Automobile exhaust gas has been a major source of environmental pollutions over the last several decades. Hydrocarbons (HCs), CO, and nitrogen oxides (NOx) are the three major toxic components of vehicle exhaust emissions. In recent years, there has been significant research directed towards the purification of exhaust gases [1, 2, 3], typically based on the use of catalysts [4, 5]. To meet emission standards that are becoming increasingly stringent, two kinds of catalysts are usually installed in the engine exhaust system [6]. One is the close-coupled catalyst (CCC) used to eliminate the majority of hydrocarbon emissions (60%-80% of the total emitted) during cold starts [4]. To ensure that this unit quickly achieves the appropriate light-off temperature (T50), the CCC is often installed close to the engine [7]. The other is the three-way catalyst (TWC), typically installed behind the CCC and used to simultaneously eliminate all three primary toxic emissions [8].
Because the CCC is installed very close to the engine, which can operate over 1050 °C, the catalyst must exhibit good thermal stability over prolonged exposure to high temperatures while still maintaining excellent low-temperature activity [7]. Therefore, the ideal support material for a CCC should show not only good hydrocarbon purification efficiency but also superior anti-aging performance. Al2O3 is regarded as a promising candidate for anti-aging supports and has been widely used as such in CCC units [9, 10]. However, at above 1000 °C, pure Al2O3 will gradually sinter and transition from a γ-phase to an α-phase, resulting in severe agglomeration of the active components. Modification of Al2O3 by one or more stabilizers is considered to be a viable means of suppressing this phase transition, and introducing alkali metals, alkaline earth metals, or rare earth elements has been shown to improve the thermal stability and surface area of Al2O3 [11, 12, 13, 14]. In the present study, La was applied as the stabilizer in an Al2O3 support. Palladium is often used in CCC as the active component due to its superior activity for the oxidation of HCs [5, 15]. Moreover, Pd currently also has lower price among the precious metals. PdOx species exhibit considerably higher activity for the purification of C3H8 compared to metallic Pd, even in the form of large particles formed by sintering at high temperatures [16, 17]. However, pure PdOx decomposes into metallic Pd particles above 750 °C, even in oxygen-containing atmospheres, which can greatly lower the catalytic efficiency [18]. As a result, it is necessary to find techniques to either maintain or increase the decomposition temperature of PdOx. One approach is to alter the support materials; according to Farrauto et al. [19], adding TiO2 or CeO2 to the support material can increase the decomposition temperature of PdOx by about 130 °C. Another approach is more direct and involves synthesizing bimetallic catalysts by introducing other noble metals such as Pt and Rh. For example, the introduction of Pt increases catalytic activity and prevents deactivation of the catalyst [20, 21]. Rassoul et al. [18] reported that the addition of Rh to form a Pd0.75Rh0.25/Al2O3 catalyst by stepwise impregnation can increase the temperature at which the PdOx species are stable by −230 °C, and demonstrated strong interactions between RhOx and PdOx on the surface of the catalyst based on temperature-programmed reduction (TPR) measurements. The addition of Rh to a Pd-based catalyst may also improve the activity of Pd during methane combustion due to the increased pore diffusion resistance at higher temperatures, while the effect of Rh-doping is most significant below 400 °C [21]. Due to its superior NOx reduction activity, Rh is widely used in CCCs and TWCs. However, there is no evidence in the literature of the additive effect of Rh on the catalytic activity of a Pd-based CCC meant primarily for the removal of C3H8, or on the specific reactions of C3H8 over a CCC.
We study the effects of introducing Rh into a Pd/Al2O3 CCC intended for the elimination of C3H8. To investigate the influence of Rh on the overall activity of the Pd catalyst, single reactions associated with the conversion of C3H8 were also studied.
La-stabilized Al2O3 was prepared by the peptizing method, dried at 105 °C overnight, and then calcined at 600 °C for 3 h in a muffle furnace. After calcination at 1000 °C for 5 h, the Al2O3 had a surface area of 138 m2/g and a pore volume of 0.47 mL/g.
