Pulse Transient Responses of NO Decomposition ... - ACS Publications

Jul 23, 2009 - The 6-port valve is used to pulse 1 cm3 NO into He or He/H2 flow at 723, 773, and 823 K. The changes in the concentration of IR-observa...
1 downloads 0 Views 2MB Size
Download PDF
J. Phys. Chem. C 2009, 113, 14963–14971

14963

Pulse Transient Responses of NO Decomposition and Reduction with H2 on Ag-Pd/Al2O3 Duane D. Miller and Steven S. C. Chuang* The UniVersity of Akron, Akron, Ohio 44325-3906 ReceiVed: May 15, 2009; ReVised Manuscript ReceiVed: July 3, 2009

Catalytic decomposition of nitric oxide (NO) over Pd/Al2O3 and Ag-Pd/Al2O3 has been studied using the pulse transient response technique coupled with in situ infrared (IR) and mass spectrometry (MS) at 723-823 K. In the absence of H2, pulsing NO over the Pd/Al2O3 catalyst produces adsorbed NO species (i.e., Pd+-NO, Pd0-NO, and Pd-NO-) as well as gaseous N2, O2, and N2O products. Transient responses of the N2 and O2 profiles show that the addition of Ag onto Pd/Al2O3 catalyst shifts the O2 profile forward, increases oxygen formation and the oxidation resistance of Pd, but did not decrease the amount of retained oxygen (Oret) and did not improve the catalytic cycle for NO decomposition. Oret on the Pd surface is not able to desorb in the temperature range of this study; however, Oret on Ag-Pd/Al2O3 can be desorbed at higher temperatures than its formation and adsorption temperature. The presence of H2 during the NO pulse allowed NO reduction to occur, producing N2, N2O, O2, NH3, and H2O. Pd/Al2O3 is a more active catalyst for the formation of NH3 and H2O than Ag-Pd/Al2O3. Comparison of the transient gaseous product responses over Pd/Al2O3 and Ag-Pd/ Al2O3 catalysts show that Ag (i) promotes the formation of N2, shifting its profile forward, and (ii) suppresses the formation of NH3 and H2O, delaying their formation. The lack of the initial activity of Ag-Pd/Al2O3 for NH3/H2O formation can be attributed to the alloy state of Ag-Pd on Al2O3. As the NO reduction proceeds in the presence of H2, adsorbed oxygen produced from N-O dissociation could cause the dealloying of Ag-Pd, producing Pd sites, which exhibited high selectivity for NH3/H2O formation. 1. Introduction Effective removal of NOx from lean-burn and diesel engines has been considered to be the most challenging task to the automotive industry. Research has been focused on developing catalysts capable of removing NOx from O2-rich exhaust either by the direct decomposition of NOx to N2 and O2 (i.e., 2NO T N2 + O2 )1-3 or by the selective catalytic reduction (SCR) with a reductant.4-6 The direct NOx decomposition is an appealing approach for the removal of NOx from an exhaust mixture. The direct NO decomposition has been proposed to proceed via the following steps:7

NO(g) f NOads

(1)

NOads f Nads + Oads

(2)

Nads + Nads f N2(g)

(3)

Oads + Oads f O2(g)

(4)

NOads + Nads f N2O(g)

(5)

NOads + Oads f NO2(g)

(6)

NO adsorbs on the catalyst surface (step 1) followed by the dissociation of adsorbed NO (NOads) to form adsorbed nitrogen (Nads) and oxygen (Oads) (step 2). Adsorbed nitrogen and oxygen can combine to form gaseous nitrogen (step 3) and oxygen (step * To whom correspondence should be addressed. E-mail: schuang@ uakron.edu.

4), respectively. Adsorbed NO can also react with adsorbed nitrogen to form gaseous N2O (step 5) or with adsorbed oxygen to form gaseous NO2 (step 6). Studies on these proposed steps over single crystals and supported catalysts3,7-9 have revealed that adsorbed oxygen from dissociation of NOads is strongly bounded on the metal catalyst surface, poisoning NO dissociation sites. The removal of Oads should increase the concentration of active NO adsorbates and increase NO conversion. Silver (Ag), known for its resistance to oxidation,10 has been shown to decrease the reduction temperature of metal oxides such as Co/Al2O3.11 These results suggest that Ag may assist in desorption of Oads on the catalyst surface. We have therefore postulated that Ag may promote desorption of adsorbed oxygen from the NO dissociation site and create free sites needed for further NO dissociation (step 2), allowing the completion of the catalytic NO decomposition cycle. An alternative and common approach for the removal of Oads is the addition of reducing agents such as H2,4,12,13 CO,14 hydrocarbons,5,15 and NH3.16 Both H2 and NH3, reducing agents, have been shown to exhibit high reactivity toward the Oads especially under lean-burn conditions.13 Adsorbed hydrogen (Hads) on the Pd surface can react not only with Oads but also with Nads. The reaction of Hads with Nads, a hydrogenation reaction step, produces NH3 which can either (i) serve as a reducing agent to further remove Oads or (ii) result in an undesirable slip from the reactor. Ag is also known for its low activity for catalyzing hydrogen-related reactions such as suppressing the hydrogenation of ethylene in the acetylene hydrogenation reaction.17 The addition of Ag may limit the formation of NH3 during NO reduction. This paper is aimed at determining the effect of Ag on Pd/ Al2O3 during the NO decomposition and reduction with H2 by in situ IR spectroscopy coupled with the NO pulse technique. The effect of Ag on Pd/Al2O3 on NO decomposition/reduction

10.1021/jp904538t CCC: $40.75  2009 American Chemical Society Published on Web 07/23/2009

14964

J. Phys. Chem. C, Vol. 113, No. 33, 2009

Figure 1. Experimental apparatus.

