Plasma-catalytic hybrid reactor: Application to methane removal

G Model

ARTICLE IN PRESS

CATTOD-9507; No. of Pages 7

Catalysis Today xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

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

Plasma-catalytic hybrid reactor: Application to methane removal Thien Pham Huu a , Sonia Gil b , Patrick Da Costa c , Anne Giroir-Fendler b , Ahmed Khacef a,∗ a

GREMI – UMR 7344, CNRS-Université d’Orléans, 14 rue d’Issoudun, BP 6744, 45067 Orléans, France Université Lyon1, CNRS, UMR 5256, IRCELYON, 2 avenue Albert Einstein, 69626 Villeurbanne, France c Sorbonne Universités, UPMC Paris 6, Institut Jean Le Rond d’Alembert, CNRS UMR 7190, 2 place de la gare de ceinture, 78210 Saint Cyr l’école, France b

a r t i c l e

i n f o

Article history: Received 29 October 2014 Received in revised form 24 February 2015 Accepted 3 March 2015 Available online xxx Keywords: Non-thermal plasma DBD Catalysis Oxidation Methane Pd/Al2 O3

a b s t r a c t Methane oxidation was investigated in a pulsed dielectric barrier discharge at atmospheric pressure coupled with Pd/Al2 O3 catalysts. Comparison between plasma, catalytic, and plasma-catalysis (both inplasma and post-plasma catalysis) systems were performed in the temperature range of 25–500 ◦ C and specific input energy up to 148 J/L. For plasma-alone experiments, CH4 conversion reached a maximum of 67% and the main products obtained were CO, CO2 , O3 , and HNO3 . The plasma catalytic treatment leads to an increase of the CH4 oxidation even at low temperature. It is evidenced that, compared to plasma alone, both Al2 O3 and Pd/Al2 O3 catalysts coupled plasma discharge increase the CH4 conversion. Moreover, for all plasma-catalytic systems, the CH4 conversion plots were shifted toward lower temperature as the specific input energy increases. Although the difference is low, the in-plasma catalysis configuration seems to be more efficient compared to post-plasma catalysis. In all cases, CH4 oxidation in presence of Pd/Al2 O3 catalyst becomes more selective in CO2 formation than the reaction in plasma alone. © 2015 Elsevier B.V. All rights reserved.

1. Introduction The emission of volatile organic compounds (VOCs) and nitrogen oxides (NOx ) by various industrial processes and human activities is a serious source of air pollution, therefore, a problem for human health and the environment in general. Among VOCs, methane (CH4 ) as a potent greenhouse gas, is the most abundant reactive trace gas in the atmosphere and arises from both natural and anthropogenic sources. Although the CH4 lifetime in the atmosphere is much shorter than that of CO2 , CH4 is more efficient at trapping radiation than CO2 . The comparative impact of CH4 on climate change is over 20 times higher than CO2 at equivalent emission rates. For these reasons, and according the current and future worldwide regulations, CH4 emissions must be necessarily reduced. Catalytic combustion of CH4 has been widely explored in the last few decades for both pollutant abatement and power generation. This environmentally friendly technology offers the possibility to produce heat and energy at much lower temperatures than the conventional thermal combustion. Currently, catalysts such as supported noble and transition metals are being extensively used [1–3]. The most active catalysts for oxidation of saturated hydrocarbons, including CH4 , are platinum and palladium, the latter is

∗ Corresponding author. Tel.: +33 238494875. E-mail address: [email protected] (A. Khacef).

considered as superior due to its high activity, thermal stability, low cost compared to the Pt-based catalysts [4]. It is often suggested that the active form of Pd in CH4 oxidation at low temperature is crystalline PdO, but it is reported that at high temperature metallic Pd may also be active [5]. While presenting interesting performances for the abatement of different type of molecules (aromatic, oxygenated or paraffinic of medium molecular weight), the conventional VOC treatment technologies are not appropriate and not cost-effective in the case of moderate gas flow rates containing low amounts of pollutant [6]. On the other hand, the severe emission tolerances are often difficult to handle using these conventional technologies [7]. Thus, it is necessary to develop new methods for CH4 conversion. As an alternative to the catalytic oxidation, non-thermal plasma (NTP) such as dielectric barrier discharges (DBD) and corona discharges have been extensively investigated for VOCs abatements [8–11]. In NTP, background gaseous species are excited and/or dissociated directly by electronic impact while the gas temperature remains relatively low, and thus the product distributions far from the chemical equilibrium may be obtained. In that case, the deposited energy controls the production of active species, which lead to the chemical conversion of VOCs. Plasmas processes offer a unique way to induce gas phase reactions by electron collisions, but they produce large amounts of by-products and are often less selective than the catalytic processes. As emphasized by different groups, a more effective use of NTP is possible by exploiting its inherent synergistic potential

http://dx.doi.org/10.1016/j.cattod.2015.03.001 0920-5861/© 2015 Elsevier B.V. All rights reserved.