The close-coupled catalysts Pd/Al2O3 and bimetallic Rh-Pd/Al2O3 were prepared by the impregnation method, employing Pd(NO3)3 and RhCl3 in the aqueous precursor solutions. The Pd and Rh contents of the finished products were 2.5 and 0.12 g/L, respectively. Deionized water was added to these materials to form homogeneous slurries, and the resulting slurry was spread over honeycomb cordierite (2.5 cm3, Corning, China). Excess slurry was blown off using compressed air. These materials were dried at 120 °C and calcined at 1000 °C for 5 h in a muffle furnace, generating the final monolithic catalysts.
The textural properties of the support materials were assessed using a Quantachrome SI instrument. To remove surface impurities, the samples were first degassed under vacuum at 300 °C for 3 h prior to the measurements.
The overall activity for C3H8 and the individual reactions of C3H8 were evaluated in a multiple fixed-bed continuum flow micro-reactor, by passing a gas mixture similar to gasoline engine exhaust through the reactor. These gases were regulated with mass flow controllers before entering a blender. The simulated exhaust gas used to assess C3H8 activity contained a mixture of O2 (adjustable), CO (0.6%), C3H8 (0.06%), NO (0.06%), CO2 (12%), H2O (10%), and N2 (balance). The gas space velocity was 40000 h−1. The isolated reactions of interest consisted of direct oxidation, reduction, and steam reforming of C3H8. The C3H8 concentration was analyzed online by a five-component analyzer (FGA-4100, Foshan Analytical Instrument Co., Ltd., China) before and after the simulated gas was passed through the micro-reactor.
H2-TPR measurements were performed in a quartz tubular micro-reactor. Each sample (100 mg, 40-60 mesh) was pretreated in N2 (20 mL/min) from room temperature to 400 °C, maintained at this temperature for 60 min, and then cooled to room temperature under N2 (20 mL/min). The reduction was carried out under a flow of a 5% H2-95% N2 (v/v) gas mixture (20 mL/min) between room temperature and 900 °C at a heating rate of 8 °C/min. The consumption of H2 was assessed by thermal conductivity detector (TCD).
X-ray photoelectron spectroscopy (XPS) data were obtained using a spectrometer (XSAM-800, Kratos Co) with Mg Kα radiation under ultra-high vacuum. XPS measurements were conducted with a resolution of 0.9 eV and a step size of 0.05 eV. The X-ray source was powered at 13 kV and 20 mA. The electron binding energy was referenced to the C 1s line of adventitious carbon at 284.8 eV to correct for the effects of charging on the XPS spectra.
The degree of Pd dispersion was determined by CO chemisorption measurements. In these tests, catalyst samples of approximately 200 mg were placed in a U-shaped quartz tube, reduced under H2 (20 mL/min) at 500 °C for 60 min, and then purged with pure He (30 mL/min). After cooling to room temperature, pulses of CO were passed through the tube, and the level of adsorption was assessed by TCD.
Figure 1 shows the total catalytic activity for the simulated C3H8-based exhaust gas over both the Pd/Al2O3 and Rh-Pd/Al2O3 catalysts. The C3H8 conversions over both catalysts increased with increasing temperature, and complete conversion was rapidly achieved. The T50 obtained with Pd/Al2O3 and Rh-Pd/Al2O3 were 356 and 333 °C, respectively, meaning that T50 decreased by 23 °C when Rh was included in the catalyst. A lower T50 indicates superior low-temperature activity of the catalyst [22]. The complete conversion temperature (T90) for the Rh-Pd/Al2O3 catalyst was 361 °C, while that of Pd/Al2O3 catalyst was 379 °C. In contrast, the C3H8 conversion over the Rh-treated catalyst was already 100% at 380 °C. In summary, the addition of Rh as a promoter to the Pd-based catalyst had a positive effect on the catalytic performance, especially with regard to the low-temperature activity. This improvement may be attributed to the ability of Rh to facilitate the start-up process [21].