and the nature of the surface sites is elucidated by the transient response (i.e., changes in IR and MS intensity) of adsorbates and gaseous species. A fundamental understanding of the reaction pathway could help in guiding the preparation of the desired catalyst structure for enhancing NO decomposition activity and the H2 selectivity in NO reduction. 2. Experimental Section 2.1. Catalyst Preparation and Characterization. The 5% Pd/Al2O3 catalyst was prepared by incipient wetness impregnation of Pd(NO3) · H2O (Aldrich) aqueous solution onto R-Al2O3 (Aldrich) and dried at 298 K. The 5% Ag-5% Pd/Al2O3 catalyst was prepared by incipient wetness coimpregnation of Pd(NO3)2 · H2O (Aldrich) and Ag(NO3) (Aldrich) aqueous solution onto R-Al2O3 (Aldrich) and dried at 298 K. The catalysts were calcined at 773 K in air for 2 h and reduced at 773 K in H2 for 1 h. The Pd particle size was determined to be 6 nm for Pd/ Al2O3 by X-ray diffraction (XRD) using the Scherer equation from line broadening.18,19 XRD analysis shows the presence of a very broad Ag-Pd peak on Ag-Pd/Al2O3 indicating the existence of an Ag-Pd alloy phase and an Ag-Pd particle size less than 3 nm.18 2.2. In Situ IR/MS Reaction Studies. The experimental apparatus, shown in Figure 1, consist of (i) a reactant metering system (Brooks Instrument 5850 mass flow controllers), (ii) a gas sampling system with 4-port and 6-port valves, (iii) a diffuse reflectance infrared fourier transform spectroscopy (DRIFTS, Harrick Scientific) reactor containing 150 mg catalyst and placed inside a Fourier transform infrared spectrometer (FTIR, Varian Inc. FTS-4000), and (iv) a mass spectrometer (MS, Pfeiffer Omnistar) for determining changes in the DRIFTS reactor effluent concentrations. The 4-port valve allows switching the inlet flow from 100% He to He/H2 (90/10 vol %) while maintaining a total flow rate of 34 cm3/min over the Pd/Al2O3 or Ag-Pd/Al2O3 catalyst. The 6-port valve is used to pulse 1 cm3 NO into He or He/H2 flow at 723, 773, and 823 K. The changes in the concentration of IR-observable adsorbates are monitored by FTIR. The IR spectra collected by DRIFTS are reported in absorbance units which are more applicable for low adsorbate surface concentrations than that of Kubelka-Munk

Miller and Chuang

Figure 2. Infrared spectra of gaseous and adsorbed species; (a) CO adsorption on Pd and Ag-Pd/Al2O3, (b) NO adsorption on Pd and Ag-Pd/Al2O3 at 298 K.

units.20 The IR absorbance spectrum of adsorbed and gaseous species is obtained by A ) -log(I0/I),21 where I0 is the background IR single beam spectrum (32 coadded scans and resolution 4 cm-1) of the catalyst under He or He/H2 flow and I is the IR single beam spectrum during the NO pulse reaction. The MS responses corresponding to N2 (m/e ) 28), NO (m/e ) 30), O2 (m/e ) 32), and N2O (m/e ) 44) are monitored. 3. Results 3.1. CO and NO Adsorption on Pd/Al2O3 and Ag-Pd/ Al2O3. Figure 2 shows the IR absorbance spectra of gaseous CO and NO as well as their adsorbed species on Pd/Al2O3 and Ag-Pd/Al2O3 catalysts at 298 K. CO adsorbed on Pd/Al2O3 as linear CO at 2099 cm-1 and bridged CO at 1949 and 1994 cm-1;17,22 CO adsorbed on Ag-Pd/Al2O3 as a intense linear CO band at 2081 cm-1 and a weak bridged CO band at 1957 cm-1.22 The low IR intensity of bridged CO on Ag-Pd/Al2O3 as compared to that on Pd/Al2O3 indicates that Ag disrupts some of the Pd0-Pd0 sites for CO adsorption.10,23 Exposure of the Pd/Al2O3 catalyst to NO flow led to an immediate formation of Pd0-NO at 1754 cm-1.24 Increasing NO exposure time produced N2O at 2223 cm-1, NO2 at 1628 cm-1, and chelating bidentate nitrate at 1527 and 1307 cm-1.25 Flowing NO over Ag-Pd/Al2O3 produced gaseous N2O at 2223 cm-1 and Pd-NO- at 1545 cm-1. The formation of NO2 at 1628 cm-1 was accompanied by that of chelating bidentate nitrate species at 1527 and 1415 cm-1. Pd0-NO has been identified as an active adsorbate species for NO dissociation, which is the required step for the formation of N2O and NO2.24 The absence of Pd0-NO during NO exposure over Ag-Pd/ Al2O3, did not prevent formation of the N2O and nitrate species. 3.2. NO Pulse Reaction in He Flow over Pd/Al2O3. Figure 3, panels a and b, shows the IR absorbance spectra and MS intensity profiles during the first and second NO pulses into He flow over Pd/Al2O3 catalyst at 773 K. The first NO pulse produced gaseous N2O at 2223 cm-1, Pd0-NO at 1735 cm-1, bent NO (Pd-NO-) at 1641 cm-1, and chelating bidentate nitrate species at 1573 cm-1.9,26 The intensities of the Pd0-NO followed closely that of gaseous NO. These adsorbed NO species showed a higher rate of disappearance

Catalytic Decomposition of Nitric Oxide

Figure 3. NO pulses over Pd/Al2O3 catalyst at 773 K. (a) Infrared spectra, (b) MS intensity spectra.

than that of the nitrate at 1573 cm-1, revealing the high reactivity of Pd0-NO, a precursor for N-O dissociation, step (2). The second NO pulse produced a lower IR intensity of Pd0-NO and a higher intensity of Pd+-NO at 1769 cm-1 than the first NO pulse. Figure 3(b) shows the second NO pulse also produced lower MS intensity profiles for N2 and N2O than the first NO pulse. The low MS intensity corresponds to the small amount of the products produced. The amount of the products formed from each pulse is calculated by integrating the area under the MS profile and summarized in Table 1. The results show that the second pulse decreased the amount of N2 produced by a factor of 2.6 and O2 by a factor of 1.2 at 773 K. These factors increased markedly with increasing temperature, showing the occurrence of rapid catalyst deactivation at higher temperatures. The decrease in the intensity of the Pd0-NO, the amount of N2, N2O, and O2 formed, and the increase in the Pd+-NO intensity provide the direct evidence to confirm that Pd0-NO is an active precursor for the dissociation of N-O, step (2).27 The Oads produced can further oxidize Pd0 to Pd+. The effect of lowering the number of Pd0 sites and increasing the number of Pd+ sites on the dynamic behavior of the product formation can be unravelled by examining the lead/lag of the N2 and O2 responses in the first and second pulses in Figure 3b and subsequent pulses (not shown here). The second NO pulse, which occurred on the catalyst with more Pd+ and less Pd0 sites, moved the N2 formation peak from a lead position in the first NO pulse to a position where it appeared together with the O2 and N2O peaks. In SSITKA (steady-state isotopic-transient kinetic analysis), the delay time in the gaseous isotopically labeled product response corresponds to the residence time (i.e., the reciprocal of the intrinsic rate constant) of intermediates leading to the formation of the specific gaseous product.28,29 This residence time concept can be extended to the single reactant reaction, i.e., NO decomposition, without complication of the second isotopically labeled reactant. Thus, the lagging N2 response can be explained by an increase in the residence time

