The different NO_x trap performance on ceria and barium/ceria containing LNT catalysts below 200 °C

Abstract

The low-temperature NO_x storage efficiency of Ce-containing lean NO_x trap catalysts was investigated using temperature programmed adsorption. The incorporation of ceria either as a NO_x storage phase or as the support for barium oxides improves the NO_x storage capacity from 200 °C to 350 °C. However, as temperature increases from 100 °C to 150 °C, the rate of NO_x storage slows down over the catalysts containing ceria as a storage phase, while this phenomenon is less obvious if ceria is used as the support for barium oxides. By investigating the individual components of catalysts using NO-TPD, we found that the NO_x release was mainly from ceria. Based on the in situ DRIFT data, we conclude that nitrite species are formed on ceria during the NO_x storage at low temperatures, and the nitrite species can be activated and transformed to nitrate species upon further oxidation, accompanied by NO release; on the other hand, NO release occurs weakly on barium/ceria as the nitrites are more stable on the more basic Ba phase than on the ceria. In addition, the dispersion of barium species on ceria is higher than that on γ-Al₂O₃, resulting in the formation of a larger amount of basic sites. The addition of barium onto ceria not only improves the stability of ad-NO_x species but also positively impacts the amounts of NO_x trapped.

---

1. Introduction

Compared with normal gasoline engines, lean burn gasoline and diesel engines are more fuel efficient and emit less CO₂. However, the conventional three-way catalyst is ineffective in eliminating NO_x under oxygen-rich (lean) conditions. There is a general consensus that a lean NO_x trap (LNT) catalyst, comprising precious metals (Pt, Rh) and alkaline-earth metal oxide (e.g. BaO) supported on a high surface area support (γ-Al₂O₃), is one of the promising solutions to reduce NO_x emissions from lean-burn engine exhaust.¹ During the lean cycle (normal engine operation), NO_x is stored on the alkaline-earth metal oxide in the form of nitrites and nitrates for a few minutes. As the LNT catalyst becomes saturated with stored NO_x, a rich transient is introduced lasting a few seconds, which reduces the stored NO_x to N₂.

One of the major challenges facing LNT catalysts is insufficient NO_x storage capacity at low temperature.² The average exhaust temperature from diesel cars is below 150 °C for the city portion (ECE, first 800 seconds) of the European driving cycle (NEDC, 1200 seconds). A current strategy of most of the car manufacturers is to trap NO_x for the entire ECE portion and regenerate the catalyst during the high-temperature, high-speed portion (EUDC, the last 400 seconds).

Because of their oxygen storage capacity and their ability to modulate oxygen partial pressure of automotive exhaust around the stoichiometric point, ceria-containing materials are widely used in three-way catalysts. Recently, ceria-containing materials were found to trap NO_x below 400 °C in a lean environment,^3–6 which increases the overall NO_x storage capacity of a Ba–NO_x trap.⁷ However, there is limited work focusing on the stability of adsorbed NO_x species below 150 °C, which is an important issue for the application of LNT catalysts because diesel car exhaust is typically below 150 °C for city driving and the desorption of NO_x species upon acceleration will decrease the overall NO_x conversion.

In this work, we report the low-temperature, NO_x storage efficiency on ceria-containing catalysts using temperature programmed adsorption (TPA) experiments. Ceria serves as either a NO_x trapping component or a support for Ba. Combined with the results of in situ DRIFT and NO-TPD experiments, we discuss the relationship between the stability of adsorbed NO_x species and NO_x trap efficiency.

2. Experimental

2.1 Catalysts preparation

Pt/Al₂O₃ (133.5 m² g⁻¹) was prepared by impregnating Pt(NO₃)₂ solution on a γ-Al₂O₃ (136.5 m² g⁻¹) support using the incipient wetness technique to reach a Pt loading of 2% by weight. The impregnated powder was dried in the air overnight at 100 °C and then calcinated in air at 500 °C for 5 h.

20 wt% of BaO was loaded on CeO₂ (BC, 95.8 m² g⁻¹) or Al₂O₃ (BA, 90.0 m² g⁻¹) by impregnating barium acetate solution on the powders. The impregnated powder was dried in the air overnight at 100 °C and subsequently calcinated in air at 500 °C for 5 h. γ-Al₂O₃ and CeO₂ (153.8 m² g⁻¹) were provided by BASF Corporation.

