Learning Center
Plans & pricing Sign in
Sign Out



1. Introduction
       Concerns on health effects and the tightening of diesel emissions regulations have
driven research and development on control technologies for the reduction of harmful diesel
exhaust emissions. Although improvement in engine technology has remarkably reduced
engine out emissions on both NOx and particulate matter, together with a diesel oxidation
catalyst (DOC) for the oxidation of CO and HC, diesel particulate filter (DPF) will probably
become a standard feature of future diesel powered vehicle [1]. The introduction of diesel
particulate filter can reduce diesel particulates emission up to 98% [2,3]. Besides the trap
itself, the accumulation of the trapped particulate matter (soot) will create a back-pressure
that has to be maintained as low as possible. An increase of 100 mbar back-pressure is
estimated to increase the fuel consumption with 1 to 4.5% [4]. To keep the back-pressure at
an acceptable level, the accumulated trapped soot has to be combusted.
       Oxidation of trapped soot is the topic of this study. During this PhD project, two types
of diesel soot oxidation systems were studied. The first reaction system was based on molten
salt catalysed soot oxidation. In this feature soot oxidation with O2 is a basic reaction. The
activity of molten salt catalyst was examined under simulated diesel exhaust conditions. The
influence of NOx on the oxidation rate of soot with molten salt was investigated, and
compared to the performance of a platinum catalysed system. Furthermore, the molten salt
catalyst-soot interaction was evaluated by means of in-situ visible microscopy. The second
reaction system studied was the soot-NO2 based reaction, which is the basis of the so-called
CRT system. Efforts have been made to optimise the reaction system in anticipation to the
trend of decreasing engine out-NOx concentration. The study was intended to find a proper
catalytic filter configuration that effectively utilises NOx. Furthermore, the potential of CeO2
catalyst for the acceleration of soot-NO2 reaction was also evaluated. At the end of the
chapter the practical option for future diesel soot abatement technologies will be discussed
and evaluated.

2. Soot oxidation with molten salt catalyst
      Studies to explore mobile type materials as candidates for diesel soot oxidation catalysts
were stimulated by the awareness that catalyst-soot contact is a very important factor for the
catalytic oxidation of diesel soot based on C-O2 reactions. The catalyst-soot contact can be
created by either evaporation or surface migration. It has been shown that a eutectic mixture
of Cs2SO4 and V2O5 (Cs2SO4.V2O5), a molten salt catalyst, showed activity in the presence of
O2 [5]. Considerably high activity is exhibited at and above the catalyst melting point (625
K). In Chapter 2, the catalyst performance evaluated under simulated diesel exhaust
conditions is described. The main conclusion is that the catalyst is hardly influenced by any
diesel exhaust gas component. It is almost completely inactive in the oxidation of CO and HC
to CO2, and of NO to NO2. As a result, if the catalyst would be applied it will not take any
advantage from NO present in the exhaust stream. Furthermore, an oxidation catalyst is still
needed for the removal of HC and CO. Its performance mainly relies on the C-O2 reaction. In
contrast, platinum catalyst requires NO to play any role in the oxidation of soot. The role of
platinum catalyst is to oxidise NO to NO2, where the formed NO2 subsequently oxidises soot.
These reaction can occur in a kind of fixed bed reactor configuration where Pt catalyst can be
located upstream of soot bed. In a proper platinum catalyst-soot configuration, for example,
in the physical mixture of catalyst and soot, the reaction sequence of both Pt catalysed NO
oxidation and soot oxidation with NO2 can be repeated in the depth of reactor configuration.
A recycle mode is created and NO is repeatedly used.
     With respect to the catalyst-soot contact, a characterisation by means of in-situ visible
microscope has been carried out. It is concluded in Chapter 3 that during the oxidation
process the soot particle is tightly attached to the molten catalyst. At the catalyst-melting
regime, where the catalyst becomes significantly more active, a catalyst-soot-gas (O2)
interface is created where the oxidation reaction takes place. As far as the experimental
method is concerned, microscope observation is an excellent way to demonstrate the catalyst
activity and to predict the oxidation rate of soot. The predicted oxidation rate is in agreement
with the rates obtained in the standard equipment such as flow reactor and TGA.
      In view of practical application, molten salt catalyst, as well as other types of mobile
materials, faces a stability issue. The stability of Cs2SO4.V2O5 catalyst has been reported to
be low at severe hydrothermal conditions (10 vol.%H2O and 875 K) [6]. Considering that the
catalyst is active above 625 K, the application of molten salt will be limited to, for example,
stationary or ships engine. It is expected that in these types of diesel engine applications such
a temperature range can be reached.

