention of chlorine in the used catalysts was mainly due to the metallic phase. The metallic phase was chlorinated to a higher extent in the Pt/Al-PILC catalysts than in their Pd/Al-PILC counterparts. Also, the metallic phase is chlorinated to a higher extent

in the low-loaded catalysts than in the high-loaded ones. Organic chlorine entities, probably associated to coke deposits, were also detected in the presence of the support and catalysts. Pd/Al-PILC catalysts, which are less active in PhCl combustion, more selective to CO, and more active in (oxy)chlorination, than their Pt/Al-PILC counterparts, are not suited for PhCl combustion. In comparison with the rest of the catalysts studied, 0.1Pt provides a high conversion of PhCl, and a low formation of CO and PhClx . By comparing 0.1Pt with the rest of the Pt/Al-PILC catalysts, it is concluded that while there are no large differences in PhCl conversion, PhClx formation with 0.1Pt is considerably less. Therefore, factors affecting the deposition of coke should be adjusted. We propose further investigation of 0.1Pt in order to develop a catalyst of interest for PhCl combustion.

Acknowledgements This work was funded by the Spanish Ministry of Science and Innovation (MICINN) and the European Regional Development Fund (FEDER) (project MAT2007-66439-C02). A. Aznarez acknowledges financial support from the Public University of Navarra through a Ph.D. fellowship.

References [1] E. Noordally, J.R. Richmond, S.F. Tahir, Destruction of volatile organic compounds by catalytic oxidation, Catal. Today 17 (1993) 359–366. [2] G.J. Hutchings, C.S. Heneghan, I.D. Hudson, S.H. Taylor, Uranium oxide based catalysts for the destruction of volatile chloro-organic compounds, Nature 384 (1996) 341–343. [3] M.L.H. Green, R.M. Lago, S.C. Tsang, Total oxidation of chlorinated hydrocarbons by copper and chlorine based catalysts, J. Chem. Soc.-Chem. Commun. (1995) 365–366. [4] A. Khaleel, A. Al-Nayli, Supported and mixed oxide catalysts based on iron and titanium for the oxidative decomposition of chlorobenzene, Appl. Catal. Environ. 80 (2008) 176–184. [5] E. Finocchio, G. Busca, M. Notaro, A review of catalytic processes for the destruction of PCDD and PCDF from waste gases, Appl. Catal. B-Environ. 62 (2006) 12–20. [6] M.H. Zheng, P.Y. Liu, Y.C. Bao, X.B. Xu, Aspects of formation and degradation of polychlorinated dibenzo-p-dioxins and dibenzofurans, Chin. Sci. Bull. 44 (1999) 1249–1257. [7] S.F. Zuo, R.X. Zhou, Al-pillared clays supported rare earths and palladium catalysts for deep oxidation of low concentration of benzene, Appl. Surf. Sci. 253 (2006) 2508–2514. [8] L.C.A. Oliveira, R.M. Lago, J.D. Fabris, K. Sapag, Catalytic oxidation of aromatic VOCs with Cr or Pd-impregnated Al-pillared bentonite: byproduct formation and deactivation studies, Appl. Clay Sci. 39 (2008) 218–222. [9] V. de Jong, M.K. Cieplik, W.A. Reints, F. Fernandez-Reino, R. Louw, A mechanistic study on the catalytic combustion of benzene and chlorobenzene, J. Catal. 211 (2002) 355–365. [10] E.C. Moretti, Reduce VOC and HAP emissions, Chem. Eng. Prog. 98 (2002) 30–40. [11] P. Papaefthimiou, T. Ioannides, X.E. Verykios, Combustion of non-halogenated volatile organic compounds over group VIII metal catalysts, Appl. Catal. BEnviron. 13 (1997) 175–184. [12] J.M. Giraudon, T.B. Nguyen, G. Leclercq, S. Siffert, J.F. Lamonier, A. Aboukais, A. Vantomme, B.L. Su, Chlorobenzene total oxidation over palladium supported on ZrO2, TiO2 nanostructured supports, Catal. Today 137 (2008) 379–384. [13] J.D. Kilgroe, Control of dioxin, furan, and mercury emissions from municipal waste combustors, J. Hazard. Mater. 47 (1996) 163–194. [14] P. Andersson, C. Rappe, O. Maaskant, J. Unsworth, S. Marklund, Low temperature catalytic destruction of PCDD/F in flue gas from water incineration, Organohalogen Comp. 36 (1998) 109–112. [15] Y. Liu, Z.B. Wei, Z.C. Feng, M.F. Luo, P.L. Ying, C. Li, Oxidative destruction of chlorobenzene and o-dichlorobenzene on a highly active catalyst: MnOx /TiO2 Al2 O3 , J. Catal. 202 (2001) 200–204. [16] K. Everaert, J. Baeyens, Catalytic combustion of volatile organic compounds, J. Hazard. Mater. 109 (2004) 113–139. [17] C. Lahousse, A. Bernier, P. Grange, B. Delmon, P. Papaefthimiou, T. Ioannides, X. Verykios, Evaluation of gamma-MnO2 as a VOC removal catalyst: comparison with a noble metal catalyst, J. Catal. 178 (1998) 214–225. [18] J.J. Spivey, Complete catalytic oxidation of volatile organics, Indus. Eng. Chem. Res. 26 (1987) 2165–2180. [19] R.W. van den Brink, R. Louw, P. Mulder, Formation of polychlorinated benzenes during the catalytic combustion of chlorobenzene using a Pt/gamma-Al2 O3 catalyst, Appl. Catal. B-Environ. 16 (1998) 219–226.