2.2 Reaction tests and characterizations

2.2.1 Temperature programmed adsorption (TPA). Pt/Al₂O₃ (PA, 0.125 g) was mechanically mixed with BaCO₃/Al₂O₃ (BA, 0.25 g), CeO₂ (Ce, 0.125 g) or BaCO₃/CeO₂ (BC, 0.25 g) to form catalysts PA–BA, PA–Ce or PA–BC. The catalyst was mixed with quartz sand, and the total amount of powder (catalyst + quartz) has been kept at 2.0 g. The total flow rate was 1 L min⁻¹, yielding a space velocity of 30 000 h⁻¹.

A NICOLET 380 FT-IR spectrometer, equipped with a two meter gas cell, was used to detect the outlet concentrations of NO, NO₂, NH₃, N₂O, CO, CO₂, and H₂O (g) at a 4 second interval. TPA was performed within the temperature range of 100–400 °C with a feed gas consisting of 120 ppm NO, 10% O₂, 5% CO₂, 6% H₂O, and N₂ balance. A typical TPA experimental consists the following steps: (1) heating a sample to 450 °C in 5% CO₂–6% H₂O–N₂ balance, (2) reducing it at the same temperature in 7.5% CO–5% CO₂–6% H₂O–N₂ balance for 5 min, (3) cooling the sample to 100 °C in 5% CO₂–6% H₂O–N₂ balance, (4) holding the sample at 100 °C for 5 min in 120 ppm NO–10% O₂–5% CO₂–6% H₂O–N₂ balance, and (5) heating the sample to 400 °C at 10 °C min⁻¹ in the same environment. The NO_x storage efficiency is defined as

NO_x storage efficiency = (120 − NO_xoutlet)/120 × 100%

where 120 is the inlet [NO_x] in ppm.

2.2.2 In situ DRIFTS. The in situ DRIFT experiments were performed on a NICOLET 6700 FT-IR spectrometer equipped with a MCT detector. Each spectrum was collected at a resolution of 4 cm⁻¹ with 10 scans. A sample was first purged with N₂ from room temperature to 500 °C at 10 °C min⁻¹ and kept at this temperature in 10% O₂–N₂ for 30 min. The sample was subsequently cooled to 100 °C in N₂, and background spectra were then collected.

NO adsorptions were conducted by introducing 120 ppm NO–10% O₂–N₂ balance into the chamber at 100 °C for 40 min, purging the sample with N₂ for 10 min, and heating the sample to 200 °C at 10 °C min⁻¹ in N₂ flow (100 mL min⁻¹). The adsorbed species were recorded as a function of time/temperature.

2.2.3 NO-TPD. The stability of adsorbed NO_x species was measured by NO-TPD measurement. A catalyst, with the same weight as that in the TPA experiment, was placed in the tube reactor. The catalyst was first pre-treated at 500 °C for 30 min in 10% O₂–N₂ flow and then cooled to 100 °C in N₂ flow. The adsorption of NO_x, with the total flow rate of 1 L min⁻¹, was carried out at 100 °C in 120 ppm NO–10% O₂–N₂ balance for 5 min. After flushing the catalyst with N₂ flow at 100 °C for 12 min, we increased the temperature to 550 °C at a rate of 10 °C min⁻¹ in N₂ flow (500 mL min⁻¹). The concentrations of NO_x species were monitored by a NICOLET 380 FT-IR spectrometer.

2.2.4 CO₂-TPD. Catalyst basicity was measured by CO₂-TPD using the tube reactor with 0.3 g of sample. Before measurement, a catalyst was pre-treated at 500 °C for 30 min in 10% O₂–N₂ balance, and subsequently the sample was reduced in 7.5% CO–N₂ balance for 5 min at 450 °C and cooled to 30 °C in N₂ flow. The CO₂ adsorption was conducted with 25% CO₂–N₂ (200 mL min⁻¹) at 30 °C until saturation. After flushing under N₂, we increased the temperature to 650 °C at 10 °C min⁻¹ in N₂ flow (40 mL min⁻¹). The CO₂ concentration was monitored by a NICOLET 380 FT-IR spectrometer.

2.2.5 XRD. The powder X-ray diffraction (XRD) patterns were obtained at room temperature using a Bruker D8 Focus, equipped with nickel-filtered Cu K_α radiation (λ = 1.54056 Å). The spectrometer, operating at 40 kV and 40 mA, scanned the samples from 10° to 90° with a 0.02 step size. The average crystallite sizes of BaCO₃ were calculated using JADE 5.

2.2.6 HR-TEM. High-resolution transmission electron microscope (HR-TEM) analyses were performed using a top entry Jeol JEM 2100 (200 kV) microscope. Catalyst powder was dissolved in absolute ethyl alcohol and deposited onto 230 mesh Cu grids coated with porous carbon film.