J. Phys. Chem. C, Vol. 113, No. 33, 2009 14965 of the intermediates leading to N2, reflecting a decrease in the intrinsic rate constant of the combination of Nads and Nads, step (3). Figure 4a illustrates the effect of reaction temperature on the IR intensity of adsorbates obtained after 0.25 min of the second NO pulse. Raising the reaction temperature from 723 to 823 K decreased the IR intensity of all the adsorbed species including partially positive charged Pd+-NO at 1769 cm-1, Pd0-NO at 1735 cm-1, and Pd-NO- at 1641 and 1573 cm-1. The adsorbate IR intensity variations result from changes in the number of Pd0/Pd+ sites, the adsorption equilibrium, and rate constants. Given a quasi-equilibrium state of these adsorbed species, the intensity ratio of Pd+-NO/Pd0-NO would reflect the molar ratio of Pd+ to Pd0 sites. Table 1 show that the amount of O2 produced is significantly less than that of N2 at higher reaction temperatures. An equal (i.e., stoichiometric amount) amount of N2 and O2 formed would be the completion of the catalytic NO decomposition cycle (2NO f N2 + O2). The amount of the retained oxygen (i.e., Oret) would correspond to the number of the incomplete NO decomposition cycle (2NO f N2 + 2Pd-Oret). The amount of Oret on the catalyst during each pulse, shown in Table 1, can be estimated by taking the difference between the amounts of N2 and O2 produced, and the amount of oxygen remaining after the N2O formation Oret ) (N2 - O2)/2 + N2O. The ratio of oxygen retained to gaseous oxygen formed increased from 2.46 (first pulse) to 3.60 (second pulse) at 723 K, shown in Table 1. These retained oxygen species could further convert Pd0 site to Pd+ sites, as evidenced by the increasing IR intensity ratio of Pd+-NO/Pd0-NO with temperature. Figure 4b shows increasing temperature slowed down the response of O2, shifting the O2 maxima of the O2 response from 0.28 min at 723 K to 0.48 min at 823 K. This shift corresponds to an increase in the intensity ratio of Pd+-NO/Pd0-NO, further suggesting that the formation of Pd+ sites could be the key factor slowing down the O2 response, decreasing the reactivity of oxygen intermediates. 3.3. NO Pulse Reaction in He Flow over Ag-Pd/Al2O3. The NO pulse on Ag-Pd/Al2O3 at 773 K produced gaseous N2O at 2223 cm-1, shown in Figure 5a, as well as the formation of N2 and O2 in Figure 5b. The absence of adsorbed NO species, in Figure 5a at 773 K, indicates that the rate of desorption and conversion of these adsorbed species is higher than those of their formation. The adsorbed NO species was also not observed during the NO adsorption study at 298 K (see Figure 2b). The absence of adsorbed NO species indicates that the surface of the Ag-Pd/Al2O3 catalysts is enriched with a layer of Ag, which is known to weakly interact with adsorbed NO at room temperature, exhibiting low temperature reactivity. Reflection IR studies have shown that NO adsorbs on the Ag(111) surface, producing a band at 1776 cm-1 at 45 K. This species can be converted to the (NO)2 dimer and then N2O at 110 K.30 Table 1 show that the amount of N2 produced was increased 5.5 times during the first pulse, and 7.4 times during the second pulse; the amount of O2 produced was increased 2.2 times during both the first and second pulse when Ag is added onto Pd/Al2O3 in the 723-823 K temperature region. In summary, the presence of Ag increased the amount of N2 and O2 formed as well as the ratio of Oret/O2. The N2 selectivity and NOx conversion during the first NO pulse over Pd/Al2O3 and Ag-Pd/Al2O3 at 723, 773, and 832 K obtained from this study generally fall in the same range as those in the literature,31-40 shown in Table 2. MS profiles in Figure 5b shows that the initial O2 profile led those of N2O and N2 on Ag-Pd/Al2O3, in contrast to the O2

14966

J. Phys. Chem. C, Vol. 113, No. 33, 2009

Miller and Chuang

TABLE 1: First and Second NO Pulse into Flowing He, Conversion, Selectivity, and Product Species Formation quantity of gas species (µmol) catalyst Pd/Al2O3

Ag-Pd/Al2O3

Pd/Al2O3

Ag-Pd/Al2O3

temp. (K)

N2

O2

723 773 823 Ea (kJ/mol) 723 773 823 Ea (kJ/mol)

0.32 0.97 3.98 62.83 4.05 5.37 7.15 14.16

0.29 0.41 0.6 18.46 0.86 0.98 1.03 4.46

1st NO Pulse 0.65 0.85 0.92 8.98 1.59 1.21 1.02 -11.05

723 773 823 Ea (kJ/mol) 723 773 823 Ea (kJ/mol)

0.29 0.38 0.93 28.16 2.4 2.81 3.1 6.41

0.22 0.35 0.52 20.90 0.7 0.78 0.92 6.74

2nd NO Pulse 0.67 0.61 0.88 6.51 1.41 1.21 0.98 -8.16

N2 O

lagging behind the N2 on Pd/Al2O3 in Figure 3b. These results reveal that the increase in O2 formation is closely related to the promotion effects of Ag on Oads desorption. This promotion effect is consistent with the observed increase in N2 formation as well as decreases in Ea (activation energy) for oxygen formation on Ag-Pd/Al2O3. The Ea for O2 was decreased from 18.46 kJ/mol on Pd/Al2O3 to 4.46 kJ/mol on Ag-Pd/Al2O3 during the first NO pulse and from 20.9 to 6.74 kJ/mol during the second pulse. The enhancement of oxygen desorption by Ag is also consistent with the reported Ag effect on the removal of oxygen from Ag bimetallic catalysts in which Ag facilitate the reduction of Co supported on Al2O3.11 Figure 6, panels a and b, shows the IR absorbance spectra and MS intensity profiles during the second NO pulse over Ag-Pd/Al2O3 at 723, 773, and 823 K. Although Ag-Pd/Al2O3 was more active than Pd/Al2O3 for the NO decomposition, Ag-Pd/Al2O3 was not able to give an appreciable IR intensity for adsorbed NO in the entire temperature range of the reaction study. The effect of Ag on the reaction must be elucidated mainly by the MS profile results. The N2 response profile show a significant trailing, revealing N2 formation is slowed down by the presence of Ag on Pd. The O2 MS profile show less