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024

G Model CATTOD-9181; No. of Pages 13

ARTICLE IN PRESS A. Aznárez et al. / Catalysis Today xxx (2014) xxx–xxx

[20] R.W. van den Brink, M. Krzan, M.M.R. Feijen-Jeurissen, R. Louw, P. Mulder, The role of the support and dispersion in the catalytic combustion of chlorobenzene on noble metal based catalysts, Appl. Catal. B-Environ. 24 (2000) 255–264. [21] R. Lopez-Fonseca, J.I. Gutierrez-Ortiz, M.A. Gutierrez-Ortiz, J.R. GonzalezVelasco, Dealuminated Y zeolites for destruction of chlorinated volatile organic compounds, J. Catal. 209 (2002) 145–150. [22] L. Becker, H. Forster, Oxidative decomposition of chlorobenzene catalyzed by palladium-containing zeolite Y, J. Catal. 170 (1997) 200–203. [23] S. Scire, S. Minico, C. Crisafulli, Pt catalysts supported on H-type zeolites for the catalytic combustion of chlorobenzene, Appl. Catal. B-Environ. 45 (2003) 117–125. [24] J.M. Giraudon, A. Elhachimi, G. Leclercq, Catalytic oxidation of chlorobenzene over Pd/perovskites, Appl. Catal. B-Environ. 84 (2008) 251–261. [25] F. Bertinchamps, C. Gregoire, E.M. Gaigneaux, Systematic investigation of supported transition metal oxide based formulations for the catalytic oxidative elimination of (chloro)-aromatics - Part I. Identification of the optimal main active phases and supports, Appl. Catal. B-Environ. 66 (2006) 1–9. [26] F. Bertinchamps, C. Gregoire, E.M. Gaigneaux, Systematic investigation of supported transition metal oxide based formulations for the catalytic oxidative elimination of (chloro)-aromatics - Part II. Influence of the nature and addition protocol of secondary phases to VOx /TiO2 , Appl. Catal. B-Environ. 66 (2006) 10–22. [27] D.P. Debecker, F. Bertinchamps, N. Blangenois, P. Eloy, E.M. Gaigneaux, On the impact of the choice of model VOC in the evaluation of V-based catalysts for the total oxidation of dioxins: furan vs. chlorobenzene, Appl. Catal. B-Environ. 74 (2007) 223–232. [28] S.C. Petrosius, R.S. Drago, V. Young, G.C. Grunewald, Low temperature decomposition of some halogenated hydrocarbons using metal oxide porous carbon catalysts, J. Am. Chem. Soc. 115 (1993) 6131–6137. [29] B. Mendyka, A. Musialikpiotrowska, K. Syczewska, Effect of chlorine compounds on the deactivation of platinum catalysts, Catal. Today 11 (1992) 597–610. [30] H. Muller, K. Deller, B. Despeyroux, E. Peldszus, P. Kammerhofer, W. Kuhn, R. Spielmannleitner, M. Stoger, Catalytic purification of waste gases containing chlorinated hydrocarbons with precious-metal catalysts, Catal. Today 17 (1993) 383–390. [31] D.M. Papenmeier, J.A. Rossin, Catalytic oxidation of dichloromethane, chloroform, and their binary mixtures over a platinum-alumina catalyst, Ind. Eng. Chem. Res. 33 (1994) 3094–3103. [32] M.A. Vicente, R. Trujillano, K.J. Ciuffi, E.J. Nassar, S.A. Korili, A. Gil, Pillared clay catalysts in green oxidation reactions, in: A. Gil, S.A. Korili, R. Trujillano, M.A. Vicente (Eds.), Pillared Clays and Related Catalysts, Springer, New York, 2010, pp. 301–318. [33] A. Aznárez, A. Gil, S.A. Korili, Palladium and platinum supported on alumina pillared clays: I. Preparation and characterization of the catalysts, 2014 (under review). [34] T. Miyake, M. Hanaya, Screening of metal chloride catalysts for oxychlorination of propene, Appl. Catal. A-Gen. 121 (1995) L13–L17. [35] M. Taralunga, J. Mijoin, P. Magnoux, Catalytic destruction of chlorinated POPs—catalytic oxidation of chlorobenzene over PtHFAU catalysts, Appl. Catal. B-Environ. 60 (2005) 163–171. [36] E.G. Derouane, Factors affecting the deactivation of zeolites by coking, in: B. Imelik, C. Naccache, G. Coudurier, Y.B. Taarit, J.C. Vedrine (Eds.), Studies in Surface Science and Catalysis, vol. 20, Elsevier, 1985, pp. 221–240.