Please cite this article in press as: T. Pham Huu, et al., Plasma-catalytic hybrid reactor: Application to methane removal, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.001

G Model CATTOD-9507; No. of Pages 7

ARTICLE IN PRESS T. Pham Huu et al. / Catalysis Today xxx (2015) xxx–xxx

2

through coupling with a heterogeneous catalyst. NTP has been extensively used for several chemical and catalytic processes and only a few studies were related to the total oxidation of CH4 . Examples are the treatment of diesel exhaust gases [12–16], abatement of volatile organic compounds [17–21], hydrocarbons reforming processes [22,23], partial oxidation [24,25], and total CH4 oxidation in natural gas combined heat power [26–28]. In this paper, NTP-DBD reactor coupled with Pd/Al2 O3 catalysts are used to investigate the oxidation of CH4 at atmospheric pressure. In order to understand the synergistic effect of NTP–catalyst coupling, CH4 oxidation has been performed using (i) thermal catalysis; (ii) NTP without catalyst; and (iii) NTP–catalyst coupling in both in-plasma catalysis (IPC) and post-plasma catalysis (PPC) configurations. Results will be presented as a function of the gas temperature and the plasma specific input energy (SIE). 2. Experimental

discharge zone. For the two arrangements, the catalyst in the form of beads (1.0 or 1.8 mm diameter) was placed in between quartz wool. The plasma-catalyst reactor temperature was increased from 22 to 500 ◦ C, at the rate of 5 ◦ C min−1 , using tubular furnace. The reactor temperature was measured by a K-type thermocouple placed on the reactor surface and the temperature uncertainty is ±3 ◦ C. The experiments are steady state experiments; we have waited at least 30 min for stabilization. The DBD reactor is driven by a homemade pulsed HV generator (20 kV, 200 Hz). Electrical characterization of the plasma was performed by current and voltage measurements using a Tektronix P6015A HV probe and a Pearson 4001 current probe, respectively. The output signals were transmitted to a Tektronix DPO 3054 oscilloscope and the energy consumption was evaluated through the specific input energy (SIE) calculated from Eq. (1) where Ep , f, and Q denote the pulse energy, pulse frequency, and gas flow rate, respectively.

2.1. Plasma-catalytic system SIE The experimental set-up comprises the mass flow controllers, a high voltage power supply, a plasma-catalytic reactor installed in a tubular furnace, and a Fourier Transform Infrared Spectrometer. The system was described in detail previously [20] and is briefly described here for clarity. The plasma reactor is a dielectric-barrier discharge (DBD) reactor in a cylindrical configuration. It consists of 0.9 mm-diameter tungsten wire used as main electrode. The tungsten wire was placed in the center of, 300 mm long Pyrex® , tube using two ceramic rings. The outer surface of the tube was wrapped with copper grid and used as a ground electrode. The discharge length was fixed as 100 mm and the corresponding discharge volume is 18 cm3 . The plasma reactor gives a possibility to combine catalyst in two ways: the catalyst in the discharge zone (in-plasma catalysis, IPC) or the catalyst downstream the discharge zone (post-plasma catalysis, PPC) as shown in Fig. 1. In IPC arrangement, the catalyst is directly in contact with the discharge and the active species are generally close vicinity of the catalyst surface (e.g., excited-state atoms and molecules, reactive radicals, photons, and electrons). In addition, in IPC configuration plasma modifies the catalyst surface. In contrast, in the PPC arrangement, the catalyst is exposed only to long-lived species that exit from the plasma zone (ozone). For IPC configuration the catalyst, 2.2 or 3.1 g, was placed in the middle of the discharge zone. It leads to the catalyst volume of 2.8 and 4.1 ml. In case of PPC, the catalyst was placed just after the plasma

J L

=

EP (J) · f (Hz) Q (L/s)

(1)

End-products detection and quantification were carried out using a FTIR (Nicolet 6700, Thermo-Scientific) equipped with a MCT detector and 10 m gas path cell. Steady-state activity, CH4 conversion, was measured in the presence and in the absence of catalyst at temperatures ranged from 22 to 500 ◦ C. The uncertainty was determined by repeating each experiment four times. After reproducibility, we can conclude that the experimental relative uncertainty error is less than 5%. 2.2. Catalysts preparation Palladium supported catalysts were prepared by impregnating alumina beads (diameters: 1.0 and 1.8 mm) (Sasol Germany GmbH), using a palladium tetramine nitrate solution (Pd(NH3 )4 (NO3 )2 ; STREM Chemicals Inc.) as the metal precursor. The concentration of the precursor was calculated from the desired metal loading of the final catalyst (0.5 and 1 wt.%). The suspension was stirred for 3 h at 50 ◦ C and at atmospheric pressure. After a complete evaporation of water under reduced pressure, the impregnated samples were dried at 120 ◦ C for 24 h. Finally, the dried samples were calcined at 500 ◦ C for 4 h at the heating rate of 3 ◦ C min−1 , under air flow to obtain the final form of supported catalysts. 2.3. Catalysts characterization 2.3.1. Chemical analysis The Pd metal loading was determined by Inductively Coupled Plasma–Optical Emission Spectrometry (ICP-OES) using an ACTIVA spectrometer from Horiba Jobin Yvon. Samples (ca. 0.5 g) were treated in H2 SO4 :HNO3 :HCl solutions. 2.3.2. Specific surface area and porosity Specific surface area/porosity measurements were conducted using a Micromeritics ASAP 2010 apparatus with N2 as the sorbate at −196 ◦ C. Prior to surface area measurement, all the samples were outgassed at 300 ◦ C and under 5 × 10−3 Torr pressure for 3 h. The total specific surface area was determined by the multipoint BET method, and the mesoporosity was evaluated using the Barrett–Joyner–Halenda (BJH) method.