The reducibility of a catalyst plays an important role in its catalytic activity. A lower reduction temperature as well as a larger area of the active component reduction peaks are associated with improved catalytic activity because these peak areas correspond to the adsorbed hydrogen consumption and thus increased areas indicate a greater amount of adsorption [23]. The TPR profiles of the Pd/Al2O3 and Rh-Pd/Al2O3 catalysts are shown in Fig. 2. Due to the low levels of RhOx species in the Rh-Pd/Al2O3 catalyst, no peaks associated with the reduction of RhOx species, expected in the vicinity of 180 °C [18], are observed. The negative α peak at about 68 °C seen solely in the profile of the Pd-only catalyst corresponds to the decomposition of palladium hydrides [7, 24]. Metallic Pd can dissociate H2 molecules into H atoms that subsequently generate large crystallites of palladium hydrides at room temperature, although this effect is greatly suppressed if the Pd is highly dispersed [25]. These results indicate that, compared with the Rh-Pd/Al2O3 catalyst, the Pd/Al2O3 contained a grea ter amount of metallic Pd, which has much lower activity for C3H8 conversion than PdOx. The β peak is attributed to the reduction of highly dispersed PdOx species, while the χ peak is ascribed to the reduction of stable PdOx species having strong interactions with the support [26]. No χ peak is evident in the profile of the Pd/Al2O3 catalyst, while a χ peak appears at approximately 90 °C in the Rh-treated catalyst, indicating that the introduction of Rh promoted interactions between the active components and the support. Other researchers have suggested that one effect of Rh doping may be the stabilization of Pd in an oxidized form, and that the re-oxidation of Pd to PdOx species may also be accelerated due to Rh doping [21]. The reduction temperatures of the β peaks in the Pd/Al2O3 and Rh-Pd/Al2O3 profiles are 85 and 79 °C, respectively. From Fig. 2 the peak area of the Rh-Pd/Al2O3 catalyst is larger than that of the Pd/Al2O3 catalyst, meaning that the addition of Rh into the Pd/Al2O3 catalyst both increased the amount of PdOx and inhibited the sintering of PdOx species, both of which favor improved catalytic activity.
Figure 3 presents the Pd 3d photoelectron peaks of the catalysts. No Rh peaks are seen on the surface of the Rh-treated catalyst due to the low concentration of Rh added. It has been reported that the binding energies of metallic Pd0 and PdO are 335.2 and 336.8 eV [27], respectively, and it is evident that the binding energies of the Pd 3d5/2 electrons of all the catalysts fall between 335.2 and 336.8 eV. This result indicates that the Pd species in these catalysts were in a partially oxidized state, Pdδ+ (0 < δ < 2), that was more active than metallic Pd [28]. Compared with the Pd-only catalyst, the Pd 3d5/2 binding energy of the Rh-treated catalyst is shifted to a higher value, likely due to the electronic effect of Rh, leading to the transferring of electrons from PdOx to RhOx. Various additives may be used to modify the active phase by changing the oxidation state and by modifying the electronic environment of the active components [29]. Considering the observed changes in the total catalytic activity, the higher oxidation state of the active PdOx species may be one of the reasons for the superior activity of the Rh-treated catalyst.
The dispersion of active metal components is an important factor affecting catalytic activity, and the dispersions of active PdOx in the Pd/Al2O3 and Rh-Pd/Al2O3 catalysts were determined by CO chemisorption. The average Pd particle size, d (nm), was approximated by 1.1/D [30], where D (%) is the dispersion of Pd. The active Pd surface area, S (m2/g), was found using the formula V0Nδ/22.4wp, where V0 is the consumption of CO (L), N is Avogadro’s constant (6.023 × 1023), δ is the atomic cross section of Pd (0.06 nm2) [31], w (g) is the mass of catalyst, and p (%) is the mass fraction of Pd.