3. Results

3.1 NO_x storage efficiency

Fig. 1 compares the NO_x storage efficiency on PA–BA, PA–Ce and PA–BC samples as a function of temperature. During temperature ramp, the highest NO_x storage efficiency over PA–BA is 24% at 305 °C, and those over PA–Ce and PA–BC are 29% at 260 °C and 43% at 330 °C, respectively. The improvement in NO_x storage efficiency by incorporating ceria into Ba-based NSR catalysts is obvious, which is in consistent with the reported results.^3,8–12

Fig. 1 NO_x storage efficiency obtained by TPA experiments. The samples were kept under 120 ppm NO–10% O₂–5% CO₂–6% H₂O–N₂ balance for 5 min, and then the temperature was increased to 400 °C at 10 °C min⁻¹ under the same atmosphere conditions.

However, there is a decrease in NO_x storage efficiency as temperature increases from 100 °C to 150 °C, resulting in a minimum net storage efficiency of 2% on PA–BA, −2% on PA–Ce, and 8% on PA–BC. Net NO_x release (catalyst outlet [NO_x] is higher than the inlet [NO_x]) was found on PA–Ce. The decrease in NO_x storage efficiency with temperature ramp arises from the desorption/decomposition of the unstable stored NO_x during the isothermal period (first 5 min at 100 °C, see Section 2.2.1). The net NO_x release on PA–Ce indicates that a large amount of unstable adsorbed NO_x species desorbs easily.

3.2 In situ DRIFTS

The stability of adsorbed NO_x species was investigated by in situ DRIFT experiments. After an oxidation pre-treatment, a sample was kept in NO–O₂ flow at 100 °C for 40 min. After purging the sample with N₂ for 10 min, we increased the temperature from 100 to 200 °C at 10 °C min⁻¹.

Fig. 2a shows the DRIFT spectra of NO_x adsorption on PA–BA. The main adsorption band appeared at 1230 cm⁻¹ on PA–BA due to the formation of Ba nitrites,¹³ and its intensity increased with time. Another band was found at 1310 cm⁻¹, with a lower intensity than that at 1230 cm⁻¹, attributed to the monodentate nitrates on BaO. A band appeared at 1532 cm⁻¹, due to the formation of bidentate nitrates.^14–16 The band at 1532 cm⁻¹ later blue shifted to 1546 cm⁻¹. The intensity of the 1546 cm⁻¹ band on PA–BA is lower than those of two other major bands (1230 and 1310 cm⁻¹), and this may be due to the low surface sensitivity of DRIFT because of the surface-scattering effects,^17–19 where the NO_x species in the bulk of storage material could not be detected. Fig. 2c shows the desorption spectra of the NO_x species adsorbed on PA–BA. After heating up to 200 °C in N₂, we found that the intensity of the 1230 cm⁻¹ (Ba nitrites) band decreased slightly as temperature increased from 100 to 200 °C, while that of the 1310 cm⁻¹ (Ba nitrates) band remained the same. The intensity of the band at 1546 cm⁻¹ was almost unchanged.

Fig. 2 DRIFT spectra of (a) NO_x adsorptions, (b) N₂ purging and (c) temperature ramp on PA–BA. The samples were kept under 120 ppm NO–10% O₂–N₂ balance at 100 °C for 40 min. Subsequently, the samples were purged with N₂ for 10 min and were heated up to 200 °C at 10 °C min⁻¹ in N₂ flow.

Fig. 3a shows the DRIFT spectra of NO_x adsorption on PA–Ce. The band emerges at 1180 cm⁻¹ with high intensity due to the asymmetric N–O stretching mode of chelating bidentate nitrites adsorbed on the surface of CeO₂.^20–22 Its intensity increased with time up to 25 min and then decreased slightly after that. The band at 1306 cm⁻¹ may be another type of nitrite species, which gradually red shifted to 1292 cm⁻¹ during NO_x adsorption.

Fig. 3 DRIFT spectra of (a) NO_x adsorptions, (b) N₂ purging and (c) temperature ramp on PA–Ce. The samples were kept under 120 ppm NO–10% O₂–N₂ balance at 100 °C for 40 min. Subsequently, the samples were purged with N₂ for 10 min and were heated up to 200 °C at 10 °C min⁻¹ in N₂ flow.

The intensities of the 1180 and 1292 cm⁻¹ bands decreased obviously during the purge period (Fig. 3b), while that of the band at 1230 cm⁻¹ on PA–BA remained unchanged (Fig. 2b). Upon increasing the temperature to 200 °C, however, the band at 1292 cm⁻¹ disappeared completely and the band at 1180 cm⁻¹ decreased significantly and red shifted to 1160 cm⁻¹ (Fig. 3c). These results indicate that the nitrite species formed on ceria are unstable.

