Mechanistic Study of 1, 3-Butadiene Formation in Acetylene

Dec 31, 2013 - Mechanistic Study of 1,3-Butadiene Formation in Acetylene Hydrogenation over the Pd-Based Catalysts Using Density Functional Calculatio...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/JPCC

Mechanistic Study of 1,3-Butadiene Formation in Acetylene Hydrogenation over the Pd-Based Catalysts Using Density Functional Calculations Bo Yang,† Robbie Burch,† Christopher Hardacre,*,† P. Hu,*,† and Philip Hughes‡ †

CenTACat, School of Chemistry and Chemical Engineering, The Queen’s University of Belfast, Belfast BT9 5AG, United Kingdom Johnson Matthey Catalysts, P.O. Box 1, Billingham, Teesside TS23 1LB, United Kingdom



S Supporting Information *

ABSTRACT: Green oil, which leads to the deactivation of the catalysts used for the selective hydrogenation of acetylene, has long been observed but its formation mechanism is not fully understood. In this work, the formation of 1,3-butadiene, known to be the precursor of green oil, on both Pd(111) and Pd(211) surfaces is examined using density functional theory calculations. The pathways containing C2 + C2 coupling reactions as well as the corresponding hydrogenation reactions are studied in detail. Three pathways for 1,3-butadiene production, namely coupling plus hydrogenation and further hydrogenation, hydrogenation plus coupling plus hydrogenation, and a two step hydrogenation followed by coupling, are determined. By comparing the effective barriers, we identify the favored pathway on both surfaces. A general understanding toward the deactivation process of the industrial catalysts is also provided. In addition, the effects of the formation of subsurface carbon atoms as well as the Ag alloying on the 1,3-butadiene formation on Pd-based catalysts are also investigated and compared with experimental results.

1. INTRODUCTION The selective hydrogenation of acetylene to ethylene is an important process in the petrochemical industry, because acetylene is the major impurity in ethylene feeds produced from naphtha cracking processes. Acetylene is also a significant poisoning species to the catalyst used for downstream ethylene polymerization.1,2 As has been addressed in our previous work and the work from other groups, the major issues in this process are the overhydrogenation of ethylene to ethane and the oligomerization of acetylene to higher molecular weight species, such as butadiene and polymers, known as green oil.3,4 The overhydrogenation results in yield loss and the accumulation of oligomers on the catalyst surface leads to accelerated catalyst deactivation. Extensive experimental and theoretical work has been carried out to improve the selectivity of acetylene hydrogenation catalysts to produce only ethylene. The critical factors determining the selectivity have also been identified, namely the adsorption energy and the hydrogenation barrier of ethylene. Because of the sensitivity of the ethylene polymerization catalyst to acetylene and the economic considerations, the selectivity/activity requirements are very strict and, although several catalysts have been developed and commercialized, new catalyst formulations are still required to provide maximum ethylene product at low temperature.5−11 In addition to the selectivity/activity of the catalyst, because of the high throughput of gas and the severe cost implications of downtime in refinery processes, these catalysts must be robust and typically must maintain their activity for 5+ years. © 2013 American Chemical Society

However, much less attention has been paid to the deactivation of the catalyst compared with the selectivity in acetylene hydrogenation processes, and the mechanism of the green oil formation on these catalysts is still unknown despite its importance in catalyst deactivation.12−24 In fact, the majority of the ethylene purification industries use the tail-end process where the hydrogen is present in stoichiometric amounts with respect to acetylene, that is, H2/acetylene is around 1.1 In this case, the coverage of acetylene, or its hydrogenation intermediates, may be very high on the catalyst surface and thus it is likely to increase the probability of green oil formation. This has been observed experimentally, where the concentration of oligomeric species increases with the decreasing H2/acetylene ratio.13 As well as deactivating the catalyst via surface adsorption, green oil can block the catalyst pores, reducing mass transfer processes such as hydrogen diffusion.12,13,15 Asplund and co-workers proposed that 1,3-butadiene may be the precursor of green oil formation.12,13,15 This has also been observed by Moon and co-workers22 and they also found that the deactivation of Pd catalyst proceeds in three stages.24 Initially, the activity of the catalyst is only slightly decreased due to the formation of light green oil (stage I). Thereafter, further polymerization takes place to form several heavier species causing increased surface deactivation (stage II). Finally, the Received: September 3, 2013 Revised: December 28, 2013 Published: December 31, 2013 1560

dx.doi.org/10.1021/jp408807c | J. Phys. Chem. C 2014, 118, 1560−1567

The Journal of Physical Chemistry C

Article

where Etotal is the energy of the system after adsorption, Eg is the energy of the gas phase adsorbent, and Eslab is the energy of slab.

heavy green oil condenses in the pores of the catalyst, leading to pore blocking and reducing the hydrogen transfer which exacerbates the polymerization and reduces the activity of the catalyst dramatically (stage III). The mechanism for the coupling of propyne and acetylene on Cu, Ni, and Pd surfaces was studied by Bridier et al.25,26 and Garcia-Mota et al.27 previously using density functional theory (DFT) calculations. They found that, on Cu(111), the coupling of propyne was slightly endothermic by 0.28 eV, while it was endothermic by 0.83 eV on Ni(111). Consequently, the lowest barrier for propyne coupling was 1.01 eV on Cu while for Ni(111) it was 1.49 eV. On the clean Pd(111) surface, the barrier for acetylene coupling was calculated to be 1.38 eV. However, the whole reaction pathway of the precursor formation was not considered. Moreover, the effect of surface structure on the formation of the precursors was also not examined. In this work, we investigate the mechanism of 1,3-butadiene formation with several pathways on both Pd(111) and Pd(211) surfaces, which are representative of typical flat and defect Pd surfaces, in order to gain an insight to the effect of catalyst surface structures on 1,3-butadiene formation. Because much evidence has shown that the Pd catalyst might be Pd-carbide phase during the reaction, the influence of the subsurface carbon formation are also examined. In addition to the “three stages” mechanism for the green oil formation on Pd catalyst surface proposed by Moon et al., they also found that the drastic decrease of the activity was not observed during period III on Pd−Ag catalyst, indicating that the lifetime of the Pd catalyst was extended by Ag promotion.24 Therefore, the influence of alloying with Ag on the 1,3-butadiene formation pathways over Pd surface has been studied.

