425
Topics in Catalysis Vols. 42–43, May 2007 ( 2007) DOI: 10.1007/s11244-007-0218-7
Pd/Co3O4 catalyst for CH4 emissions abatement: study of SO2 poisoning effect L.F. Liottaa,*, G. Di Carlob, G. Pantaleob, A.M. Veneziaa, G. Deganelloa,b, E. Merlone Borlac, and M.F. Pidriac a ISMN-CNR, via Ugo La Malfa, 153, 90146 Palermo, Italy Dipartimento di Chimica Inorganica e Analitica ‘‘Stanislao Cannizzaro’’, Universita` di Palermo, Viale delle Scienze, Parco d’Orleans II, 90128 Palermo, Italy c Centro Ricerche FIAT, Strada Torino, 50, 10043 Orbassano (To), Italy
b
A catalyst with 0.7 wt% Pd load supported over Co3O4 oxide was investigated in the methane oxidation by operating under CH4/O2 stoichiometric conditions. The effect of the noble metal addition on the activity of bare Co3O4 was evaluated. Samples were characterized by BET, XRD, TPR and XPS analyses. The SO2 poisoning of Pd catalyst and Co3O4 was studied by performing CH4 oxidation tests under stoichiometric conditions in SO2 (1 ppm or 10 ppm). Experiments evidenced that in our conditions the low amount of SO2 doesn’t influence the Pd behaviour, whereas in presence of 10 ppm of SO2 some deactivation occurs that becomes more evident above 450 C at which the catalyst doesn’t reach 100% of methane conversion. Catalytic tests performed over Co3O4 and the Pd supported catalyst, after a treatment at 350 C for 15 h in 10 ppm SO2/He, suggest that Co3O4 is a sulphating support, as confirmed by XPS analysis. Therefore, an important role in lowering the sulphur poisoning of Pd may be played by Co3O4. KEY WORDS: methane oxidation under stoichiometric conditions; Pd/Co3O4; sulphur oxide effect.
1. Introduction Nowadays natural gas, which consists primarily of methane, represents a promising alternative energy source in automotive and heavy-duty field, because of the low negative impact on air quality [1]. Nevertheless, emissions of unburned CH4 (a much more powerful greenhouse gas than CO2) must be reduced by an effective catalytic after-treatment of the exhausts [2]. Methane combustion at relatively low temperatures can be achieved only by using highly active noble metal catalysts [3]. Pd and Pt supported systems are among the most effective catalysts for the abatement of unburned methane from natural gas fuelled vehicles, depending on the operation conditions, lean engine or stoichiometric air/fuel ratio, respectively [4]. However, Pd supported catalysts exhibit a strong sensitivity to SO2 poisoning under lean conditions due to the formation of PdSO4 species [5–7]. The rate of PdO poisoning is found to depend on the nature of the support, being a sulphating or non-sulphating support. In particular a good interaction between SO2 and the support would lower the palladium sulphation and promote regeneration [5, 7–9]. Recently, we have reported that pure Co3O4 along with Co3O4/CeO2 composite oxide containing Co/Ce atomic ratio about 1:1 [10, 11] are effective catalysts for methane oxidation. Moreover, studies of water vapour poisoning have shown that such oxides exhibit good resistance to water vapour poisoning [11]. In this work a * To whom correspondence should be addressed. E-mail:
[email protected]
catalyst Pd/Co3O4 containing a low amount (0.7 wt%) of noble metal has been tested, as powder, in the methane oxidation by operating in CH4/O2 stoichiometric ratio that is of interest for bi-fuel engines application [12]. To our knowledge, while several works have focused the critical effect of sulphur poisoning in the catalytic oxidation of methane in lean conditions [5–9], no studies were devoted to the catalysts sensitivity to SO2 under stoichiometric air/fuel ratio. Therefore the aim of the present paper is to analyze in such conditions the influence of SO2 poisoning (1, 10 ppm) on the activity of Co3O4 and the supported Pd catalyst. The data here reported are preliminary results of a comprehensive study aiming to investigate a monolithic catalyst of Co3O4 with low-Pd content for CO and CH4 emissions abatement [13].