3. NOx-assisted soot oxidation
     The second catalytic oxidation system studied is the soot-NO2 based reaction system.
The goal was to explore the optimal NOx-assisted soot oxidation system in anticipation of
decreasing engine-out NOx concentration. In Chapter 4, a comparative study in order to
understand the characteristic of soot-NO2 based reaction between two types of realistic soot
samples (the idle soot and the full load soot) and a soot model (Printex-U) is presented. A
number of interesting conclusions that can be drawn from this study are the following. The
reactivity of Printex-U and diesel soot (idle and full load soot) to NO2 and NO2/O2 mixture is
similar. With respect to O2 alone, the reactivity of the real diesel soot samples is much higher
and follows the order of Printex-U < full load soot < idle soot. This sequence correlates with
the hydrogen-content in the samples. The role of NO2 and O2 in the reaction mechanism was
examined. In the soot-NO2/O2 reaction system, the oxidation of soot is initiated by the soot-
NO2 reaction to create Surface Oxygen Complexes (SOC's) as intermediates. These
intermediates are reactive to oxygen, yielding less stable intermediates or they directly
decompose to CO and/or CO2. As a result, soot oxidation with NO2 is enhanced by the
presence of O2. The reason for the low reactivity of O2 is that it only slowly generates SOC’s
at low temperature. It is important to emphasise that for both soot-O2 and soot-NO2 based
reaction Printex-U is a reliable and realistic soot model.
     A concept of multifunctional catalytic reactor filter is proposed in Chapter 5. The filter,
TU Delft (TUD) filter, consists of Pt/ceramic foam upstream of a wall-flow monolith filter.
This filter configuration is able to overcome the drawback of two standard filter
configurations; the so-called an integrated system (only/Pt ceramic foam) and a separated
(CRT-like) system. In the integrated system, Pt/ceramic foam acts as a reasonably efficient
soot trap, NO2 generator, as well as a reactor for the oxidation of soot with NO2. Recycle
reactions are observed; NO is repeatedly used and a very high reactivity in soot oxidation is
determined. The use of Pt/ceramic foam as NO2 generator and reactor for the oxidation of
soot tends to produce a secondary emission of NO2 (so-called NO2-slip). The separated
system consists of NO2 generator upstream of a wall-flow monolith as soot trap. Comparing
the separated system with the integrated system, the former produces less NO2-slip with a
low oxidation rate. The integrated system will, however, lead to a high soot oxidation rate at
the expense of a relatively high NO2-slip.
      An improvement on the utilisation of NO2 has been observed in the TUD catalytic filter.
In this filter soot oxidation takes place in a Pt/ceramic foam as a kind of integrated system
and in the wall-flow monolith as a CRT system. The oxidation rate of the TUD catalytic filter
is just about 90 % of that of the integrated system with effectively low NO2-slip. In the TUD
catalytic filter, as in all diesel particulate filters, secondary CO emissions take places. The CO
emission of the TUD filter is 2-3 times lower than the CO emissions of the CRT-like
configuration at identical diesel soot oxidation rates.
      The oxidation of soot in the wall-flow monolith part both in the CRT and the TUD filter
apply only a non-catalytic soot-NO2 reaction in the system. To improve the system, CeO2
catalyst has been evaluated as a catalyst for the soot-NO2 reaction. It is shown in Chapter 6
that CeO2 has the potential to accelerate the NOx-assisted soot oxidation rate. The interaction
of NO2 with CeO2 will form initially surface nitrites and nitrate, which is followed by oxygen
transfer to CeO2. This oxygen is stored on CeO2 and, the desorption of it, leads to the release
of 'active oxygen'. This active stored oxygen is postulated to play a major role in the
acceleration of soot oxidation. Decomposition of surface nitrates result in a gas-phase NO2
that led to a minor effect in soot oxidation. Unfortunately, N2O emission is observed from the
system. In the process of the storage of “active oxygen” around 10% of NOx is converted into
N2O, a green house and ozone depletion gas, which is highly unwanted.

