Monatshefte für Chemie 136, 91 -105 (2005)
`DOI 10.1007/s00706- 004 -0261 -z
`
`Monatshefte für Chemie
`Chemical Monthly
`Printed in Austria
`
`Catalytic Coatings for Active and Passive
`Diesel Particulate Filter Regeneration
`
`Claus Görsmann*
`
`Johnson Matthey PLC, Environmental Catalysts and Technologies, Royston,
`SG8 5HE, United Kingdom
`
`Received July 15, 2004; accepted (revised) October 27, 2004
`Published online January 3, 2005 © Springer- Verlag 2005
`
`Summary. This paper will give an overview how catalytic coatings are applied in diesel particulate
`filter systems to support filter regeneration by soot oxidation with nitrogen dioxide or oxygen. Catalytic
`coatings can be placed on a catalyst substrate in front of a diesel particulate filter, on a filter, or in a
`combined system on both. Strategies and conditions for successful filter regeneration of those systems
`will be discussed.
`
`Keywords. Heterogeneous catalysis; Kinetics; Oxidations; Nitrogen dioxide; Soot oxidation.
`
`Introduction
`Diesel engines combust diesel fuel to convert its chemical energy into mechanical
`power. The primary combustion products of diesel fuel are carbon dioxide and
`water. More than 99% of diesel emissions consist of CO2, H2O, and the portion
`of air left over from the combustion process. As real diesel combustion is not ideal,
`a fraction of a percent of the total diesel emissions are unwanted combustion by-
`products [1]. These pollutants are mainly particulate matter (PM), nitrogen oxides
`(NO, = NO and NO2), hydrocarbons (HC), and carbon monoxide. Of those, PM
`emissions are seen as perhaps the most critical. They are of major concern regard-
`ing negative health [2 -6] and possible climate effects [7 -10].
`The most effective method to control PM emissions from diesel engines is
`applying diesel particulate filter (DPF) systems. They are very effective in filtering
`PM of all sizes out of diesel emissions [11 -17]. The main technical challenge
`for DPF systems in automotive applications is their regeneration from the soot
`they retain from the exhaust gases. For a large variety of applications ranging from
`passenger cars, city busses, garbage trucks, delivery trucks, in -door fork lifters,
`construction machines, tunnel- and mining equipment, to locomotives, various
`
`* E-mail: goersc@matthey.com
`
`BASF-2003.001
`
`

`
`92
`
`C. Görsmann
`
`DPF systems with soot regeneration technologies are in use today [13, 18]. What-
`ever the regeneration method, catalytic coatings play a major role in supporting it.
`
`Fundamental Principles of DPF System Regeneration
`Particulate matter consists of an agglomeration of primary soot particulates
`(mainly elemental carbon), adsorbed hydrocarbons (VOF), water, sulphuric and
`nitric acid, and engine oil derived components like Ca, P, or Zn [19]. The ratio
`of these components, the microstructure and properties of PM depend on the com-
`bustion conditions, the fuel and oil used, and whether and which aftertreatment
`device is present in the exhaust system.
`When trapped in a DPF all of the components present in particulates except the
`normally very small amount of inorganic oil derived residues can be converted into
`gaseous products by evaporation (water), decomposition (H2SO4, HNO3), or oxi-
`dation (VOF, soot (= mainly C)). Therefore all technologies commonly used for
`DPF regeneration are based on some heat treatment and the oxidation of soot to
`CO2. From each regeneration a small amount of "oil ash ", which typically consists
`mainly of zinc phosphate and calcium sulphate, remains in the filter and has to be
`mechanically removed during vehicle service.
`The microstructure of particulate matter influences its reactivity against oxida-
`tion [20]. However, a much greater influence on the rate at which PM can be
`oxidised comes from the concentration and the reactivity of the applied oxidising
`agent. In diesel exhaust emissions there are two oxidising gases present, which are
`suitable for soot oxidation: oxygen and nitrogen dioxide. Both play an important
`role in DPF regeneration. There is much more 02 present in diesel exhaust emis-
`sions than NO2, but NO2 is much more active than 02. The difference in reactivity
`between 02 and NO2 as oxidising agent for soot can be seen in Fig. 1, which shows
`the temperature -ranges at which NO2 and 02 start to combust soot.
