SiO2 catalysts in acetylene hydrogenation

SiO2 catalysts in acetylene hydrogenation

Applied Catalysis A: General 308 (2006) 75–81 www.elsevier.com/locate/apcata Performance of La2O3- or Nb2O5-added Pd/SiO2 catalysts in acetylene hydr...

321KB Sizes 3 Downloads 20 Views

Applied Catalysis A: General 308 (2006) 75–81 www.elsevier.com/locate/apcata

Performance of La2O3- or Nb2O5-added Pd/SiO2 catalysts in acetylene hydrogenation In Young Ahn, Woo Jae Kim, Sang Heup Moon * School of Chemical and Biological Engineering and Institute of Chemical Processes, Seoul National University, San 56-1, Shillim-dong, Kwanak-ku, Seoul 151-744, Republic of Korea Received 23 February 2006; received in revised form 6 April 2006; accepted 10 April 2006

Abstract Pd catalysts promoted by La2O3 and Nb2O5, which were selected among metal oxides showing a strong metal-support interaction (SMSI), were tested for acetylene hydrogenation and their kinetic behaviors were compared with those of Pd-only and TiO2-added catalysts. The La2O3-added catalyst showed lower acetylene conversions but a higher ethylene selectivity and slower deactivation rates than the Pd-only catalyst. The improvement in ethylene selectivity and catalyst lifetime was significant, particularly when the catalysts were reduced at a temperature of 500 8C. On the other hand, Nb2O5-added catalysts showed higher acetylene conversions as well as both improved ethylene selectivity and catalyst lifetime when compared with the Pd-only catalyst. In this case, the improved selectivity attained by reducing the catalyst at 500 8C was significant when the reaction was conducted at low temperatures, i.e. 40 8C and 50 8C instead of 60 8C. Based on the characterization of catalysts by temperature-programmed reduction (TPR), H2 chemisorption, and the temperature-programmed desorption (TPD) of ethylene, the La2O3 added to the catalyst modified the Pd surface in a manner similar to TiO2, which interacted strongly with Pd, both geometrically and electronically, after reducing the catalyst at 500 8C. La2O3 interacted with Pd more strongly and, consequently improved catalyst performance to a greater extent than TiO2. Nb2O5 interacted with Pd similar to the other two oxides but also showed activity for hydrogenation, which contributed to the beneficial performance of Nb2O5-added catalysts: a higher activity for acetylene hydrogenation than that of Pd-only catalysts; remarkably low deactivation rates compared with the other catalysts. Among the three types of catalysts studied, La2O3-added catalysts that were reduced at 500 8C showed the highest ethylene selectivity at a reaction temperature of 60 8C. On the other hand, the Nb2O5-added catalysts showed the highest activity and longest catalyst lifetime due to the additional activity of Nb oxides for hydrogenation. Consequently, the former catalysts would be advantageous for use at high temperatures and the latter at low temperatures. # 2006 Elsevier B.V. All rights reserved. Keywords: Acetylene hydrogenation; Pd catalyst; La2O3; Nb2O5; SMSI; Ethylene selectivity; Deactivation

1. Introduction The selective hydrogenation of acetylene is used in naphtha crackers in the manufacture of polymer-grade ethylene, which is required to contain less than 5–10 ppm of acetylene as an impurity [1–5]. Supported Pd catalysts are typically used in the process but they need to be further improved, particularly in terms of increasing their lifetime and ethylene selectivity [2]. Many additives have been proposed as promoters of Pd

* Corresponding author. E-mail address: [email protected] (S.H. Moon). 0926-860X/$ – see front matter # 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.apcata.2006.04.027

catalysts, which include Ag [6,7], Ni [8], Cu [9], K [10,11], Au [12], Ti [13,14], and Si [15–17]. We previously reported that Pd catalysts containing TiO2 as a promoter showed improved performance in acetylene hydrogenation, particularly when the catalysts were reduced at high temperatures, e.g. 500 8C [13,14]. The promotional effect of TiO2 was attributed to a strong interaction between the Ti species and Pd, similar to the case of the so-called ‘‘strong metal-support interaction (SMSI)’’ [20,21]. The added Ti species modified the Pd surface, both geometrically and electronically, such that the characteristic adsorption of chemicals involved in the reaction became changed. This paper reports on results of our continuing studies of Pd catalysts modified by oxides that show the SMSI phenomenon.