Figure 4. NO pulses over Pd/Al2O3 catalyst at 723, 773, and 823 K. (a) Infrared spectra, (b) MS intensity spectra for gaseous N2 and O2 species.

conversion and selectivity (%) Oret/O2

XNO

SN 2

SN2O

2.46 4.78 12.76

4.7 9.4 24.6 41.00 28.4 31.5 40.1 8.52

32.9 53.2 81.2

67.1 46.8 18.8

71.8 81.6 87.5

28.2 18.4 12.5

4.3 4.5 9.2 18.40 18.5 19.6 20.1 2.04

30.6 30.7 51.4

69.4 69.3 48.6

62.9 69.9 75.2

37.1 30.1 24.8

9.21 10.24 12.84

3.60 1.92 3.24 6.85 6.73 5.83

variation with respect to temperature on Ag-Pd/Al2O3 than on Pd/Al2O3, further confirming low Ea for oxygen formation on Ag-Pd/Al2O3. However, this Ea value is lower than those commonly observed for a catalytic reaction. Furthermore, the amount of O2 is less than that of N2, showing the majority of the catalytic NO decomposition cycle was not completed and a large fraction of adsorbed oxygen was retained on the catalyst. To understand the nature of the O2 formation, we further examined the O2 MS intensity profiles (i.e., response) while increasing the reaction temperature from 773 to 823 K in Figure 7, panels a and b. Heating the Ag-Pd/Al2O3 catalyst from 773 to 823 K, shown in Figure 7b, caused an increase in the O2 MS intensity profile (i.e., oxygen desorption) compared to the absence of variation over Pd/Al2O3 in Figure 7a. The different dependence of oxygen desorption on temperature on Pd and Ag-Pd/Al2O3 indicates that the presence of Ag on Pd/Al2O3

Figure 5. NO pulses over Ag-Pd/Al2O3 catalyst at 773 K. (a) Infrared spectra, (b) MS intensity spectra.

Catalytic Decomposition of Nitric Oxide

J. Phys. Chem. C, Vol. 113, No. 33, 2009 14967

TABLE 2: Literature Values for NOx Conversion and Selectivity Data during NO Decomposition Study on Pd Based Catalysts reactants (cm3/min) catalyst

NO

H2

C xH y

temp. (K)

NOx conv. (%)

SN2 (%)

SN2O (%)

a

0

0

723 773 873 673 673 673 1073 473 673 1173 1173 673 723 573 973

16.05 16.99 18.60 7.2 8.7 25 86.7 80 30 75 70 100c 100c 92c 97c

1.18 3.15 5.93 0

52.36 53.57 55.52 46.78

723 773 873 723 773 873

4.7 9.4 24.6 28.4 31.5 40.1

Tb-Pt/Al2O3

1

Cu-ZSM-5

0 1a 0 0 0 0 0 0 0 0 0

0

Pd-Mo/Al2O3 Pd/Al2O3 Pd/Al2O3 Pd/Al2O3 Pd/Al2O3 Ag-Pd/Al2O3 PdMOR Pd/Al2O3 /SiO2 Pd/NaY

1a 2.5b 0.11b 2b 1b 0.11b 1.2b 1.2b 3.6b 6b 6b

Our Study Pd/Al2O3

1

0

0

Ag-Pd/Al2O3

1

0

0

a

0 1b 0 0 0 0 0 0 0

SV (h-1) 3

34, 35 40 000 6500

94.2

40 000 15 000 30 000 81 17 32.9 53.2 81.8 71.8 81.6 87.5

67.1 46.8 18.8 28.2 18.4 12.5

ref

32 33 38 39 40 36 31 37

53 992 53 992

Pulse data. b Steady state data. c Deactivation after time exposed to reactants.

Figure 6. NO pulses over Ag-Pd/Al2O3 catalyst at 723, 773, and 823 K. (a) Infrared spectra, (b) MS intensity spectra of gaseous N2 and O2 species.

Figure 8. NO pulses into H2 over Pd/Al2O3 catalyst at 773 K. (a) Infrared spectra, (b) MS intensity spectra.

Figure 7. First and second NO pulse into He at 823 K, (a) Pd/Al2O3 catalyst, (b) Ag-Pd/Al2O3 catalyst.

shift the O2 desorption equilibrium at each specific temperature, lowering the capacity of retained oxygen (i.e., Oret) on Pd/Al2O3. 3.4. NO Pulse Reaction in He/H2 Flow over Pd/Al2O3. Figure 8a shows that the first NO pulse at 773 K produced gaseous NH3 IR bands at 3336 and 1624 cm-1, and a broadband at 1847 cm-1. This broadband which is overlapped with gaseous NO band may be related to either a H2O-NO complex41 or a

dinitrosyl species.42 The gaseous NH3 IR band at 1624 cm-1 overlapped with that of adsorbed H2O at 1641 cm-1. The amount of N2 produced is about 4 to 8 times more in the presence of H2 than in the absence of H2, by comparing Tables 1 and 3. In the absence of H2, N2 can be produced by (i) dissociation of adsorbed NO (step 2) over Pd0 followed by (ii) the combination of adsorbed nitrogen and desorption of N2 (step 3). In the presence of H2, N2 can also be produced via the reaction of Pd0-NO with adsorbed NH3 which is formed by the reaction of Nads with Hads. The transient response of the NO pulse in Figure 8b showed that the formation of NH3 lagged that of N2. This lag time between the initial N2 and NH3 formation