`
`Oxygen Based Soot Oxidation
`The oxygen comes from the charge air left over from the diesel combustion process
`and its concentration in diesel exhaust emissions is in the order of around 10%.
`Exhaust temperatures above 550 °C are required for soot oxidation by 02. With the
`
`NO2
`
`02
`(cat.
`reaction)
`
`2
`
`2.5
`
`2.0
`
`-
`
`1.5
`
`1.0
`
`0.5
`
`0.0
`
`o
`
`100
`
`200
`
`300
`400
`Tl°C
`
`500
`
`600
`
`700
`
`Fig. 1. Soot combustion (to CO2) by NO2 and 02 as a function of temperature
`
`BASF-2003.002
`
`

`
`Diesel Particulate Filter Regeneration
`
`93
`
`application of so- called "fuel borne catalysts" (FBCs), which precursors are added
`to the diesel fuel, the required temperature can be lowered to around 400 -450 °C
`[21 -33]. Alternatively oxygen could be "activated" by applying non -thermal
`plasma to generate ozone [34], which could then be used to combust soot at low
`temperatures [35]. However, this approach has not been used in commercial appli-
`cations yet.
`
`Chemistry of Oxygen Based Soot Oxidation
`If the temperature exceeds 500 °C soot oxidation by oxygen becomes significant
`(Eq. (1)).
`
`C + 02 > CO2
`(1)
`The kinetics of soot oxidation is also influenced by the presence of water
`vapour and adsorbed species like hydrocarbons on the soot surface. A possible
`influence from water vapour could come from the steam reforming reaction
`(Eq. (2)) which could be followed by the oxidation of CO and H2 to CO2 and
`water according to Eqs. (3) and (4).
`C + H2O *-> CO + H2
`CO + 2 02 > CO2
`H2+202>H20
`(4)
`The CO of Eq. (2) could also react with water in the water gas shift reaction (Eq. (5)).
`CO + H2O <---* CO2 + H2
`The combination of Eqs. (2) and (5) results in Eq. (6).
`+2H204- >CO2 +2 H2
`C
`(6)
`The presence of CO2 could lead to carbon removal according to Eq. (7) towards the
`Boudouard equilibrium [36], which could be followed by the oxidation of CO to
`CO2 (Eq. (3)).
`
`(2)
`
`(3)
`
`(5)
`
`CO2+C4->2CO
`(7)
`Adsorbed hydrocarbons could help to ignite the soot by its easier oxidation.
`The oxidation of these adsorbed HCs could lead to local exotherm generation
`heating up the soot locally to its combustion temperature. Additionally, when
`HCs are desorbing or burning off from soot particulates, the surface area of the
`soot particulates accessible to oxygen increases. This activates the soot for its
`easier combustion.
`Whereas sulphur in the fuel has a negative effect on the NO2 formation and PM
`emissions (see later), it might have also beneficial effects for the soot oxidation
`itself. Sulphuric acid adsorbed on particulate matter becomes more concentrated,
`when the temperature comes close to its boiling point (338 °C). H2SO4 is a strong
`oxidant and might support the soot oxidation according to Eq. (8).
`H2SO4 + C -> H20 + SO2 + CO
`
`(8)
`
`BASF-2003.003
`
`

`
`94
`
`C. Görsmann
`
`This reaction might become more important, after a long period of passive regen-
`eration at low temperatures, when some H2SO4 has been accumulated within the
`soot layer.
`
`Nitrogen Dioxide Based Soot Oxidation
`NO is formed under the high -temperature, high -pressure in- cylinder conditions
`from the nitrogen and oxygen present in the charge air during the diesel combus-
`tion. NO2 forms from NO by reaction with excess oxygen. During the expansion
`stroke of the piston the temperature decreases rapidly within the combustion cham-
`ber, but the NO concentration does not decrease to the NO /O2 /NO2 equilibrium
`concentration as NO is kinetically relatively stable under these conditions [37].
`Figure 2 shows the NO2 concentration at thermodynamical NO /NO2 /O2 equilib-
`rium (Eq. (9)) for the temperature range from 0 -700 °C under typical diesel 02
`concentration levels.
`
`NO + 2 02 <--> NO2
`(9)
`The concentration of nitrogen oxides in diesel exhaust gas is around 0.01-
`0.1 %. That is 2 -3 orders of magnitude
`the 02 concentration.