3. RESULTS AND DISCUSSION 3.1. Schemes of 1,3-Butadiene Formation. A C2 + C2 coupling mechanism was examined to investigate the formation of 1,3-butadiene on the surface of Pd catalyst, as shown in Scheme 1. Propagation of this mechanism may also be applied Scheme 1. Three Possible Pathways Containing Coupling Reactions and Hydrogenation Reactions to Produce 1,3Butadiene on Pd Surface

to study the reactions to form heavier green oil as it has been reported that the green oil contains an even number of carbon atoms.18 Our recent work showed that the Horiuti−Polanyi mechanism,38−41 which describes a sequential addition of hydrogen atoms to the substrate in heterogeneous catalytic hydrogenation reactions, are not favored in some cases when the catalysts used are Ag or Au, but it is still the preferred mechanism on many transition metal (e.g., Pd and Pt) catalyst surfaces.42 In the current work, there are three species on the Pd surface, which are important in the formation of 1,3butadiene: H, C2H2, and C2H3. Three pathways involving these species can lead to 1,3-butadiene involving three coupling reactions, that is, C2H2 + C2H2, C2H2 + C2H3 and C2H3 +C2H3. The three possible pathways shown in Scheme 1 can be generalized as coupling → hydrogenation → hydrogenation; hydrogenation → coupling → hydrogenation; and hydrogenation → hydrogenation → coupling. It should be noted that there are some competing steps not shown in Scheme 1, which are the hydrogenation of the different carbon atoms in C4H4 and C 4 H 5 . For example, for the hydrogenation of CHCHCHCH* (* denotes the surface adsorption state), all four carbon atoms might be attacked by hydrogen, depending on the symmetry of its surface adsorption configurations, forming CH2CHCHCH* or CHCH2CHCH*. Scheme 1 is used to investigate the formation of 1,3-butadiene on all the surfaces studied in the current work. 3.2. Pathways of 1,3-Butadiene Formation on Pd Surfaces. 3.2.1. 1,3-Butadiene Formation on Pd(111). In the current work, we consider the diffusion of acetylene on the Pd surface as an equilibrated process and will not influence the reaction rate of the whole process, since experimental results have shown that for the adsorbed acetylene species on Pd(111) surface, the diffusion occurs at temperatures as low as 70 K.43 The calculated highest diffusion barrier of acetylene is only 0.36 eV,43 which is much lower than the hydrogenation barrier of the adsorbed acetylene (0.96 eV) reported in our previous work.4 The corresponding transition state structures of coupling reactions, that is, C2H2 + C2H2, C2H2 + C2H3, and C2H3 + C2H3, are shown in Figure 1. In addition, the adsorption geometries of the C4H4, C4H5, and C4H6 on Pd(111) are illustrated in this figure. The adsorption energy of C4H6 on Pd(111) was calculated to be −1.64 eV. We should mention

2. COMPUTATIONAL DETAILS In this work, the DFT calculations were performed within a periodic slab model using the plane wave based VASP package.28−31 The generalized gradient approximation (GGA) was used with the exchange-correlation functional PW91.32 The atomic cores and electrons were described with the projector augmented wave (PAW) method.33,34 A four-layer slab with a (4 × 4) surface supercell was employed for the clean Pd(111) surface, the Pd(111) surface with subsurface carbon and surface Pd−Ag alloy, the two uppermost layers were relaxed with the adsorbates while the bottom two layers were fixed during the optimization. The calculations for Pd(211) used 12 layers (1 × 4) slabs with the upmost 6 layers relaxed with 4 Pd atoms in the direction parallel to the step edge. We tested the adsorption energy of butadiene over Pd(211) surface with a larger unit cell (2 × 8) and found that the difference is within 0.01 eV, indicating that the size of the unit cell used in the current work is sufficient to describe the processes involving butadiene. A vacuum cutoff of more than 12 Å was set to minimize the interactions between slabs. A 2 × 2 × 1 k-point grid was used in the calculations for the Pd(111) surface as well as the carbide and alloyed surfaces and that used for the Pd(211) surface was 4 × 2 × 1. We used an energy cutoff of 500 eV and the force threshold in structural optimization was set to 0.05 eV/Å. The transition states were located with a constrained minimization method.35−37 The adsorption and binding energies were calculated as follows Ead = Etotal − (Eg + Eslab) 1561

dx.doi.org/10.1021/jp408807c | J. Phys. Chem. C 2014, 118, 1560−1567

The Journal of Physical Chemistry C

Article

Table 1. Reaction Barriers (Ea) of Hydrogenation Reactions of C4H4 and C4H5 on Pd(111), Pd(211), Pd(111)-0.25C and Pd-0.25Ag/Pd(111) Surfacesa C4H4

C4H5

C1 C2 C3 C4 C1 C2 C3 C4

Pd(111)

Pd(211)

Pd(111)-0.25C

Pd-0.25Ag/Pd(111)

0.67 0.91

0.56 0.90 0.79 0.60 0.91 1.34 1.15 0.76

0.51 1.00 0.78 0.53 0.75 1.22 1.01 0.55

0.78 1.60

1.04 1.24 1.13 0.89

1.13 1.35 1.30 0.79

a

The definitions of C1, C2, C3, and C4 are shown in Figures 1, 3, 4 and 5. The unit is eV.