2. Experimental Co3O4 oxide has been prepared by precipitation method by adding at r.t. a solution of (NH4)2CO3 (1 M) to Co(NO3)2 Æ 6H2O dissolved in distilled water. The precipitate thus obtained was aged at r.t. for 3 h, filtered and washed with distilled water. Then, the powder was dried at 120 C overnight and calcined at 650 C for 5 h. The Pd catalyst, with loading 0.7 wt%, was prepared by impregnation of the so prepared Co3O4 oxide (specific surface area of 12 m2/g) with Pd(NH3)4(NO3)2 aqueous solution under stirring for 1 h at 60 C. The catalyst was then dried at 120 C and calcined at 400 C for 4 h (as prepared). The real weight loading of the Pd catalyst 1022-5528/07/0500-0425/0 2007 Springer Science+Business Media, LLC
426
L.F. Liotta et al./CH4 oxidation by Pd/Co3O4
was checked by X-ray fluorescence (S2 Ranger, Bruker) and corresponded to the nominal one. The as prepared samples, bare Co3O4 and the Pd supported catalyst, were characterized by specific surface area measurements (BET method), X-ray diffraction (XRD), Temperature programmed reduction (TPR) and X-ray photoelectron spectroscopy (XPS). In order to investigate the nature of sulphur species as well as the oxidation state of palladium, XPS technique was used to characterize also the samples after reaction. Details on the characterization techniques are reported elsewhere [10, 11]. Before the catalytic tests the as prepared samples were pre-treated ‘‘in situ’’ under flowing O2 (5 vol% in He) at 350 C for ½ h. The standard reagent gas mixture consisting of 0.3 vol% of CH4 + 0.6 vol% O2 in He was led over the catalyst (50 mg) at a flow rate of 50 mL/min (STP), equivalent to a weight hourly space velocity (WHSV) of 60000 mL g)1 h)1. Activities were measured by increasing the temperature from 200 to 600 C (by steps of 50 C). Experiments of methane oxidation in presence of SO2 were performed by co-feeding 1 or 10 vol ppm of SO2, respectively. On the as prepared samples treatments with flowing 10 vol ppm of SO2/He at 350 C were also carried out, then after cooling to 200 C the catalytic activity (0.3 vol% of CH4 + 0.6 vol% O2/He) was measured. The gas composition at the outlet of the reactor was analysed by on line mass quadrupole (ThermostarTM, Balzers) and an IR analyser (ABB Uras 14).
3. Results and discussion
TCD signal (A.U.)
In figure 1 the TPR profiles of Co3O4 and the Pd supported catalyst are shown. As reported in the literature, the reduction profile of Co3O4 shows two defined peaks at relatively low temperature, 260 and 315 C, respectively, according to a step-wise process via Co3+ fi Co2+ fi Co0 [10, 11]. The quantitative evaluation of the overall hydrogen consumption
(b) (a)
-100
0
100 200 300 400 500 600 700 800 900
Temperature (°C) Figure 1. H2-TPR profiles of (a) bare Co3O4 and (b) Pd(0.7 wt%)/ Co3O4.