76

I.Y. Ahn et al. / Applied Catalysis A: General 308 (2006) 75–81

La2O3 and Nb2O5 were selected because the former is difficult to reduce, being sometimes referred to as a non-reducible oxide [18,19], while the latter is reducible at relatively low temperatures [20,21]. We compared the performance of La2O3- and Nb2O5-added Pd catalysts with that of a TiO2added one, with respect to their initial activity, ethylene selectivity, and deactivation rates. We also attempted to explain the results based on characterizing the catalysts by H2 chemisorption and the temperature-programmed desorption (TPD) of ethylene.

pulses to the He stream until the catalyst surface was saturated with ethylene, as evidenced by detecting ethylene signals of a constant intensity at the reactor outlet. The samples were then flushed with He for 20 min to remove weak adsorbed species from the surface. TPD was performed by heating the sample from room temperature to 400 8C at a rate of 10 8C/min. The effluent gas was analyzed with a mass spectrometer (VG Sensorlab).

2. Experimental

Acetylene hydrogenation was conducted in a pyrex microreactor using a 0.03 g catalyst sample. A gas mixture containing 1.02% acetylene in ethylene was used as the reactant. Catalyst samples were reduced at 300 8C or 500 8C in flowing H2 prior to use in the reaction. Reaction products were analyzed by on-line GC (HP model 5890 series II with FID) using Porapak N as a column material. The flow rate of the reactant mixture was varied from 20 ml/min to 120 ml/min (space time: 1.01  10 3 min to 1.68  10 4 min) to change the acetylene conversion at a fixed H2/acetylene ratio of 2.

2.1. Catalyst preparation A 1 wt.% Pd/SiO2 sample was prepared as a reference catalyst using Pd(NH3)4(OH)2 as a Pd precursor using an ion exchange method reported in the literature [22,23]. TiO2-, La2O3-, and Nb2O5-added Pd/SiO2 were prepared by impregnating Pd/SiO2 in a hexane solution of diisopropoxide dipivaloylmethanato titanium [Ti(O-iPr)2(DPM)2], an aqueous solution of lanthanum nitrate hydrate [La(NO3)3xH2O], and a hexane solution of tetrakis(2,2,6,6-tetramethyl-3,5-heptanedionato)niobium [Nb(C11H19O2)2], respectively. The atomic ratio of promoter/Pd was typically adjusted to 1. After drying at 110 8C overnight, all catalysts were calcined in air at 300 8C for 2 h, and then reduced in H2 at either 300 8C or 500 8C in 1 h prior to use. 2.2. Temperature-programmed reduction and H2 chemisorption A quartz reactor was loaded with a 0.3 g sample, which was then dried at 110 8C overnight and purged with Ar. The sample temperature was raised from 30 8C to 900 8C at a rate of 10 8C/ min, while 5 mol% H2/Ar was allowed to flow through the reactor at a rate of 30 ml/min. The H2 concentration at the outlet of the reactor was monitored using a thermal conductivity detector (TCD). The amounts of hydrogen chemisorbed to the catalysts were measured at 35 8C using an ASAP 2010 instrument (Micrometrics Co.). Prior to the measurements, the catalysts were reduced at 300 8C or 500 8C for 2 h in H2, and the cell was then evacuated at the same temperature for 30 min. The amounts of chemisorbed hydrogen were determined using a back-sorption method described by Benson et al. [24]. In this procedure, the sample cell was evacuated at 35 8C for 30 min after the initial isotherm measurements, and a second isotherm was subsequently obtained by introducing hydrogen into the cell. H2 isotherms were measured over the pressure range of 2–50 Torr at 35 8C. 2.3. Temperature-programmed desorption of ethylene For the ethylene-TPD experiments, 0.3 g samples of catalysts were reduced in a micro-reactor at a pre-determined temperature for 1 h, flushed with He for 1 h, and then cooled to room temperature in flowing He. Ethylene was injected as

