- Email: [email protected]
Effects of hydrothermal aging on SiC-DPF with metal oxide ash and alkali metals
Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
Contents lists available at ScienceDirect
Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec
Effects of hydrothermal aging on SiC-DPF with metal oxide ash and alkali metals Byungchul Choi a,b,*, Bao Liu b, Jong-Woo Jeong a a b
School of Mechanical Systems Engineering, Chonnam National University, Gwangju 500-757, Republic of Korea Graduate School of Mechanical Engineering, Chonnam National University, Gwangju 500-757, Republic of Korea
A R T I C L E I N F O
A B S T R A C T
Article history: Received 31 December 2008 Accepted 7 February 2009
The silicon carbide used to make diesel particulate filters (SiC-DPF) has a maximum temperature of use, which is not the melting point of the filter material itself but rather the eutectic melting points of the ash materials and alkali metals deposited on the filter wall. Chemical reactions between the SiC filter and the other materials, i.e. ash materials and/or alkali metals, decrease the filtration efficiency and catalytic reactivity of engine out emission. The objective of this study is to understand the effect of hydrothermal aging on the SiC-DPF, and on the SiC-CDPF (catalyzed diesel particulate filter) deposited with ash materials and/or alkali metals. Hydrothermal aging simulated for the extreme condition of uncontrolled regeneration in DPF is carried out by using H2O at high temperature. The surface change of the SiC filter was characterized in terms of the geometric microstructure and metal composites of the filter by using the SEM-EDS, BET and XRD. The accumulated ash materials and alkali metals in the SiC-DPF were an admixture, and the SiC-DPF after-treatment system always contained H2O. According to the results, H2O in the after-treatment system can be regarded as an influential factor of SiC-DPF durability even though the SiC itself has a very high melting point. The regeneration temperature has to be controlled under a critical value to ensure the durability of SiC-DPF in the after-treatment system, considering the fact that large quantities of ash materials, alkali metals and H2O components are included in the exhaust gas. ß 2009 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Keywords: Diesel engine Silicon carbide (SiC) Diesel particulate filter (DPF) Hydrothermal aging
1. Introduction Diesel particulate matter (PM) discharged from diesel vehicles is harmful to both human health and the environment and has been an obstacle to the development of diesel vehicles. An effective technology against PM is the diesel particulate filter (DPF) in the exhaust after-treatment system [1,5]. The materials of the DPF have to have considerable endurance because the filter has to trap the PM from the diesel exhausts during severe operating conditions. At present, silicon carbide (SiC) is one of the most promising candidates as the base material of DPF or catalyzed diesel particulate filter (CDPF) because it has good mechanical performance, excellent heat-resistant properties, and wide applicability in diesel emission control. SiC also has high thermal conductivity, which allows the higher regeneration of the filter by both thermal oxidation and catalytic PM oxidation at the SiC-DPF or CDPF system. With these advantages, SiC-DPF or CDPF has been developed and used on diesel vehicles [2,4,6]. The PM accumulated in the filter system needs to be oxidized or removed physically from the system. In the oxidation method, inorganic oxide materials are added into the diesel fuel as additives
* Corresponding author. Tel.: +82 62 530 1681; fax: +82 62 530 1689. E-mail address: [email protected] (B.C. Choi).
admixture to remove the accumulated PM. On the other hand, the physical removal method is hardly used because of its ineffectiveness [3,7]. Ash materials of sulfates, phosphates, or other oxides of calcium and zinc are formed from the burned additives of the engine lubricant in the engine’s combustion chamber. Some oxides of metals such as iron, manganese, and cerium, which are presented as additives of catalyst to CDPF to reduce the activation temperature of the catalytic reaction, are required to initiate the catalytic reaction of the soot in the CDPF. In addition, iron-containing oxides may result from the diesel engine friction, and from the erosion of the exhaust system [3,8]. However, the oxidation method may still have some disadvantages because it uses additives as admixtures (Fe, Ce, etc.) in the diesel fuel to decrease the temperature of PM oxidation. For example, the filter will become clogged with ash materials that result from the burning of inorganic materials. Furthermore, a second contamination or slip by the admixture of additives may occur. The accumulated carbon ingredients of the PM are removed during DPF’s regeneration process, but the inorganic materials are not removed and can accumulate continuously [3,4,9]. As a result, the pressure drop can increase, gradually, and in turn the temperature of the DPF can increase greatly, and this would ultimately result in the reaction of materials with DPF. Finally, DPF can experience partial thermal aging or can malfunction [4,11,12]. Alkali metals such as sodium and cesium are used as a type of