14968

J. Phys. Chem. C, Vol. 113, No. 33, 2009

Miller and Chuang

TABLE 3: First and Second NO Pulse into Flowing He/H2, Conversion, Selectivity, and Product Species Formation quantity of gas species (µmol) catalyst Pd/Al2O3 Ag-Pd/Al2O3

Pd/Al2O3 Ag-Pd/Al2O3 a

conversion and selectivity (%) c

S N 2a

SN2Ob

temp. (K)

N2

O2

N 2O

NH3

XNO

723 773 823 723 773 823

2.81 3.99 5.15 8.07 8.8 9.78

0.13 0.12 0.12 0.41 0.44 0.47

1st NO Pulse 0.42 0.3 0.27 1.96 1.22 1

5.54 13.04 11.34 0.18 4.23 6

29.3 52.9 54.2 49.4 59.3 67.4

87 92.9 94.9 80.5 87.8 90.7

13 7.1 5.1 19.5 12.2 9.3

723 773 823 723 773 823

2.93 3.29 4.48 6.97 8.03 9.29

0.11 0.1 0.11 0.29 0.3 0.32

2nd NO Pulse 0.4 0.29 0.27 1.87 0.99 0.37

4.43 13.55 11.59 0.1 5.45 7.37

27.2 50.7 51.6 45.8 57.4 65.3

87.9 91.8 94.4 78.8 89 96.2

12.1 8.2 5.6 21.2 11 3.8

XNO(%) )

mol Oin - mol Oout 100 mol Oin

b

SN2(%) )

mol N2 100 mol N2 + mol N2O + mol O2

SN2O(%) )

mol N2O 100 mol N2 + mol N2O + mol O2

c

represents the difference in the intrinsic activity for N2 and NH3 formation. This delay is shortened during the second NO pulse, Figure 8b, because the build up of adsorbed NH3 after the first NO pulse facilitate NH3 desorption. The NH3 produced appears to participate in the NO reduction, causing the NO MS profile to decrease to the baseline within 0.5 min, in contrast to the NO profile dragging for more than 1.5 min during the NO pulse in He, shown in Figure 3b. The effect of temperature on the IR intensity of adsorbed species is shown in Figure 9a. Changing the reaction temperature did not shift the wavenumber of the IR observable species, but caused a variation in IR intensities as a result of greater or lesser amounts of products formed (i.e., NH3 and H2O) in this temperature region. Figure 9b shows that the major effect of raising the reaction temperature is to shift the NH3, O2, and H2O profiles forward. The most dramatic shift is observed for NH3 profiles when increasing temperature from 723 to 773 K. Further increasing temperature to 823 K does not advance the NH3 profile closer to the initial N2 elution time. 3.5. NO Pulse Reaction in He/H2 Flow over Ag-Pd/Al2O3. Figure 10a shows that the first NO pulse produced low IR intensities of NH3 at 3336 and 1624 cm-1 on Ag-Pd/Al2O3 as compared to those over Pd/Al2O3, shown in Figure 8a. The NH3 IR band emerged at 0.36 s; the NH3 MS profile increased at 0.4 s in Figure 8b. The significant delay in NH3 over Ag-Pd/ Al2O3 compared with Pd/Al2O3 indicates that the N-H bond formation is inhibited by the presence of Ag. Figure 10b shows that the O2, NO, and N2O responses appear prior to the emerging of the NH3 and H2O responses. The profiles for the O2 and N2 responses here resemble those produced from the NO pulse in Figure 5b. Figure 11a shows that raising the temperature increases the IR intensity of both gaseous NH3 and H2O bands. Figure 11b shows increasing the temperature did not alter the contours and the maxima of NO, N2, and O2, but shifted the NH3 response forward. The major shift occurred when the temperature increased from 723 K where NH3 was not formed to 773 K where NH3 emerged.

Figure 9. NO pulses into H2 over Pd/Al2O3 catalyst at 723, 773, and 823 K. (a) Infrared spectra, (b) MS and intensity spectra of gaseous NO, N2, O2, NH3, and H2O species.

4. Discussion Vibrational spectroscopy has provided evidence for the adsorption of NO over single crystals and supported catalysts.27,42-47 NO adsorbs on the Pd(111) single crystal in the 1735-1755 cm-1 region as Pd0-NO from 150 to 350 K.42,45,46 Pd0-NO adsorbs on supported Pd catalysts at 1741-1754 cm-1 between 303 and 673 K during NO-CO reaction.27,47 Pd0-NO were observed on Pd/Al2O3 at 1735 cm-1 in the temperature range of 723-823 K during NO decomposition, shown in Figure 3 and 4. The presence of this species under such a high temperature, indicates that (i) the rate of its conversion (i.e.,

Catalytic Decomposition of Nitric Oxide

Figure 10. NO pulses into H2 over Ag-Pd/Al2O3 catalyst at 773 K. (a) Infrared spectra, (b) MS intensity spectra.

Figure 11. NO pulses into H2 over Ag-Pd/Al2O3 catalyst at 723, 773, and 823 K. (a) Infrared spectra, (b) MS intensity spectra of gaseous NO, N2, O2, NH3, and H2O species.

desorption and decomposition) is not sufficiently high for its depletion and (ii) Pd0 sites remain on the Pd surface. It can also be further inferred that Pd(111) is the dominant surface structure45,48 of Pd/Al2O3 used in this study which has a Pd particle size of about 6 nm. The wavenumbers of bridged/ multiple bonded NO (i.e., NO on 3-fold hollow sites) and bent NO fall in the range of the nitrate species. The overlapping of these bands makes it difficult to identify the dynamic behavior