2.4. Acetylene hydrogenation

2.5. Deactivation test and green-oil analysis The reaction conditions used for the deactivation test were more severe than those used for the activity and selectivity measurements. That is, the H2/acetylene ratio was 1 instead of 2, the temperature was 90 8C, and the reactant stream, flowing at 31 ml/min (space time: 1.08  10 3 min), contained 4.1% acetylene in ethylene. The products were analyzed with an online GC (HP model 6890 series with FID) using a capillary column (HP-AL/S). The amounts of green oil deposited on the catalyst after use for extended periods were determined by measuring the amounts of gaseous CO2 generated by oxidizing the used catalysts in a 10% O2/He stream. During the oxidation, the temperature was increased from 30 8C to 700 8C at a rate of 5 8C/min, and the effluent gas was analyzed with a mass spectrometer (VG Sensorlab). 3. Results 3.1. Interaction between added oxides and Pd 3.1.1. TPR Fig. 1 shows the TPR profiles for Pd catalysts either or not containing the metal oxides. The first peak in the range of 50– 170 8C, which corresponds to the reduction of PdO [25–27], is shifted to higher temperatures for catalysts that contain added oxides. Among the oxides, La2O3 caused the largest shift. The second peak, which appears above 200 8C, corresponds to the reduction of added oxide. Separate TPR experiments using the pure metal oxides of this study, indicated that none of the oxides were reduced at temperatures lower than 400 8C. Accordingly, the results shown in Fig. 1 suggest that the temperatures for the reduction of metal oxides are lowered by the presence of Pd, obviously due to interactions between the

I.Y. Ahn et al. / Applied Catalysis A: General 308 (2006) 75–81

Fig. 1. TPR profiles for sample catalysts: (a) Pd/SiO2; (b) Pd-1Ti/SiO2; (c) Pd1Nb/SiO2; (d) Pd-1La/SiO2; (e) Pd-3La/SiO2.

oxides and Pd. In fact, the same trend was reported by Yang et al. [27], who observed the partial reduction of La2O3 at 350 8C in the presence of Pd. The reducibility of the oxides, estimated from the position of the second peak, changes in the order, Nb2O5 > TiO2 > La2O3. Fig. 1 also shows the TPR profile of Pd-3La/SiO2, which contains La2O3 at a La/Pd atomic ratio of 3. Compared with the case of Pd-1La/SiO2, the position of the first peak is shifted to higher temperatures and the second peak is intenser. Based on the above results, it can be concluded that, in the presence of added oxides, the reduction of PdO is retarded, possibly due to an interaction between PdO and the oxides. The extent of the retardation, as estimated from the shift in the first peak, is the smallest for Pd-Nb/SiO2 and the largest for Pd-La/ SiO2, which is the same sequence as for the reducibility of the oxides. Information from the second peak indicates that the metal oxides of this study can be reduced at temperatures lower than their intrinsic reduction temperatures, which is also believed to be due to an interaction between Pd and the oxide species [18–21]. 3.1.2. H2 chemisorption One of the unique characteristics of the SMSI phenomenon is the suppression in the amounts of hydrogen chemisorbed to catalysts, which have been reduced at high temperatures, e.g. 500 8C [20,21,28]. Table 1 summarizes the uptake of H2 by Table 1 The amounts of H2 uptake by sample catalysts reduced at different temperatures

