NOx storage/reduction over lean-burn automotive catalysts

This paper shows the behavior of a Pt/Ba/γ–Al₂O₃ automotive catalyst in a fixed bed reactor during cyclic operation at lean and rich gas phase conditions at short (seconds) and long (hours) cycling times at different temperatures. Reactor exit gas phase concentrations have been measured and catalyst properties have been determined before and after selective cycling experiments. The experimental results indicate that: (i) Upon 9 h lean and 15 h rich cycling, the NO oxidation efficiency of the catalyst decreases with time while incomplete regeneration is seen, even after 15 h rich exposure with H₂. The cyclic steady state is reached after 3 lean/rich cycles, at which only 60% of the available barium is involved in the NO_x storage/reduction. (ii) The BET surface area, pore volume, and Pt dispersion decrease by approximately 40%, which may be a result of masking of Pt sites or blocking of pores of the barium clusters as BaCO₃ becomes Ba(NO₃)₂. Experiments with catalyst pellet sizes of 180 and 280 μm along with XPS measurements show that blocking of catalyst pellet pores is not taking place. (iii) When applying lean/rich cycling in the order of seconds, it appears that catalyst history and lean/rich timing affect the number of cycles required to arrive at a closed N balance. XRD results after lean exposure confirm the formation of barium nitrate in the bulk of the barium cluster.     

NOx storage and reduction characteristics of Pd/MnOx–CeO2 at low temperature

The reaction between hydrogen and NO was studied over 1 wt.% Pd supported on NO_x-sorbing material, MnO_x–CeO₂, at low temperatures. The result of pulse mode reactions suggest that NO_x adsorbed as nitrate and/or nitrite on MnO_x–CeO₂ was reduced by hydrogen, which was spilt-over from Pd catalyst. The NO_x storage and reduction (NSR) cycles were carried out over Pd/MnO_x–CeO₂ in a conventional flow reactor at 150 °C. In a storage step, NO was removed by the oxidative adsorption from a stream of 0.04–0.08% NO, 5–10% O₂, and He balance. This was followed by a reducing step, where a stream of 1% H₂/He was supplied to ensure the conversion of nitrate/nitrite to N₂ and thus restore the adsorbability. It was revealed that the NSR cycle is much more suitable for the H₂–deNO_x process in excess O₂, compared to a conventional steady state reaction mode.     

Reactivity of Mg-Ba catalyst in NOx storage/reduction

A new catalyst for NO_x storage/reduction was prepared to improve the activity of Ba-Pt/γ-Al₂O₃ by replacing Ba with a mixture of Ba and Mg. The catalyst was prepared by impregnating 1 wt.% Pt and then the alkaline-earth metals (Mg, Ba) on commercial γ-Al₂O₃. The tests have been carried out in a wide temperature range (ca. 200–400 °C) in order to understand the role of the mixture of alkaline-earth metals as a function of temperature. The behaviour of the two catalysts was different and indicated a synergetic effect between Mg and Ba.     

The reduction phase in NOx storage catalysis: Effect of type of precious metal and reducing agent

The effect of different reducing agents (H₂, CO, C₃H₆ and C₃H₈) on the reduction of stored NO_x over PM/BaO/Al₂O₃ catalysts (PM = Pt, Pd or Rh) at 350, 250 and 150 °C was studied by the use of both NO₂-TPD and transient reactor experiments. With the aim of comparing the different reducing agents and precious metals, constant molar reduction capacity was used during the reduction period for samples with the same molar amount of precious metal. The results reveal that H₂ and CO have a relatively high NO_x reduction efficiency compared to C₃H₆ and especially C₃H₈ that does not show any NO_x reduction ability except at 350 °C over Pd/BaO/Al₂O₃. The type of precious metals affects the NO_x storage-reduction properties, where the Pd/BaO/Al₂O₃ catalyst shows both a high storage and a high reduction ability. The Rh/BaO/Al₂O₃ catalyst shows a high reduction ability but a relatively low NO_x storage capacity.     