`lower than
`80 -95% of the nitrogen oxides leave the engine as NO and only 5 -20% as NO2.
`In exhaust systems without oxidation catalysts NO is oxidised to NO2 mainly
`only after leaving the tailpipe as the reaction of NO with 02 is very slow under
`these conditions. The presence of an oxidation catalyst can speed up the NO
`oxidation reaction and increase the NO2 -share of NO downstream of the catalyst,
`where the NO2 can then be utilised for soot oxidation. An alternative, but not yet
`
`...,
`
`N. \ i \
`
`\i, \
`
`&
`
`,
`
`s
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`, \
`
`,
`
`w \
`
`w
`
`10
`
`_ ..
`
`.. NO2 concentration at 1% 02 concentration
`NO2 concentration at 5% 02 concentration
`NO2 concentration at 12% 02 concentration
`
`o
`
`., ` r,
`
`o
`
`50
`
`100 150 200 250 300 350 400 450 500 550 600 650 700 750 800
`
`Temperature / °C
`
`Fig. 2. Equilibrium NO2 concentration for various (diesel emissions typical -) oxygen concentrations
`as a function of temperature
`
`BASF-2003.004
`
`

`
`Diesel Particulate Filter Regeneration
`
`95
`
`commonly applied way to increase the rate of the NO oxidation is the application
`of a non -thermal- plasma reactor [34, 39 -42].
`NO2 oxidises soot to CO2 at reasonable rates at temperatures as low as ca.
`250 °C [43].
`
`Design of Diesel Oxidation Catalysts (DOC) for Nitrogen Dioxide
`Based Soot Regeneration Strategies
`A DOC typically consists of a ceramic (most common is cordierite) or metallic (most
`common is stainless steel) substrate with geometric surfaces of a few square meters.
`To increase the surface area for a good distribution of the precious metal catalyst
`material (most common is Pt for DOCs) a so- called "washcoat" consisting of a
`thermally stable high surface area metal oxide (with a few hundred square meters
`surface area per gram, e.g. A1203) is used. The main function of the washcoat is to
`stabilise the fine dispersed catalyst material against sintering, but can also be used to
`enhance the catalytic activity or protect it against poisoning. The aim is to use as
`much washcoat material as possible without compromising the backpressure behav-
`iour of the catalyst too much. The typical range for washcoat loading is between 3
`and 300 g /dm3 catalyst volume with typical metal loadings of 0.1 -10 g /dm3 catalyst.
`
`Reactions Catalysed by DOCs for NO2 Based Soot Regeneration
`Besides of the NO oxidation further reactions taking place over the DOCs are the
`oxidation of CO according to Eq. (3) and hydrocarbons (HCs) according to Eq. (10).
`C,IH,n + (n + )02
`n CO2 + 2 H2O
`(10)
`As CO and HC compete with NO for the active sites on the catalyst, a very high
`activity for all these oxidation reactions is desired. When sulphur -containing spe-
`cies like SO2 are present in the diesel exhaust, they will also be oxidised over the
`DOC to eventually form sulphuric acid and sulphates (Eqs. (11)- (13)).
`SO2 + 2 02
`SO3
`
`(11)
`
`SO3 + H2O -4 H2SO4
`
`(12)
`
`Al2O3 + 3 SO3 -> Al2(SO4)3
`(13)
`Equation (11) competes with Eq. (9) and therefore limits the NO2 formation. The
`formation of sulphuric acid (Eq. (12)) can block the active sites and the formation
`of washcoat component sulphates (e.g. Eq. (13)) can destroy its functionality.
`Another disadvantage of sulphuric acid and sulphate formation is that sulphuric
`acid and sulphates can pass DPFs and are measured as particulate matter in legis-
`lated emission tests. Therefore the sulphur contents of the diesel fuel as well as the
`sulphur contents of the engine oil have to be minimised.
`
`Chemistry of Nitrogen Dioxide Based Soot Oxidation
`The chemistry of nitrogen dioxide based soot oxidation has been subject to several
`studies [20, 44, 45]. In recent synthetic gas studies with carbon black exposed to a
`
`BASF-2003.005
`
`

`
`96
`
`C. Görsmann
`
`mixture of NO2, 02, and water [46] two types of C- oxidation reactions have been
`identified: direct C -NO2 reactions (Eqs. (14) and (15)) and cooperative C -NO2-
`02 reactions (Eqs. (16) and (17)) in which ( +NO2) refers to the occurrence of an
`unstable C- (02)
`- NO2 complex, which promotes the desorption from the solid
`carbon surface.