For nitrate species shown in Fig. 3a, the bands at 1541 and 1255 cm⁻¹ are due to the formation of chelating bidentate nitrates.²³ And the band at 1541 cm⁻¹ gradually blue shifted to 1548 cm⁻¹ as the adsorption proceeded. The discrete frequency shifts to higher wavenumbers can be explained by the repulsive interactions among the adsorbed molecules at a higher surface coverage.²¹ Further, we found a new pair of bands at 1527 and 1225 cm⁻¹ and assign them to another kind of nitrates.^22,23

During N₂ purging (Fig. 3b), the band at 1545 cm⁻¹ increased unexpectedly but decreased slightly with temperature (Fig. 3c). On the other hand, the bands at 1527 and 1223 cm⁻¹ grew with temperature in N₂.

Fig. 4 depicts the evolution of nitrites and nitrates on PA–Ce under NO–O₂ conditions at 100 °C for 40 min. The intensities of the bands at 1180 cm⁻¹ and 1541–1548 cm⁻¹ reflect the surface concentrations of nitrites and nitrates, respectively. The intensity of the 1180 cm⁻¹ band increases quickly for the initial 20 min, gradually reaching to its peak between 25 and 30 min, and then decreases slightly after 30 min. The nitrite species strongly adsorbs on the surface of ceria and proceeds with the process of adsorption → saturation → decomposition. Meanwhile, the formation of chelating bidentate nitrates (1545 cm⁻¹) is relatively slow during the first 20 min but increases quickly after 35 min when the rate of formation of nitrites (1180 cm⁻¹) is near constant or decreasing. It is possible that the decomposition of nitrites (1180 cm⁻¹) creates O*, which facilitates the formation of nitrates (1545 cm⁻¹). Moreover, the DRIFT spectra collected during N₂ purging (Fig. 3b) show that PA–Ce displays the opposite trends in surface concentration between nitrites (1180 cm⁻¹) and nitrates (1545 cm⁻¹). Based on this observation, we can propose a pathway for the formation of the chelating bidentate nitrites. As shown in Scheme 1 (pathway I), surface saturation leads to the formation of sites with two neighboring adsorbed nitrites; one surface nitrite is activated by weakening one of its N–O bonds, which decomposes easily, leaving O* on the surface and releasing NO into gas phase; the other nitrite then reacts with O* forming a surface nitrate.

Fig. 4 The evolution of the intensities for nitrites and nitrates on PA–Ce by in situ DRIFT experiments. The sample was kept under 120 ppm NO–10% O₂–N₂ balance at 100 °C for 40 min.

Scheme 1 The transformation of nitrite to nitrate species under NO–O₂ conditions at low temperatures.

On the other hand, many authors reported the direct oxidation of nitrites by surface O* of the catalyst^22,24,25 and observed weak vibration of nitrites due to enhanced transformation from nitrite to nitrate by fast oxygen mobility at high temperatures.²⁶ During the temperature ramp in N₂ (Fig. 3c), unlike other bands, the bands at 1527 and 1225 cm⁻¹ increased, with temperature, perhaps due to the formation of another kind of nitrate species.²² Therefore, another pathway for the formation of nitrates can be proposed. As shown in Scheme 1 (pathway II), higher temperature increases oxygen mobility on ceria and promotes the formation of intrinsic O*, and O* can oxidize the nitrite species forming nitrates.

Fig. 5a shows the DRIFT spectra of NO_x adsorption on PA–BC. The intensive nitrite bands that appeared at 1198 cm⁻¹ and 1130 cm^{−110,27–30} remained unchanged during N₂ purging (Fig. 5b) but decreased slightly with increasing temperature (Fig. 5c). Comparing with the nitrite species on PA–Ce, the nitrite species on both the PA–BA and PA–BC samples exert the higher stability.

Fig. 5 DRIFT spectra of (a) NO_x adsorptions, (b) N₂ purging and (c) temperature ramp on PA–BC. The samples were kept under 120 ppm NO–10% O₂–N₂ balance at 100 °C for 40 min. Subsequently, the samples were purged with N₂ for 10 min and were heated up to 200 °C at 10 °C min⁻¹ in N₂ flow.

3.3 NO-TPD

The stability of adsorbed NO_x species was also investigated by NO-TPD experiments. Fig. 6a shows NO-TPD profiles on the three components: BA, Ce and BC. After an oxidation treatment, the sample had been kept under NO–O₂ atmospheres for 5 min at 100 °C. The sample was then purged for 12 min with N₂ and heated up to 400 °C at 10 °C min⁻¹ in N₂.

