The present invention relates to a spark ignition engine comprising an exhaust system comprising a catalyst and an engine control unit programmed to control the air-to-fuel ratio of the engine to run at the stoichiometric air-to-fuel ratio during normal running conditions and to run lean of the stoichiometric air-to-fuel ratio during a defined portion of an engine speed/load map. In particular, the present invention relates to such an engine wherein the catalyst is a three-way catalyst (TWC) including a NOx storage component.
A heterogeneous catalyst capable of simultaneous conversion of nitrogen oxides (NOx), carbon monoxide (CO) and unburnt hydrocarbons (HC) in exhaust gas from a stoichiometrically operated, spark-ignited combustion engine is known as a three-way catalyst (TWC). NOx reduction readily occurs over the TWC when the air-to-fuel ratio is rich of stoichiometric, whereas CO and HC reactions are hindered by insufficient oxygen (O2). On the lean side, the CO and HC conversions are high, but NOx reduction is difficult because of the excess of oxidising species. Accordingly, effective three-way conversion occurs in a relatively narrow air-to-fuel ratio window. In practice an oxygen sensor is used to detect the lambda composition of the exhaust gas upstream of the TWC and to adjust the air-to-fuel ratio accordingly to equilibrate the exhaust gas.
A typical TWC comprises platinum (Pt) and/or palladium (Pd) as an oxidation catalyst and rhodium (Rh) as a reduction catalyst on a suitable high surface area oxide support, such as alumina (AM2O3), and an oxygen storage component (OSC), e.g. a ceria-zirconia mixed oxide. Various minor amounts of base metal catalyst promoters, stabilisers and hydrogen sulphide suppressers can be included. For flirter details, see WO 98/03251 (incorporated herein by reference).
A consequence of using detected oxygen concentration to control the air-to-fuel ratio is that there is a time lag associated in adjusted air-to-fuel ratio. This results in perturbation around the control set point. Thus, when operating rich, there is a need to provide a small amount of O2 to consume the unreacted CO and HC. Conversely, when the exhaust gas goes slightly oxidising, the excess O2 needs to be consumed. One development in TWC technology adopted in order to reduce the problem of emissions associated with perturbations was to incorporate an O2 storage component in the TWC composition. This component adsorbs (or absorbs) O2 in the lean environment and releases it in the rich environment, thus effectively extending the time at which the exhaust gas is at the set point Where more significant amounts of HC fuel are required to maintain the air-to-fuel ratio, such as during acceleration, this can be provided e.g. by adjusting the fuel injection period.
More recently, there has been a move towards running gasoline combustion engines lean, for example, in gasoline direct injection engines. The rationale is to improve fuel economy (thus decreasing the emissions of CO2) by running lean of stoichiometry. The principal problem in pursuing this strategy is that the lean environment inhibits NOx reduction in the TWC. One technology which has been developed to meet this problem is called, variously, a NOx absorber/catalyst, lean NOx trap (LNT) or NOx trap and is based on acid-base washcoat chemistry. It involves adsorption (or absorption) and storage of NOx in the catalyst washcoat during lean driving conditions and release under rich operation. The released NOx is catalytically converted to nitrogen just as it is in TWCs.
A typical NOx trap composition comprises Pt and Rh on a high surface area oxide support, such as Al2O3, and a NOx storage component such as barium oxide (BaO (see, e.g. EP 0758713, incorporated herein by reference)). Generally, loadings of NOx storage components in NOx trap washcoats can be up to 50 wt % or even higher
A major problem with the use of NOx trap technology is that it requires very careful and complicated control of the engine in order to provide for rich regeneration of the NOx storage component Additionally, a number of feedback sensors are used to control the function of the NOx trap NOx storage capacity, e.g. sensors to estimate cumulative engine-out NOx production utilising stored engine maps and NOx trap temperature sensors, because the efficiency of the NOx storage component to absorb NOx is temperature dependent.
It is known in the art of TWCs, to stabilise a gamma-Al2O3 component from sintering during high temperature ageing using small amounts (1-3 percent) of BaO or lanthanum oxide (La2O3). Base metal catalyst promoters, such as barium (Ba), cerium, lanthanum, magnesium, calcium and strontium can also be used (see WO 98/03251, mentioned above).