hydrogenation of the terminal CH (carbon 4) in C4H5 is favored and thus C4H6 (1,3-butadiene) is formed. Therefore, it is evident that 1,3-butadiene can be readily formed from C4H4 and C4H6 hydrogenation. The corresponding reaction barriers of the elementary steps in Scheme 1, taking place on Pd(111), are listed in Table 2 Table 2. Reaction Barriers (Ea) and the C−C (Coupling Reactions) and C−H (Hydrogenation Reactions) Distances at the Transition States on Pd(111), Pd(211), Pd(111)0.25C, and Pd-0.25Ag/Pd(111) Surfaces, According to Scheme 1 Figure 1. Top: Energy profiles of the three pathways shown in Scheme 1 from 2C2H2 + 2H to C4H6 on Pd(111). TS1, TS2 and TS3 are the same as those in Table 2. Bottom: Transition state structures of coupling reactions of C2H2+C2H2, C2H2+C2H3 and C2H3+C2H3(from left to right) on Pd(111). The adsorption structures of the C4H4, C4H5 and C4H6 produced, are also shown. Different carbon atoms are marked as 1, 2, 3, and 4. Blue, gray, and white balls are Pd, C and H, respectively. This notation is used throughout the paper.

pathway I

TS1

TS2

TS3

C2H2 + C2H2

C4H4 + H

C4H5 + H

Ea/eV Pd(111) Pd(211) Pd(111)-0.25C Pd-0.25Ag/Pd(111) pathway II

that a wide range of different transitions states and adsorption geometries were calculated and the structures shown in Figure 1(as well as those structures shown in the figures, hereafter) are the most stable adsorption geometries and the transition state structures with the lowest barriers. One can see from Figure 1 that, at the transition states of the coupling reactions, one C2 species (either C2H2 or C2H3) adsorbs at the fcc site while the other C2 species sits at an adjoining hcp site with the two sites sharing one Pd atom. The arrangement is thought to facilitate the coupling reactions. One can also see from Figure 1 that C4H4, C4H5 and C4H6 interact with 6, 5, and 4 Pd atoms, respectively, due to the increasing saturation of the terminal carbon atoms. It should be mentioned that the favored adsorption geometry of C4H6 shown in Figure 1, as well as its adsorption energy on Pd(111), is in good agreement with previously reported DFT results.44 Because both the pathways I and II include the hydrogenation steps of C4H4 and C4H5 and the product distribution of these steps is diverse, the hydrogenation on each different carbon atom in C4H4 and C4H5 was studied on the Pd(111) surface. The corresponding hydrogenation barriers are listed in Table 1. Because C4H4 adsorbs symmetrically on Pd(111) (see Figure 1), we only located the transition states for the hydrogenation of the carbon atoms marked as 1 and 2 in Figure 1. The hydrogenation of C4H5 is more complex since all the carbon atoms are different. From Table 1, it is clear that the

0.97 1.86 1.10 1.10 C2H2 Ea/eV

Pd(111) Pd(211) Pd(111)-0.25C Pd-0.25Ag/Pd(111) pathway III

Pd(111) Pd(211) Pd(111)-0.25C Pd-0.25Ag/Pd(111)

0.96 0.97 1.00 0.98 C2H2

DC−C/ Ǻ

Ea/eV

DC−H/ Ǻ

1.98 1.97 1.97 1.96 +H

0.67 0.56 0.51 0.78 C2H2 +

1.85 1.84 1.92 1.83 C2H3

DC−H/ Ǻ

Ea/eV

DC−C/ Ǻ

1.67 1.64 1.66 1.66 +H

0.81 1.59 0.75 0.91 C2H2

Ea/eV 0.89 0.76 0.55 0.79 C4H5 Ea/eV

DC−H/ Ǻ 1.82 1.78 1.85 1.83 +H DC−H/ Ǻ

2.05 2.04 2.09 2.05 +H

0.89 0.76 0.55 0.79 C2H3 +

1.82 1.78 1.85 1.83 C2H3

Ea/eV

DC−H/ Ǻ

Ea/eV

DC−H/ Ǻ

Ea/eV

DC−C/ Ǻ

0.96 0.97 1.00 0.98

1.67 1.64 1.66 1.66

0.96 0.97 1.00 0.98

1.67 1.64 1.66 1.66

0.83 0.54 0.26 0.74

2.17 2.12 2.33 2.19

together with the C−C and C−H distances at the transition states. The hydrogenation barrier of C2H2 and the C−H distance at the transition state on Pd(111) were obtained from our previous work.4 It should be noted that the hydrogenation barriers of C4H4 and C4H5 listed in Table 2 are the most favored among the hydrogenation reactions of all four carbon atoms. The corresponding transition state structures of the hydrogenation of C4H4 and C4H5 with the lowest barriers are shown in Figure 2. It can be seen from Figure 2 that in all of the transition states calculated the hydrogen atoms interact with the same Pd atom to which the carbon atoms bind at the transition 1562

dx.doi.org/10.1021/jp408807c | J. Phys. Chem. C 2014, 118, 1560−1567

The Journal of Physical Chemistry C

Article

Figure 2. Transition state structures of the favorite hydrogenation of C4H4 and C4H5 on the Pd(111), Pd(211), Pd(111)-0.25C, and Pd-0.25Ag/ Pd(111) surfaces. The light blue atoms are silver atoms and this notation is used throughout the paper.