The results, listed in Table 1, show that the Rh-Pd/Al2O3 catalyst exhibited higher Pd dispersion and a larger Pd active surface area compared with the Pd-only catalyst, and that the active Pd particle size in Rh-Pd/Al2O3 was smaller than that in the Pd/Al2O3 catalyst. Higher dispersion and smaller active component particle size both produce a higher active surface area, and thus improve the activity of the catalyst. The measured extents of Pd dispersion are in good agreement with the catalytic activity, and the lower dispersion of Pd in the Pd-only catalyst is one reason for its poor activity in this study. These data indicate that the addition of Rh to the Pd/Al2O3 catalyst increased the stability and dispersion of PdOx species.
To fully investigate the effects of Rh on the overall catalytic activity of the Pd catalyst, specific reactions of C3H8 were assessed and compared at different temperatures. The main single reactions relevant to the elimination of HCs over a CCC consist of direct oxidation (HC + O2 → CO2 + H2O), reduction (NOx + HC → N2 + CO2 + H2O), and steam reforming (SR, HC + H2O → CO + H2). The oxidation reaction of C3H8 was studied in a simulated exhaust gas consisting of a mixture of O2 (adjustable), C3H8 (0.06%), and N2 (balance).
Figure 4 shows the extent of C3H8 conversion by direct oxidation over the two catalysts, as well as the C3H8 conversion achieved upon the injection of NO and/or H2O. The concurrent changes in T50 and T90 are summarized in Table 2. C3H8 conversion by direct oxidation did not increase as rapidly as the overall reaction with increasing temperature: the ΔT values (ΔT = T90 - T50) obtained for the Pd/Al2O3 and Rh-Pd/Al2O3 catalysts were 46 and 48 °C, respectively. Shinjoh et al. [32] suggested that either C3H8 or intermediate species (such as partial oxidation products) strongly adsorbed on the catalyst surface can result in self-inhibition by limiting the adsorption of oxygen. However, the Rh-Pd/Al2O3 catalyst still presented better activity than the Pd/Al2O3. The T50 of the Pd/Al2O3 and Rh-Pd/Al2O3 catalysts were 404 and 385 °C, respectively, indicating the superior low-temperature activity of the Rh-treated catalyst. The T90 of the Pd/Al2O3 and Rh-Pd/Al2O3 catalysts were 450 and 433 °C, respectively. In addition, the conversion of C3H8 over the Rh-treated catalyst was already 100% at 450 °C, while at 450 °C the Pd/Al2O3 catalyst was only 90%. With regard to the oxidation reaction, therefore, the Rh-Pd/Al2O3 catalyst presented superior catalytic activity and low-temperature properties compar ed to the Pd/Al2O3 catalyst, because the addition of Rh increased the oxidation state of the PdOx species and/or accelerated the re-oxidation of Pd to PdOx [21], both of which are helpful to the adsorption of O2.
To investigate the effect of NO on the direct oxidation reaction of C3H8, conversions were assessed in a C3H8/O2/NO/N2 atmosphere. Figure 4 shows that the effect of NO on the oxidation reaction over both the Pd/Al2O3 and Rh-Pd/Al2O3 catalysts is negative; the T50 of the Pd/Al2O3 catalyst under a C3H8/O2/NO/N2 atmosphere was 425 °C, and thus was increased by 21 °C, while the conversion at 450 °C was only about 70% with the inclusion of NO. However, from Fig. 4(b) the T50 and T90 of the Rh-Pd/Al2O3 catalyst under the same conditions were 396 and 440 °C, respectively, and thus increased by only 11 and 7 °C subsequent to the addition of NO. These results indicate that the presence of NO inhibits the oxidation of C3H8 due to the competition between NO and O2, and that the inhibiting effect of NO on the Rh-treated catalyst was less than that on the Pd-only catalyst, which may result from the superior activity of Rh for NOx.
Figure 4 summarizes the positive effect of H2O on the oxidation reaction over both Pd/Al2O3 and Rh-Pd/Al2O3 catalysts. The T50 and T90 values of both catalysts decreased while the conversions at similar temperatures were higher with the addition of H2O. The Rh-treated catalyst also showed superior low-temperature activity. The T50 of the Pd/Al2O3 and Rh-Pd/Al2O3 catalysts were 361 and 351 °C, respectively. After H2O was injected, the T50 and T90 of the Pd/Al2O3 catalyst decreased by 43 and 56 °C, respectively, while the T50 and T90 of the Rh-Pd/Al2O3 catalyst decreased by 34 and 56 °C. The presence of H2O enhanced the direct oxidation of C3H8 by weakening the adsorption of C3H8 or its intermediates, leading to lower complete conversion temperatures. The SR reaction also takes place under these conditions.