In our WO 99/00177 (incorporated herein by reference) we describe a catalytic converter for a lean burn internal combustion engine, such as a direct injection gasoline engine, comprising a catalyst component capable of storing NOx. In one embodiment, the catalytic converter comprises a supported layered catalyst having a first, inner layer containing a first platinum group metal (PGM), e.g. Pt, and a NOx storage component, e.g. Ba, and a second, outer layer containing a second different PGM, such as Rh supported on a non-Al2O3 support, and optionally an OSC such as a mixed oxide of ceria and zirconia.
Japanese Unexamined Patent Publication (KOKAI) No. 5-317,652 (incorporated herein by reference) describes a catalyst comprising a substrate, and an alkaline-earth metal compound and Pt loaded on the substrate. The description observes that during urban driving a vehicle is frequently accelerated and decelerated. Consequently, this can cause the air-to-fuel ratio to vary frequently from the range of values adjacent to the stoichiometric point during more steady conditions, e.g. idling, to the fuel rich side. In order to lower the fuel consumption during e.g. urban driving conditions, the gasoline engine is run on the fuel lean side, such as an air-to-fuel ratio of up to 23:1 (wt./wt.). The catalyst is designed to adsorb (or absorb) NOx on the alkaline-earth metal during lean running conditions and to use the natural fluctuation of the air-to-fuel ratio to the rich side to enable stored NOx to be released and reduced, thus regenerating the NOx storage capacity of the alkaline-earth metal compound.
U.S. Pat. No. 5,575,983 (incorporated herein by reference) observes that the NOx absorbing capacity of the catalyst of Japanese Unexamined Patent Publication No. 5-317,652 is poisoned by sulfate. In order to combat this, it proposes a catalyst comprising Pt or Pd and alkali metals, alkaline-earth metals and rare-earth elements including lanthanum (La) supported on lithium-stabilised Al2O3.
We believe that the use of lean burn gasoline engines other than direct injection gasoline engines, such as those described in Japanese Unexamined Patent Publication (KOKAI) No. 5-317,652 and U.S. Pat. No. 5,575,983 has not received widespread acceptance in the vehicle industry and that one reason for this is the difficulty in controlling NOx emissions to meet existing and future emission legislation.
We have now found, very surprisingly, that it is possible to operate a spark ignition engine, such as a port fuel injection gasoline engine, which engine comprising a TWC including a NOx storage component, in such a way as to benefit from the increased fuel economy available during lean running conditions whilst avoiding the requirement for expensive and complicated control systems.
According to one aspect, the invention provides A spark ignition engine comprising an exhaust system comprising a catalyst and an engine control unit programmed to control the air-to-fuel ratio of the engine to run at the stoichiometric air-to-fuel ratio during normal running conditions and to run lean of the stoichiometric air-to-fuel ratio during a defined portion of an engine speed/load map, which catalyst comprising a three-way catalyst (TWC) including a NOx storage component, characterised in that the engine control unit is further programmed to determine the amount of NOx contacting the TWC during lean running operation in response to data input from sensor means, thereby to determine the remaining NOx storage capacity of the TWC and to return the air-to-fuel ratio to stoichiometry when the remaining NOx storage capacity is below a pre-determined value, the arrangement being such as to substantially prevent passing more NOx to atmosphere during an engine cycle compared with a spark ignition engine run continuously in stoichiometric mode.
By “engine cycle” herein we mean the period between key on and key off.
The present invention takes advantage of the natural fluctuation in the composition of an exhaust gas of a spark ignition engine operated at the stoichiometric air-to-fuel ratio to rich lambda values, e.g. during acceleration, to regenerate the NOx storage component. NOx stored on the NOx storage component is generally not released, and the NOx storage component is not regenerated, under lambda=1 conditions, i.e. conditions rich of lambda=1 are required. We have also devised a catalyst that facilitates the regeneration of a NOx storage component in a TWC, which catalyst is used in a preferred embodiment according to the invention.