states. This is in agreement with previously reported transition state structures for hydrogenation reactions.3,4,45−50 Those C− C distances of the reacting carbon atoms are found to increase from the reactions of C2H2 + C2H2, C2H2 + C2H3 to C2H3 + C2H3, while the reaction barriers lie in the order of C2H2 + C2H2 > C2H3 + C2H3 > C2H2 + C2H3. Using these energetic values, we can obtain the energy profiles of these three pathways as shown in Figure 1. It should be mentioned that in Pathway III the two hydrogenation reactions (C2H2 hydrogenation) are not sequential and, therefore, only one transition state (TS1) is represented in the energy profile of this pathway. From Figure 1, one can see that pathway I shows the lowest energies for almost all the transition states and intermediate states compared with the other two pathways, indicating that this is the favored pathway, which is different with the order of the reaction barriers of the coupling reactions. Furthermore, on combining all the steps in the pathway to obtain the effective barriers,4 it is found that the effective barriers of pathways I, II, and III are 0.97, 0.99, and 1.19 eV, respectively. Therefore, pathway I is found to be relatively favored on Pd(111), but pathway II has a similar effective barrier and is also likely to be involved in the reaction. 3.2.2. 1,3-Butadiene Formation on Pd(211). An analogous study was undertaken for the formation of 1,3-butadiene on a stepped Pd surface, Pd(211). Previous studies on the adsorption of acetylene and vinyl group on Pd(211) surface showed that the favored adsorption sites of them are the 4-fold site under the step edge (B5 site) and the step edge sites, respectively.4 The corresponding coupling reaction transition state structures for C2H2 + C2H2, C2H2 + C2H3, and C2H3 + C2H3 and the adsorption geometries of the products C4H4, C4H5, and C4H6 are shown in Figure 3. The adsorption energy of C4H6 on Pd(211) was calculated to be −1.94 eV, which is much stronger than on Pd(111) surface (−1.64 eV). The C−C distances of the reacting carbon atoms are also found to increase following the order of C2H2 + C2H2 < C2H2 + C2H3< C2H3 + C2H3 while the reaction barriers follow the opposite order. This is different from the trends observed on Pd(111) where the reaction barriers are similar. All the C4 species formed, that is, C4H4, C4H5, and C4H6, are found to adsorb at the step edge on Pd(211). The hydrogenation of C4H4 and C4H5 at the step sites of Pd(211) was also studied. As the adsorption geometry of C4H4 is no longer symmetric, each of the carbon atoms is inequivalent and must be considered separately. Therefore,

Figure 3. Top: Energy profiles of the three pathways shown in Scheme 1 from 2C2H2 + 2H to C4H6 on Pd(211). TS1, TS2, and TS3 are the same as those in Table 2. Bottom: Transition state structures of coupling reactions of C2H2 + C2H2, C2H2 + C2H3, and C2H3 + C2H3 (from left to right) on Pd(211). The adsorption structures of the C4H4, C4H5, and C4H6 produced are also shown. Different carbon atoms are marked as 1, 2, 3, and 4.

we considered the hydrogenation of all of the carbon atoms in C4H4, which are marked as 1, 2, 3, and 4 in Figure 3. The corresponding hydrogenation barriers for C4H4 and C4H5 on Pd(211) are also summarized in Table 1. It is found that hydrogenation of terminal carbon atom 1 in C4H4 to produce C4H5 is still favored. For the hydrogenation of C4H5, the hydrogenation of carbon 4 is favored and 1,3-butadiene is again readily formed from C4H5 hydrogenation on Pd(211). 1563

dx.doi.org/10.1021/jp408807c | J. Phys. Chem. C 2014, 118, 1560−1567

The Journal of Physical Chemistry C

Article

butadiene on Pd(111)-0.25C are shown in Figure 4. The adsorption geometries for the products C4H4, C4H5, and C4H6

The activation barriers of these reactions and C−C and C−H distances at the transition states for the coupling and hydrogenation reactions, respectively, on Pd(211) surface are listed in Table 2. The transition states of the hydrogenation of C4H4 and C4H5 on Pd(211) are shown in Figure 2. Interestingly, we find that the C−C and C−H distances at the transition states for the coupling and hydrogenation reactions, respectively, on Pd(211) are similar to those found on Pd(111), although the reaction barriers and the structures are quite different. The reaction barriers for coupling reactions follow the order of C2H2 + C2H2 > C2H2 + C2H3 > C2H3 + C2H3 on Pd(211). In order to compare the energetics of each pathway on Pd(211), the energy profiles of them are shown in Figure 3, and the effective barriers for pathway I, pathway II and pathway III are estimated. These are found to be 1.86, 1.80, and 0.97 eV, respectively, indicating that, on Pd(211), pathway III should be favored for 1,3-butadiene formation. 3.2.3. General Discussion on the Surface Structure Effect. The above results provide a basis for a full description of 1,3butadiene and heavier green oil formation on Pd surfaces. First, as H2/acetylene ratio is rather low in tail-end systems (commonly around 1) compared with the high ratio (commonly higher than 10) in front-end process, the surface coverage of acetylene is high and thus increases the probability of C2H2 + C2H2 coupling. The effective barrier of ethylene formation (1.07 eV) is higher than that of 1,3-butadiene formation (0.97 eV) on Pd(111), while on Pd(211), the two effective barriers are the same (both are 0.97 eV), suggesting that butadiene formation will be much easier at the flat sites than at the step sites. Subsequent polymerization of C4H6 may also take place on Pd(111) surface due to its strong adsorption (−1.64 eV). In addition to the formation of the coupled products, hydrogenation of C2H2 will still occur at the same time at the stepped sites. However, the green oil formed will first block flat sites and then diffuse to the step sites of Pd catalyst and the support, when a large amount of green oil are accumulated on the catalyst. As we showed previously, although the flat sites prefer to selectively produce ethylene while the step sites prefer to produce ethane,4 the undesired overhydrogenation product, it can be reasoned that the selectivity of ethylene on the Pd catalyst will decrease during the formation of green oil, which is consistent with the experimental work of Asplund and co-workers.12,13 Eventually, the pores of catalyst will be blocked once the amount of the heavier green oil formed is too much, leading to the reaction becoming mass transfer limited. 3.3. Effect of the Formation of Subsurface Carbon. The Pd(111) surface with subsurface carbon coverage of 0.25 monolayer (ML), which is denoted as Pd(111)-0.25C, is considered in the current work due to the following reasons. First, the formation of butadiene has been concluded to be favored on Pd(111) over Pd(211). Second, the subsurface coverage of carbon on the Pd catalyst under the reaction condition was determined to be ∼0.15 ML experimentally,51−53 which is close to the coverage studied, herein. Furthermore, in our previous work we calculated the thermodynamics of the carbon adsorption process at the subsurface sites of Pd(111) and found that the decomposition of acetylene to form carbon atoms and hydrogen is exothermic until the subsurface carbon coverage is ∼0.50 ML, indicating that the subsurface carbon coverage cannot be higher than 0.50 ML. The corresponding transition state structures for the C2H2 + C2H2, C2H2 + C2H3, and C2H3 + C2H3 coupling reactions for the formation of 1,3-