catalysts reduced at two different temperatures, 300 8C and 500 8C. In the case of the Pd-only catalyst, the uptake of hydrogen, represented by H/Pd is decreased when the reduction temperature is increased from 300 8C to 500 8C. This is due to the sintering of Pd particles during the reduction at 500 8C, which has been reported previously [13] and is confirmed by XRD analyses in this study (the results are not shown here). Oxide-modified Pd catalysts show lower H/Pd ratios than unmodified Pd catalysts because the Pd surface is partially covered with added oxides [13]. Table 1 indicates that after reduction at 500 8C, the oxide-added catalysts lose the initial H2 uptake to greater extents, by 22–33%, than the Pd-only catalyst does, by 18%. The additional loss of H2 uptake by the former catalysts is believed to arise from the interaction between the added oxides and Pd, similar to the SMSI [20,21,28]. Among the three oxides tested in this study, La2O3 appears to show the strongest interaction with Pd, as indicated by the largest decrease in H2 uptake, 33%. In fact, this trend is in agreement with that observed from the TPR result (Fig. 1), which shows that the peak representing the reduction of PdO is shifted to the largest extent for the La2O3-added catalyst. 3.2. Acetylene hydrogenation 3.2.1. Activity and ethylene selectivity Table 2 shows conversions obtained using different catalysts under the same reaction conditions. The conversion is lower for Pd/500 than for Pd/300 because Pd particles undergo sintering at 500 8C as indicated by H2 chemisorption. Pd-Ti/300 and PdLa/300 show slightly lower conversions than Pd/300 because the Pd surface is partially covered with the added oxides. The activities of the oxide-added catalysts are further lowered after reduction at 500 8C due to the SMSI-like phenomenon, which was confirmed in this study by TPR and H2 chemisorption experiments. Nevertheless, it is noteworthy that the conversions for Nb2O5-added catalysts, Pd-Nb/SiO2, are slightly higher than those for the Pd-only catalyst, regardless of the catalyst reduction temperatures used. Independent reaction tests carried out using Nb2O5/SiO2, which contained only Nb2O5 in the same amounts as for Pd-Nb/SiO2 and was reduced at 300 8C, showed about a 3–5% conversion of the feed acetylene. Accordingly, the characteristic behavior of the Nb2O5-added catalyst is Table 2 Typical acetylene conversions obtained using different sample catalysts under the same reaction conditions Catalyst a

Catalyst

H/Pd

Pd/300a Pd/500 Pd-Ti/300 Pd-Ti/500 Pd-La/300 Pd-La/500 Pd-Nb/300 Pd-Nb/500

0.45 0.37 0.18 0.13 0.24 0.16 0.22 0.17

a

For example, Pd/300 denotes Pd/SiO2 reduced at 300 8C.

77

Pd/300 Pd/500 Pd-Ti/300 Pd-Ti/500 Pd-La/300 Pd-La/500 Pd-Nb/300 Pd-Nb/500

Acetylene conversion 0.70 0.68 0.69 0.67 0.59 0.54 0.78 0.75

Reaction condition: temperature = 60 8C, H2/acetylene = 2, total flow rate = 60 ml/min. a For example, Pd/300 denotes Pd/SiO2 reduced at 300 8C.

78

I.Y. Ahn et al. / Applied Catalysis A: General 308 (2006) 75–81

Fig. 2. Changes in the ethylene selectivity with conversion in acetylene hydrogenation for different sample catalysts (H2/acetylene = 2, reduction temperature = 300 8C or 500 8C, reaction temperature = 60 8C): (a) Pd/300; (b) Pd/ 500; (c) Pd-Ti/300; (d) Pd-Ti/500; (e) Pd-La/300; (f) Pd-La/500; (g) Pd-Nb/300; (h) Pd-Nb/500.

Fig. 3. Changes in the ethylene selectivity with conversion in acetylene hydrogenation for catalysts either or not containing Nb2O5 at various reaction temperatures (H2/acetylene = 2, reduction temperature = 300 8C for Pd/SiO2 or 500 8C for Pd-Nb/SiO2): (a) Pd/300/40; (b) Pd-Nb/500/40; (c) Pd/300/50; (d) Pd-Nb/500/50; (e) Pd/300/60; (f) Pd-Nb/500/60.

believed to originate from its additional hydrogenation activity contributed by the partially reduced Nb oxides [29]. Fig. 2 shows a plot of ethylene selectivity versus acetylene conversion for a reaction at 60 8C using catalyst samples reduced at either 300 8C or 500 8C. The selectivity of Pd/SiO2 is lowered when the reduction temperature is increased from 300 8C to 500 8C, due to the sintering of the Pd particles [30]. The oxide-added catalysts show a higher selectivity than the Pd-only catalyst, and the improvement in selectivity is greater when the catalysts are reduced at 500 8C instead of 300 8C. Among the three oxides, La2O3 shows the largest promotion of selectivity, followed by TiO2 and Nb2O5. It is noteworthy that Pd-Nb/SiO2, which has an additional activity for hydrogenation, does not show a significant increase in selectivity compared with the other catalysts and, furthermore, the selectivity is decreased after reduction at 500 8C instead of 300 8C. To further examine the kinetic behavior of the Nb2O5added catalyst, we measured its ethylene selectivity at different reaction temperatures. Fig. 3 indicates that selectivity improvement by added Nb2O5 is more significant at 40 8C and 50 8C, than at 60 8C. Accordingly, it can be concluded that the promotional effect of Nb2O5 is strongly dependent on the reaction temperature used.

activity of the partially reduced Nb2O5. This subject will be discussed in Section 4.