Fig. 1. Schematic overview of the plasma-catalytic reactor: (a) in-plasma catalysis (IPC) and (b) post-plasma catalysis (PPC).

2.3.3. X-ray diffraction (XRD) XRD analyses were conducted with a Bruker D5005 powder diffractometer scanning using Cu-K␣ radiation. Samples were scanned (0.02◦ /step) in the 2 range of 4–80◦ (scan time = 2 s/step). Diffraction patterns were assigned using Joint Committee on

Please cite this article in press as: T. Pham Huu, et al., Plasma-catalytic hybrid reactor: Application to methane removal, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.001

G Model CATTOD-9507; No. of Pages 7

ARTICLE IN PRESS T. Pham Huu et al. / Catalysis Today xxx (2015) xxx–xxx

3

Table 1 Physico-chemical properties of the Pd/Al2 O3 catalysts. Al2 O3 diameter (mm)

a b

Catalysts

BET surface area (m2 g−1 )

Total pore volume (cm3 g−1 )

Average pore diameter (nm)

Pd loading Theor/Expa (wt.%)

dPdO

dPd

XRDb (nm)

1.0

0.5 wt.% Pd/Al2 O3 1 wt.% Pd/Al2 O3

153 162

0.46 0.47

8.6 8.7

0.5/0.36 1/0.73

3.8 4.8

5.2 6.3

1.8

0.5 wt.% Pd/Al2 O3 1 wt.% Pd/Al2 O3

191 178

0.45 0.5

7.4 7.6

0.5/0.35 1/0.72

4.1 5.4

5.6 6.5

Actual Pd loading determined by ICP analysis. Average Pd particle sizes determined by XRD for the half-width of the main PdO and Pd0 peaks.

Powder Diffraction Standards (JCPDS) cards supplied by the International Center for Diffraction Database (ICDD). The average crystallite sizes of Al2 O3 support and Pd supported catalysts were estimated using the Scherrer equation. 2.3.4. X-ray photoelectron spectroscopy (XPS) Chemical states of the atoms in the catalyst surface were investigated by XPS spectrometer (AXIS Ultra DLD spectrometer, KRATOS ANALYTICAL) equipped with Al (K␣) radiation. XPS data were calibrated using the binding energy of C 1s (284.6 eV) as the standard. The XPS core level spectra were analyzed with a fitting routine which decomposes each spectrum into individual, mixed Gaussian–Lorentzian peaks using a Shirley background subtraction over the energy range of the fit. The surface composition was calculated from the integrated peaks, using empirical cross-section factors for XPS. 3. Results and discussion 3.1. Catalysts characterization The physico-chemical properties of Al2 O3 supports show that the packed bulk density decreases with increasing the bead diameter. Both the BET surface area and the total pore volume increases with the alumina bead diameter, being in the range commonly accepted by the IUPAC for mesoporous materials. Table 1 reports the physico-chemical properties of the Pd/Al2 O3 catalysts. Compared to fresh Al2 O3 , the surface area and the pore volume decrease with increasing Pd loading, and could be attributed to the partial pore blockage of the support by the metal species. N2 adsorption–desorption isotherms and pore size distributions are shown in Fig. 2. The N2 adsorption/desorption profile of the catalysts can be assigned to the type IV isotherm (Fig. 2a), showing a hysteresis loop due to capillary condensation, which is a typical phenomenon of a mesoporous material. The average pore size of the catalysts 0.5 wt.% Pd/Al2 O3 and 1 wt.% Pd/Al2 O3 were around 8.6 and 8.7 nm ( = 1.0 mm) and, 7.4 and 7.6 nm ( = 1.8 mm), respectively (Fig. 2b). XRD patterns corresponding to the Pd supported catalysts are presented in Fig. 3. The results suggest the formation of alumina phase with presence of the characteristic XRD peaks for ␥-Al2 O3 phase. Diffraction peaks at 19.4◦ , 37.2◦ , 45.6◦ and 66.9◦ , corresponding to the (1 1 1), (3 1 1), (4 0 0) and (4 4 0) planes, respectively, are attributed to the mesoporous ␥-Al2 O3 support (JCPDS-ICDD Card No. 00-010-0425). The widths of the XRD peaks allow us to calculate the average crystallite size using the Debye–Scherrer equation (results derived from the (4 0 0) and (4 4 0) reflections are averaged). This analysis shows the average crystallite size of 0.5 wt.% Pd/Al2 O3 and 1 wt.% Pd/Al2 O3 catalysts to be approximately 5.3 and 6.8 nm ( = 1.0 mm), and 4.9 and 5.4 nm ( = 1.8 mm), respectively. A broad diffraction lines of PdO phase were detected for all catalysts after calcinated at 500 ◦ C for 4 h. Peaks at 32.9◦ , 54.7◦ and 60.5◦ , correspond to (0 0 2), (1 1 2) and (2 0 0) planes, respectively

Fig. 2. (a) N2 adsorption/desorption isotherms and (b) pore size distributions associated with 0.5% Pd/Al2 O3 (solid line) and 1% Pd/Al2 O3 (dashed line), 1 mm diameter (black lines) and 1.8 mm diameter (gray lines).

Fig. 3. XRD patterns of 0.5 and 1 wt.% Pd/Al2 O3 (1.0 and 1.8 mm diameter) and the corresponding patterns of ␥-Al2 O3 (--), PdO (--) and Pd0 (--).