J. Phys. Chem. C, Vol. 113, No. 33, 2009 14969 of these species on supported Pd catalysts which give the prominent nitrate IR bands during NO-CO reaction24 and NO decomposition, shown in Figure 3a. The decrease in the IR intensity for Pd0-NO at 1735 cm-1 as well as the formation of N2, in Figure 3 and 4, confirm that reduced Pd0 is needed for the conversion of NO to N2. The subsequent formation of Pd+-NO at 1769 cm-1 along with the amount of Oret (listed in Table 1) reveals that Oads, which do not desorb as oxygen molecules, are retained at Pd0 sites oxidizing Pd0 to Pd+. The lagging O2 and leading N2 profiles in Figure 3b further reveal that the O-O formation, step (4), is slower than the N-N formation, step (3). The weak dependence of O2 formation on temperature over both Pd/Al2O3 and Ag-Pd/ Al2O3 indicates that the O2 formation may not be a catalytic process in nature. Ag was added on Pd/Al2O3 with the intention of increasing catalytic NO decomposition activity by (i) enhancing the oxidation resistance of Pd0 sites, (ii) decreasing Oret, and (iii) increasing the rate of O2 desorption. It is indeed surprising to observe that the addition of Ag shifted the O2 response forward and increased the amount of N2 and O2 formation, but did not decrease the ratio of Oret/O2. The ratio corresponds to that of the number of incomplete catalytic NO decomposition cycle to the complete NO decomposition cycle. The leading response of O2 on Ag-Pd/Al2O3 indicates the high reactivity of these Oads for desorption. Small Pd particle have a higher tendency to be oxidized than large Pd particle.49 A similar trend has been observed for NO decomposition study over Pt/Al2O3, which shows decreasing the Pt particle size results in higher activity for NO dissociation and higher temperature for oxygen desorption, indicating the higher tendency to be oxidized.50 NO-CO reaction studies over Pd/Al2O348,51 have shown that smaller Pd particles are more active for NO dissociation. Small Pd particles allow oxygen to move to subsurface sites,49 exhibiting higher oxygen uptake and desorbing oxygen at higher temperature than large Pd particles. Our XRD results show the Ag-Pd particles on Al2O3 are smaller than 3 nm which is less than half of the Pd particle size on Pd/Al2O3. The smaller Ag-Pd particles show a higher oxygen uptake capacity and thus give a higher ratio of Oret/O2. However, these Oret on the Ag-Pd surface are significantly different from those on Pd, exhibiting a unique property for its desorption. The Oret on the Pd particle oxidizes Pd0 to Pd+ and are not able to desorb from the Pd surface in the 723-823 K region; the Oret produced on Ag-Pd particles can be desorbed at higher temperatures (i.e., Oret produced at 723 K can be desorbed at 773 K and those produced at 773 K can be desorbed at 823 K) as shown in Figure 7b. Furthermore, these Oret species are not able to oxidize Pd0 on Ag-Pd/Al2O3, as evidenced by the absence of Pd+-NO. Thus, it can be concluded that increasing oxidation resistance of Pd by Ag through the removal of Oret is an unnecessary requirement for the enhancement of catalytic NO decomposition. Further studies are needed to look into (i) the nature of oxygen desorption sites, (ii) the approaches for inhibiting conversion of adsorbed oxygen from dissociated NO to Oret, the type which oxidize Pd0 to Pd+, and (iii) the effect of Pd particle size on these oxygen-related sites. The effectiveness of H2 in the reduction of NO can be inferred from the formation of N2, N2O, O2, NH3, and H2O. Adsorbed hydrogen not only accelerates the NO conversion and N2 formation but also reacts with adsorbed nitrogen and oxygen to produce NH3 and H2O, respectively. During the in situ formation of NH3 and H2O, the IR spectra in Figure 8a show a broadband at 1845 cm-1 that has been assigned to the dinitrosyl

14970

J. Phys. Chem. C, Vol. 113, No. 33, 2009

species having a symmetric vibration at 1779 cm-1 and an antisymmetric band at 1855 cm-1.52 The dinitrosyl species is known to exhibit IR adsorption bands at 1855 and 1826 cm-1 at relatively high temperature on the Pd(111) single crystal.42 Interestingly, this species is absent on the Ag-Pd/Al2O3 catalyst. It has been proposed that the dinitrosyl species is an active species for N2O formation.53 The absence of dinitrosyl species as well as the high rate of N2O formation on Ag-Pd/Al2O3 (see table 3) suggests that Ag may promote the rapid conversion of the dinitrosyl species to N2O. Figure 8b and 10b shows gaseous NH3 emerged prior to H2O, revealing that adsorbed hydrogen is more reactive toward Nads than Oads on both Pd/Al2O3 and Ag-Pd/Al2O3 catalysts. The major effect of Ag is to delay the formation of NH3 and H2O. The MS profile in Figure 10b shows that the initial NH3 formation occurred 0.3 min after that of N2, O2, and N2O on Ag-Pd/Al2O3. These observations suggest that the presence of Ag on Pd/Al2O3 suppresses initial hydrogenation of adsorbed nitrogen and adsorbed oxygen, allowing these species to recombine to form molecular N2 and O2, respectively. Ag has long been known to exhibit very low activity for hydrogen chemisorption due to a high activation barrier for H2 dissociation.54 H2 temperature-programmed desorption (TPD) studies show that the addition of Ag on Pd/Al2O3 significantly suppresses hydrogen adsorption on the Pd catalyst surface.55 Acetylene hydrogenation study has also shown that addition of Ag on Pd inhibits the formation of Pd hydride.56 Low activity of Ag for hydrogen chemisorption is suggested as a key factor in suppressing hydrogenation of ethylene and enhancing the selectivity of acetylene hydrogenation to ethylene.17 Ag-Pd has been shown to form an alloy in a hydrogen reducing environment and dealloy in an oxygen environment.57,58 The dealloying process by oxygen causes the enrichment of Ag on the Ag-Pd surface.58,59 The enrichment process has been shown to be driven by the high adsorption energy of oxygen on a specific metal,60 such as Ag. The surface enrichment of an alloying species and realloy process has also been reported for Au-Pt61 and Ni-Cu62 alloys. This surface behavior may explain why NO was first converted to N2, O2, and N2O in the NO decomposition pathway during the first 0.25 min and then converted to NH3 and H2O, shown in Figure 10b. The initial NO decomposition can be attributed to the presence of Ag-Pd alloy on Ag-Pd/Al2O3 due to its exposure to steady state H2 flow. The presence of this Ag-Pd alloy is supported by the absence of bridged CO on Ag-Pd/Al2O3 in Figure 2a. The Ag-Pd could facilitate desorption of N2 and O2, providing sites for NO dissociation. The remaining oxygen (Oret) on Ag-Pd exhibit a peculiar behavior which (i) give a low Ea for O2 formation during NO decomposition and (ii) show a significant O2 desorption profile upon increasing temperature. These Oret on the catalyst surface may interact with Ag on Pd through dealloying, allowing the formation of Pd metal sites for the adsorption of Pd0-NO63 which react with hydrogen to form NH3 and H2O. The dealloying process is a solid state atomic diffusion process which possesses high Ea.64 Thus, dealloying could only occur at 773 and 823 K at a very limited extent and does not occur at 723 K. Furthermore, the presence of hydrogen would remove the retained oxygen, produced from NO dissociation, allowing the catalyst surface to return to the alloying state. The lack of NH3 and H2O formation on Ag-Pd/Al2O3 at 723 K suggests that the dealloying process did not occur at this temperature. Ag-Pd/Al2O3 is more selective for the formation of N2 from dissociated NO than that of NH3 compared to Pd/