3.2.2. Deactivation rates Fig. 4 shows the deactivation characteristics of the catalysts, presented by changes in conversion normalized to the initial value, as a function of the total amounts of converted acetylene. The deactivation of Pd/300 is slow in the initial stage but is accelerated when the amounts of converted acetylene are larger than 0.04 mol. The oxide-added catalysts show a similar trend but their deactivation rates are lower than for the Pd-only catalyst. In particular, the Nb2O5-added catalyst shows extremely slow deactivation rates in all stages of the reaction. We conclude that the characteristic behavior of Nb2O5-added catalyst, either in selectivity or deactivation rates, is closely related to the additional hydrogenation

3.3. Surface properties of oxide-added catalysts 3.3.1. Ethylene-TPD The ethylene-TPD peaks shown in Fig. 5 represent different types of ethylenic species, which are adsorbed to the catalyst surface. Peak I, observed at temperatures below 100 8C, originates from p-bonded ethylene species, which are weakly adsorbed to and desorbed from Pd without decomposition [31]. Peak II, near 150 8C, is assigned to di-s-bonded ethylene species, which are desorbed dissociatively followed by the recombination of the surface hydrocarbon species to produce either ethylene or ethane. The intensities of the two main peaks are decreased in the cases of oxide-added catalysts because their Pd surfaces are partially covered with oxides. In particular, Peak II of the

Fig. 4. Deactivation of sample catalysts vs. the total amounts of converted acetylene (H2/acetylene = 1, reduction temperature = 300 8C for Pd/SiO2 or 500 8C for oxide-added Pd/SiO2, reaction temperature = 90 8C): (a) Pd/300; (b) Pd-Ti/500; (c) Pd-La/500; (d) Pd-Nb/500.

I.Y. Ahn et al. / Applied Catalysis A: General 308 (2006) 75–81

Fig. 5. TPD of ethylene from sample catalysts (m/e = 27): (a) Pd/300; (b) PdTi/300; (c) Pd-Ti/500; (d) Pd-La/300; (e) Pd-La/500; (f) Pd-Nb/300; (g) Pd-Nb/ 500.

79

hydrocarbons adsorbed to the catalyst surface or absorbed in the catalyst pores. Peak II, observed at temperatures between 300 8C and 500 8C, represents green oil on or in the vicinity of Pd, and Peak III, above 500 8C, indicates green oil produced on the support without the influence of Pd. There are two notable changes in the peak intensity and position as a result of oxide addition. One is that the overall peak intensity is decreased, which is significant for Peaks II and III although the intensity of Peak I is increased only slightly. The results indicate that the amounts of green oil, particularly the species represented by Peaks II and III are decreased on the oxide-added catalysts. The other is that the positions of three peaks are shifted to lower temperatures as the result of oxide addition. The peak shift suggests that green-oil species are more volatile, with lower molecular weights, or present as smaller particles on the catalyst surface. A similar trend was observed for TiO2-added catalysts in our previous study, as evidenced by DTGA and IR results [14,17]. 4. Discussion

oxide-added catalysts is suppressed to an insignificant level, indicating that the majority of multiply coordinated Pd sites are blocked by the added oxides [13]. In addition to changes in the peak intensity, the position of the peaks, particularly Peak I, is affected by the added oxides. That is, Peak I is shifted to lower temperatures, by 10–20 8C, when oxides are added to the catalyst. The peak shift is slightly increased when the catalysts are reduced at 500 8C instead of 300 8C. Accordingly, it can be concluded that the ethylene species, adsorbed to Pd either in the p-bonded or the di-sbonded state, are desorbed more easily from the Pd surface when the latter is modified by added oxides. 3.3.2. TPO analysis Fig. 6 shows the amounts of CO2 produced by TPO of catalysts that were used for converting the same total amounts of acetylene, 0.05 mol. According to Larsson et al. [32], who reported results similar to those found in this study, the peak below 300 8C, designated as Peak I, is due to heavy

Fig. 6. TPO of carbonaceous species formed on used catalysts (m/e = 44): (a) Pd/300; (b) Pd-La/500; (c) Pd-Nb/500.