Please cite this article in press as: T. Pham Huu, et al., Plasma-catalytic hybrid reactor: Application to methane removal, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.001

G Model

ARTICLE IN PRESS

CATTOD-9507; No. of Pages 7

T. Pham Huu et al. / Catalysis Today xxx (2015) xxx–xxx

4

Table 2 XPS data for the Pd/Al2 O3 catalysts (alumina 1.0 mm diameter). Catalysts

0.5 wt.% Pd/Al2 O3 1 wt.% Pd/Al2 O3

Binding energy (eV) Pd 3d5/2

Pd phase surface composition (at.%)

PdO

PdO2

PdO

PdO2

336.5 336.6

338.5 338.6

63.6 77.6

36.4 22.4

Surface Pd/Al ratio (at.%/at.%)

0.0003 0.0037

the average Pd metal particle size derived from the (1 1 1) reflection, of 0.5 and 1 wt.% Pd/Al2 O3 catalysts was calculated to be approximately 5.2 nm and 6.3 nm ( = 1.0 mm), and 5.6 and 6.5 nm ( = 1.8 mm), respectively. Obtained results show that the crystal particle size, both the PdO and Pd0 , increases with increasing metal loading and the specific surface area.

3.2. Methane oxidation over Pd/Al2 O3

Fig. 4. XPS of the Pd 3d region associated to 0.5 wt.% Pd/Al2 O3 (solid line) and 1 wt.% Pd/Al2 O3 (dashed line) catalysts, 1.0 mm diameter.

(JCPDS-ICDD Card No. 00-046-1211). XRD analysis confirms a small Pd metal peak to be present along with the major PdO peaks. The peak observed at 39.4◦ correspond to the (1 1 1) of metallic palladium (JCPDS-ICDD Card No. 87-0638). However, XPS analysis (Table 2 and Fig. 4) of 0.5 and 1.0 wt.% Pd/Al2 O3 ( = 1.0 mm), show only the Pd2+ peak. Then, it can be concluded that only PdO is present on the catalysts surface. The XPS Pd 3d spectrum obtained from these two catalysts confirm that the main oxidation state of Pd is Pd2+ , a doublet with peaks at 336.7 and 341.8 eV, corresponding to the PdO phase. Thus, the exchange or equilibration should occur on the surface of PdO at lower temperatures. XPS results have also shown the formation of palladium species with high oxidation state, such as PdO2 (338 eV), according to the results reported by Otto et al. [29]. This study evidenced the fact that, on 0.5% Pd/Al2 O3 catalysts PdO and PdO2 majorly present on the surface. The presence of PdO2 could be attributed to oxygen transfer from the promoter to the PdO interface, inducing the formation of new interfacial sites for the oxidation reaction. These results are in agreement with the results reported by various authors [30–32], which show that the PdO decomposition temperature is near 700 ◦ C. Based on the experimental studies of Baylet et al. [30] the decomposition of PdO, on Pd/Al2 O3 catalysts, occurs only above 700 ◦ C in inert conditions. Datye et al. [33] using XRD and XPS analysis demonstrated that PdO is the only phase detected on Pd/Al2 O3 catalysts (calcined, reduced at 400 ◦ C and next heated in air at 770 ◦ C). Li et al. [34] using TEM micrographs and XRD analysis evidenced that PdO is the only phase found on the Pd/SiO2 catalysts after reducing at room temperature under H2 /Ar (5/95) flow. The PdO species in the catalysts was completely reduced by H2 to metallic palladium when temperature was raised to 600 ◦ C. Therefore, PdO phase could be defined as the major phase in all the catalysts, where the crystal particle size can be calculated using XRD peaks. This analysis shows the average PdO particle size, derived from the (2 0 0) and (2 2 0) reflection, of 0.5 and 1 wt.% Pd/Al2 O3 catalysts to be approximately 3.8 and 4.8 nm ( = 1.0 mm), and 4.1 and 5.4 nm ( = 1.8 mm), respectively. In the same way,