4. Results significance and an outlook to state-of-the-art systems
     In view of practical application, there are two significant outcomes of this study viz. the
concept of the TUD filter (multifunctional catalytic reactor) and the finding of "active
oxygen" as a result of NO2 interaction with CeO2. In relation to the existing filter systems, the
TUD filter is concluded to out perform the CRT system by providing soot trapping and
recycle reaction in Pt/catalysed ceramic foam. As the postulated "active oxygen" has been
shown to increase the oxidation rate dramatically, the system has a potential for practical
application. Discussion on active oxygen will be presented at the end of this chapter in
relation to a newly developed filter system [7,8].
       Figure 1 summarises the concepts of TUD filter in comparison with the CRT system. It
is shown that all emission reduction processes of the CRT system (CO, HC, and soot
oxidation) can be accomplished in the TUD filter. The advantage of TUD filter is that soot
filtration and soot oxidation takes place in both the Pt/ceramic foam and the wall-flow
monolith leading to a high oxidation rate.

                            Functions:                                     Reactions:
   HC’s, CO,NO,
     Soot, O2
                  CRT System:

                  Pt coated flow-through monolith :           NO                  1
                                                                                     O2   
                                                                                            NO2                      (1)

                  1. NO2 generator                            CO                  2 O2  
                                                                                   1      Pt
                                                                                              CO2                    (2)

                  2. Reactor for CO and HC oxidation         CxH y                    1
                                                                                       2          
                                                                                           O 2   C O 2  H 2O

                  Wall-flow monolith:
                  1. Soot filter (total filtration)                    
                                                             NO2  C   CO  NO                                      (4)
                  2. CRT system utilising NO2-slip

    CO, NO, NO2

   HC’s, CO,NO,
     Soot, O2     TU Delft Filter
                                                             NO               1             
                                                                                            P t
                                                                                       O 2   N O 2
                  Pt/ceramic foam :                                            2
                  1. Soot filter (40-70% efficiency)                   
                                                             NO2  C   CO  NO
                  2. NO2 generator
                  3. Reactor for soot, CO and HC oxidation   CO               1
                                                                               2         
                                                                                   O 2  Pt  C O 2                   (2)
                                                             CxH   y                  1
                                                                                       2         
                                                                                           O 2  P t  C O 2  H 2O   (3)
                      NO2-slip: non-reacted NO2
                      emitted from Pt/ceramic foam

                  Wall-flow monolith:
                  1. Soot filter (total filtration)                    
                                                             NO2  C   CO  NO                                      (4)

                  2. CRT system utilising NO2-slip
   CO, NO, NO2

Figure 1. Schematic representation of the TUD filter and the CRT system; functions and reactions