`
`-
`
`-C + 2 NO2
`CO2 + 2 NO
`-C + NO2 -7 CO + NO
`
`(14)
`
`(15)
`
`CO2 + NO ( +NO2)
`
`(16)
`
`-C + 2 02 + NO2( +NO2)
`-C + 2 02( +NO2)
`CO ( +NO2)
`(17)
`The contribution of the cooperative C- N0, -02 reaction to the overall oxida-
`tion rate increases when the temperature increases. The presence of water vapour
`promotes the direct C -NO2 reactions. The main products from these direct (Eqs.
`(14) and (15)) and cooperative (Eqs. (16) and (17)) reactions are CO2 and NO, the
`products of the reactions shown in Eqs. (14) and (16). The amount of CO formed
`(Eqs. (15) and (17)) is only a few percent of the total amount of products, but
`increases with temperature increase.
`In engine and vehicle tests of CRT® applications it is often found that tailpipe
`NO emissions are up to 2 -10% lower than engine out NO emissions [11]. The
`mechanism of this NO conversion is not fully understood yet. Possible NO con-
`version products could be HNO3 or nitrates (which might be stored on the partic-
`ulates or the filter), N20, or N2. PM can contain nitrates, but the amount found on
`PM is usually very small. N20 formation could occur even without the presence of
`a reducing agent between 200 and 500 °C according to Eq. (18).
`3 NO2 -+NO +N2O +2 02
`(18)
`The possible equilibrium amount of N2O under diesel exhaust conditions (ca.
`500 ppm NO and 5% 02) lies in the lower percent region. However, significant
`amounts of N20 have not been observed. As N2 is present in large amounts in
`diesel exhaust a possible NO reduction to N2 would not be measurable under
`engine conditions and cannot be excluded. Possible NO reaction pathways could
`be as shown in Eqs. (19) -(22).
`NO2 +C -+ CO2 +;N2
`
`(19)
`
`NO2 + 2C
`
`2 CO + ; N2
`
`(20)
`
`CO2 + N2
`2 NO + C
`NO +C- CO +2 N2
`(22)
`There might also be a small contribution of NO reduction by so called "lean -
`NO," reactions with HCs over the DOC. Its ideal pathway can be described by
`Eq. (23) [47].
`
`(21)
`
`(2m + 2 n) NO + C,nH,, - (m + n) N2 + in CO2 + 2 n H2O
`
`(23)
`
`BASF-2003.006
`
`

`
`Diesel Particulate Filter Regeneration
`
`97
`
`However, real lean -NO, reactions between HCs and NO over Pt- containing cata-
`lysts are very complex [48, 49] and usually lead to a product mix with a major
`share of N20.
`To verify, which of the above or alternative reactions lead to some NO con-
`version, model gas tests and possible isotopic "labelling" of N or O in NO might
`help to characterise the reaction pathways.
`As shown, there are numerous possible reaction pathways for soot oxidation to
`regenerate DPFs. Model gas tests can help to verify the conditions, under which
`these possible reactions might occur, but it seems to be very difficult to fully
`characterise real world DPF regenerations.
`
`Passively Regenerating Diesel Particulate Filter Systems
`In passively regenerating DPF systems the soot oxidation happens during the
`normal operation of the vehicle without any action required by either the operator
`or the engine control system. For a passive regeneration strategy using oxygen, the
`exhaust temperatures have to be reliably frequent above 400 °C in case of FBCs
`present or even 500 -550 °C without FBCs present. For a passive regeneration
`strategy using NO2 the exhaust temperatures have to be reliably frequent above
`250 °C.
`Exhaust temperatures of passenger car applications span a wide range during
`"normal" operation as their duty cycles can be extremely different depending on
`vehicle is driven. At high speed driving exhaust temperatures
`how
`might exceed 500 °C, but at low load city driving, temperatures might not even
`reach 200 °C. Because of the unpredictable drive cycle of passenger cars, passenger
`car DPF regeneration cannot rely on any passive regeneration strategy.