CH4 catalytic oxidation was carried out in different experimental steady-state conditions after a stabilization of 30 minutes at each studied temperature. Thus, we considered in this study, the influence of the support size, the amount of Pd loading, and the GHSV on CH4 conversion. The first set of experiments was carried out at GHSV = 15 000 h−1 i.e. a constant catalyst volume for a given flow rate. The second one was performed at constant catalyst weight leading to the two different GHSV such as 21 500–15 000 h−1 for support diameters of 1.0 and 1.8 mm. Before comparing the effect of thermal catalysis, plasmacatalysis, and plasma alone on the CH4 conversion efficiency, it was evidenced that in the absence of plasma, ␥-Al2 O3 does not exhibit any catalytic activity for CH4 conversion at temperature lower than 450 ◦ C. These results are in agreement with the work of Marques et al. [28] who reported a very poor ␥-Al2 O3 activity at 450 ◦ C (8 and 2% for GHSV = 20 000 and 40 000 h−1 , respectively) in more complex gas mixture representative CHP emissions, typically 150 ppm NO, 8% O2 , 7% CO2 , 1000 ppm CH4 in N2 as balance. In Table 3, the catalytic activity of Pd/Al2 O3 for the total oxidation of CH4 was followed as a function of the reaction temperature for various sizes of alumina beads and palladium loading. For GHSV = 15 000 h−1 and according to the values of T10 , T30 and T50 reported, it can be seen that for the same loading of PGM Pd/Al2 O3 ( = 1.0 mm) catalyst appeared slightly more active than Pd/Al2 O3 ( = 1.8 mm) catalyst in the overall range temperature explored. These results could be correlated to the high residence time of the active species in the reactive zone when 1.0 mm of Al2 O3 is packed into the reactor. Moreover, it could also be correlated to the better dispersion of the palladium particles on the small diameter support, related to the high surface exposure of the active sites. When comparing the CH4 conversion activity at constant GHSV, the 1 wt.% Pd/Al2 O3 ( = 1.0 mm) catalyst exhibits a higher activity than 0.5 wt.% Pd/Al2 O3 ( = 1.0 mm) according to the values of T10 , T30 and T50 reported in the Table 3. This latter result was expected since the loading of Pd was higher and well dispersed on the support. In addition to that, the experiment was conducted with 1 wt.% Pd/Al2 O3 ( = 1.0 mm) by changing the GHSV from 15 000 to 21 000 h−1 . As expected, since the GHSV increases the catalytic activity decreases in all the range of the temperature. Fig. 5 shows the CH4 conversion over Pd supported on Al2 O3 beads ( = 1.0 mm) at initial concentrations of 500 and 1000 ppm. For both 500 and 1000 ppm of inlet concentrations, CH4 conversion increases with increasing the Pd loading from 0.5 to 1 wt.%. The T50 values were shifted to lower temperatures by about 27 ◦ C when comparing activities of 0.5 and 1 wt.% Pd/Al2 O3 for both CH4 initial concentrations 500 and 1000 ppm. This fact could be attributed to the greater number of active sites present on the catalyst to increase

Please cite this article in press as: T. Pham Huu, et al., Plasma-catalytic hybrid reactor: Application to methane removal, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.001

G Model CATTOD-9507; No. of Pages 7

ARTICLE IN PRESS T. Pham Huu et al. / Catalysis Today xxx (2015) xxx–xxx

5

Table 3 Catalytic activity of the Pd/Al2 O3 catalysts.  (Al2 O3 ) (mm)

1.0 1.8 1.0 1.0 1.8

GHSV (h−1 )

15 000 15 000 15 000 21 000 15 000

Pd loading (wt.%) theor/expa

0.5/0.36 0.5/0.35 1.0/0.73 1.0/0.73 1.0/0.72

Catalyst weight (g)

3.1 2.2 3.1 2.2 2.2

Catalytic activityb T10 (◦ C)

T30 (◦ C)

T50 (◦ C)

T90 (◦ C)

278 285 247 262 255

320 328 291 305 300

342 350 315 330 324

397 400 373 383 380

The catalysts were calcined in air at 500 ◦ C before reaction. Reaction conditions: 1000 ppm in air. a Actual Pd loading determined by ICP analysis. b Temperature at which CH4 conversion is reached 10, 30, 50, and 90%.

Fig. 5. Influence of Pd loading (0.5 and 1%) and CH4 inlet concentrations (500 and 1000 ppm) on conversion as a function of temperature.

the amount of Pd loaded in agreement with the literature [35,36]. XPS analysis (Fig. 4 and Table 2) shows that the amount of PdO in the surface is higher for the catalyst with high amount of Pd (high Pd/Al ratio). This fact could explain the higher catalytic activity, because it is known that the PdO is the active phase in the CH4 oxidation at low temperature. Thus, the higher ratio of the surface PdO amounts changes also the number of the surface PdO sites that are thought as responsible for different activity of PdO/Al2 O3 catalysts in the CH4 oxidation reaction, as described by several authors [31,37]. Moreover, it is clearly seen from Fig. 5 that the total conversion of CH4 is achieved at 400 ◦ C and FTIR results show that only CO2 and H2 O were detected as expected. 3.3. Plasma and plasma-catalytic oxidation of methane The effect of the temperature and the specific input energy on the CH4 oxidation (1000 ppm) will be presented for 1 wt.% Pd/Al2 O3 catalyst ( = 1.0 mm), since we evidenced that it is the most active catalyst under our experimental conditions. Results presented as a function of temperature for the maximum SIE are shown in Fig. 6. The performances of both catalytic and plasmacatalytic reactors for CH4 conversion are higher compared to those of plasma reactor used alone. Firstly, using plasma alone (without catalyst) we observed a quasi-linear dependence of the CH4 destruction above ∼250 ◦ C, and the maximum of about 67% conversion is reached at 400 ◦ C. The main products of CH4 conversion are CO and CO2 . However, plasma produced products such as O3 and HNO3 at temperature below 200 ◦ C. Above 250 ◦ C, the HNO3 concentration decreases with increasing temperature. It can be suggested that, HNO3 molecules decompose and produce NO and NO2 . These results are in line with the work of Pringle et al. [38] who observed the similar species formation at 130 ◦ C. Baylet et al. [39] investigated the CH4 total oxidation, at 200 ◦ C using DBD reactor in a gas mixture of N2 /O2 /CO2 /H2 O (75/18.7/4/2 vol.%) with the SIE

Fig. 6. CH4 conversion as a function of temperature at SIE = 148 J/L for plasma alone and plasma-catalysis (IPC vs. PPC). The CH4 concentration was fixed as 1000 ppm at the reactor inlet.