13

[37] P. Marecot, A. Akhachane, C. Micheaud, J. Barbier, Deactivation by coking of supported palladium catalysts. Effect of time and temperature, Appl. Catal. AGen. 169 (1998) 189–196. [38] F. Wattimena, W.M.H. Sachtler, T. Seiyama, K. Tanabe, Catalyst research for the shell chlorine process, Studies in Surface Science and Catalysis, vol. 7, Part B, Elsevier, 1981, pp. 816–827. [39] P.M. Sommeling, P. Mulder, R. Louw, D.V. Avila, J. Lusztyk, K.U. Ingold, Rate of reaction of phenyl radicals with oxygen in solution and in the gas phase, J. Phys. Chem. 97 (1993) 8361–8364. [40] E. Finocchio, G. Ramis, G. Busca, A study on catalytic combustion of chlorobenzenes, Catal. Today 169 (2011) 3–9. [41] J.F.M. Aarts, K.G. Phelan, Dehalogenation of fluorobenzene and chlorobenzene on Pd (111) studied by electron-energy loss spectroscopy, Surf. Sci. 222 (1989) L853–L860. [42] J. Cioslowski, G. Liu, D. Moncrieff, Energetics of the homolytic C–H and C–Cl bond cleavages in polychlorobenzenes: the role of electronic and steric effects, J. Phys. Chem. A 101 (1997) 957–960. [43] H. Lieske, G. Lietz, H. Spindler, J. Volter, Reactions of platinum in oxygenand hydrogen-treated Pt/gamma-Al2 O3 catalysts: I. Temperature-programmed reduction, adsorption, and redispersion of platinum, J. Catal. 81 (1983) 8–16. [44] G. Lietz, H. Lieske, H. Spindler, W. Hanke, J. Volter, Reactions of platinum in oxygen-treated and hydrogen-treated Pt-Gamma-Al2 O3 catalysts. 2. Ultraviolet visible studies, sintering of platinum, and soluble platinum, J. Catal. 81 (1983) 17–25. [45] M.A. Larrubia, G. Busca, An FT-IR study of the conversion of 2-chloropropane, o-dichlorobenzene and dibenzofuran on V2 O5 -MoO3 -TiO2 . SCR-DeNO(x) catalysts, Appl. Catal. B-Environ. 39 (2002) 343–352. [46] J. Tsou, L. Pinard, P. Magnoux, J.L. Figueiredo, M. Guisnet, Catalytic oxidation of volatile organic compounds (VOCs)—oxidation of o-xylene over Pt/HBEA catalysts, Appl. Catal. B-Environ. 46 (2003) 371–379. [47] L. Becker, H. Forster, Investigations of coke deposits formed during deep oxidation of benzene over Pd and Cu exchanged Y-type zeolites, Appl. Catal. A-Gen. 153 (1997) 31–41. [48] S. Ordonez, H. Sastre, F.V. Diez, Thermogravimetric determination of coke deposits on alumina-supported noble metal catalysts used as hydrodechlorination catalysts, Thermochim. Acta 379 (2001) 25–34. [49] R. Benoit, LaSurfacecom. Available from: www.lasurface.com [50] E.M. Vass, M. Hävecker, S. Zafeiratos, D. Teschner, A. Knop-Gericke, R. Schlögl, The role of carbon species in herogeneous catalytic processes: an in-situ soft X-ray photoelectron spectroscopy study, J. Phys.: Condens. Matter 20 (2008) 1–13. [51] R.W. van den Brink, P. Mulder, R. Louw, Catalytic combustion of chlorobenzene on Pt/gamma-Al2 O3 in the presence of aliphatic hydrocarbons, Catal. Today 54 (1999) 101–106. [52] A.A. Barresi, G. Baldi, Deep catalytic-oxidation of aromatic hydrocarbon mixtures—reciprocal inhibition effects and kinetics, Ind. Eng. Chem. Res. 33 (1994) 2964–2974. [53] M.X. Yang, M. Xi, H.J. Yuan, B.E. Bent, P. Stevens, J.M. White, NEXAFS studies of halobenzenes and phenyl groups on Cu (111), Surf. Sci. 341 (1995) 9–18. [54] P. Marecot, A. Fakche, B. Kellali, G. Mabilon, M. Prigent, J. Barbier, Propane and propene oxidation over platinum and palladium on alumina—effects of chloride and water, Appl. Catal. B-Environ. 3 (1994) 283–294. [55] N.W. Cant, D.E. Angove, M.J. Patterson, The effects of residual chlorine on the behaviour of platinum group metals for oxidation of different hydrocarbons, Catal. Today 44 (1998) 93–99.

Please cite this article in press as: A. Aznárez, et al., Catalysts based on pillared clays for the oxidation of chlorobenzene, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.024