Transient behaviour of catalytic monolith with NOx storage capacity

Transient behaviour of catalytic monolith converter with NO_x storage is studied under conditions typical for automobiles with lean-burn engines (i.e., diesel and advanced gasoline ones). Periodical alternation of inlet concentrations is applied—NO_x are adsorbed on the catalyst surface during a long reductant-lean phase (2–3 min) and then reduced to N₂ within a short reductant-rich phase (2–6 s). Samples of industrial NO_x storage and reduction catalyst of NM/Ba/CeO₂/γ-Al₂O₃ type (NM = noble metal), washcoated on 400 cpsi cordierite substrate, are used in the study. Effects of the rich-phase length and composition on the overall NO_x conversions are examined experimentally. Reduction of NO_x by CO, H₂ and unburned hydrocarbons (represented by C₃H₆) in the presence of CO₂ and H₂O is considered.

Effective, spatially 1D, heterogeneous mathematical model of catalytic monolith with NO_x and oxygen storage capacity is described. The minimum set of experiments needed for the evaluation of relevant reaction kinetic parameters is discussed: (i) CO, H₂ and HC oxidation light-off under both lean and rich conditions, including inhibition effects, (ii) NO/NO₂ transformation, (iii) NO_x storage, including temperature dependence of effective NO_x storage capacity, (iv) water gas shift and steam reforming under rich conditions, i.e., in situ production of hydrogen, (v) oxygen storage and reduction, including temperature dependence of effective oxygen storage capacity, and (vi) NO_x desorption and reduction under rich conditions. The experimental data are compared with the simulation results.     

NOx sorption–desorption study: application to diesel and lean-burn exhaust gas (selective NOx recirculation technique)

NO_x adsorption/desorption capacities of barium aluminates and BaSnO₃ were measured under representative exhaust gas mixture at temperatures below 550°C and compared to those of bulk BaO. The capacities are high and the test of sorption–desorption is reproducible on barium aluminate and BaSnO₃, while this is not the case on BaO. The difference is due to the electronic environment of barium oxide. If BaO is not engaged in a chemical bond, progressive formation of high stability carbonates is observed. This is not the case with barium aluminate and BaSnO₃, where carbonation does not take place because the competition between nitrate and carbonate formation is in favour of the nitrate due to its chemical nature. An N-bounded nitrate, with IR frequencies at 1360 and 1415 cm⁻¹, is formed on barium aluminate and BaSnO₃ and not on bulk BaO.     

Influence of CO2 and H2O on NOx storage and reduction on a Pt–Ba/γ-Al2O3 catalyst

In this paper, the effect of CO₂ and H₂O on NO_x storage and reduction over a Pt–Ba/γ-Al₂O₃ (1 wt.% Pt and 30 wt.% Ba) catalyst is shown. The experimental results reveal that in the presence of CO₂ and H₂O, NO_x is stored on BaCO₃ sites only. Moreover, H₂O inhibits the NO oxidation capability of the catalyst and no NO₂ formation is observed. Only 16% of the total barium is utilized in NO storage. The rich phase shows 95% selectivity towards N₂ as well as complete regeneration of stored NO. In the presence of CO₂, NO is oxidized into NO₂ and more NO_x is stored as in the presence of H₂O, resulting in 30% barium utilization. Bulk barium sites are inactive in NO_x trapping in the presence of CO₂·NH₃ formation is seen in the rich phase and the selectivity towards N₂ is 83%. Ba(NO₃)₂ is always completely regenerated during the subsequent rich phase. In the absence of CO₂ and H₂O, both surface and bulk barium sites are active in NO_x storage. As lean/rich cycling proceeds, the selectivity towards N₂ in the rich phase decreases from 82% to 47% and the N balance for successive lean/rich cycles shows incomplete regeneration of the catalyst. This incomplete regeneration along with a 40% decrease in the Pt dispersion and BET surface area, explains the observed decrease in NO_x storage.     