(367 mL/gcatalyst) accounts for the complete reduction (99%) of the Co3O4 to cobalt metal. The reduction of Pd/Co3O4 takes place at significantly lower temperature, indicating that the addition of palladium enhances substantially the reducibility of the support, according to the literature [14]. During the TPR process, the reduction of PdOx species is expected also to occur. Accordingly, the peak at 18 C with a hydrogen uptake of 2.9 mL/gcatalyst may be attributed to the reduction of PdO2 crystallites, highly dispersed on the support. Palladium, once reduced, catalyzes the reduction of the Co3O4, the higher the interaction with the support, the lower the Co3O4 reduction temperature. Indeed, the TPR profile exhibits above 50 C two main peaks, centred at 115 C and 245 C, respectively, (overall hydrogen uptake of 369 mL/g) that are likely associated with the complete reduction of different portion of Co3O4, more or less interacting with the Pd metal. Characterization of pure Co3O4 and Pd/Co3O4 by XRD showed only reflections of Co3O4 spinel (ICSD, n 24210), nor Pd neither PdOx phases were identified, indicating that the metal loading (0.7 wt%) is too low and/or that the particle sizes are below the detection limit of the instrument (3 nm). The crystallite sizes of the Co3O4 phase, calculated by XRD using the Scherrer equation are of 72 nm in the bare oxide and Pd supported catalyst as well. The effect of SO2 addition (1, 10 ppm) to the feed (0.3 vol% CH4 + 0.6 vol% O2/He) was investigated over Pd/Co3O4 and bare Co3O4, as reported in figures 2(a, b). All the reported curves have been recorded over the as prepared samples, so they represent first runs. In table 1, the temperatures of 50% conversion (T50) are listed. It can be clearly seen that 1 ppm of SO2 in the reaction stream doesn’t influence the activity of Pd catalyst and neither of the Co3O4 oxide. The latter sample seems to tolerate also higher sulphur content, up to 10 ppm of SO2 fed. On the contrary, the Pd catalyst deactivates in presence of 10 ppm of SO2. The T50 increases of almost 40 C and the poisoning effect becomes more evident above 450 C at which the catalyst is unable to reach 100% of methane conversion. Interestingly, when the methane conversion versus temperature was reported for Pd/Co3O4 and Co3O4, after a treatment with flowing 10 ppm SO2 for 15 h at 350 C, an opposite trend was noticed. The treatment with SO2 has a drastic effect on the activity of bare Co3O4, approximately 120 C increase of T50 was observed. On the contrary, for the Pd supported catalyst an increase of T50 as small as 20 C was found (see table 1). In order to get insight in the deactivation process of Co3O4, the effect of thermal ageing at 350 C, without sulphur, was checked and no significant decrease of activity was noticed. Therefore the observed deactivation, after prolonged exposure to 10 ppm of SO2, could be attributed to sulphur poisoning of the catalytic sites.
427
L.F. Liotta et al./CH4 oxidation by Pd/Co3O4
CH4 conversion (%)
100
CH4 conversion (%)
80 60 40 20 0 200
CH4 conversion (%)
100
(a)
100
300
400
500
600
80
60
20
0 200
(b)
test1 2 3 4
40
300
400
500
600
Temperature (°C)
80
Figure 3. CH4 conversion (%) as a function of the temperature over Pd/Co3O4 in four successive runs performed in different conditions. Filled symbols refer to tests without SO2; open symbols are related to tests with SO2 (1 ppm for tests 1–3; 10 ppm for test 4).
60 40 20 0 200
300
400
500
600
Temperature (°C)
Figure 2. CH4 conversion (%) as a function of the temperature over (a) Pd/Co3O4 and (b) Co3O4 in different conditions: (n) test without SO2, (d) test with 1 ppm SO2, (m) test with 10 ppm SO2, (r) test without SO2, after a treatment with 10 ppm SO2 for 15 h at 350 C. All the reported curves are first runs performed over portions of as prepared samples.