`Exhaust temperatures of heavy duty diesel (HDD) engine applications are in
`the region between 200 and 400 °C during normal operation and most of their drive
`cycles are repeating and very predictable. Under "normal" operation conditions
`HDD exhaust gas temperatures don't reach the areas, which would be required for
`oxygen -based regeneration, but most vehicles would reach the temperature region
`above 250 °C frequently enough to allow a passive, NO2 based regeneration strat-
`egy. The following commercially available systems are based on NO2 based
`(= Continuously Regenerating Trap: Diesel Oxi-
`passive PM regeneration: CRT
`dation Catalyst (DOC) followed by DPF) [50, 51], CSF (= Catalysed Soot Filter:
`Catalyst Coated DPF) [52, 53], and CCRTTM ( =DOC followed by CSF) [54].
`They all require ultra low sulphur fuel (max. 50 ppm S, ideally <10 ppm S), the
`knowledge of the operation temperature profile, and a minimum NO
`to PM
`ratio.
`
`Function Principle of CRT®, CSF, and CCRTTM
`The first step for carbon burning at low temperatures by NO2 is generating NO2.
`This can be done by oxidising NO over an oxidation catalyst upstream of a partic-
`ulate filter (CRT® system) or over a catalytic coating placed on a diesel particulate
`filter (CSF). NO2 oxidises soot (mostly carbon) collected on top of and inside the
`walls of the particulate filter to CO2. As a result NO2 is reduced back to NO.
`
`BASF-2003.007
`
`

`
`k;
`
`98
`
`C. Görsmann
`
`The main functional requirements for a passive regenerating DPF system are a
`sufficient temperature profile and a minimum NO
`to PM ratio. The required
`values are system dependent and interdependent of each other. The basic factors
`involved are:
`
`1. NO mass flow, depending on speed and load of the engine, combustion
`conditions.
`2. NO to NO2 oxidation kinetics, depending mainly on the catalyst used and its
`ageing status, temperature, and mass flows of NO, NO2, 02, CO, and HC.
`3. PM mass flow, depending on speed and load of the engine, combustion quality.
`4. Soot oxidation kinetics, depending mainly on temperature, NO2 mass flow, soot
`reactivity, soot loading of the filter, and soot distribution within the filter.
`A DOC followed by a DPF forms 'a CRT ° -- system. Figure 3 shows its typical
`set -up.
`Figure 4 shows the pollutant conversion of field aged CRT® systems. Even after
`very long operating distances the systems tested show still very high CO, HC, and
`PM conversion similar to the performance of a new one.
`
`Fig. 3. CRT('''-schematic
`
`CO
`
`111HC PM
`
`airport
`loco-
`motive
`bus
`600k km 550k km
`
`express
`bus
`500k
`km
`
`mail
`city
`truck
`bus
`450k km 250k km
`
`garbage garbage
`truck A
`truck B
`200k
`100k
`km
`km
`
`new
`
`Fig. 4. Pollutant conversion of field aged CRT-systems
`
`BASF-2003.008
`
`

`
`Diesel Particulate Filter Regeneration
`
`NO }. isia
`
`CO2 + 11,0
`
`O - ,
`
`99
`
`r-
`
`sCO2+ NO
`
`602+ NO
`
`Flow Through Monolith
`
`''
`
`Wall Flow Filter
`
`r
`
`COZ+ NO
`
`NO2
`
`Fig. 5. Scheme of the working principles of CRTC, CSF, and CCRTTM
`
`In a "coated DPF" or "CSF" (Catalysed Soot Filter) the role of the oxidation
`catalyst (CO, HC, and NO oxidation) has been transferred to the filter. As NO? is
`formed along the length of the filter, some NO? is only formed downstream of
`some of the soot. When this NO? hits soot particulates collected on the filter, these
`are oxidised to CO2, while NO? is reduced to NO. Over the catalytic coating of the
`filter NO can be re- oxidised to NO? by present oxygen, and is therefore available
`for further soot oxidation.
`Combining CRT® and CSF results in a CCRTTM- system, which allows the
`maximum use of the NO for soot regeneration. Figure 5 shows the principle of
`CRT `t, CSF, and CCRTTM. The key reactions shown in Fig. 5 are those given in
`Eqs. (9) and (14).