The interaction between Pd and added oxides was evidenced by the results of H2 uptake (Table 1) and TPR experiments (Fig. 1). Among the three oxides tested in this study, La2O3 interacts the strongest with Pd, as indicated by the greatest suppression in H2 uptake after reduction at 500 8C and the largest shift in the peak for PdO reduction in the case of the La2O3-added catalyst. Unlike La2O3, Nb2O5 interacts relatively weakly with Pd. However, Nb2O5 is the most easily reduced among the three oxides, as indicated by the appearance of a second TPR peak at the lowest temperatures and with the highest intensity (Fig. 1). The role of added La2O3 in modifying the performance of Pd catalysts in acetylene hydrogenation is similar to that of TiO2 [13,14]. That is, Pd catalysts modified by La2O3 show a lower activity (Table 2) but an enhanced ethylene selectivity and decreased deactivation rates (Figs. 2 and 4) compared with the case of unmodified Pd catalysts. The activity is lowered because the Pd surface is partially covered with La species, particularly after reduction at 500 8C, as indicated by H2 uptake (Table 1). The improved ethylene selectivity and deactivation behavior of the La2O3-added catalysts also arises from the decoration of the Pd surface with partially-reduced oxides, resulting in a decrease in the strength of ethylene adsorption (Fig. 5) and in the amounts of green oil deposited (Fig. 6) on the Pd surface. The same mechanism is attributed to the improved behavior of TiO2-added catalysts in acetylene hydrogenation [13,14]. Among all the catalysts of this study, Pd-La2O3/SiO2, reduced at 500 8C (Pd-La/500), shows the highest selectivity. This can be explained based on the strength of the interaction between Pd and added La2O3, as mentioned above. Nevertheless, a strong interaction between La2O3 and Pd is somewhat unusual considering the fact that La2O3 is the most difficult to reduce among three oxides of this study [18,19] and that the SMSI phenomenon is typically observed between noble metals and reducible metal oxides [20,21,28]. As one of the factors responsible for the interaction between La2O3 and Pd, we

80

I.Y. Ahn et al. / Applied Catalysis A: General 308 (2006) 75–81

propose the low acidity of La2O3 [33]. In fact, Yang et al. [27] observed a strong interaction between Pd and La2O3 in their study of La2O3-added Pd/Al2O3. Unlike catalysts containing La2O3 and TiO2, Nb2O5-added catalysts show an activity higher than that of Pd-only catalysts, even after reduction at 500 8C (Table 2). Ethylene selectivity is improved when Pd catalysts are modified by Nb2O5. However, the selectivity, measured at 60 8C, is decreased when Nb2O5added catalysts are reduced at 500 8C instead of 300 8C (Fig. 2), which is different from the behavior of catalysts containing other oxides. Nevertheless, Pd-Nb2O5/SiO2 reduced at 500 8C (Pd-Nb/500) shows a higher selectivity than for Pd-Nb/300 when the reaction is conducted at lower temperatures, below 50 8C (Fig. 3). Among all catalysts of this study, Pd-Nb/500 shows the slowest deactivation rates (Fig. 4), which is supported by the presence of smaller amounts and relatively volatile species of green oil on the catalyst (Fig. 6). The above performance of Nb2O5-modified catalysts in acetylene hydrogenation can be explained based on the property of Nb2O5 to facilitate the absorption and desorption of hydrogen. For example, Barkhordarian et al. [34,35] reported that Nb2O5 is an efficient catalyst for hydrogen absorption by nano-crystalline Mg and attributed the property to the presence of multiple valance states in the oxides. Borgschlte et al. [36,37] also observed the catalytic activity of Nb2O5 in the dissociation and absorption of hydrogen and attributed the activity to surface vacancies in the oxide. Although uncertainty remains concerning the active species responsible for the dissociation and absorption of hydrogen, it is evident that Nb oxides, including Nb2O5 and sub-oxides, are efficient catalysts for the activation of hydrogen. In this respect, it is not surprising that the Nb2O5-added catalysts of this study exhibit an activity higher than that of Pd catalysts. However, the excessive supply of hydrogen by the Nb oxides lowers the ethylene selectivity of catalysts, as observed for Pd-Nb/500 in the reaction at 60 8C (Fig. 2). The selectivity decrease due to an excess hydrogen supply is relieved when the reaction proceeds at lower temperatures, 40 8C and 50 8C, in which case the rates of hydrogenation would be expected to be decreased to greater extents than the rates of ethylene desorption [36,37]. This is so because hydrogenation is a chemical process while desorption is a physical one. As is well known, the former has higher activation energy, accordingly is more strongly dependent on temperature, than the latter. The additional hydrogenation activity of Nb oxides improves the catalyst lifetime (Fig. 4) by retarding surface polymerization leading to the formation of green oil and by producing relatively volatile carbon species that are included in the green oil. Based on the above findings concerning the performance of Pd catalysts modified by either La2O3 or Nb2O5, we can summarize the merits as well as limitations of the oxide-added catalysts as follows. La2O3-added catalysts would be advantageous for use at high temperatures, e.g. 60 8C, because they show the highest ethylene selectivity among the three types of catalysts of this study. The slight decrease in their activity, caused by added La2O3, can be compensated for by running the reaction at elevated temperatures. On the other hand, Nb2O5 has