Miller and Chuang Al2O3 which is not selective between N2 and NH3 formation at 723 K. Thus, it should be feasible to increase the activity and selectivity of Ag-Pd/Al2O3 for catalyzing the NO decomposition pathway by controlling the reaction temperature and the amount of hydrogen exposure. 5. Conclusions The pulse transient responses show that NO decomposes over Pd/Al2O3 to form products in the following sequence: N2, N2O, and then O2. During the NO pulse, the intensity of adsorbed Pd0-NO at 1735 cm-1 decreased while the intensity for Pd+-NO at 1769 cm-1 and the amount of gaseous N2 product increased. Subsequent NO pulses increased the intensity of Pd+-NO and slowed down the N2 product response, confirming that the reaction occurs on the Pd0 sites and oxygen retained on the catalyst oxidizes Pd0 to Pd+ sites. The addition of Ag to Pd/Al2O3 shifted the O2 response forward. This leading behavior for oxygen desorption from Ag-Pd/Al2O3 appears to be related to its low Ea for O2 formation. Oret on Ag-Pd exhibits a unique activity where those Oret produced from dissociated NO at 723 K can be desorbed at 773 K and those produced at 773 K can be desorbed at 823K. These Oret species are not able to oxidize Pd0 on Ag-Pd/Al2O3 to Pd+ in contrast to those on Pd. Improving oxidation resistance of Pd did not result in increasing the catalytic cycle for NO decomposition. NO pulse into H2 (i.e., NO reduction) over Pd/Al2O3 produced N2, N2O, O2, NH3, and H2O. The addition of Ag allowed the formation of N2, O2, and N2O in the absence of NH3 and H2O responses at 723 K indicating Ag delayed and suppressed NH3/ H2O formation. This observation indicates that adsorbed hydrogen on Ag-Pd/Al2O3 is effective in maintaining Pd in the reduced state for NO decomposition; adsorbed hydrogen species is not able to react with adsorbed nitrogen and oxygen for NH3 and H2O formation at 723 K. Increasing the temperature above 723 K decreased the extent of Ag suppression for NH3 and H2O formation. Acknowledgment. We acknowledge partial support from Ohio Board of Regents, Ohio Coal Development Office, and FirstEnergy Corp. for supporting this research. We also thank Dr. Abhaya Datye and Dr. Hugo Zea at the Center for Microengineered Materials and Department of Chemical and Nuclear Engineering, University of New Mexico, for the preparation of the Pd and Ag-Pd catalysts. We thank Dr. Khalid Almusaiteer for his suggestions and comments. References and Notes (1) Li, L. D.; Yu, J. J.; Hao, Z. P.; Xu, Z. P. J. Phys. Chem. C 2007, 111, 10552. (2) Banas, J.; Najbar, M.; Tomasic, V. Catal. Today 2008, 137, 267. (3) Chuang, S. S. C.; Tan, C.-D. J. Phys. Chem. B 1997, 101, 3000. (4) Costa, C. N.; Efstathiou, A. M. J. Phys. Chem. C 2007, 111, 3010. (5) Arve, K.; Svennerberg, K.; Klingstedt, F.; Eraenen, K.; Wallenberg, L. R.; Bovin, J. O.; Capek, L.; Murzin, D. Y. J. Phys. Chem. B 2006, 110, 420. (6) Chi, Y.; Chuang, S. S. C. Catal. Today 2000, 62, 303. (7) Hightower, J. W.; Leirsburg, D. A. V. The Catalytic Chemistry of Nitrogen Oxides; Plenum Press: New York, 1975. (8) Haneda, M.; Kintaichi, Y.; Nakamura, I.; Fujitani, T.; Hamada, H. J. Catal. 2003, 218, 405. (9) Ozensoy, E.; Hess, C.; Goodman, D. W. J. Am. Chem. Soc. 2002, 124, 8524. (10) Niemantsverdriet, J. W. Spectroscopy in Catalysis; Wiley-VCH: New York, 2000. (11) Simionato, M.; Assaf, E. M. Mater. Res. (Sao Carlos, Braz.) 2003, 6, 535. (12) Li, X.; Zhu, P.; Wang, F.; Wang, L.; Tsang, S. C. J. Phys. Chem. C 2008, 112, 3376.