of 1823 J/L, and evidenced that CO is produced as a major product. Secondly, conventional thermal catalysis, i.e. in the absence of plasma, shows a very slow increase in CH4 conversion until 250 ◦ C, and thereafter it increases rapidly and reaches about 100% at 400 ◦ C. Finally, when the plasma is combined with the catalyst, the shape of the curve is similar to thermal catalysis. However, for all the investigated temperature, it has shown more CH4 conversion. Although, plasma–catalyst coupling does not show synergetic effect, but it highlights the fact that plasma increases the CH4 conversion and total oxidation. For a given CH4 conversion, for instance to attain 50% conversion, compared to thermal catalysis, plasma-catalysis required only 310 ◦ C. This is 30 ◦ C lower than the temperature required for conventional thermal catalysis and 70 ◦ C lower than the plasma alone. This emphasizes the fact that using plasma-catalysis high CH4 conversion could be achieved at low temperature. On the other hand, in the plasma-catalysis at high temperature, CO was oxidized to CO2 by the Pd/Al2 O3 catalyst. This latter results show that at temperature higher than 300 ◦ C the major effect is the catalytic effect, since the differences in CH4 abatement are slight between the systems plasma ON and plasma OFF, which is agreement with the studies of Da Costa et al. [26] and Marques et al. [28]. The comparison with plasma-alone system show that CH4 conversion efficiency over 1 wt.% Pd/Al2 O3 in the plasma discharge zone (IPC) was improved by 40% at 300 ◦ C and 50% at 400 ◦ C. The difference of CH4 conversions when plasma was switched-off suggest that the catalyst could be affected differently by the ON/OFF sequences of plasma operation depending on its position relative to the plasma. In IPC configuration, the catalyst may affect the discharge properties and influence the production of excited and short lifetime reactive plasma species [20,40]. On the other hand, a fraction of unstable reactive species recombines to form more stable species (e.g. ozone, NO, NO2 , HNO2 , HNO3 ). Since these long lifetime

Please cite this article in press as: T. Pham Huu, et al., Plasma-catalytic hybrid reactor: Application to methane removal, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.001

G Model CATTOD-9507; No. of Pages 7 6

ARTICLE IN PRESS T. Pham Huu et al. / Catalysis Today xxx (2015) xxx–xxx

Fig. 7. FTIR spectra showing specie features recorded after processing of air–CH4 at energy deposition of 87 J/L: (a) Pd/Al2 O3 catalyst at 300 ◦ C, (b) plasma at 300 ◦ C, (c) plasma–Pd/Al2 O3 (PPC) at 30 ◦ C, and (d) plasma–Pd/Al2 O3 (PPC) at 300 ◦ C.

Fig. 8. Ozone concentration as a function of specific input energy at T = 200 ◦ C for plasma alone and plasma-catalysis (IPC vs. PPC): [CH4 ]0 = 1000 ppm.

species are able to reach the catalyst material positioned downstream the discharge zone (PPC). This indicates that during plasma operation, catalyst surfaces might be modified differently depending on the relative position of the plasma and the catalyst. These modifications, due to the dual effect of plasma discharges and temperature, could results to enhance the dispersion of active catalytic components [41], modify the oxidation state of the material [42,43], and enhance the specific surface area or change of catalytic structure [44]. Consequently, the CH4 conversion is affected. When the plasma was switched-on, we measure a better CH4 conversions in IPC compared to PPC configuration. Although the difference is low (<6%), the amounts of the reaction products were higher in IPC compared to those obtained in PPC. Although obtained in different experimental conditions (plasma energy deposition, gas mixture, catalyst support and loading), these results are compared with those of Baylet et al. [39] who reported that the highest CH4 conversion rate was obtained with an initial methane concentration of 0.3 vol.% and 1458 J/L without CO production when the catalyst was in post-plasma position. These authors attributed the decrease of CH4 conversion when the catalyst was located in the plasma zone to the decrease of the residence time of the reactant in the reactive zone. Fig. 7 presents examples of FTIR spectra recorded at 300 ◦ C for catalyst alone, plasma alone, catalyst post-plasma, and at 30 ◦ C for catalyst post-plasma. The analysis of these spectra can help to understand the mechanisms involved in these complex systems. In addition to CH4 oxidation products, other products such as O3 , N2 O, NO, NO2 , HNO3 could be observed and their concentrations are strongly temperature-dependent. The comparison of spectra obtained in the case of plasmacatalysis at 30 and 300 ◦ C (Fig. 7(c) and (d)) indicate the oxidation of CH4 and the formation of CO2 and H2 O. Moreover, at high temperature O3 and HNO3 disappear in favor of NO and NO2 . It was also observed that the performance of plasma reactor combined with Pd/Al2 O3 catalyst was found to be better in terms of less formation of CO compared especially to IPC configuration. Fig. 8 presents the concentration of ozone as a function of SIE for plasma and plasma–catalyst (IPC and PPC) systems at 200 ◦ C. For all the investigated systems, the ozone concentration decreases with increasing the SIE. Indeed, for simple DBD reactor, the decrease in ozone concentration is predominant as compared to IPC and PPC configurations. This phenomenon can be attributed to the gas phase ozone consumption by NO/NO2 and H2 O. At SIE = 55 J/L, the concentration of ozone at the reactor outlet, packed with Pd/Al2 O3 , was decreased by nearly 80%. The combination of plasma and catalyst not only improved the CH4 oxidation but also considerably