the advantage of additional hydrogenation activity, which is responsible for enhanced acetylene conversions and retarded deactivation rates. Nb2O5 has another advantage as a selectivity promoter, particularly when the catalysts are used at relatively low temperatures. 5. Conclusions Pd catalysts containing added La2O3 or Nb2O5 exhibited improved ethylene selectivity and catalyst lifetime in acetylene hydrogenation, as was also observed for TiO2-added catalysts previously. The improved performance of oxide-added catalysts originates from a strong interaction between Pd and the added oxides, which suppresses the chemisorption of hydrogen on Pd and facilitates the desorption of ethylene from Pd, particularly after reduction at 500 8C. Nevertheless, the strength of the interactions is different depending on the oxide and, consequently, catalyst performance is affected to different extents by individual oxides. Based on the results of reaction tests and catalyst characterization, the following conclusions can be made concerning the effects of the oxides. (1) Among the three oxides examined in this study, La2O3 showed the strongest interaction with Pd and, as a result, La2O3-added catalysts showed the highest ethylene selectivity when the reaction proceeded at 60 8C. The La2O3-added catalyst was deactivated more slowly than the TiO2-added catalyst. (2) Nb2O5-added catalysts showed a higher activity than the Pd-only catalyst, unlike the cases of the La2O3- and TiO2added catalysts, because partially reduced Nb oxides had additional hydrogenation activity. The hydrogenation activity of the Nb oxides also contributed to the production of relatively volatile green-oil species and eventually to remarkably low rates of catalyst deactivation. (3) The selectivity of the Nb2O5-added catalyst was improved after reduction at 500 8C, but this was observed only when the reaction was carried out at low temperatures, below 50 8C. That is, the selectivity did not exceed that of 300 8C-reduced catalyst, for a reaction at 60 8C. This is clearly different from the case of the other catalysts, in which the selectivity was improved after reduction at 500 8C, even when the reactions were conducted at 60 8C. This characteristic behavior of the Nb2O5-added catalyst arises because an increase in the hydrogenation activity of the catalyst lowers the selectivity at temperatures higher than 60 8C. (4) Considering the characteristic behaviors of La2O3- and Nb2O5-added catalysts, we propose that the former catalysts are advantageous for use in acetylene hydrogenation at high temperatures and the latter at low temperatures. Acknowledgements This study was supported by BASF Aktiengesellschaft, Germany, the Brain Korea 21 project and the National Research Laboratory program of Ministry of Science and Technology, Korea.