Catalytic Decomposition of Nitric Oxide (13) Shimizu, K. I.; Sugino, K.; Kato, K.; Yokota, S.; Okumura, K.; Satsuma, A. J. Phys. Chem. C 2007, 111, 6481. (14) MacLeod, N.; Lambert, R. M. Appl. Catal. B: EnViron. 2002, 35, 269. (15) Gervasini, A.; Manzoli, M.; Martra, G.; Ponti, A.; Ravasio, N.; Sordelli, L.; Zaccheria, F. J. Phys. Chem. B 2006, 110, 7851. (16) Parvulescu, V. I.; Grange, P.; Delmon, B. Catal. Today 1998, 46, 233. (17) Jin, Y.; Datye, A. K.; Rightor, E.; Gulotty, R.; Waterman, W.; Smith, M.; Holbrook, M.; Maj, J.; Blackson, J. J. Catal. 2001, 203, 292. (18) Miller, D. D.; Chuang, S. S. C. J. Taiwan Inst. Chem. Eng. 2009, accepted for publication. (19) Klug, H. P.; Alexander, L. E. X-ray Diffraction Procedures, 2nd ed.; John Wiley & Sons: New York, 1974. (20) Sirita, J.; Phanichphant, S.; Meunier, F. C. Anal. Chem. 2007, 79, 3912. (21) Skoog, D. A.; Holler, F. J.; Crouch, S. R. Principles of Instrumental Analysis, 6th ed.; Thomson Brooks/Cole: Belmont, 2007. (22) Eischens, R. P.; Pliskin, W. A.; Francis, S. A. J. Chem. Phys. 1954, 22, 1986. (23) Gao, F.; Lundwall, M.; Goodman, D. W. J. Phys. Chem. C 2008, 112, 6057. (24) Almusaiteer, K.; Chuang, S. S. C. J. Catal. 1998, 180, 161. (25) Chi, Y.; Chuang, S. S. C. J. Catal. 2000, 190, 75. (26) Grill, C. M.; Gonzalez, R. D. J. Phys. Chem. 1980, 84, 878. (27) Almusaiteer, K.; Chuang, S. S. C. J. Catal. 1999, 184, 189. (28) Nwalor, J. U.; Goodwin, J. G., Jr. Top. Catal. 1994, 1, 285. (29) Chuang, S. S. C.; Guzman, F. Top. Catal. 2009, accepted for publication. (30) Brown, W. A.; Gardner, P.; King, D. A. J. Phys. Chem. 1995, 99, 7065. (31) Consul, J. M. D.; Peralta, C. A.; Benvenutti, E. V.; Ruiz, J. A. C.; Pastore, H. O.; Baibich, I. M. J. Mol. Catal. A: Chem. 2006, 246, 33. (32) Eberhardt, A. M.; Benvenutti, E. V.; Moro, C. C.; Tonetto, G. M.; Damiani, D. E. J. Mol. Catal. A: Chem. 2003, 201, 247. (33) Gervasini, A.; Carniti, P.; Ragaini, V. Appl. Catal., B 1999, 22, 201. (34) Konduru, M. V.; Chuang, S. S. C. J. Catal. 1999, 187, 436. (35) Konduru, M. V.; Chuang, S. S. C. J. Catal. 2000, 196, 271. (36) Marins de Oliveira, A.; Crizel, L. E.; Silveira da Silveira, R.; Pergher, S. B. C.; Baibich, I. M. Catal. Commun. 2007, 8, 1293. (37) Pergher, S. B. C.; Dallago, R. M.; Veses, R. C.; Gigola, C. E.; Baibich, I. M. J. Mol. Catal. A: Chem. 2004, 209, 107. (38) Sica, A. M.; Dos Santos, J. H. Z.; Baibich, I. M.; Gigola, C. E. J. Mol. Catal. A: Chem. 1999, 137, 287.

J. Phys. Chem. C, Vol. 113, No. 33, 2009 14971 (39) Tonetto, G. M.; Ferreira, M. L.; Damiani, D. E. J. Mol. Catal. A: Chem. 2003, 193, 121. (40) Wu, R. J.; Chou, T. Y.; Yeh, C. T. Appl. Catal., B 1995, 6, 105. (41) Dozova, N.; Krim, L.; Alikhani, M. E.; Lacome, N. J. Phys. Chem. A 2006, 110, 11617. (42) Hess, C.; Ozensoy, E.; Yi, C.-W.; Goodman, D. W. J. Am. Chem. Soc. 2006, 128, 2988. (43) Consul, J. M. D.; Peralta, C. A.; Ruiz, J. A. C.; Pastore, H. O.; Baibich, I. M. Catal. Today 2008, 133-135, 475. (44) Smeets, P. J.; Groothaert, M. H.; van Teeffelen, R. M.; Leeman, H.; Hensen, E. J. M.; Schoonheydt, R. A. J. Catal. 2007, 245, 358. (45) Ozensoy, E.; Hess, C.; Goodman, D. W. Top. Catal. 2004, 28, 13. (46) Nakamura, I.; Fujitani, T.; Hamada, H. Surf. Sci. 2002, 514, 409. (47) Dujardin, C.; Twagirashema, I.; Granger, P. J. Phys. Chem. C 2008, 112, 17183. (48) Rainer, D. R.; Vesecky, S. M.; Koranne, M.; Oh, W. S.; Goodman, D. W. J. Catal. 1997, 167, 234. (49) Penner, S.; Bera, P.; Pedersen, S.; Ngo, L. T.; Harris, J. J. W.; Campbell, C. T. J. Phys. Chem. B 2006, 110, 24577. (50) Furusawa, T.; Aika, K.-i. Bull. Chem. Soc. Jpn. 2000, 73, 795. (51) Holles, J. H.; Davis, R. J.; Murray, T. M.; Howe, J. M. J. Catal. 2000, 195, 193. (52) Hadjiivanov, K. I. Catal. ReV. - Sci. Eng. 2000, 42, 71. (53) Queeney, K. T.; Friend, C. M. J. Chem. Phys. 1997, 107, 6432. (54) Cantini, P.; Mattera, L.; De Kieviet, M. F. M.; Jalink, K.; Tassistro, C.; Terreni, S. Surf. Sci. 1989, 872, 211–212. (55) Ngamsom, B.; Bogdanchikova, N.; Avalos Borja, M.; Praserthdam, P. Catal. Commun. 2004, 5, 243. (56) Zhang, Q.; Li, J.; Liu, X.; Zhu, Q. Appl. Catal., A 2000, 197, 221. (57) Gonzalez, S.; Neyman, K. M.; Shaikhutdinov, S.; Freund, H.-J.; Illas, F. J. Phys. Chem. C 2007, 111, 6852. (58) Moss, R. L.; Thomas, D. H. J. Catal. 1967, 8, 151. (59) Bouwman, R.; Lippits, G. J. M.; Sachtler, W. M. H. J. Catal. 1972, 25, 350. (60) Tao, F.; Grass Michael, E.; Zhang, Y.; Butcher Derek, R.; Renzas James, R.; Liu, Z.; Chung Jen, Y.; Mun Bongjin, S.; Salmeron, M.; Somorjai Gabor, A. Science 2008, 322, 932. (61) Bouwman, R.; Sachtler, W. M. H. J. Catal. 1970, 19, 127. (62) Elford, L.; Mueller, F.; Kubaschewski, O. Ber. Bunsenges. Phys. Chem. 1969, 73, 601. (63) Miller, D. D.; Chuang, S. S. C. Catal. Commun. 2009, 10, 1313. (64) Smith, W. F.; Hashemi, J. Foundations of Materials Science and Engineering, 4th ed.; McGraw-Hill: New York, 2006.

JP904538T