The present invention provides a number of very substantial advantages. One such advantage is that it enables a vehicle powered by a spark ignition engine to be run at a fuel saving over a similar vehicle operated continuously at substantially stoichiometric conditions. Such increased efficiency can result in lower CO2 emissions in a legislative test cycle for a vehicle. Lower CO2 emissions in a legislative test cycle translates to lower CO2 emissions in “real world” driving conditions. Accordingly, a vehicle according to the invention can be more “environmentally friendly”. Furthermore, in countries where vehicles are taxed depending on the amount of CO2 they emit (a so-called “green tax”), such as the UK, it can reduce the tax burden to the consumer.
A second such advantage is that it can allow existing vehicles including spark ignition engines to receive the benefit of the invention by retrofitting certain components. This can be done by simply replacing the existing TWC with a TWC including sufficient NOx storage component and the engine control unit for one programmed: (i) to run the engine lean of the stoichiometric air-to-fuel ratio during a defined portion of an engine speed/load map; and (ii) to determine the amount of NOx contacting the TWC during lean running operation in response to data input from sensor means, thereby to determine the remaining NOx storage capacity of the TWC and to return the air-to-fuel ratio to stoichiometry when the remaining NOx storage capacity is below a pre-determined value, thereby substantially to prevent passing more NOx to atmosphere during an engine cycle compared with a spark ignition engine run continuously in stoichiometric mode.
According to a particularly preferred embodiment, the defined portion of the engine speed/load map is engine idle. This is a particularly advantageous arrangement in that it provides a fail-safe system for regenerating the NOx storage component This is because the only thing that can happen to the engine after idle is that it is accelerated, following which the exhaust gas contacting the TWC is temporarily rich before the engine control unit returns the exhaust gas to equivalence. Even if the engine is switched off following idle, NOx can be stored to key on, following which the engine will be accelerated. Accordingly, this preferred arrangement takes advantage of the natural fluctuations in the composition of the exhaust gas to the rich side during acceleration to regenerate the NOx storage component For similar reasons, the defined portion of the engine speed/load map can comprise low speed driving wherein the level of NOx emitted by the engine is up to ten times more, such as five times or twice more, than at engine idle.
A very substantial advantage of this preferred arrangement is that it avoids the requirement for complicated and expensive sensors and controls in order to meet emission legislation that presently burdens the adoption of NOx traps.
A number of means for inputting data to the engine control unit to determine the amount of NOx contacting the TWC, and hence the remaining NOx storage capacity of the NOx storage component, can be used either singly or in any mechanically/electronically viable combination. Many of the sensor means required to collect this information are already included in the engine and/or vehicle fitted with the engine and are used by the engine control unit for controlling other functions of the engine and/or vehicle. This is one reason why it is possible to adopt the invention by retrofitting the vehicle engine control unit, together with a TWC including a NOx storage component.
Such detected data that can be used to monitor remaining NOx storage capacity in the TWC of the invention include: predetermined or predicted time elapsed from the start of lean running operation, by sensing the status of a suitable clock means; airflow over the TWC or manifold vacuum; ignition timing; engine speed; throttle position; exhaust gas redox composition, for example using a lambda sensor, preferably a linear lambda sensor; quantity of fuel injected in the engine; where the vehicle includes an exhaust gas recirculation (EGR) circuit, the position of the EGR valve and thereby the detected amount of EGR; engine coolant temperature; and where the exhaust system includes a NOx sensor, the amount of NOx detected upstream and/or downstream of the TWC. Where the clock embodiment is used, the predicted time can be subsequently adjusted in response to data input.
The spark ignition engine can be any capable of operating during normal running conditions at the stoichiometric air-to-fuel ratio. In one embodiment, the engine can be powered by gasoline and the engine can be of the port fuel injection or direct injection type. Optionally, the engine can be fuelled using an alternative fuel such as liquid petroleum gas (LPG), natural gas (NG), methanol, hydrocarbon mixtures including ethanol or hydrogen gas. The invention can be used on all grades of sulfur-containing fuel, but with particular efficiency with grades containing less that 50 ppm by weight of sulfur, and most preferably less than 10 ppm by weight of sulfur.
The present invention can utilise any known TWC composition provided that sufficient NOx storage component is included in order to perform the desired function.
A typical TWC composition comprises at least one PGM, and can be selected from the group consisting of Pt, Pd, Rh, ruthenium, osmium and iridium and any combination of two or more thereof.