I.Y. Ahn et al. / Applied Catalysis A: General 308 (2006) 75–81

References [1] C.M. Pradier, M. Mazina, Y. Berthier, J. Oudar, J. Mol. Catal. 89 (1994) 211. [2] A. Molna´r, A. Sa´rka´ny, M. Varga, J. Mol. Catal. A 173 (2001) 185. [3] A.H. Weiss, B.S. Gambhir, R.B. La Pierre, W.K. Bell, Ind. Eng. Chem. Proc. Des. Dev. 16 (1977) 352. [4] H.R. Aduriz, P. Bodnariuk, M. Dennehy, C.E. Gigola, Appl. Catal. 58 (1990) 227. [5] C.E. Gigola, H.R. Aduriz, P. Bodnariuk, Appl. Catal. 27 (1986) 133. [6] D.C. Huang, K.H. Chang, W.F. Pong, P.K. Tseng, K.J. Hung, W.F. Hung, Catal. Lett. 53 (1998) 155. [7] Q. Zhang, J. Li, X. Liu, Q. Zhu, Appl. Catal. A 197 (2000) 221. [8] P. Miegge, J.L. Rousset, B. Tardy, J. Massardier, J.C. Bertolini, J. Catal. 149 (1994) 404. [9] S. Leviness, V. Nair, A.H. Weiss, Z. Schay, L. Guczi, J. Mol. Catal. 25 (1984) 131. [10] Y.H. Park, G.L. Price, Ind. Eng. Chem. Res. 31 (1992) 469. [11] W.J. Kim, J.H. Kang, I.Y. Ahn, S.H. Moon, Appl. Catal. A 268 (2004) 77. [12] A. Sa´rka´ny, A. Horva´th, A. Beck, Appl. Catal. A 229 (2002) 117. [13] J.H. Kang, E.W. Shin, W.J. Kim, J.D. Park, S.H. Moon, J. Catal. 208 (2002) 310. [14] W.J. Kim, J.H. Kang, I.Y. Ahn, S.H. Moon, Appl. Catal. A 226 (2004) 226. [15] E.W. Shin, C.H. Choi, K.S. Chang, Y.H. Na, S.H. Moon, Catal. Today 44 (1998) 137. [16] E.W. Shin, J.H. Kang, W.J. Kim, J.D. Park, S.H. Moon, Appl. Catal. A 223 (2002) 161. [17] W.J. Kim, E.W. Shin, J.H. Kang, S.H. Moon, Appl. Catal. A 251 (2003) 305.

[18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37]

81

T.H. Fleisch, R.F. Hicks, A.T. Bell, J. Catal. 87 (1984) 398. R.F. Hicks, Q. Yen, A.T. Bell, J. Catal. 89 (1984) 498. S.J. Tauster, S.C. Fung, J. Catal. 55 (1978) 29. S.J. Tauster, S.C. Fung, R.T.K. Baker, J.A. Horsley, Science 211 (1981) 1121. L.F. Liotta, G. Deganello, P. Delichere, C. Leclercq, G.A. Martin, J. Catal. 164 (1996) 334. L.F. Liotta, G..A. Martin, G. Deganello, J. Catal. 164 (1996) 322. J.E. Benson, H.S. Hwang, M. Boudart, J. Catal. 30 (1973) 146. N.S. Figoli, P.C. L’Argentiere, A. Arcoya, X.L. Seoane, J. Catal. 155 (1995) 95. X.L. Seoane, N.S. Figoli, P.C. L’Argentiere, J.A. Gonzalez, A. Arcoya, Catal. Lett. 47 (1997) 213. C. Yang, J. Ren, Y. Sun, Catal. Lett. 84 (2002) 123. S.J. Tauster, S.C. Fung, R.L. Garten, J. Am. Chem. Soc. 100 (1978) 170. R.B. Moyes, P.B. Wells, J. Grant, N.Y. Salman, Appl. Catal. 229 (2002) 251. A. Sa´rka´ny, A.H. Weiss, L. Guczi, J. Catal. 98 (1986) 550. E.M. Stuve, R.J. Madix, J. Phys. Chem. 89 (1985) 105. M. Larsson, J. Jansson, S. Asplund, J. Catal. 178 (1998) 49. G. Busca, Phys. Chem. Chem. Phys. 1 (1999) 723. G. Barkhordarian, T. Klasen, R. Bormann, Scripta Mater. 49 (2003) 213. G. Barkhordarian, T. Klasen, R. Bormann, J. Alloys Compd. 364 (2004) 242. A. Borgschulte, J.H. Rector, B. Dam, R. Griessen, A. Zuttel, J. Catal. 235 (2005) 353. A. Borgschulte, W. Lohstroh, R.J. Westerwaal, H. Schreuders, J.H. Rector, B. Dam, R. Griessen, J. Alloys Compd. 404–406 (2005) 699.