A new NOx storage-reduction electrochemical catalyst

A new NO_x storage-reduction electrochemical catalyst has been prepared from a polycrystalline Pt film deposited on 8 mol% Y₂O₃-stabilized ZrO₂ (YSZ) solid electrolyte. BaO has been added onto the Pt film by impregnation method. The NO_x storage capacity of Pt-BaO/YSZ system was investigated at 350 °C and 400 °C under lean conditions. Results have shown that the electrochemical catalyst was effective for NO_x storage. When nitric oxides are fully stored, the catalyst potential is high and reaches its maximum. On the other hand, when a part of NO and also NO₂ desorb to the gas phase, the catalyst potential remarkably drops and finally stabilizes when no more NO_x storage occurs but only the reaction of NO oxidation into NO₂. Furthermore, the investigation has clearly demonstrated that the catalyst potential variation versus temperature or chemical composition is an effective indicator for in situ following the NO_x storage-reduction process, i.e. the storage as well as the regeneration phase. The catalyst potential variations during NO_x storage process was explained in terms of oxygen coverage modifications on the Pt.     

Dynamics and selectivity of NOx reduction in NOx storage catalytic monolith

Several nitrogen compounds can be produced during the regeneration phase in periodically operated NO_x storage and reduction catalyst (NSRC) for conversion of automobile exhaust gases. Besides the main product N₂, also NO, N₂O, and NH₃ can be formed, depending on the regeneration phase length, temperature, and gas composition. This contribution focuses on experimental evaluation of the NO_x reduction dynamics and selectivity towards the main products (NO, N₂ and NH₃) within the short rich phase, and consequent development of the corresponding global reaction-kinetic model. An industrial NSRC monolith sample of PtRh/Ba/CeO₂/ -Al₂O₃ type is employed in nearly isothermal laboratory micro-reactor. The oxygen and NO_x storage/reduction experiments are performed in the temperature range 100–500 °C in the presence of CO₂ and H₂O, using H₂, CO and C₃H₆ as the reducing agents.The spatially distributed NSRC model developed earlier is extended by the following reactions: NH₃ is formed by the reaction of H₂ with NO_x and it can further react with oxygen and NO_x deposited on the catalyst surface, producing N₂. Considering this scheme with ammonia as an active intermediate of the NO_x reduction, a good agreement with experiments is obtained in terms of the NO_x reduction dynamics and selectivity. A reduction front travelling in the flow direction along the reactor is predicted, with the NH₃ maximum on the moving boundary. When the front reaches the reactor outlet, the NH₃ peak is observed in the exhaust gas. It is assumed that the ammonia formation during the NO_x reduction by CO and HCs at higher temperatures proceed via the water gas shift and steam reforming reactions producing hydrogen. It is further demonstrated that oxygen storage effects influence the dynamics of the stored NO_x reduction. The temperature dependences of the outlet ammonia peak delay and the selectivity towards NH₃ are correlated with the effective oxygen and NO_x storage capacity.     

Sulphur dioxide interaction with NOx storage catalysts

The effect of SO₂ on the NO_x storage capacity and oxidation and reduction activities of a model Pt/Rh/BaO/Al₂O₃ NO_x storage catalyst was investigated. Addition of 2.5, 7.5 or 25 vol. ppm SO₂ to a synthetic lean exhaust gas caused deactivation of the NO_x storage function, the oxidation activity and the reduction activity of the catalyst. The degree of deactivation of the NO_x storage capacity was found to be proportional to the total SO₂ dose that the catalyst had been exposed to. SO₂ was found to be accumulated in the catalyst as sulphate.     