Figure 4. Top: Energy profiles of the three pathways shown in Scheme 1 from the adsorption state of 2C2H2 + 2H to the adsorption state of C4H6 on Pd(111)-0.25C surface. Bottom: Transition state structures (first column) of the coupling reactions of C2H2 + C2H2, C2H2 + C2H3, and C2H3 + C2H3 (from left to right) on Pd(111)-0.25C surface and the adsorption structures (second row) of the C4H4, C4H5, and C4H6 produced. Different carbon atoms are marked as 1, 2, 3 and 4.

are also shown in Figure 4. From Figure 4, one can find that C2H2 and C2H3 adsorb at two adjacent face-centered cubic (fcc) and hexagonal close-packed (hcp) sites without any carbon atoms beneath the surface. A comparison with Figure 1 shows that the adsorption geometry of C4H4, C4H5, and C4H6 on Pd(111)-0.25C are also similar to those on Pd(111). The barriers of hydrogenation reactions on different carbon atoms of C4H4 and C4H5 on Pd(111)-0.25C surface are listed in Table 1. It should be mentioned that the adsorption of C4H4 is asymmetric on Pd(111)-0.25C due to the presence of subsurface carbon atoms (see Figure 4) and thus it is necessary to calculate the barriers of hydrogenation on all four carbon atoms. From Table 1, it is clear that the hydrogenation of C1 in C4H4 is favored over Pd(111)-0.25C surface, while the hydrogenation of the terminal CH(C4) in C4H5 is favored, leading to 1,3-butadiene formation. The adsorption energy of C4H6 on Pd(111)-0.25C was found to be −1.11 eV, which is much lower than on the Pd(111) and Pd(211) surfaces. The reaction barriers for the coupling reactions and hydrogenation reactions on Pd(111)-0.25C, according to Scheme 1, are listed in Table 2. The corresponding C−C and C−H distances at the transition states are also listed (the hydrogenation barrier of C2H2 and the C−H distance at the transition state on Pd(111)-0.25C surface was obtained from 1564

dx.doi.org/10.1021/jp408807c | J. Phys. Chem. C 2014, 118, 1560−1567

The Journal of Physical Chemistry C

Article

our previous work4). The favored transition state structures of the hydrogenation of C4H4 and C4H5 are shown in Figure 2 and the energy profiles of these three pathways on Pd(111)0.25C are shown in Figure 4. The effective barriers of pathway I, pathway II, and pathway III were calculated to be 1.10, 1.00, and 1.00 eV, respectively, indicating that pathway II and III have the same priority for 1,3-butadiene formation on Pd(111)0.25C. 3.4. Effect of the Formation of Surface Pd−Ag Alloy. Similar calculations were performed on a Ag-alloyed Pd(111) surface with low coverage of Ag(0.25 ML), which is denoted as Pd-0.25Ag/Pd(111). Figure 5 illustrates the transition state

down-shift of the d-band of the surface Pd atoms away from the Fermi level.4,54 The hydrogenation barriers of C4H4 and C4H5 are listed in Table 1, and the corresponding structures are shown in Figure 2. Only the hydrogenation barriers of C1 and C2 in C4H4 were calculated due to the symmetric adsorption geometry of C4H4 on Pd-0.25Ag/Pd(111) surface. As found on the other Pd surfaces studied, 1,3-butadiene is readily formed from the hydrogenation of C4H4 and C4H5 on Pd-0.25Ag/ Pd(111). The reaction barriers and the C−C and C−H distances at the transition states of corresponding coupling and hydrogenation reactions, respectively, on Pd-0.25Ag/Pd(111) surface are also listed in Table 2. It is found that the C−C and C−H distances on Pd-0.25Ag/Pd(111) are similar to those found on Pd(111). The coupling barriers of C2H2 + C2H2, C2H2 + C2H3, and C2H3 + C2H3 decrease, following the order of C2H2 + C2H2 > C2H2 + C2H3 > C2H3 + C2H3. With these calculated reaction barriers, the energy profiles of the three pathways shown in Scheme 1 from the adsorption state of 2C2H2 + 2H to the adsorption state of C4H6 on Pd-0.25Ag/Pd(111) surface were obtained and are shown in Figure 5. From these energy profiles, we can determine that the effective barriers of pathways I, II, and III to be 1.13, 1.23, and 1.30 eV, respectively. Therefore, pathway I should be relatively favored which is also found on unmodified Pd(111). 3.5. General Discussion. It is worth discussing the deactivation by green oil formation with respect to the ethylene formation. To this end, the effective barriers for the formation of 1,3-butadiene must be compared with those for the formation of ethylene from the hydrogenation of acetylene. The effective barriers of ethylene formation on Pd(111), Pd(211), and Pd(111)-0.25C can be obtained from our previous work.4 For the hydrogenation of acetylene on Pd0.25Ag/Pd(111), the adsorption energy of C2H2 is found to be −1.96 eV while the hydrogenation barriers of C2H2 and C2H3 are 0.98 and 0.82 eV, respectively. Therefore, from the energy profile of acetylene hydrogenation on Pd-0.25Ag/Pd(111) shown in Supporting Information Figure S1, the effective barrier of ethylene formation is estimated to be 1.14 eV. The effective barriers of ethylene and 1,3-butadiene formation on Pd(111), Pd(211), Pd(111)-0.25C, and Pd0.25Ag/Pd(111) surfaces are shown in Figure 6. In the