The C3H8 conversions before and after the injection of NO and H2O to the oxidation reaction environment are compared in Fig. 4. Compared to the C3H8 conversions obtained from the direct oxidation reaction, the conversions increased at similar temperatures with the addition of NO and H2O due to the positive effect of H2O in terms of weakening the adsorption of C3H8 or its intermediates. The T50 and T90 of the Pd/Al2O3 catalyst decreased by 26 and 53 °C after NO and H2O were injected, and the T50 and T90 of the Rh-Pd/Al2O3 catalyst decreased from 385 and 433 °C to 368 and 379 °C, respectively. Under the same conditions, the Rh-treated catalyst thus exhibited better activity than the Pd-only catalyst. The positive effect of H2O on the oxidation reaction plays an important role at higher temperatures and results in complete conversion being achieved rapidly. Comparing the C3H8 conversions in C3H8/O2/H2O/N2 and C3H8/O2/NO/H2O/N2 atmospheres, we conclude that, at similar temperature below 380 °C, the addition of NO dramatically decreases conversion due to the inhibiting effect of NO.
In summation, the presence of Rh increases the promotional effect of H2O and also inhibits the negative effects of NO on the oxidation reaction. The catalytic activity of the Rh-treated catalyst under different atmospheres remains superior to that of the Pd-only catalyst, which is in good agreement with the overall activity results.
To study the reduction reaction of C3H8, catalysts were assessed in a C3H8/NO/N2 atmosphere at different temperatures. According to Fig. 5, the reduction activity is weak compared with the direct oxidation reaction. Mannila et al. [33] have reported that the oxidation of alkenes proceeds through one or more cracking steps to give surface precursors such as CH3* and CH2*, and that the partial oxidation product CO may be produced with increasing temperatures. CO is able to strongly adsorb on active sites, and this results in a self-inhibition effect that can inhibit the reduction reaction of C3H8. The conversion over the Pd/Al2O3 catalyst at 450 °C was only about 35%, while the T50 of the Rh-Pd/Al2O3 catalyst was 431 °C and the conversion at 450 °C was about 68%, a value that is much higher than that of the Pd/Al2O3 catalyst. This result is attributed to the superior activity of Rh for NOx as well as the higher oxidation state of the PdOx species in the Rh-Pd/Al2O3 catalyst.
It is evident from Fig. 5 that the conversion of C3H8 increases at similar temperatures in the presence of O2 because O2 can react with C3H8 or its intermediates strongly adsorbed on active metal sites more easily than NO, and thus regenerate the active sites. As an example, the intermediate species CO can react with O2 below 200 °C. The conversions over the Pd/Al2O3 and Rh-Pd/Al2O3 catalysts at 450 °C reached 75% and 96%, respectively, and the T50 of the catalysts were 425 and 396 °C.
The activity increased dramatically at similar temperatures after the injection of H2O to the C3H8/NO/N2 atmosphere because H2O weakens the adsorption of reactants. The T50 of the Rh-Pd/Al2O3 catalyst decreased from 431 to 354 °C, and the conversion was already 100% at 400 °C. However, the T50 and T90 of the Pd/Al2O3 catalyst remained at 376 and 409 °C, both of which are much higher than the Rh-additive values.
Unfortunately, C3H8 conversions over the Pd/Al2O3 catalyst below 385 °C and over Rh-Pd/Al2O3 below 375 °C decrease after O2 was included in the C3H8/NO/H2O/N2 atmosphere. Keeping in mind the previous discussion of the oxidation activity of C3H8, we conclude that the low-temperature activities of both catalysts are reduced when O2 and NO are both present in the C3H8/O2/H2O/NO/N2 atmosphere.