       In Table 1, the dependency of the CRT system to engine out NO (NOx) is described in
comparison with TUD filter. Using the listed assumptions, it can be seen that in CRT system
100 unit quantities of NO, for example 100 NO molecules, will yield about 40 CO molecules,
which is equivalent to 40 C atoms combusted. In the TUD filter this dependency is lower,
100 NO molecules will yield 72 (40+72) C atoms combusted. In other words, to oxidise the
same amount of soot, engine out NOx concentration needed by TUD filter can be about 35%
lower than that of the CRT system. Furthermore, the additional CO emission resulted from
soot oxidation will be lower for the TUD filter.
Table 1. CRT and TUD filter dependency to engine out NOx
         Gas in          Reactions and location                          Gas out
         100 NO          Pt/catalysed flow-through monolith
                          NO       1
                                   O2  NO2                            50 NO2
CRT                      (Assumed that 50% NO is converted to NO2)       50 NO
         50 NO2          Wall-flow monolith:                             40 CO,
         50 NO                     
                          NO2  C  CO  NO                             (50+40) NO, 10 NO2
                         (Assumed that 80% NO2 reacts)                   (100 NOX out)
                                                                         (40 C is oxidised)
         100 NO          Pt/ceramic foam:                                50 CO2
                          NO       1
                                             P t
                                        O 2   N O 2                   40 NO2-slip

                          NO2  C   CO  NO
                         CO        1
                                    2          
                                        O2  CO2
filter                                                                   (40 C is oxidised)
                         (Assumed that 50% NO is converted to NO2, and
                         CO produced is equivalent to NO2)*
         40 NO2-slip     Wall-flow monolith:                             32 CO
                         NO2  C  CO  NO
                         (Assumed 80% NO2 reacting; 40 C is oxidised)    (32 C is oxidised)
*) Note: By the occurrence of recycle reaction more C can be oxidised in the Pt/ceramic

      The TUD filter has only been tested in the laboratory experiment. For further study, it is
recommended to test the filter in the real exhaust conditions to evaluate the performance of
the filter such as back-pressure and balance temperature characteristics as a function of
engine out NOx. Since the system is thought to work with low NOx, the application of the
TUD filter for light-duty vehicle might be relevant. For heavy-duty vehicle where higher NOx
is normally available, the system might work better than in the light-duty vehicle. As with the
CRT system the TUD filter will be a sulfur sensitive system. The fuel regulation towards
sulfur free diesel fuel in many countries can boost the application of the "CRT-based" filter
      With respect to the application of the CeO2 catalyst, there are several possibilities that
can be considered. CeO2 catalyst might be loaded on the DPF wall-flow monolith part of the
CRT system or TUD filter. In this system NO2 is produced from the oxidation of NO on
Pt/ceramic foam or Pt catalysed flow-through monolith. This NO2, un-reacted NO, and O2
will interact with CeO2 catalyst. As presented in Chapter 6, the interaction results in surface
nitrate and stored oxygen that can be released as “active oxygen”. The release of “active
oxygen” gives an additional soot oxidation route leading to a higher soot oxidation rate. Since
the role of active oxygen has been shown only with powdered catalyst, it is recommended to
test the system on a suitable support. The choice for an appropriate filter material and loading
methods of the active storage materials are, among others, the recommended studies. The
interaction of CeO2 with NO2 also tends to yield N2O, which is highly undesired. Therefore,
efforts to suppress the formation of N2O or to decompose it have to be considered.
Furthermore, an extensive screening of metal oxides, or the combination thereof, that can
produce active oxygen following the interaction with NO2 without N2O production will be an
interesting and practical topic.
      Although generated from different system, “active oxygen” is postulated as a species
that plays a role in the newly developed system, the Diesel Particulate and NOx Reduction
(DPNR) Toyota Motors system [7,8]. In the DPNR system a layer of an “active oxygen”
storage alkali metal oxide is deposited along diesel soot filtration surface areas. On this layer
platinum is dispersed. The “active oxygen” is created by the conversion of gas-phase NO
over the platinum into surface nitrate species. These surface nitrates will be decomposed at
the interface between the soot and the active oxygen layer into very reactive adsorbed oxygen
atom and NO. The NO can be re-oxidised once more to surface nitrate and the adsorbed
oxygen atom is able to oxidise the deposited soot at 575 K and higher. The active oxygen
storage material acts at the same time as a NOx-trap. When all the soot has been oxidised the
active oxygen storage material is fully converted into nitrates. CO and HC’s can decompose
these nitrates into nitrogen. These CO and HC’s are generated by running the engine rich or
by fuel addition into the exhaust stream. The introduced or generated CO and HC’s will be
converted into CO2 by the surface nitrates and the nitrates mainly to N2 and to some extent
also to NO. In other words this type of soot oxidation trap acts as a soot abatement
technology, but at the same time it acts as a NOx abatement technology. Figure 2 illustrates
the chemical processes of the Toyota system.