Figure 5. Top: Energy profiles of the three pathways shown in Scheme 1 from the adsorption state of 2C2H2 + 2H to the adsorption state of C4H6 on Pd-0.25Ag/Pd(111) surface. Bottom: Transition state structures(first column) of the coupling reactions of C2H2 + C2H2, C2H2 + C2H3, and C2H3 + C2H3(from left to right) on Pd-0.25Ag/ Pd(111) surface and the adsorption structures(second column) of the C4H4, C4H5, and C4H6 produced. Each of the different carbon atoms are marked as 1, 2, 3, and 4.

structures of the coupling reactions of C2H2 + C2H2, C2H2 + C2H3, and C2H3 + C2H3, and the adsorption geometries of the products from the coupling reactions on Pd-0.25Ag/Pd(111). All the species, including C2H2, C2H3, C4H4, C4H5, and C4H6, adsorb at the Pd sites either at the coupling transition states of C2 + C2 or the adsorption states of C4 species and without direct interaction with the Ag atoms. From Figure 5, one can see that the 1,3-butadiene adsorbs at three Pd atoms in a row on Pd-0.25Ag/Pd(111) with an adsorption energy of −1.40 eV. When the adsorption energies of 1,3-butadiene on Pd-0.25Ag/ Pd(111), Pd(111)-0.25C(−1.11 eV), and Pd(111)(−1.64 eV) surfaces are compared, it can be found that the adsorption on Pd(111) is the strongest. This can be understood from the

Figure 6. Comparison of the effective barriers of the formation of ethylene and 1,3-butadiene on Pd(111), Pd(211), Pd(111)-0.25C, and Pd-0.25Ag/Pd(111) surfaces. 1565

dx.doi.org/10.1021/jp408807c | J. Phys. Chem. C 2014, 118, 1560−1567

The Journal of Physical Chemistry C

Article

Notes

presence of subsurface carbon on Pd(111), the effective barrier of 1,3-butadiene formation slightly increases while the effective barrier of ethylene formation slightly decreases indicating that the presence of subsurface carbon to some extent will inhibit the formation of 1,3-butadiene and will improve the selectivity to ethylene. On alloying Pd with Ag, both the effective barriers of ethylene and 1,3-butadiene formation are found to be slightly increased, indicating that both reaction channels are hindered, leading to a lower active catalyst, which is in accordance with the experimental results reported previously.24 When the effective barrier difference are compared for the same type of reactions, we find that the effective barrier of ethylene formation is increased by 0.07 eV while that of 1,3-butadiene formation is increased by 0.16 eV upon Ag doping, indicating that the presence of Ag atom on Pd(111) will, to a large extent, reduce the reaction rate of 1,3-butadiene formation. Furthermore, the adsorption of 1,3-butadiene on Pd(111)-0.25C and Pd-0.25Ag/Pd(111) surfaces are much weaker than that on Pd(111), which will further prevent the formation of green oil by reducing the coverage of 1,3-butadiene and thus decreasing the possibility of coupling of the 1,3-butadiene. These results are found to be consistent with the experimental results reported before,24 where it was reported that the deactivation of Pd catalyst was significantly retarded when Ag was added.

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work is financially supported by EPSRC and Johnson Matthey through the CASTech programme. The authors would like to thank The Queen’s University of Belfast for computing time. B.Y. also acknowledges the financial support of the Dorothy Hodgkin Postgraduate Award (DHPA) studentship jointly funded by EPSRC and Johnson Matthey.



(1) Borodziński, A.; Bond, G. C. Selective Hydrogenation of Ethyne in Ethene-Rich Streams on Palladium Catalysts. Part 1. Effect of Changes to the Catalyst During Reaction. Catal. Rev.: Sci. Eng. 2006, 48, 91−144. (2) Borodziński, A.; Bond, G. C. Selective hydrogenation of ethyne in ethene-rich streams on palladium catalysts, Part 2: Steady-state kinetics and effects of palladium particle size, carbon monoxide, and promoters. Catal. Rev.: Sci. Eng. 2008, 50, 379−469. (3) Yang, B.; Burch, R.; Hardacre, C.; Headdock, G.; Hu, P. Origin of the Increase of Activity and Selectivity of Nickel Doped by Au, Ag, and Cu for Acetylene Hydrogenation. ACS Catal. 2012, 2, 1027−1032. (4) Yang, B.; Burch, R.; Hardacre, C.; Headdock, G.; Hu, P. Influence of surface structures, subsurface carbon and hydrogen, and surface alloying on the activity and selectivity of acetylene hydrogenation on Pd surfaces: A density functional theory study. J. Catal. 2013, 305, 264−276. (5) Bailey, S.; Bonne, R. L. C.; Booth, J. S.; Griffiths, C.; Watson, M. J. Selective Hydrogenation Process and Catalyst Therefor. WO/2004/ 108638, December 16, 2004 (6) Studt, F.; Abild-Pedersen, F.; Bligaard, T.; Sørensen, R. Z.; Christensen, C. H.; Nørskov, J. K. Identification of non-precious metal alloy catalysts for selective hydrogenation of acetylene. Science 2008, 320, 1320−1322. (7) Osswald, J.; Giedigkeit, R.; Jentoft, R. E.; Armbrüster, M.; Girgsdies, F.; Kovnir, K.; Ressler, T.; Grin, Y.; Schlögl, R. Palladiumgallium intermetallic compounds for the selective hydrogenation of acetylene - Part I: Preparation and structural investigation under reaction conditions. J. Catal. 2008, 258, 210−218. (8) Osswald, J.; Kovnir, K.; Armbrüster, M.; Giedigleit, R.; Jentoft, R. E.; Wild, U.; Grin, Y.; Schlögl, R. Palladium-gallium intermetallic compounds for the selective hydrogenation of acetylene - Part II: Surface characterization and catalytic performance. J. Catal. 2008, 258, 219−227. (9) Armbrüster, M.; Kovnir, K.; Behrens, M.; Teschner, D.; Grin, Y.; Schlögl, R. Pd-Ga Intermetallic Compounds as Highly Selective Semihydrogenation Catalysts. J. Am. Chem. Soc. 2010, 132, 14745− 14747. (10) Bridier, B.; Pérez-Ramírez, J. Cooperative Effects in Ternary CuNi-Fe Catalysts Lead to Enhanced Alkene Selectivity in Alkyne Hydrogenation. J. Am. Chem. Soc. 2010, 132, 4321−4327. (11) López, N.; Vargas-Fuentes, C. Promoters in the hydrogenation of alkynes in mixtures: insights from density functional theory. Chem. Commun. 2012, 48, 1379−1391. (12) Asplund, S. Coke formation and its effect on internal mass transfer and selectivity in Pd-catalysed acetylene hydrogenation. J. Catal. 1996, 158, 267−278. (13) Larsson, M.; Jansson, J.; Asplund, S. Incorporation of deuterium in coke formed on an acetylene hydrogenation catalyst. J. Catal. 1996, 162, 365−367. (14) Pena, J. A.; Herguido, J.; Guimon, C.; Monzon, A.; SantamarIa, J. Hydrogenation of acetylene over Ni/NiAl2O4 catalyst: Characterization, coking, and reaction studies. J. Catal. 1996, 159, 313−322. (15) Larsson, M.; Jansson, J.; Asplund, S. The role of coke in acetylene hydrogenation on Pd/alpha-Al2O3. J. Catal. 1998, 178, 49− 57.