To study the SR reaction of C3H8, the conversions at different temperatures were determined under a C3H8/H2O/N2 environment, with the results shown in Fig. 6. The T50 of the Pd/Al2O3 and Rh-Pd/Al2O3 catalysts were 374 and 363 °C, respectively (Table 2). The Rh-treated catalyst exhibited far superior low-temperature activity compared to the Pd-only catalyst. The T90 of the Pd/Al2O3 catalyst was 399 °C, while the conversion of the Rh-treated catalyst at this temperature was already 100%. The addition of Rh to the catalyst thus dramatically increases the catalytic activity for the SR reaction. The support is the principal site for water activation, and the metal sites are responsible for the activation of hydrocarbons in SR reactions [34]. In addition, H2O weakens the adsorption of C3H8 or its intermediates. With the addition of Rh, not only is the activation of C3H8 adsorbed on Pd sites enhanced, but the conversion of C3H8 due to the SR reaction increases.
It is interesting to see from Fig. 6(b) that the inclusion of NO promotes the SR reaction over the Rh-Pd/Al2O3 catalyst, especially for low-temperature activity (T50 decreases from 363 to 354 °C). However, the effect of NO on the SR reaction over the Pd/Al2O3 catalyst is negative (Fig. 6(a)): T90 increases from 399 to 409 °C. The superior activity of Rh for NOx and the higher dispersion of PdOx species may result in the promotional effect of NO on the SR reaction over the Rh-Pd/Al2O3 catalyst.
It can also be concluded from Fig. 6 that the C3H8 conversion due to the SR reaction increases dramatically with the addition of O2, such that the C3H8 conversions over these two catalysts are much higher than with O2 at similar temperatures. The T50 and T90 of the Rh-Pd/Al2O3 catalyst decreased by 12 and 15 °C, respectively, after O2 was injected, while the T50 and T90 of the Pd/Al2O3 catalyst decreased by 13 and 5 °C. Thus, in the case of these close-coupled catalysts, the oxidation and SR reactions appear to proceed synergistically.
According to Fig. 6, the low-temperature performance of both catalysts declined with the injection of NO to the C3H8/H2O/O2/N2 atmosphere. The conversion over the Pd/Al2O3 catalyst at 370 °C fell from 41.9% to 26.7% after NO was added, and the conversion over the Rh-treated catalyst at 360 °C decreased from 43.7% to 14.4%. This may be the reason for the poor overall catalytic activity for C3H8 at temperatures below T50. These results indicate that the inhibiting effect of NO plays a major role in the low-temperature activity of these catalysts, while the positive effect of O2 has a significant effect in terms of achieving complete conversion at higher temperatures.
In summary, two main factors affect the catalyst activity. Firstly, the textural properties of the support material play an important role in the dispersion of active components and also affect both mass and heat transfer [35]. The support materials for both the Pd/Al2O3 and Rh-Pd/Al2O3 catalysts were the same, so the effects of the support texture on the dispersion of noble metals are similar. Secondly, the high dispersion and stability of PdOx species directly improve the catalytic activity, and the addition of Rh increases the catalytic performance of a Pd-based catalyst by inducing superior dispersion and higher oxidation states of these PdOx species.
The addition of Rh plays an important role in the total catalytic activity for C3H8 elimination and in the performance of specific reactions of C3H8 over Pd/Al2O3 catalyst. The introduction of Rh can increase the dispersion of active PdOx species and change this electronic states of these species. Comparing the performance of single reactions between the Pd-only catalyst and the Rh-treated catalyst, we conclude that the inclusion of Rh enhances the conversion of C3H8 due to specific reactions, especially those reactions involving NO.
Acknowledgments We gratefully thank Xin Chuan (Analysis and Testing Center Chengdu Branch, Chinese Academy of Sciences) for the XPS measurements.
2015, Vol. 36 Issue (2): 229-236
PDF (561 KB)
2015, Vol. 36 Issue (2): 229-236
PDF (561 KB)