Fig. 6 The evolution of NO_x concentration during NO-TPD experiments. The samples were kept under 120 ppm NO–10% O₂–N₂ balance for 5 min at 100 °C. After N₂ purging at 100 °C for 12 min, the temperature was increased to 550 °C with 10 °C min⁻¹ under N₂ flow.

Table 1 shows the amount of NO_x adsorbed during the adsorption period, the amounts of NO_x desorbed during purge and TPD periods. The integrated amount of NO_x in (purging + TPD) period is slightly higher than that in (NO_x adsorption) period. This is because the NO_x species in the gas phase cannot be cleaned up immediately at the initial period of N₂ purging. However, the mass balance of adsorbed and desorbed NO_x is better than 95% for all samples. The amounts of adsorbed NO_x on CeO₂ (12.4 μmol) and BC (12.8 μmol) are higher than that on BA (9.5 μmol). During the purge period, the amount of desorbed NO_x follows the order, CeO₂ (3.4 μmol) > BA (2.5 μmol) ≈ BC (2.4 μmol). The TPD results show that ceria enhances NO_x adsorption, and the ad-NO_x is more stable when barium is also present.

Table 1 The adsorption–purge–desorption amount of NO_x by NO-TPD experiments. The samples were kept under 120 ppm NO–10% O₂–N₂ balance for 8 min at 100 °C. After N₂ purging at 100 °C for 8 min, the temperature was increased to 550 °C with 10 °C min⁻¹ under N₂ flow

Sample	NOx adsorption (μmol)	Purging (μmol)	TPD (μmol)
BA	9.5	2.5	7.0
CeO2	12.4	3.4	9.4
BC	12.8	2.4	11.2

---

On BA, the rate of NO_x desorption was constant below 400 °C, and as temperature increased further, NO_x desorption peaked at 470 °C. The two ceria containing components show very different desorption features from BA. Ceria has three desorption peaks, a weaker one at 160 °C and two strong ones at 250 and 420 °C. Similarly, BC has a weaker peak at 160 °C, but the two stronger peaks appear at higher temperatures (330 and 490 °C) compared with those on CeO₂. This suggests that the NO_x species adsorbed on BC is more stable than that on CeO₂.

Fig. 6b shows the comparison of TPD profiles between PA–Ce and PA–BC with PA–BA as reference. On PA–Ce, a large amount of NO desorbs around 160 and 230 °C, while on PA–BC the first desorption peak appears around 280 °C. Our DRIFT spectra show that the nitrite species (1292 cm⁻¹ and 1180 cm⁻¹) are almost disappeared below 200 °C on PA–Ce (Fig. 3c), while they (1198 cm⁻¹ and 1130 cm⁻¹) are still at a relatively high level on PA–BC (Fig. 5c). This confirms that barium oxide enhances the stability of surface nitrite species.

Both the PA–BA and PA–BC samples exhibit two main NO desorption peaks around 300 °C and 475 °C, i.e. the stability of adsorbed NO_x species on PA–BC is similar to that on PA–BA. However, the amount of adsorbed NO_x species on PA–BC is about 1.5 times of that on PA–BA.

3.4 CO₂-TPD

The basicity of the catalysts is crucial to the stability of the adsorbed NO_x species. Fig. 7 compares the basicity of BA, CeO₂ and BC measured by CO₂-TPD. The integrated quantity of CO₂ desorbed decreases in the following order: BC (222.0 μmol g⁻¹) > BA (116.7 μmol g⁻¹) ≈ CeO₂ (112.6 μmol g⁻¹). The BC sample possesses the largest amount of basic sites among the samples investigated in this work.

Fig. 7 The evolution of CO₂ concentration during CO₂-TPD experiments. The samples were kept under 25% CO₂–N₂ until saturation was achieved. After purging under N₂, the temperature was increased from 30 °C to 650 °C at 10 °C min⁻¹ in N₂ flow.

Moreover, we can attribute different basic strengths depending on the CO₂ desorption temperature. Both the CeO₂ and BC samples exhibit one intense peak around 100 °C and a long CO₂ desorption tail up to 550 °C. While for the BA sample, there is a CO₂ desorption peak with a lower intensity around 70 °C, and an intensive desorption peak at 600 °C. According to our previous work, the peak around 100 °C is due to the decomposition of bicarbonates.³¹ All of the CeO₂, BC and BA samples exhibit the similar amount of such kinds of weak basic sites. The evolution of CO₂ in the 200–500 °C range can be attributed mainly to the decomposition of amorphous BaCO₃.^32,33 The BC sample possesses the largest amount of such kinds of medium basic sites, while it is limited to detection on the BA sample. And the intensive CO₂ peak at 600 °C can be attributed to the decomposition of crystalline BaCO₃.³³ The BA sample exerts the highest amount of such strong basic sites.