4. CONCLUSIONS This paper provides a mechanistic understanding on the longstanding problem of green oil formation in the petrochemical industry from first principles calculations. In summary, we have investigated the formation mechanism of 1,3-butadiene, a potential precursor of green oil formation, in the selective hydrogenation of acetylene on both the flat Pd(111) and stepped Pd(211) surfaces using DFT calculations in this work. Among many possible pathways, we have found that C4H6 can readily form through the pathway of C2H2 + C2H2 + 2H → C4H4 + 2H → C4H5 + H → C4H6 on Pd(111) surface while the favorite pathway on Pd(211) surface is the one of C2H2 + C2H2 + 2H → C2H2 + C2H3 + H → C2H3 + C2H3→ C4H6. Furthermore, we have also investigated the formation of 1,3butadiene at different sites by comparing the effective barriers. It has been found that the favorable site for C4H6 formation is the site on the flat surface. The formation of 1,3-butadiene on Pd−C and Pd−Ag surfaces, that is, Pd(111)-0.25C and Pd-0.25Ag/Pd(111), have also been studied in order to gain an insight into the influence of subsurface carbon and Ag alloying on Pd. We have found that the presence of subsurface carbon on Pd(111) will to some extent prevent the formation of 1,3-butadiene from happening and the formation rate of 1,3-butadiene on Pd-0.25Ag/Pd(111) surface is slightly lower than that on Pd(111), which is consistent with the experimental results.



ASSOCIATED CONTENT

S Supporting Information *

Energy profile of acetylene hydrogenation on Pd-0.25Ag/ Pd(111) surface. This material is available free of charge via the Internet at http://pubs.acs.org.



REFERENCES

AUTHOR INFORMATION

Corresponding Authors

*E-mail: (C.H.) [email protected]. *E-mail: (P.H.) [email protected] 1566

dx.doi.org/10.1021/jp408807c | J. Phys. Chem. C 2014, 118, 1560−1567

The Journal of Physical Chemistry C

Article

(16) Molero, H.; Bartlett, B. F.; Tysoe, W. T. The hydrogenation of acetylene catalyzed by palladium: Hydrogen pressure dependence. J. Catal. 1999, 181, 49−56. (17) Albers, P.; Pietsch, J.; Parker, S. F. Poisoning and deactivation of palladium catalysts. J. Mol. Catal. A: Chem. 2001, 173, 275−286. (18) Molnár, A.; Sárkány, A.; Varga, M. Hydrogenation of carboncarbon multiple bonds: chemo-, regio- and stereo-selectivity. J. Mol. Catal. A: Chem. 2001, 173, 185−221. (19) Sárkány, A. Formation of C-4 oligomers in hydrogenation of acetylene over Pd/Al2O3 and Pd/TiO2 catalysts. React. Kinet. Catal. Lett. 2001, 74, 299−307. (20) Kim, W. J.; Shin, E. W.; Kang, J. H.; Moon, S. H. Performance of Si-modified Pd catalyst in acetylene hydrogenation: catalyst deactivation behavior. Appl. Catal, A 2003, 251, 305−313. (21) Kim, W. J.; Kang, J. H.; Ahn, I. Y.; Moon, S. H. Deactivation behavior of a TiO2-added Pd catalyst in acetylene hydrogenation. J. Catal. 2004, 226, 226−229. (22) Ahn, I. Y.; Lee, J. H.; Kum, S. S.; Moon, S. H. Formation of C4 species in the deactivation of a Pd/SiO2 catalyst during the selective hydrogenation of acetylene. Catal. Today 2007, 123, 151−157. (23) Trimm, D. L.; Liu, I. O. Y.; Cant, N. W. The oligornerization of acetylene in hydrogen over Ni/SiO2 catalysts: Product distribution and pathways. J. Mol. Catal. A: Chem. 2008, 288, 63−74. (24) Ahn, I. Y.; Lee, J. H.; Kim, S. K.; Moon, S. H. Three-stage deactivation of Pd/SiO2 and Pd-Ag/SiO2 catalysts during the selective hydrogenation of acetylene. Appl. Catal., A 2009, 360, 38−42. (25) Bridier, B.; López, N.; Pérez-Ramírez, J. Partial hydrogenation of propyne over copper-based catalysts and comparison with nickel-based analogues. J. Catal. 2010, 269, 80−92. (26) Bridier, B.; López, N.; Pérez-Ramírez, J. Molecular understanding of alkyne hydrogenation for the design of selective catalysts. Dalton Trans. 2010, 39, 8412−8419. (27) García-Mota, M.; Bridier, B.; Pérez-Ramírez, J.; López, N. Interplay between carbon monoxide, hydrides, and carbides in selective alkyne hydrogenation on palladium. J. Catal. 2010, 273, 92−102. (28) Kresse, G.; Hafner. Ab Initio Molecular-Dynamics for LiquidMetals. J. Phys. Rev. B 1993, 47, 558−561. (29) Kresse, G.; Hafner, J. Ab-Initio Molecular-Dynamics Simulation of the Liquid Metal Amorphous-Semiconductor Transition in Germanium. Phys. Rev. B 1994, 49, 14251−14269. (30) Kresse, G.; Furthmuller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci. 1996, 6, 15−50. (31) Kresse, G.; Furthmuller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 1996, 54, 11169−11186. (32) Perdew, J. P.; Wang, Y. Accurate and Simple Analytic Representation of the Electron-Gas Correlation-Energy. Phys. Rev. B 1992, 45, 13244−13249. (33) Blöchl, P. E. Projector Augmented-Wave Method. Phys. Rev. B 1994, 50, 17953−17979. (34) Kresse, G.; Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 1999, 59, 1758−1775. (35) Alavi, A.; Hu, P.; Deutsch, T.; Silvestrelli, P. L.; Hutter, J. CO oxidation on Pt(111): An ab initio density functional theory study. Phys. Rev. Lett. 1998, 80, 3650−3653. (36) Liu, Z. P.; Hu, P. General Rules for Predicting Where a Catalytic Reaction Should Occur on Metal Surfaces: A Density Functional Theory Study of C-H and C-O Bond Breaking/Making on Flat, Stepped, and Kinked Metal Surfaces. J. Am. Chem. Soc. 2003, 125, 1958−1967. (37) Michaelides, A.; Liu, Z. P.; Zhang, C. J.; Alavi, A.; King, D. A.; Hu, P. Identification of General Linear Relationships between Activation Energies and Enthalpy Changes for Dissociation Reactions at Surfaces. J. Am. Chem. Soc. 2003, 125, 3704−3705. (38) Horiuti, J.; Polanyi, M. A Catalysed Reaction of Hydrogen with Water. Nature 1933, 132, 819−819.