`The principle difference between CRT ®, CSF, and CCRTTM is the location,
`where these reactions happen: In a CRT® system Eq. (9) takes place over the
`DOC and Eq. (14) in the soot layer of the DPF. As there is no catalytic coating
`on the DPF the NO from Eq. (14) cannot be re -used in the DPF. Due to the presence
`of oxidation catalyst coating within the filter walls of the CSF -only and CCRTTM-
`systems the NO formed in Eq. (14) can be re- oxidised according to Eq. (9) and can
`be made available for further C- oxidation reactions according to Eq. (14). Depend-
`ing on the temperature and the availability of C and active sites to re- oxidise NO,
`this cycle can occur several times. However, in reality the systems are more complex
`than described above and not all reaction pathways are well understood yet.
`
`Comparison of CSF, CRT °, and CCRTTM Systems
`
`Comparison CSF -CRT®
`A CSF system is more compact than a CRT® system as it requires only one
`component instead of two. But its compactness requires compromises regarding
`exhaust backpressure and catalyst loading.
`
`BASF-2003.009
`
`

`
`100
`
`C. Görsmann
`
`The soot layer on top of the filter walls burns less effectively in a CSF than a
`CRT° system. This is because in a CSF NO is oxidised to NO2 underneath the soot
`layer whereas in a CRT° system NO2 is formed upstream of the DPF and has
`therefore a higher chance to get in contact with the soot layer on top of the filter
`walls. For the same reason a CSF system has a larger NO2 slip than a CRT° system
`with the same amount of precious metals applied and requires higher temperatures
`( >300 °C) for significant carbon regeneration.
`A CSF system requires less installation space but higher operating temperatures
`than a CRT° system. In a CSF NO can potentially be used more than once for soot
`oxidation. At high enough temperatures ( >350 °C) the NOt:PM ratio requirements
`can be lower for CSFs than for CRT® systems.
`Due to its design a CRT° system is less sensitive to ash accumulation and more
`robust against sintering of the precious metal components. Ash accumulation in
`CSF reduces efficiency due to blocking of gas pathways as the ash plug builds up.
`Precious metal sintering is a high temperature ageing mechanism leading to a
`reduced number of catalytically active sites. To minimise the increase of back -
`pressure due to the catalyst coating, the amount of washcoat, which stabilises the
`precious metal components from sintering, has to be minimised. This leaves the
`precious metals in a CSF slightly less protected against sintering than in a DOC. In
`case of spontaneous thermal regenerations with oxygen the carbon burning creates
`heat within the DPF. Such heat would not effect an upstream DOC (like in the
`CRT® set -up), but the catalytic coating of the CSF would be fully exposed to it. If
`the temperature on the CSF surface becomes too high, this could potentially cause
`some precious metal sintering resulting in catalyst de- activation.
`Both systems have advantages and disadvantages. Thus, which system to choose
`depends on the conditions for its application. In HDD applications, where the low
`temperature activity is most important, the CRT° system seems to be the better choice.
`
`Comparison of CCRTTM with CSF and CRT° Systems
`Figure 6 compares the pressure rise of an un- coated DPF, a CSF, a CRT® and a
`CCRTTM system with each other in a low temperature cycle (T< 270 °C) to assess
`
`Bare Filter Alone
`..-CSF Alone
`
`-- CCRT'"
`
`0.16
`0.14 -
`0.12
`0.1 -
`0.08
`
`0.06
`
`0.04
`
`0.02
`
`0
`
`CSF Alone
`
`CRT`'
`
`CCRT"
`
`0
`
`5
`
`10
`
`15
`
`25
`
`30
`
`35
`
`40
`
`45
`
`20
`t/h
`
`Fig. 6. Un- coated DPF, CSF, CRT®, and CCRTTM; backpressure behaviour over time during low
`temperature cycle at T< 270°C
`
`BASF-2003.010
`
`

`
`e1
`
`Diesel Particulate Filter Regeneration
`
`101
`
`their low temperature behaviour. The backpressure of these particulate filter sys-
`terns relates to its soot loading. A higher soot loading results in a higher back -
`pressure. The change
`in backpressure relates therefore to
`the regeneration
`efficiency compared to the soot accumulation.