Response surface study of the performance of lean NOx storage catalysts as a function of reaction conditions and catalyst composition

NO_x storage and reduction (NSR) catalysts containing Pt, Ba and Fe were studied as a function of reaction conditions and catalyst composition using response surface methodology combined with high-throughput experimentation. The concentrations of the reactant gases and the reactor temperature were varied to probe their effect on catalyst performance, as quantified by lean NO_x storage and N₂O production. An empirical model relating the catalyst performance to five reaction condition variables and three metal weight loading variables has also been developed. It was found that the temperature and the concentrations of the reducing agents, i.e. carbon monoxide and ethylene, had the strongest effect on the lean NO_x storage. It was also found that the Pt and Ba weight loadings had a much greater effect than Fe weight loadings on the performance of NSR catalysts. This model provides insight about the factors controlling the NO_x conversion by NSR catalysts and also predicts the optimum catalyst composition for given reaction conditions and vice versa. As an additional study, the relationship between sulfur poisoning, nitrous oxide production, and exotherm generation was also explored.     

Mean field modelling of NOx storage on Pt/BaO/Al2O3

A mean field model, for storage and desorption of NO_x in a Pt/BaO/Al₂O₃ catalyst is developed using data from flow reactor experiments. This relatively complex system is divided into five smaller sub-systems and the model is divided into the following steps: (i) NO oxidation on Pt/Al₂O₃; (ii) NO oxidation on Pt/BaO/Al₂O₃; (iii) NO_x storage on BaO/Al₂O₃; (iv) NO_x storage on Pt/BaO/Al₂O₃ with thermal regeneration and (v) NO_x storage on Pt/BaO/Al₂O₃ with regeneration using C₃H₆. In this paper, we focus on the last sub-system. The kinetic model for NO_x storage on Pt/BaO/Al₂O₃ was constructed with kinetic parameters obtained from the NO oxidation model together with a NO_x storage model on BaO/Al₂O₃. This model was not sufficient to describe the NO_x storage experiments for the Pt/BaO/Al₂O₃, because the NO_x desorption in TPD experiments was larger for Pt/BaO/Al₂O₃, compared to BaO/Al₂O₃. The model was therefore modified by adding a reversible spill-over step. Further, the model was validated with additional experiments, which showed that NO significantly promoted desorption of NO_x from Pt/BaO/Al₂O₃. To this NO_x storage model, additional steps were added to describe the reduction by hydrocarbon in experiments with NO₂ and C₃H₆. The main reactions for continuous reduction of NO_x occurs on Pt by reactions between hydrocarbon species and NO in the model. The model is also able to describe the reduction phase, the storage and NO breakthrough peaks, observed in experiments.     

Regeneration of NOx trap catalysts

We have investigated the regeneration of a nitrated or sulphated model Pt/Ba-based NO_x trap catalyst using different reductants. H₂ was found to be more effective at regenerating the NO_x storage activity especially at lower temperature, but more importantly over the entire temperature window after catalyst ageing. When the model NO_x storage catalyst is sulphated in SO₂ under lean conditions at 650 °C almost complete deactivation can be seen. Complete regeneration was not achieved, even under rich conditions at 800 °C in 10% H₂/He. Barium sulphate formed after the high temperature ageing was partly converted to barium sulphide on reduction. However, if the H₂ reduced sample was exposed to a rich condition in a gas mixture containing CO₂ at 650 °C, the storage activity can be recovered. Under these rich conditions the S²⁻ species becomes less stable than the CO₃²⁻, which is active for storing NO_x. Samples which were lean aged in air containing 60 ppm SO₂ at <600 °C, after regeneration at λ=0.95 at 650 °C, have a similar activity window to a fresh catalyst. It is, therefore, important that CO₂ is present during the rich regenerations of the sulphated model samples (as of course it would be under real conditions), as suppression of carbonate formation can lead to sulphide formation which is inactive for NO_x storage.     