      On CeO2 catalyst, besides from the surface nitrate, active oxygen can be stored as a
result of NO2 interaction with metal centres of CeO2 that may contain oxygen vacancies
( Ce  ).

        NO2 ( gas)  Ce   Ce  O  NO( gas)
The quantity of indirectly measured stored oxygen was larger than surface nitrate (Chapter 6).
The potential of this phenomenon needs to be explored further. For example, since a large
quantity of stored oxygen on CeO2 is difficult to release, a study to find a way to release the
stored “active oxygen” for the purpose of soot oxidation acceleration is prerequisite.
                        NOx                                                Rich

               NO                                           HC            CO2 H2O
                              NO2 + O*                     CO             N2
              O2                          NOx storage
                                           material                           NO + O*
                        Pt                                          Pt

                      Substrate                                   Substrate

                    Storage of NOx                          Reduction of NOx

        NO + O2       O* PM       CO2                                             CO2
                                          NOx storage
              NO2                          material
  PM                                                                Pt
                     Substrate                                    Substrate
            Continuous oxidation of PM                   Continuous oxidation of PM
            by active oxygen and O2                      by active oxygen

Figure 2. NOx and PM reaction mechanism in DPNR Toyota system. After [7]

      Total reduction of diesel exhaust emissions (CO, HC's, PM, and NOx) is preferably to
be achieved in a single filter system as the DPNR system. But, the system may encounter
several problems such as engine ash deposit, the complexity of data-logging, and the
effectiveness of engine-out NOx concentration. It is reported that the fresh DPNR system
reduce 80% of NOx and PM emissions and might meet the US tier 2 bin 5 or 6 emissions
standards using low sulfur diesel fuel [9]. However, a fleet test has to demonstrate the
efficiency and the robustness of the system. A combination of PM reduction and a NOx
reduction system is still a rational option. Such an option is demonstrated in a compact
arrangement of the CRT system and a SCR system [10] and in a filter system comprising Pt
and Ce fuel additives, a Pt-impregnated wall-flow monolith soot filter and a vanadia-type
monolithic NH3-SCR catalyst [11,12]. So, a separate PM system, like TUD and CRT
systems, is still relevant for future research and application.
      For the passenger cars market, the PSA Peugeot Citroën system has been on the market
about 3 years. The system control particulate matter emission down to 0.004 g/km [13]. For
the new type diesel engine with less soot emission Rhodia has developed a new type of
additive based on cerium and iron [14]. The claimed advantages are that a lower additive
dosage rate of only 10 ppm can be used, the regeneration starts at a temperature of around
650 K, and the time needed for regeneration is much shorter. This new development might
keep the system as a leading particulate control method for diesel passenger car, since the
operation is independent on the amount of sulfur in the diesel fuel. One disadvantage is the
periodic cleaning of the filter in connection to the deposited cerium after approximately 300
000 km. Integration of NOx abatement system to the PSA Peugeot Citroën system will be a
challenge. Perhaps, modifying the trap to function not only as soot trap but also as a NO x-
storage system can be an option to consider.
       It is realised that after-treatment system is strongly dependent on the engine out
conditions. Therefore, a fast development in engine technology might shift the relevant
catalytic abatement technology to study in the future. Advance in injection system, for
example, very high injection pressure provided by the Siemens Automotive AG system called
Piezo Common Rail (PCR), was predicted to reduce emission to a level where no particulate
filter is needed [16]. The heart of this third PCR generation is the newly designed injector
with the significantly smaller piezo actuator integrated directly into the valve shaft. In the
common-rail injection system, all the cylinders of the diesel engine are supplied with fuel by
a high-pressure pump via a single fuel line, the "common rail". The fuel is currently injected
with pressures of up to 1600 bar, with the new PCR generation at pressures of even up to
1800 bar, so that a particularly fine fuel/air mixture is formed in the cylinder that burns
quickly, efficiently and cleanly. It is claimed that numerous injection models could comply
with the proposed emissions limits required by Euro V from 2008, probably without the need
for a particulate filter. If these claims are valid, new directions on diesel soot the after-
treatment technology have to be considered.