`The CCRTTM system combines properties of CRT© and CSF. It allows a more
`efficient use of the NO emitted from the engine for carbon burning and is even at low
`CSF precious metal loadings superior to CRT®- and CSF- systems. This is especially
`true for low temperature applications and applications with a low NQ to PM ratio.
`NO2 -slip can be minimised by applying optimised metal loadings and distributions.
`
`Active Regenerating Diesel Particulate Filter Systems
`If applications are too cold (e.g. garbage trucks or some city busses) or driving
`cycles do not reliably frequent cause the exhaust temperature to reach at least
`250 °C (e.g. passenger cars) to enable passive regeneration, active regeneration is
`required. If the NON- emissions of the engine are high enough, the temperature can
`be raised to burn soot by NO2. If the NOR-emissions of the engine are low, the soot
`combustion has to be done by oxygen.
`Increasing the temperature to 300 -350 °C for NO2 -based regeneration can be
`done in a number of ways, e.g. by fuel injection into the exhaust or by engine
`means. Active regeneration via NO2 requires a relatively long time (due to low
`NO2 concentration in the exhaust) but is a safe strategy, as high temperatures are
`avoided.
`For a temperature increase to 550- 600 °C, which is required for regeneration
`with 02, "post injection" with a high rate is required. "Post injection" is a diesel
`fuel injection into the engine combustion chamber at a very late time of the com-
`bustion stroke of the engine so that the fuel is not burned as part of the usual
`combustion process. The aim of a post injection is to put large amounts of
`unburned hydrocarbons into the exhaust. These hydrocarbons burn over the oxida-
`tion catalyst and raise the exhaust temperature via the heat generated. Due to the
`higher 02 concentration the reaction rate of 02 -based soot regeneration is higher
`than the rate of the NO2 concentration limited NO2 -based regeneration. However,
`there is a risk of destroying the filter. This risk is especially high in situations where
`the carbon load of the filter is high and the exhaust mass flow low. If a 02 -based
`regeneration starts under these conditions, carbon burning creates a lot of heat in
`the filter, which might not be sufficiently removed from the filter. If too much heat
`is retained in the filter, it might melt (cordierite, sinter metal) or oxidise (SiC). If
`the temperature gradient along the diameter or the axis of the filter exceeds the
`filter material limits, the thermal stress could cause filter cracking. Therefore
`applying this strategy requires care.
`
`Active Regeneration of a CRI® System via NO2 in a Field Trial
`In a field trial to study active regeneration by NO2 [55] the exhaust temperature has
`been controlled by varying the turbo charger efficiency. Active regeneration was
`triggered when the system backpressure reached a critical value. The temperature
`increased and the backpressure dropped. The success of this trial demonstrated that
`
`BASF-2003.011
`
`

`
`102
`
`C. Görsmann
`
`this approach represents a promising strategy. Active regeneration via NO, is a safe
`strategy, but requires a long time dùe to the low mass flow of NOR.
`CCRTTM- systems offer significant advantages for the NO2 -based active regen-
`eration compared to CRT ® -- and CSF- systems.
`
`Active Regeneration via 02
`Active Regeneration via 02 is applied successfully in more than 500.000 LDD
`applications [56] and is the most likely solution for all US HDD- applications (with
`low temperatures and /or low NOR to PM ratio) to meet US 2007 legislation limits.
`The first LDD- systems with 02 -based active regeneration used a highly active
`DOC in front of a SiC -filter and a cerium based additive to lower the 02 -C -reac-
`tion temperature (see Fig. 1). To avoid excessive ash accumulation HDD- applica-
`tions would have to do without additives. HDD engines are built to run for more
`than one million kilometres and even with oil ash accumulation only there would
`be too much ash to store in the filter for the lifetime of the vehicle. Oil ash
`accumulated in a DPF needs to be cleaned every 100000 -200000 km. If FBCs
`would be used, the frequency of ash cleaning intervals, which cause downtime
`for the vehicle, would increase, which is highly undesirable.
`For the non -catalysed oxygen -carbon reaction (Eq. (1)) temperatures of
`550 -600 °C are required. These temperatures can be obtained by post injection
`in combination with an oxidation catalyst. Here the unburned fuel is being oxidised
`over the oxidation catalyst and produces the heat required to heat up the exhaust
`gases to the desired temperature. The heat production can be done over an oxida-
`tion catalyst, which is functioning as a catalytic burner or over a (precious metal
`containing) filter coating. In both cases a high resistance of the coating against high
`temperature exposure is required.