Operando systems for the evaluation of the catalytic performance of NOx storage and reduction materials

Operando measurements were carried out in a quartz reactor to evaluate the catalytic performance of NO_x storage and reduction materials containing Pt and Ba supported on Al₂O₃. Carbonates present on the surface after activation were removed after the first exposure of the sample to the nitration flow. Nitrite species bound to barium were observed at low temperatures. Barium nitrates are the predominant species in the studied temperature range under wet and dry conditions. These species are not stable at temperatures above 723 K. The presence of water inhibits the formation of alumina nitrates favoring the coordination of nitrates with the barium sites.     

Pt-Ba/alumina NOx storage-reduction catalysts: Influence of Ba loading on NOx storage behavior

The NO_x storage behavior of a series of Pt-Ba/Al₂O₃ catalysts, prepared by wet impregnation of Pt/Al₂O₃ with Ba(Ac)₂, has been investigated. The catalysts with Ba loadings in the range 4.5–28 wt.% were calcined at 500 °C in air and subsequently exposed to NO pulses in 5 vol.% O₂/He atmosphere. Catalysts were characterized by means of thermogravimetry (TG) combined with mass spectroscopy (MS) and XRD before and after exposure to NO pulses. Characterization of the calcined catalysts corroborated the existence of three Ba-containing phases which are discernible based on their different thermal stability: BaO, LT-BaCO₃ and HT-BaCO₃. Characterization after NO_x exposure showed that the different Ba-containing phases present in the catalysts possess different reactivity for barium nitrate formation, depending on their interfacial contact. The different Ba(NO₃)₂ species produced upon NO_x exposure could be distinguished based on their thermal stability. The study revealed that during the NO_x storage process a new thermally instable BaCO₃ phase formed by reaction of evolved CO₂ with active BaO. The fraction of Ba-containing species that were active in NO_x storage depended on the Ba loading, showing a maximum at a Ba loading of about 17 wt.%. Lower and higher Ba loading resulted in a significant loss of the overall efficiency of the Ba-containing species in the storage process. The loss in efficiency observed at higher loading is attributed to the lower reactivity of the HT-BaCO₃, which becomes dominant at higher loading, and the increased mass transfer resistance.     

Investigations of sulphur deactivation of NOx storage catalysts: influence of sulphur carrier and exposure conditions

The influence of SO₂, H₂S and COS in low concentrations on the deactivation of Pt/Rh/BaO/Al₂O₃ NO_x storage catalysts was investigated. Different samples of the catalyst were exposed to synthetic gas mixtures mimicking lean/rich engine cycling in a mixed lean application at 400 °C. The lean gas mixture contained 8 vol.% O₂, 500 vol-ppm C₃H₆ and 400 vol-ppm NO balanced to 100 vol.% with Ar. The rich excursions were performed by switching off the oxygen supply. Sulphur, 25 vol-ppm of either SO₂, H₂S or COS, was added to the gas flow either during the lean, the rich or both periods. This procedure aimed at investigating the influence of the exposure conditions and therefore the lean and rich periods were kept equally long (5 min). In addition, thermodynamical calculations for the prevailing conditions were performed.

It was concluded that all sulphur compounds investigated, i.e. SO₂, H₂S and COS, had similar, negative impact on the NO_x storage ability of the catalyst and that they all showed increased deactivation rates during rich exposure compared to lean. During lean exposure, all sulphur carriers showed similar behaviour, while H₂S and COS caused severe loss of noble metal activity during rich exposure.     

New insight into the interaction of sulfur with diesel NOx storage catalysts

A commercial NO_x-storage catalyst (NSC) has been subjected to different aging procedures on the engine bench simulating 100,000 km mileage. The aging consisted of cyclical sulfur exposure, subsequent sulfur removal and testing of the catalytic activity. More aggressive desulfation procedures result in more efficient sulfur removal and consequently good high temperature NO_x-conversion. However, low temperature NO_x-performance is lower than for agings employing more moderate desulfation conditions.