[1] A. König, G. Herding, B. Hupfeld, Th. Richter and K. Weidmann, Current task and
    challenges for exhaust after-treatment research. A viewpoint from automotive industry,
    Topics in Catal. 16/17, Nos 1-4 (2001) 23.
[2] G.A. Merkel, D.M. Beall, D.L. Hickman, and M.J. Vernacotola, Effect of microstructure
    and cell geometry on performance of cordierite diesel particulate filter, SAE paper 2001
    01-0193 (2001).
[3] K. Pattas, Z. Samaras, N. Kyriakis, T. Manikas, P. Pistikopoulos, T. Seguelong, and P.
    Rouveirolles, Comparative assessment of DPFs of different materials: A Case study on a
    EURO I light Duty Truck, SAE paper 2001-01-1287 (2001).
[4] T.V. Johnson, Diesel emission control in review-The last 12 months, SAE paper 2003-01-
    0039 (2003).
[5] S.J. Jelles, B.A.A.L. van Setten, M. Makkee and J.A. Moulijn, Molten salts as promising
    catalysts for oxidation of diesel soot: importance of experimental conditions in testing
    procedures, Appl. Catal. B 21 (1999) 35.
 [6] B. A. A. L. van Setten, C. G. M Spitters, J. Bremmer, A. M. M. Mulders, M. Makkee, and
     J. A. Moulijn, Stability of catalytic foam diesel-soot filters based on Cs2O, MoO3, and
     Cs2SO4.V2O5 molten-salt catalysts, Appl. Catal. B 42, 4 (2003) 337.
 [7] K. Nakatani, S. Hirota, S. Takeshima, K. Itoh, and T. Tanaka, Simultaneous PM and NOx
     reduction system for diesel engines, SAE paper 2002-01-0957 (2002).
 [8] K. Itoh, T. Tanaka, S. Hirota, T. Asanuma, K. Kimura, and K. Nakatani, Exhaust
     purifying method and apparatus of an internal combustion engine, US patent, US
     6,594,911, assigned to Toyota Motors Ltd (2003).
 [9] J. McDonald and B. Bunker, Testing of the Toyota Avensis DPNR at U.S. EPA-NVFEL,
     SAE paper 2002-01-2877 (2002)
[10] A.P. Walker, R. Allansson, P.G. Blakeman, M. Lavenius, S. Erkfeldt, H Landdallv, B.
     Ball, P. Harrord, D. Manning, and L. Bernegger, The development and performance of
     the compact SCR-Trap system: A 4-way diesel emission control system, SAE paper 2003-
     01-0078 (2003)
[11] M. Makkee, H.C. Krijnsen, S.S. Bertin, H.P.A. Calis, C.M. van den Bleek, and
    J.A. Moulijn, Bench-scale demonstration of an integrated deSoot–deNOx system, Catal.
    Today 75 (2002) 459
[12] H.C. Krijnsen, S.S. Bertin, M. Makkee, C.M. van den Bleek, J.A. Moulijn, and
    H.P.A. Calis, Bench-scale demonstration of an integrated deSoot–deNOx system, SAE
    paper 2001-01-0515 (2001)
[13] accessed July 2003
[14] accessed September 2003
[15] accessed September 2003

To top