`The filter coating could also have a role in the catalytic carbon oxidation by
`oxygen. However, compared to uncoated DPFs coated DPFs have shown only
`small advantages in supporting 02 based soot regenerations [30]. These results
`indicate that catalytic DPF coatings have no major influence on the 02 -C- reaction.
`However, catalytic CSF coatings resulted in a faster, more complete regeneration in
`the temperature range of 540- 640 °C.
`Active regeneration is a fast process and represents a very promising strategy
`for the filter regeneration.
`
`Diesel Particulate Filter Systems as Part of 4- Way- Systems
`For additional nitrogen oxide reduction DPF- systems can be combined with NOR
`emission control technologies. The currently most common technology for NO
`emission reduction is exhaust gas recirculation (EGR). By combining a CRT
`with EGR very high ( >90%) PM -, CO -, and HC- reduction and ca. 40% NO,-
`reduction can be achieved [57]. If a more effective NOR- reduction is required, a
`NO,-storage catalyst coating or an SCR -system can be applied. A DPF with a NOI
`storage catalyst coating is known under the term "DPNR" (created by Toyota) [58,
`59]. By combining a CRT® with an SCR -system a so- called SCRT® system is
`obtained [51, 60]. Depending on engine out emissions as well as speed /load and
`
`BASF-2003.012
`
`

`
`Diesel Particulate Filter Regeneration
`
`103
`
`Table 1. Emissions data of various SCRT® systems obtained by ESC (g kW-111-1, %) [48, 57]
`
`System
`
`HC
`
`CO
`
`NO,.
`
`PM
`
`SCRT [48] /Eng. out
`(conversion)
`SCRT [57] /Eng. out
`(conversion)
`
`0.003/0.162
`(98 %)
`0.002/0.123
`(98 %)
`
`0.000/0.989
`(100 %)
`0.000/0.324
`(100 %)
`
`1.061/7.018
`(85 %)
`0.562/6.926
`(92 %)
`
`Euro V Limits
`
`0.460
`
`1.500
`
`2.000
`
`0.007/0.163
`(96 %)
`0.008/0.022
`(64%)
`
`0.020
`
`temperature conditions, the efficiency of such systems can exceed 90% for all four
`pollutants, CO, HC, PM, and CO. Table 1 shows emission data of SCRT® systems
`obtained by the ESC tests.
`
`Conclusions
`A variety of DPF systems are available to reduce soot emissions from diesel engine
`applications. The main challenge associated with DPF systems is the regeneration
`of the soot retained in them. Soot regeneration is based on soot oxidation. This can
`be done by oxygen or nitrogen dioxide. If the driving cycle and the exhaust con-
`ditions allow, it is .possible to use passively regenerating systems. For all other
`cases active regenerating systems are available. Whatever the regeneration method,
`catalytic coatings play a key role in supporting them.
`DPF systems can also be combined with NO control technologies to form 4-
`way systems. With such systems more than 90% of PM, CO, HC and NO emis-
`sions can be removed to meet current and future legislation requirements.
`
`Acknowledgements
`
`I would like to thank numerous colleagues for helpful discussions during the writing of this paper,
`notably Dr. A. P. Walker. I would like to thank Johnson Matthey for permission to publish this paper.
`
`References
`
`[1] Majewski WA (2004) What are Diesel emissions (Ecopoint Inc. Revision 1999.09a) In:
`DieselNet Technology Guide www.dieselnet.com /tech /emi_into.hmtl (accessed 26. 09. 04)
`[2] Maynard RL, Howard CV (eds) (1999) Particulate Matter: Properties and Effects upon Health,
`Oxford
`[3] Seaton A, MacNee W, Donaldson K, Godden D (1995) Lancet 345: 176
`[4] Brüske -Hohlfeld I, Cyrys J, Peters A, Dockery DW (2004) In: Steinmetz E (ed) Minimierung der
`Partikelemissionen von Verbrennungsmotoren, p 58
`[5] Wichmann HE et al (2000) Daily mortality and fine and ultrafine particles in Erfurt, Germany.
`Part I: Role of particle number and particle mass. Health Effects Institute Research Report 98
`[6] Ibald -Mu