Sulfur post mortem analyses reveal a slight decrease of residual sulfur concentration over the length of all catalysts after completion of the aging. BET and CO-chemisorption data are in line with the increase of temperature from catalyst inlet to outlet during the desulfation. The conversion of BaCO₃ to BaSO₄ during the sulfur poisoning was followed by IR, TPD and TPR. A quantitative analysis of the data shows that at the end of the agings all residual sulfur is mainly located at barium sites as opposed to other oxide components like e.g. alumina or ceria. TPR data suggest that prolonged rich purges of the sulfated catalyst lead to an efficient decomposition of sulfates however some sulfur is being trapped in the form of BaS which seems difficult to remove under constant rich conditions.

XPS data suggest that the bulk sulfur amounts in the catalyst may be decoupled from the actual concentration at the catalyst surface. In that sense, the residual sulfur concentration might be limited in some cases as a criterion to assess the performance of a NSC. More reducing desulfation conditions cause the residual sulfur to be present in the form of more reduced sulfur species (sulfites, sulfides) on the catalyst.     

Mn/MFI catalyzed reduction of NOx with alkanes

Mn/MFI catalysts were prepared by different methods and probed as catalysts for the catalytic reduction of NO_x with CH₄ or iso-butane in a gas flow containing excess O₂. Mn/MFI with high manganese loading was obtained by solid state ion exchange (SSI). The intensity of an IR band at 957 cm⁻¹, which is due to the perturbation of zeolite lattice vibrations by Mn ions attached to cage walls is proportional to the Mn content of the catalysts. Conversely, the intensity of the 3610 cm⁻¹ band, assigned to Brønsted acid sites decreases linearly with the Mn loading. A catalyst obtained by exchanging Na/MFI with an aqueous solution of Mn acetate is found most active for NO_x reduction with methane. Transport by surface diffusion of Mn ions from MnI₂ to exchange positions in MFI is more efficient than their transport through the gas phase. High NO conversion over proton-free catalysts indicates that protons are not instrumental in NO_x reduction over Mn/MFI.     

An infrared study of the behavior of SO2 and NOx over carbon and carbon-supported catalysts

The heterogeneous reactivity of NO_x and SO₂ with carbon has been investigated with FT-IR spectroscopy. The interaction between NO and SO₂ on carbon surface have been studied in the presence and in the absence of oxygen. Thermal stabilities of surface structures, formed as a result of NO_x and SO₂ chemisorption have been determined by means of FT-IR spectroscopy. During the reaction of NO/O₂ mixture with carbon the surface species, including C–NO₂, C–ONO, C–NCO and anhydride structures are formed. It has been found that SO₂ retards the oxidation reaction of carbon by oxygen. The oxidation of SO₂ on carbon was found to be greatly enhanced by the presence of NO + O₂ mixture. The adsorption capacity of cellulose based carbon, catalytic NO_x decomposition and TPD was studied using a fixed bed flow reactor.     

The removal of NOx from a lean exhaust gas using storage and reduction on H3PW12O40·6H2O

NO_x sorption and reduction capacities of 12-tungstophosphoric acid hexahydrate (H₃PW₁₂O₄₀·6H₂O, HPW) were measured under representative alternating conditions of lean and rich exhaust-type gas mixture. Under lean conditions, the sorption of NO_x is large and is equivalent to 37 mg of NO_x/g_HPW. Although a part of these NO_x remains unreduced, HPW is able to reduce some of the NO_x to produce N₂ by a reaction between the sorbed NO₂ and hydrocarbon (HC), but this process is slow. The addition of 1% Pt affects strongly the chemical behaviour occurring during the course of a rich operation. The NO desorption observed at the beginning of the rich phase is strongly accelerated. The direct correlation between NO₂ consumption and CO₂ production shows that the principal pathway is the reaction CO+NO₂→CO₂+NO. In a mixture of reducing gas (CO, HC, H₂), the competition is strongly in favour of CO though in its absence the reaction observed was the hydrogenation of propene to propane.