enhanced the O3 decomposition. One of the mechanisms suggested explaining the observed catalyst improvement by plasma discharges include the ozone-catalyst surface reactions. In addition, the adsorption of ozone on the surface of the catalyst and its thermal decomposition lead to its subsequent dissociation into atomic oxygen which could enhance the NO/NO2 oxidation. N2 O was also observed in these experiments and the result is not discussed here. The reaction of CH4 conversion over 1 wt.% Pd/Al2 O3 was carried out by varying the energy deposition from 0 to 148 J/L. Comparison of in-plasma catalytic results for CH4 conversion in air is presented in Fig. 9 as a function of temperature and specific input energy. In absence of plasma discharges, one can see that the Pd/Al2 O3 catalyst is oxidizing the CH4 at temperature above 250 ◦ C with a 50% conversion at 332 ◦ C. Similar observations have been reported in the literature [4,25,45]. In presence of plasma discharges, CH4 conversion increases with increasing the specific input energy. Slight changes were observed for SIE lower than 55 J/L whereas the effect of the plasma really significant when high SIE were used. Compared to catalytic conversion, the plasma significantly improves the CH4 conversion by a factor of 3 at 300 ◦ C and 1.4 at 350 ◦ C for the highest SIE used in these studies. The reduction of temperature T50 is about 30 ◦ C. For comparison, the T50 reduction for propene and toluene obtained at the same experimental conditions are 160 and 90 ◦ C, respectively, although dramatically lower temperatures are required for complete removal: 150 ◦ C for propene and 250 ◦ C for toluene as reported by Pham Huu et al. [46].

Fig. 9. Effect of plasma energy deposition on the CH4 conversion as a function of temperature: 1 wt.% Pd/Al2 O3 catalyst in IPC configuration, [CH4 ]0 = 1000 ppm.

Please cite this article in press as: T. Pham Huu, et al., Plasma-catalytic hybrid reactor: Application to methane removal, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.001

G Model CATTOD-9507; No. of Pages 7

ARTICLE IN PRESS T. Pham Huu et al. / Catalysis Today xxx (2015) xxx–xxx

4. Conclusion Methane oxidation has been investigated in a plasma-catalytic hybrid reactor. A comparison between thermal catalysis, catalyst after-plasma (PPC) and catalyst in-plasma (IPC) was systematically made at the operating temperature below 500 ◦ C and specific input energy below 148 J/L. Pd-supported on alumina beads were used as catalyst. It is evidenced that the CH4 conversion increases with increasing the specific input energy and the gas temperature. It was found that, the plasma–Al2 O3 coupling significantly enhances the catalytic activity. CH4 conversion improvements by a factor 3 at 300 ◦ C and 1.4 at 350 ◦ C were obtained in plasma-catalytic system compared to thermal catalytic experiments for the highest specific input energy used. Pd/Al2 O3 catalyst exhibits a high catalytic activity at low-temperature when plasma is switched-on for the both IPC and PPC systems. At high temperature, the catalyst itself leads to 100% conversion and the plasma is not needed and can be switchedoff. Moreover, performance was slightly better in IPC configuration. In addition to CH4 oxidation products (CO, CO2 , H2 O), FTIR analysis shows other products such as O3 , N2 O, NO, NO2 , and HNO3 with concentrations strongly temperature-dependent. Acknowledgments This work was performed within International Group of Research (GDRI) “Catalysis for polluting emissions aftertreatment and production of renewable energies”. The authors would like to thank the scientific service of IRCELYON for assistance in XRD, XPS, N2 adsorption and ICP analysis and for stimulating discussions. References [1] T. Garcia, B. Solsona, D.M. Murphy, K.L. Antcliff, S.H. Taylor, J. Catal. 229 (2005) 1–11. [2] C.H. Wang, S.S. Lin, C.L. Chen, H.S. Weng, Chemosphere 64 (2006) 503–509. [3] Y. Li, X. Zhang, H. He, Y. Yu, T. Yuan, Z. Tian, J. Wang, Y. Li, Appl. Catal. B: Environ. 89 (2009) 659–664. [4] P. Gélin, M. Primet, Appl. Catal. B: Environ. 39 (2002) 1–37. [5] D. Ciuparu, M.R. Lyubovsky, E. Altman, L. Pfefferle, A. Datye, Catal. Rev. 44 (2002) 593–649. [6] K. Urashima, J.S. Chang, IEEE Trans. Dielectr. Electr. Insul. 7 (5) (2000) 602–614. [7] D.N. Tran, C.L. Aardahl, K.G. Rappe, P.W. Park, C.L. Boyer, Appl. Catal. B: Environ. 48 (2004) 155–164. [8] H.H. Kim, Plasma Process. Polym. 1 (2) (2004) 91–110. [9] A.M. Vandenbroucke, R. Morent, N. De Geyter, C. Leys, J. Hazard. Mater 195 (2011) 30–54. [10] A. Khacef, J.M. Cormier, J.M. Pouvesle, J. Phys. D: Appl. Phys. 35 (2002) 1491–1498.