Many NOx storage components are disclosed in the prior art, and any can be utilised in the present invention. Typical NOx storage components comprise alkali metals, such as potassium or caesium, alkaline-earth metals, e.g. magnesium, calcium, strontium or Ba, rare-earth metals lanthanide group metal, preferably La, or any viable combination, e.g. a mixed oxide, of any two or more thereof.
A common component of state-of-the-art TWCs is the OSC and these too can be included in the TWC, with utility, according to the present invention. Indeed, it is trite knowledge of the person skilled in the art that NOx trap compositions do not include OSC, because an OSC assists in the combustion of HC at the stoichiometric point and slightly rich thereof. This property runs counter to the requirement of a system including a NOx trap composition which is to regenerate the NOx storage component using reducing species, such as HCs in the exhaust, resulting from air-to-fuel ratio modulation. Accordingly, the presence of an OSC in a NOx trap composition would cause increased fuel consumption for the same amount of NOx storage component regeneration relative to an OSC-free NOx trap composition.
Known OSC include optionally stabilised ceria, perovskites, NiO, MnO2, manganese-based compounds supported on Al2O3-containing mixed oxides (see PCT/GB01/05124, incorporated herein by reference) a mixed oxide of manganese and zirconium (see WO 99/34904, incorporated herein by reference), Pr2O3 or a combination of any two or more thereof. The ceria stabiliser can be zirconium, lanthanum, aluminium, yttrium, praseodymium or neodymium.
A preferred TWC for use in the present invention comprises a first PGM, preferably Pt, and the NOx storage component in a first, inner layer and an OSC and a second PGM, preferably Rh, in a second, outer layer.
This arrangement is advantageous for the following reasons. During stoichiometric running the Rh/OSC component is active for NOx reduction and other reactions while the Pt component is active for oxidation reactions. During oxygen rich conditions the Rh is relatively inactive, the Pt is active for NO, HC and CO oxidation while the NO2 produced is stored in the adsorber as nitrate.
On subsequent return to lambda=1, the OSC component in the second layer prevents the stored NOx “seeing” reducing gas so that the NOx remains stored as nitrate. The Rh is active for NOx reduction by CO and the Pt is active for oxidation of HC and CO.
On acceleration the gas mix becomes rich biased, reducing the OSC material so that the stored NOx is released, as the nitrate becomes thermodynamically unstable. The released NOx is then reduced by the Rh layer with the excess reductant,
According to a further aspect the invention provides a vehicle comprising an engine according to the present invention.
TWC are generally placed in one or both of two positions in a vehicle according to the intended purpose: the close-coupled position, in which the TWC is disposed as close to the exhaust manifold as possible; and the underfloor position. The reason for placing a TWC in the close-coupled position is to control emissions immediately following cold start, as much of the controlled emissions are emitted during the legislative test cycle immediately following cold-start. By positioning a TWC close to the engine, the catalyst is contacted by hot exhaust gases immediately after key on and accordingly reaches the light-off temperature for CO and HC oxidation sooner than a TWC in the cooler, underfloor location. However, once the exhaust system is up to temperature for efficient three-way conversion, the underfloor catalyst shoulders much of the burden of treating the exhaust gas. Meanwhile the close-coupled TWC is exposed to very high temperatures e.g. up to 1000° C. Indeed one vehicle manufacturer requires testing of close-coupled TWCs for 50 hours at catalyst bed temperatures of at least 970° C. and up to 1010° C. At these sort of temperatures, catalysts can lose activity as materials lose their surface area through sintering events, migration of active species into pores and component interactions. Accordingly, a TWC in the close-coupled position can be expected to lose some of its activity compared with a fresh catalyst. The NOx storage component can also lose NOx storage capacity through this high temperature ageing by loss-of surface area.
Whilst loss of NOx storage capacity through high temperature ageing, particularly in TWCs located in the close-coupled position, is undesirable in the present invention, the benefit of the invention is still obtained provided that a proportion of NOx storage activity is retained. Accordingly, in an embodiment of the invention, the fresh TWC includes sufficient of the NOx storage component to retain sufficient NOx storage capacity after high temperature ageing, for example in the close-coupled position.