Piacentini et al. found that the thermal stability of supported BaCO₃ species depends on their interaction with the alumina support:³³ (1) the BaCO₃ species, being in intimate contact with the support, are unstable. They can be decomposed during calcination at 500 °C. (2) The most stable are bulk-like BaCO₃ species, which are not in contact with the alumina support. Therefore, the different CO₂ desorption behavior obtained for the BC and BA samples indicates that the barium species are in close contact with ceria, however, the BaCO₃ species on alumina is more bulk like, although BaO-loading (20 wt%) on CeO₂ and Al₂O₃ is the same.

3.5 XRD

Fig. 8 shows the XRD patterns of the BC, CeO₂, BA and γ-Al₂O₃ samples. The presence of micro-crystalline γ-Al₂O₃ (JCPDS 75-0921) is evident (Fig. 8, curve d). The crystalline BaCO₃ whiterite phase (JCPDS 45-1471) appears on the BA sample (Fig. 8, curve c), and the crystalline size of BaCO₃ equals to 12.8 nm. The results are in agreement with a previous study³³ that the presence of crystalline BaCO₃ occurs for Ba-loadings higher than 16.7 wt%.

Fig. 8 XRD patterns of (a) BC, (b) CeO₂, (c) BA and (d) γ-Al₂O₃.

In the case of the BC sample, the peak profiles show four main peaks between 20° and 80°, indicating the presence of micro-crystalline CeO₂ (JCPDS 65-5923) (Fig. 8 curve a). Ba-containing species are X-ray amorphous (Fig. 8, curve b). It suggests that the Ba phase is well-dispersed over the CeO₂ support. In our previous work,³⁴ we have found that higher Ce content of Ce_xZr_1−xO₂ favors small BaCO₃ particles, which is probably due to Ba–Ce contact hindering crystallization of BaCO₃.³⁵

3.6 HR-TEM

The dispersion of BaCO₃ on the BA and BC samples is supported by the HR-TEM analysis presented in Fig. 9. The presence of (111) planes at 0.37 nm and (122) planes at 0.22 nm on the BA sample is shown by the relevant selected area diffraction pattern of the crystallites, indicating the traces of crystalline BaCO₃. While only the feature of CeO₂ surface lattice strips, i.e. (111) planes at 0.31 nm, (220) planes at 0.19 nm and (222) planes at 0.15 nm, is discernible on the BC sample.

Fig. 9 Transmission electron micrograph of BA (a) and BC (b). Inset illustrates the relevant selected area diffraction pattern of the crystallites.

4. Discussion

For ceria-containing samples, where ceria is a NO_x storage component, there was a net decrease in NO_x storage efficiency (−2%) as temperature increased from 100 °C to 150 °C. When ceria is used as a support for barium (PA–BC), the drop in NO_x storage efficiency (8%) is less. It indicates that the stability of adsorbed NO_x species is crucial to NO_x trapping at low temperatures.

Having exposed ceria-containing samples to NO–O₂ for 40 min at 100 °C, we observed large amounts of nitrites formed on PA–Ce (high-intensity bands at 1180 and 1292 cm⁻¹) and on PA–BC (1198 cm⁻¹ and 1130 cm⁻¹). The rate of formation of chelating bidentate nitrates (1180 cm⁻¹) on PA–Ce is the fastest, the band achieves saturation in 25 min. This suggests that nitrite formed on ceria is a major contributor of its NO_x storage capacity at 100 °C.

However, bidentate nitrites disappeared upon increasing temperature to 200 °C. The net NO release was found below 200 °C on PA–Ce during the TPA experiments (Fig. 1). The decrease in the NO_x storage efficiency coincides with the decomposition/transformation of the bidentate nitrites observed by in situ DRIFT. This suggests that the instability of the bidentate nitrites (1180 cm⁻¹) is the main cause of slowing down the NO_x storage process during temperature ramp. Indeed, as shown in Fig. 3c, the rate of decrease in band intensity at 1180 cm⁻¹–1160 cm⁻¹ (bidentate nitrites) is obviously higher than the rate of the increase at 1550 cm⁻¹ (nitrates). Although, a large amount of nitrites forms on PA–Ce in NO–O₂ at 100 °C, not all the nitrites are transformed to nitrates during our TPA experiment. Therefore, the main cause for NO desorption is the decomposition of nitrite species. The defective ceria surface facilitates the decomposition of nitrites. We found that the band at 1180 cm⁻¹ red shifted to 1160 cm⁻¹ upon increasing temperature from 100 °C to 200 °C. As reported in the literature for both experimental³⁶ and theoretical data,³⁷ the red shift of the 1180 cm⁻¹ band indicates the elongation of the N–O bond in nitrite species formed at oxygen vacancy sites of CeO₂, which is the initiation step for nitrites decomposition.^21,38 In this work, the nitrite species with an elongated N–O bond dissociate into O* and NO, and the latter species release into the gas phase. However, the activation of nitrites (frequency red shift) does not occur on the BC sample (Fig. 5c), indicating that the adsorbed nitrites are not in close contact with ceria, which makes the dissociation of nitrites difficult to happen. Moreover, it is not surprising that nitrites are more stable on PA–BC than on PA–Ce since nitrites would be more stable on the more basic Ba phase than on the support.¹²