For the Pd catalyst the poisoning effect is much less pronounced probably due to the migration of SO2 from PdO to the support [7]. With aim to discriminate between ageing and SO2 poisoning effects, the catalytic activity of Pd/Co3O4 was studied by performing 4 successive runs (from 100 to 600 C) in presence of SO2 and in a sulphur-free reacting mixture, respectively. Figure 3 shows the consecutive curves of methane conversion versus the temperature. In table 2 the T50 are reported. As previously reported (see figure 2(a)) the presence of 1 ppm of SO2 in the reaction feed doesn’t affect the activity of the as prepared catalyst (figure 3, tests 1 with and without SO2, respectively). After the first run, the catalyst deactivates irrespective of the presence of 1 ppm of SO2 or not (almost 40 C increase of T50). Further reaction cycles did not influence the activity significantly (see table 2, runs 3–4 Table 1 Temperatures of 50% of CH4 conversion in the first catalytic run performed over Pd(0.7 wt%)/Co3O4 and Co3O4 as a function of the experimental conditions Sample
No SO2
1 ppm SO2
10 ppm SO2
No SO2, after 15 h at 350 C with 10 ppm SO2
Pd/Co3O4 Co3O4
346 443
346 447
383 442
366 562
without SO2). A more pronounced deactivation above 450 C was observed during the fourth run performed with 10 ppm SO2. In order to get insight in the deactivation process, XPS analyses were carried out over the samples, Pd/ Co3O4 and Co3O4 after the catalytic tests. The data are summarized in table 3. The Co 2p spectrum of the support is typical of the mixed oxide Co3O4, containing some Co3+ and Co2+ species, as previously observed [11]. In the case of the Pd catalyst as prepared, palladium is present as PdO2 according to the high value of Pd 3d binding energy [15] and in agreement with the TPR results. Moreover, the metal oxide is well dispersed on the surface (Pd at% ¼ 4) as deduced by XRD. After four catalytic runs without SO2 the amount of palladium oxide at the surface decreases by a factor of 10 (see table 3), at the same time a decreasing of activity occurs (figure 3, table 2). This finding suggests that in absence of SO2 the catalyst deactivation may be related to sintering or diffusion of PdO2 species. By comparing the results after 4 cycles without and with SO2 a further decrease of palladium content at the catalyst surface is observed, being the Pd (at%) of 0.4 and 0.2, respectively (table 3). Moreover exposure of Pd/Co3O4 to SO2 under stoichiometric conditions induces a reduction of palladium. The surface metallic Pd is thought to be formed Table 2 Temperatures of 50% of CH4 conversion in four successive tests over Pd(0.7 wt%)/Co3O4 by performing the test in presence and in absence of SO2, respectively Test
No SO2
1, 10 ppm SO2
1 2 3 4
346 382 393 399
345a 382a 393a 412b
a b
1 ppm SO2. 10 ppm SO2.
428
L.F. Liotta et al./CH4 oxidation by Pd/Co3O4
Table 3 XPS data of Pd/Co3O4 and Co3O4 as prepared and after catalytic tests. The uncertainty on the atomic concentration is of the order of 10% Sample
Co 2p3/2 (eV)
Pd 3d5/2 (eV)
Pd/Co3O4
779.6 781.4 779.5 780.8 779.9 781.3 780.0 781.4 779.5 780.8 779. 9 781.1 779. 7 781.1
Pd/Co3O4 After 4 runs without SO2 Pd/Co3O4 After 4 runs with SO2 (1,10 ppm) Pd/Co3O4 after 15 h at 350 C with 10 ppm SO2 and 1 run without SO2 Co3O4 Co3O4 after 15 h at 350 C with 10 ppm SO2 and 1 run without SO2 Co3O4 after 15 h at 350 C with 10 ppm SO2 and 3 runs without SO2