According to a further aspect, the invention provides An engine control unit for a spark ignition engine comprising an exhaust system comprising a TWC including a NOx storage component, which engine control unit is programmed to control the air-to-fuel ratio of the engine to run at the stoichiometric air-to-fuel ratio during normal running conditions and to run lean of stoichiometry during a defined portion of an engine speed/load map and to determine the amount of NOx contacting the TWC during lean running operation in response to data input from sensor means, thereby to determine the remaining NOx storage capacity of the TWC and to return the air-to-fuel ratio to stoichiometry when the remaining NOx storage capacity is below a pre-determined value, the arrangement being such as to substantially prevent passing more NOx to atmosphere during an engine cycle compared with a spark ignition engine run continuously in stoichiometric mode.
According to a further aspect, the invention provides A method of treating exhaust gas of a spark ignition engine run at the stoichiometric air-to-fuel ratio during normal running conditions, which engine comprising an exhaust system comprising a TWC including a NOx storage component, which method comprising the steps of controlling the engine air-to-fuel ratio to run lean of stoichiometry during a defined portion of an engine speed/load map and determining the amount of NOx contacting the TWC during lean running operation in response to data input from sensor means, thereby to determine the remaining NOx storage capacity of the TWC and to return the air-to-fuel ratio to stoichiometry when the remaining NOx storage capacity is below a pre-determined value, the arrangement being such as to substantially prevent passing more NOx to atmosphere during an engine cycle compared with a spark ignition engine run continuously in stoichiometric mode.
In order that the invention may be more fully understood, the following Examples are provided by way of illustration only and with reference to the accompanying drawings, in which:
FIG. 1 is a graph showing HC conversion against temperature after ageing of a TWC including a NOx storage component according to the invention, a state-of-the-art TWC and a NOx trap; and
Total Hydrocarbon Light Off
FIGS. 2, 3 and 4 are graphs showing % conversion of CO, HC and NOx against lambda for a TWC including a NOx storage component according to the invention, a state-of-the-art TWC and a NOx trap formulation respectively.
Three catalyst washcoats were tested. Comparative catalyst A is a state-of-the art Pt/Rh TWC on a thermally stable, high surface area support at a ratio of 5Pt:1Rh and a total precious metal loading of at 60 g ft−3.
Catalyst B is a TWC including a NOx storage component according to the present invention supported on an identical substrate. It comprised a first, inner layer of a high surface area Al2O3 impregnated with Pt and a NOx storage component, such as BaO, and a second, outer layer of a mixed oxide OSC impregnated with Rh. The ratio of Pt:Rh and the total precious metal loading was the same as for catalyst A.
Comparative catalyst C is a NOx trap composition comprising a high surface area Al2O3-based mixed oxide support impregnated with Pt, Rh and a NOx storage component. The ratio of Pt:Rh was 6:1 and the total precious metal loading was 70 g ft−3.
Each washcoat was coated on a 4.66×6 inch (11.9×15.2 cm) ceramic substrate of 400 cells per square inch ((cpsi) 62 cells cm−2) of 0.15 mm wall thickness and the resulting coated substrate was hydrothermally aged at 800° C. for 5 hours under 10% O2/0% H2O balance nitrogen.
The catalysts were fitted to the exhaust of a four cylinder 2.0 litre Port Fuel Injection bench mounted engine controlled by a Bosch ME7 control system. The catalyst temperature was increased by adjustment of a heat exchanger fitted to the exhaust line before the catalyst. The temperature ramp rate was 14° C./minute.
The results for the HC light-off (the temperature at which the reaction is catalysed to 50% efficiency) are shown in FIG. 1, from which it can be seen that the HC light-off temperature for catalyst B is similar to that of the comparative catalyst A. It can also be seen that the HC light-off temperature of comparative catalyst C is approximately 30° C. higher than comparative catalyst A.
- EXAMPLE 2
Perturbed Lambda Scan
This result shows that the TWC including a NOx storage component (catalyst B) has very similar activity for HC activity compared with a state-of-the-art TWC (comparative catalyst A), despite the presence of the NOx storage component. The NOx trap (comparative catalyst C) performs less well than either catalyst B or comparative catalyst A, despite having a higher Pt loading.