The introduction of barium inhibits NO release significantly with temperature ramp, however, the NO_x storage efficiency varies by substitution of alumina for ceria. The PA–BC sample exerts higher NO_x storage efficiency than that on PA–BA within the whole investigated temperature range (100–400 °C). The results of CO₂-TPD, XRD and HR-TEM characterizations manifest that barium species are well-dispersed on ceria while the BaCO₃ phases on alumina are bulk like, although BaO-loading (20 wt%) on CeO₂ and Al₂O₃ is the same. It is well-known that the well-dispersed BaCO₃ provides a larger amount of basic sites and exerts the higher reactivity for NO_x trapping.³³ While the bulk like BaCO₃ exhibits a poor NO_x storage activity, which is affected by diffusional limitation.^39,40 Therefore, the addition of barium onto ceria not only improves the stability of ad-NO_x species but also positively impacts the amounts of NO_x trapped.

Additionally, it is worth noting that the enhanced ad-NO_x stability is beneficial for efficient NO_x storage during lean periods but may decrease the rate of NO_x reduction during rich periods.⁴¹ Considering the practical application, there would be a trade-off between efficient NO_x storage and NO_x reduction determined by the nature of the NO_x storage component.

5. Conclusions

NO_x storage efficiency has been evaluated using TPA experiments. We found that the incorporation of ceria either as a NO_x storage component or as a support for barium increased the NO_x storage capacity of the catalyst significantly between 150 °C and 350 °C. If ceria is used as a NO_x storage component, the NO storage efficiency decreases with increasing temperature between 100 °C and 150 °C, if ceria is used as a support for Ba, however, the decrease in NO storage efficiency is small. Using in situ DRIFT, we can conclude that NO release from the ceria surface is a consequence of the conversion process from surface nitrites to nitrates, accompanied by nitrites decomposition and NO release. The well-dispersed barium phase on ceria is crucial in stabilizing the adsorbed nitrite species and providing more basic sites for NO_x trapping.

Acknowledgements

The authors are grateful to the financial support from the National High Technology Research and Development Program of China (863 Program, 2011AA03A405), the Program of Natural Science Foundation of China (No. 50972104).