during the catalytic tests by oxidation of SO2 to SO3 according with the literature [6]. Therefore the observed deactivation of the Pd/Co3O4 catalyst in the presence of sulphur (figure 3, test 4 open symbols) seems to proceed through reduction of PdO2 to Pd0 and aggregation of the reduced metal (the Pd at% is as low as 0.2). An interesting observation is the reversibility of the PdO2 M Pd process in the sense that upon treatment with SO2, Pd appears re-oxidized during a successive catalytic test without sulphur (see table 3). Regarding the interaction of SO2 with Pd and/or Co3O4, the S 2p binding energy in all the samples analysed after reaction cycles up to 600 C is typical of chemisorbed SO3 or sulfate (SO42)) [6]. Preliminary regeneration treatments of the S-poisoned Co3O4 have been carried out by performing, after a treatment for 15 h at 350 C with 10 ppm of SO2, successive oxidation tests without poisoning. The XPS data are reported in table 3 for the sample processed, respectively, with one and three consecutive tests without sulphur. A progressive decrease of the S at%, from 5.0 to 1.1, respectively, is observed. Accordingly a partial recovery of the catalytic activity occurs, the temperature of 50% of methane conversion shifting from 562 (first run, see table 1) to 495 C in the third run (conversion curves for 2 and 3 runs not reported). Regeneration treatments of S-poisoned Pd/ Co3O4 and Co3O4 via SO2 removal under reducing and/or oxidation conditions [7–9] are in progress. 4. Conclusions Preliminary results are indicating Pd/Co3O4 as an active and sulphur resistant catalyst for methane oxidation. An important role in lowering the sulphur poisoning of Pd could be played by Co3O4 that seems to behave as a sulphating support. The Pd catalyst sulphation under stoichiometric conditions results in the reduction of palladium oxide to
S 2p (eV)
Co (at%)
Pd (at%)
337.5
38.8
4.0
337.6
36.3
0.4
S (at%)
335.6
169.5
34.8
0.2
1.9
337.7
169.5
33.5
0.4
2.4
38 169.6
36.6
5.0
169.6
37.5
1.1
metallic Pd, which tends to aggregate. However, the PdO2 M Pd process seems to be reversible.
Acknowledgment Support by the European Community, Network of Excellence (NoE) IDECAT (Integrated Design of Catalytic Nanomaterials for a Sustainable Production) is acknowledged.
References [1] C. Bach, C. La¨mmle, R. Bill, P. Soltic, D. Dyntar, P. Janner, K. Boulouchos, C. Onder, T. Landenfeld, L. Kercher, O. Seel and J.D. Baronick, SAE 010645 (2004). [2] J. White, J. Carroll, M. Brady, W. Burkmyre, W. Liss and M. Church, SAE 930222 (1993). [3] P. Ge´lin and M. Primet, Appl. Catal. B 39 (2002) 1. [4] R. Burch and P.K. Loader, Appl. Catal. B 5 (1994) 149. [5] J.K. Lampert, M.S. Kazi and R.J. Farrauto, Appl. Catal. B 14 (1997) 211. [6] D.L. Mowery and R.L. McCormick, Appl. Catal. B 34 (2001) 287. [7] P. Ge´lin, L. Urfels, M. Primet and E. Tena, Catal. Today 83 (2003) 45. [8] F. Arosio, S. Colussi, G. Groppi, A. Trovatelli, Catal. Today 117 (2006) 569. [9] F. Arosio, S. Colussi, G. Groppi, A. Trovatelli, Preceedings of the 7th International Congress on Catalysis and Automotive Pollution Control, August 30-September 1, 2006-Brussels, Belgium, p. 225. [10] L.F. Liotta, G. Di Carlo, G. Pantaleo and G. Deganello, Catal. Comm. 6 (2005) 329. [11] L.F. Liotta, G. Di Carlo, G. Pantaleo, A.M. Venezia and G. Deganello, Appl. Catal. B 66 (2006) 217. [12] M. Gambino, S. Iannaccone, M.F. Pidria, G. Miletto and M. Rollero, ‘‘Natural gas engine and catalyst deactivation: accelerated ageing tests’’, Preceedings of Fisita, World Automotive Congress 23–27 May 2004, Barcelona. [13] L.F. Liotta, G. Di Carlo, G. Pantaleo, A.M. Venezia, G. Deganello, E.Merlone Borla and M. Pidria, submitted to Appl. Catal. B. [14] P.-Y. Lin, M. Skoglundh, L. Lo¨wendahl, J.-E. Otterstedt, L. Dahl, K. Jansson and M. Nygren, Appl. Catal. B 6 (1995) 237. [15] K. Persson, K. Jansson, S.G. Ja¨ra˚s, J. Catal. 245 (2007) 401.