The same engine as in Example 1 was used to give a catalyst inlet temperature of 450° C. The lambda scans were done using a 10% perturbation (cycling between a value 10% below lambda=1, i.e. less 0.147 lambda, to a value of 10% above lambda=1 (1.147 lambda)) and at a frequency of 1 Hz. These conditions were chosen to simulate the exhaust gas composition at the inlet to a TWC disposed in an exhaust system including a lambda sensor upstream of the catalyst inlet providing feedback to the engine control unit in order to maintain lambda=1 conditions.
- EXAMPLE 3
The results are shown in FIGS. 2, 3 and 4. As can be seen, the NOx trap (comparative catalyst C) is poorer for lambda scan performance (showing poorer conversion), despite having more Pt while the TWC including a NOx storage component (catalyst B) performs similarly to the TWC (comparative catalyst A).
In a bench test cell, a 4 cylinder, 1.8 litre, 1997 model year, Mitsubishi direct injection engine from a vehicle calibrated for the Japanese market was installed with a direct current dynamometer. A catalyst substrate prepared according to Example 1 was fitted in the close-coupled position approximately 30 cm from the engine exhaust manifold. The substrate volume represented of 22% engine swept volume (ESV). Concentrations of NOx, HC, CO2, CO and O2 were measured using a dual bank of MEXA (Motor Exhaust Gas Analyser) 9500 sensors to allow continuous measurement of gas concentrations upstream and downstream of the catalyst The catalyst inlet temperature was measured by thermocouple.
The engine was operated from one of two sets of maps: one for the homogeneous mode and the other for the lean, stratified mode. Basic maps for ignition and injection timing and duration were generated by firstly logging data from the ECU of a vehicle including the same model of engine as the one used for the bench test and then basing the maps on this information by reverse engineering. In the homogeneous mode, the engine was run at a range of engine speeds and loads and a supplementary map to the basic map was generated for the best emissions of NOx, CO and HC emissions under λ=1 operation. The lean, stratified mode was mapped by matching the torque achieved in homogeneous mode at the same speed and load demand.
Prior to testing, the engine was thoroughly warmed up in idling condition. In homogeneous mode, the engine was then run so that the inlet temperature to the close-coupled catalyst was 300° C. It was then switched to lean, stratified operation and the EGR valve position was adjusted until the engine-out NOx was 300 ppm. The EGR valve position was recorded and was referred to as the lean set point The engine was switched back to homogeneous mode, and the EGR valve was closed. A rich set point was obtained by increasing the fuel injector pulse width to obtain lambda 0.80. A series of lean/rich cycles were run as follows. In the lean mode, the EGR valve was at the lean set-point position until the NOx efficiency of the system had dropped below 75%. The engine was then switched back to homogeneous mode for 15 seconds, with injector duration at the rich set point. The cycles were repeated five times and the results obtained for each of the cycles were logged. Exhaust systems were fitted to the engine and the protocol above was followed and data collected in the lean, stratified mode at 300° C. catalyst inlet temperature, a typical catalyst inlet temperature for a close-coupled TWC in the exhaust system of a port fuel injection during idling.
Table 1 shows the NOx storage efficiency with which each of comparative catalysts A and C and catalyst B store NOx, and in particular, how the efficiency with which each catalyst stores 30, 40, 50 and 60 mg of NOx. Thus the NOx trap (comparative catalyst C) stores 60 mg NOx with 97% efficiency, i.e. 97% of the NOx contacting the catalyst is stored, whereas the TWC (comparative catalyst A) stores 60 mg of NOx with 9% efficiency, i.e. for the time it takes to store 60 mg of NOx, the catalyst has slipped 91% of the NOx contacting it.
| ||TABLE 1 |
| || |
| || |
| ||NOx storage efficiency |
| || ||Catalyst B || |
| ||Comparative ||(TWC including ||Comparative |
| ||Catalyst C ||a NOx storage ||Catalyst A |
|NOx stored (mg) ||(NOx Trap) ||component) ||(TWC) |
|30 ||98% ||77% ||18% |
|40 ||98% ||73% ||16% |
|50 ||97% ||68% ||12% |
|60 ||97% ||64% || 9% |
The results of Examples 1, 2 and 3 show that the TWC including a NOx storage component maintains TWC performance despite including a NOx storage component, and that NOx storage capacity is several times higher than a state-of-the-art TWC composition.