Notes and references

1. N. Takahashi, H. Shinjoh, T. Iijima, T. Suzuki, K. Yamazaki, K. Yokota, H. Suzuki, N. Miyoshi, S.-i. Matsumoto, T. Tanizawa, T. Tanaka, S.-s. Tateishi and K. Kasahara, Catal. Today, 1996, 27, 63.
2. M. V. Twigg, Catal. Today, 2011, 163, 33.
3. M. Haneda, T. Morita, Y. Nagao, Y. Kintaichi and H. Hamada, Phys. Chem. Chem. Phys., 2001, 3, 4696.
4. A. Martínez-Arias, J. Soria, J. C. Conesa, X. L. Seoane, A. Arcoya and R. Cataluña, J. Chem. Soc., Faraday Trans., 1995, 91, 1679.
5. S. H. Overbury, D. R. Mullins, D. R. Huntley and L. Kundakovic, J. Catal., 1999, 186, 296.
6. J. A. Rodriguez, T. Jirsak, S. Sambasivan, D. Fischer and A. Maiti, J. Chem. Phys., 2000, 112, 9929.
7. Y. Ji, C. Fisk, V. Easterling, U. Graham, A. Poole, M. Crocker, J.-S. Choi, W. Partridge and K. Wilson, Catal. Today, 2010, 151, 362.
8. M. Crocker, Y. Y. Ji, T. J. Toops, U. M. Graham and G. Jacobs, Catal. Lett., 2006, 110, 29.
9. S. Philipp, A. Drochner, J. Kunert, H. Vogel, J. Theis and E. S. Lox, Top. Catal., 2004, 30–31, 235.
10. A. Drochner, M. O. Symalla, H. Vogel, S. Philipp, U. Gobel and W. Muller, Top. Catal., 2007, 42–43, 199.
11. E. Rohart, V. Belliere-Baca, K. Yokota, V. Harle and C. Pitois, Top. Catal., 2007, 42–43, 71.
12. M. Crocker, Y. Y. Ji and T. J. Toops, Catal. Lett., 2007, 119, 257.
13. J. Szanyi, J. H. Kwak, D. H. Kim, X. Wang, R. Chimentao, J. Hanson, W. S. Epling and C. H. F. Peden, J. Phys. Chem. C, 2007, 111, 4678.
14. G. Centi, S. Perathoner, D. Biglino and E. Giamello, J. Catal., 1995, 152, 75.
15. B. A. Morrow, R. A. McFarlane and L. E. Moran, J. Phys. Chem., 1985, 89, 77.
16. E. Ito, Y. J. Mergler, B. E. Nieuwenhuys, H. van Bekkum and C. M. van den Bleek, Microporous Mater., 1995, 4, 455.
17. A. Urakawa, N. Maeda and A. Baiker, Top. Catal., 2009, 52, 1746.
18. A. Urakawa, N. Maeda and A. Baiker, Angew. Chem., Int. Ed., 2008, 47, 9256.
19. A. Urakawa, E. Roedel, S. Kureti and A. Baiker, Phys. Chem. Chem. Phys., 2008, 10, 6190.
20. B. Azambre, L. Zenboury, F. Delacroix and J. V. Weber, Catal. Today, 2008, 137, 278.
21. K. I. Hadjiivanov, Catal. Rev.: Sci. Eng., 2000, 42, 71.
22. B. Azambre, L. Zenboury, A. Koch and J. V. Weber, J. Phys. Chem. C, 2009, 113, 13287.
23. M. Crocker, Y. Y. Ji, T. J. Toops and J. A. Pihl, Appl. Catal., B, 2009, 91, 329.
24. A. B. Lopez, I. Atribak, B. Azambre and A. Garcia-Garcia, Appl. Catal., B, 2009, 92, 126.
25. A. Garcia-Garcia, I. Atribak, B. Azambre and A. Bueno-Lopez, Top. Catal., 2009, 52, 2092.
26. B. Levasseur, A. M. Ebrahim and T. J. Bandosz, Langmuir, 2011, 27, 9379.
27. J. Libuda, A. Desikusumastuti, T. Staudt and H. Gronbeck, J. Catal., 2008, 255, 127.
28. J. Szanyi, C. W. Yi and J. H. Kwak, J. Phys. Chem. C, 2007, 111, 15299.
29. P. T. Fanson, M. R. Horton, W. N. Delgass and J. Lauterbach, Appl. Catal., B, 2003, 46, 393.
30. I. Nova, L. Castoldi, F. Prinetto, V. Dal Santo, L. Lietti, E. Tronconi, P. Forzatti, G. Ghiotti, R. Psaro and S. Recchia, Top. Catal., 2004, 30–31, 181.
31. X. Q. Wang, M. Q. Shen, L. Y. Song, Y. G. Su and J. Wang, Phys. Chem. Chem. Phys., 2011, 13, 15589.
32. M. Piacentini, M. Maciejewski, T. Burgi and A. Baiker, Top. Catal., 2004, 30–31, 71.
33. M. Piacentini, M. Maciejewski and A. Baiker, Appl. Catal., B, 2005, 59, 187.
34. C. Lei, M. Shen, M. Yang, J. Wang and J. Wang, Appl. Catal., B, 2011, 101, 355.
35. M. Piacentini, M. Maciejewski and A. Baiker, Appl. Catal., B, 2006, 66, 126.
36. B. Azambre, I. Atribak, A. Bueno-Lopez and A. Garcia-Garcia, J. Phys. Chem. C, 2010, 114, 13300.
37. M. Nolan, S. C. Parker and G. W. Watson, J. Phys. Chem. B, 2006, 110, 2256.
38. W. S. Epling, L. E. Campbell, A. Yezerets, N. W. Currier and J. E. Parks II, Catal. Rev.: Sci. Eng., 2004, 46, 163.
39. U. Tuttlies, V. Schmeisser and G. Eigenberger, Chem. Eng. Sci., 2004, 59, 4731.
40. R. L. Muncrief, P. Khanna, K. S. Kabin and M. P. Harold, Catal. Today, 2004, 98, 393.
41. N. Maeda, A. Urakawa and A. Baiker, J. Phys. Chem. C, 2009, 113, 16724.

---