(39) Horiuti, J.; Polanyi, M. Catalysed Reaction of Hydrogen with Water and the Nature of Over-voltage. Nature 1933, 132, 931−931. (40) Horiuti, J.; Polanyi, M. Catalytic Interchange of Hydrogen between Water and Ethylene and between Water and Benzene. Nature 1934, 134, 377−378. (41) Horiuti, J.; Polanyi, M. Exchange reactions of hydrogen on metallic catalysts. Trans. Faraday Soc. 1934, 30, 1164−1172. (42) Yang, B.; Gong, X.-Q.; Wang, H.-F.; Cao, X.-M.; Rooney, J. J.; Hu, P. Evidence To Challenge the Universality of the Horiuti-Polanyi Mechanism for Hydrogenation in Heterogeneous Catalysis: Origin and Trend of the Preference of a Non-Horiuti-Polanyi Mechanism. J. Am. Chem. Soc. 2013, 135, 15244−15250. (43) Dunphy, J. C.; Rose, M.; Behler, S.; Ogletree, D. F.; Salmeron, M.; Sautet, P. Acetylene structure and dynamics on Pd(111). Phys. Rev. B 1998, 57, R12705−R12708. (44) Valcarcel, A.; Clotet, A.; Ricart, J. M.; Delbecq, F.; Sautet, P. Selectivity Control for the Catalytic 1,3-Butadiene Hydrogenation on Pt(III) and Pd(III) Surfaces: Radical versus Closed-Shell Intermediates. J. Phys. Chem. B 2005, 109, 14175−14182. (45) Yang, B.; Wang, D.; Gong, X.-Q.; Hu, P. Acrolein hydrogenation on Pt(211) and Au(211) surfaces: a density functional theory study. Phys. Chem. Chem. Phys. 2011, 13, 21146−21152. (46) Yang, B.; Cao, X.-M.; Gong, X.-Q.; Hu, P. A density functional theory study of hydrogen dissociation and diffusion at the perimeter sites of Au/TiO2. Phys. Chem. Chem. Phys. 2012, 14, 3741−3745. (47) Michaelides, A.; Hu, P.; Alavi, A. Physical origin of the high reactivity of subsurface hydrogen in catalytic hydrogenation. J. Chem. Phys. 1999, 111, 1343−1345. (48) Gong, X. Q.; Raval, R.; Hu, P. CHx Hydrogenation on Co(0001): A density functional theory study. J. Chem. Phys. 2005, 122, 024711. (49) Crawford, P.; McAllister, B.; Hu, P. Insights into the Staggered Nature of Hydrogenation Reactivity over the 4d Transition Metals. J. Phys. Chem. C 2009, 113, 5222−5227. (50) Cao, X.-M.; Burch, R.; Hardacre, C.; Hu, P. Reaction Mechanisms of Crotonaldehyde Hydrogenation on Pt(111): Density Functional Theory and Microkinetic Modeling. J. Phys. Chem. C 2011, 115, 19819−19827. (51) Ziemecki, S. B.; Jones, G. A. Interstitial Carbon in Palladium. J. Catal. 1985, 95, 621−622. (52) Ziemecki, S. B.; Jones, G. A.; Swartzfager, D. G.; Harlow, R. L. Formation of Interstitial Pd-C Phase by Interaction of Ethylene, Acetylene, and Carbon-Monoxide with Palladium. J. Am. Chem. Soc. 1985, 107, 4547−4548. (53) Ziemecki, S. B.; Jones, G. A.; Swartzfager, D. G. Coexistence of Hydrogen and Carbon Solutes in the Palladium Lattice. J. LessCommon Met. 1987, 131, 157−162. (54) Sheth, P. A.; Neurock, M.; Smith, C. M. First-Principles Analysis of the Effects of Alloying Pd with Ag for the Catalytic Hydrogenation of Acetylene-Ethylene Mixtures. J. Phys. Chem. B 2005, 109, 12449− 12466.

1567

dx.doi.org/10.1021/jp408807c | J. Phys. Chem. C 2014, 118, 1560−1567