7

[11] M.A. Malik, Y. Minamitani, K.H. Schoenbach, IEEE Trans. Plasma Sci. 33 (2005) 50–56. [12] G. Djéga-Mariadassou, F. Baudin, A. Khacef, P. Da Costa, in: V.I. Parvulescu, M. Magureanu, P. Lukes (Eds.), Plasma Chemistry and Catalysis in Gases and Liquids, Wiley VCH, 2012. [13] U. Roland, F. Holzer, F.D. Kopinke, Appl. Catal. B: Environ. 58 (2005) 217–226. [14] A. Leray, A. Guy, M. Makarov, K. Lombaert, J.M. Cormier, A. Khacef, Top. Catal. 56 (2013) 222–226. [15] O. Gorce, H. Jurado, C. Thomas, G. Dega-Mariadassou, A. Khacef, J.M. Cormier, J.M. Pouvesle, G. Blanchard, S. Calvo, Y. Lendresse, SAE Technical Paper Series 2001-01-3508, 2001. [16] M.J. Kirkpatrick, E. Odic, J.P. Leininger, G. Blanchard, S. Rousseau, X. Glipa, Appl. Catal. B: Environ. 106 (2011) 160–166. [17] F. Holtzer, U. Roland, F.D. Kopinke, Appl. Catal. B: Environ. 38 (2002) 163–181. [18] S. Delagrange, L. Pinard, J.M. Tatibouet, Appl. Catal. B: Environ. 68 (2006) 92–98. [19] Z. Jia, A. Vega-Gonzales, M. Ben Amar, K. Hassouni, S. Tieng, S. Touchard, A. Kanaev, X. Duten, Catal. Today 208 (2013) 82–89. [20] H.T. Quoc An, T. Pham Huu, T. Le Van, J.M. Cormier, A. Khacef, Catal. Today 176 (2011) 474–477. [21] J. Van Durme, J. Dewulf, C. Leys, H. Van Langenhove, Appl. Catal. B: Environ. 78 (2008) 324–333. [22] J.J. Zou, Y.P. Zhang, J. Liu Ch, Y. Li, B. Eliasson, Plasma Chem. Plasma Process. 23 (2003) 69–82. [23] B. Pietruszka, M. Heintze, Catal. Today 90 (2004) 151–158. [24] T. Hammer, T. Kappes, M. Baldauf, Fuel Chem. Div. 46 (1) (2001). [25] M. Heintze, B. Pietruszka, Catal. Today 89 (1-2) (2004) 21–25. [26] P. Da Costa, R. Marques, S. Da Costa, Appl. Catal. B: Environ. 84 (2008) 214–222. [27] J.L. Hueso, J. Cotrino, A. Caballero, J.P. Espinos, A.R. Gonzalez-Elipe, J. Catal. 247 (2007) 288–297. [28] R. Marques, S. Da Costa, P. Da Costa, Appl. Catal. B: Environ. 82 (2008) 50–57. [29] K. Otto, L.P. Haack, J.E. DeVries, Appl. Catal. B: Environ. 1 (1992) 1–12. [30] A. Baylet, P. Marécot, D. Duprez, P. Castellazzi, G. Groppi, P. Forzatti, Phys. Chem. Chem. Phys. 13 (2011) 4607–4613. [31] Z. Li, G. Xua, G.B. Hoflund, Fuel Process. Technol. 84 (2003) 1–11. [32] K. Fujimoto, F.H. Ribeiro, M. Avalos-Borja, E. Iglesia, J. Catal. 179 (1998) 431–442. [33] A.K. Datyea, J. Bravo, T.R. Nelson, P. Atanasova, M. Lyubovsky, L. Pfefferle, Appl. Catal. A: Gen. 198 (2000) 179–196. [34] B. Li, H. Li, W.Z. Weng, Q. Zhang, C.J. Huang, H.L. Wan, Fuel 103 (2013) 1032–1038. [35] L.S. Escandon, S. Ordonez, A. Vega, F.V. Dıez, Chemosphere 58 (2005) 9–17. [36] T. Maillet, C. Solleau, J. Barbier Jr., D. Duprez, Appl. Catal. B: Environ. 14 (1997) 85–95. [37] B. Stasinska, A. Machocki, K. Antoniak, M. Rotko, J.L. Figueiredo, F. Gonc¸alves, Catal. Today 137 (2008) 329–334. [38] K.J. Pringle, J.C. Whitehead, J.J. Wilman, J. Wu, Plasma Chem. Process. 24 (2004) 421–434. [39] A. Baylet, P. Marécot, D. Duprez, X. Jeandel, K. Lombaert, J.M. Tatibouët, Appl. Catal. B: Environ. 113-114 (2012) 31–36. [40] V. Demidyuk, J.C. Whitehead, Plasma Chem. Plasma Process. 27 (2007) 85–94. [41] W. Hua, L. Jin, X. He, J. Liu, H. Hu, Catal. Commun. 11 (2010) 968–972. [42] M. Foix, C. Guyon, M. Tatoulian, P. Da Costa, Catal. Commun. 12 (2010) 20–24. [43] M. Foix, C. Guyon, M. Tatoulian, P. Da Costa, Mod. Res. Catal. 2 (2013) 68–82. [44] C.J. Liu, K. Yu, Y. Zhang, X. Zhu, F. He, B. Eliasson, Appl. Catal. B: Environ. 47 (2004) 95–100. [45] H. Widjaja, K. Sekizawa, K. Eguchi, H. Arai, Catal. Today 35 (1997) 197–202. [46] T. Pham Huu, M. Adamowka, P. Da Costa, J.M. Cormier, A. Khacef, Proceedings of the XXth International Conference on Gas Discharges and their Applications, 2014, pp. 586–589, ISBN: 978-2-9548207-1-2.

Please cite this article in press as: T. Pham Huu, et al., Plasma-catalytic hybrid reactor: Application to methane removal, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.001