1226-086X/$ – see front matter ß 2009 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jiec.2009.09.050
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
708
Table 1 Prepared sample matrix for inter-reaction between ash/alkali materials and SiC powder. Material
Components (wt.%)
Ash material Ash material mixture Alkali metal Alkali metal mixture Ash and Alkali metal mixture
CeO2
Fe2O3
SO3
CaO
P2O5
MgO
Na
Cs
100 58 – – 29
100 12 – – 6
100 12 – – 6
100 8 – – 4
100 8 – – 4
100 2 – – 1
– – 100 50 25
– – 100 50 25
Specification of SiC substrate Melting point (8C)
Porosity (%)
Mean pore diameter (fan)
Purity (%)
Thermal conductivity (W/mK)
2400
55
12
99
60
catalytic washcoat for CDPF [5,13–15,17]. Therefore, when these ash materials or alkali metals join with the DPF, the SiC material of the DPF interacts with the ash materials or/and alkali metals. The effect of this interaction on the temperature at which sintering and adherence to the filter wall occur is studied because this effect can decrease the filtration efficiency and catalytic reactivity of the DPF and CDPF, respectively [1,3,10,16]. The present study aims to understand the behavior and the effect of ash material and alkali metal depositions on the damage of SiC-CDPF monolith during hydrothermal aging upon simulated uncontrolled regeneration. Moreover, hydrothermal aging behavior of catalyzed SiC-DPF with Pt/g-Al2O3 was observed. 2. Experimental 2.1. Sample preparation The SiC substrate used in this experiment is a commercial product from IBIDEN Co. Engine lubrication oil (Hyundai 5W30) includes compositions of Zn (950 ppm), P (768 ppm), Ca (1850 ppm), S (1700 ppm) and Mg (17 ppm). Al (600 ppm), Ca (400 ppm), Fe (600 ppm), Mg and Mn (50 ppm), Na (150 ppm), Zn (300 ppm) were detected by measurements of wear metals in used engine oil [18]. SiC-DPF system uses fuel-borne catalysts as CeO2, Fe2O3 [4]. The composition of an actual ash of DPF
consisted of CaO (29.6 wt.%), ZnO (9.9 wt%), MgO (5.5 wt.%), P2O5 (15.8 wt%) and Fe2O3 (0.41%) [3]. The mixing portion of ash materials and alkali metals considered this composition. The properties of a SiC substrate are shown in Table 1 [3,10]. An SiC sample extracted from SiC-DPF was fine-crushed into powder, and the catalyst (3 wt.% Pt/g-Al2O3) was added to the SiC-DPF powder to produce a catalyzed DPF. A CDPF was obtained by gently shaking the both the catalyst and the DPF powder in a vessel and by sonification using an ultrasonic cleaner. Subsequently, to initiate the inter-reaction between the SiC substrate and ash materials and/or alkali metals, which was accumulated and deposited in performing the roles of DPF, the samples were pressed into a pellet of 2 mm in thickness and 13 mm in diameter at a special weight ratio (1:10) of ash material and/or alkali metal to SiC powder. The simplified structure of the experiment sample is showed in Fig. 1(d). Fig. 1 shows the powder sample, the equipment used for pressing the sample (pressure is 400 kg/cm2), the pellet sample, and the experiment sample. CeO2, Fe2O3, CaSO4, CaO, P2O5 and MgO powders were mixed with SiC powder at the weight ratio of 58:12:12:8:8:2. Also, the alkali metals, Cs and Na, were mixed at the weight rate of 1:1. The inter-reaction between SiC substrate and ash material mixture (or alkali metal mixture) was observed, and each mixture was prepared at 1:1 weight ratio. The sample components used in the experiment are shown in Table 1.
Fig. 1. Images of (a) mixture of Fe2O3/SiC powder, (b) equipment used for pressing sample, (c) the pellet shape sample and (d) experiment sample.
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
709
Fig. 2. Images of Fe2O3/CaO sample at 1100 V: (a) fresh sample (b) aged under air condition (c) aged under H2O/air. The results are magnified 100 times by microscope.
2.2. Hydrothermal aging The experiment was carried out at a high temperature in atmosphere of air (500 cm3/min) with (thermal aging) or without 10% H2O content (hydrothermal aging) for 10 h. The samples were heated up to 800–1100 8C at the heating rate of 10 8C/min by using an electric furnace. The experimental temperature was set higher than that of real applications to confirm the stability of the DPF against high temperature during regeneration. The results of the chemical reaction in during the hydrothermal aging process were classified as ‘‘staining–pitting–surface glazin– pinhole–bulk melting’’ to represent the inter-reaction [1,9]. The
concept of adherence declares the degree of binding force derived from the sample and SiC-base, which was cut from SiC-DPF. The aged samples were investigated microscopically by using SEM-EDS, and the chemical composition was inspected by physicochemical (BET isotherm) analysis and phase analysis (XRD). 3. Results and discussion 3.1. Inter-reaction of ash materials with SiC and Pt catalyzed SiC Fig. 2 shows the samples of Fe2O3/SiC and CaO/SiC aged at 1100 8C in atmosphere of air with or without H2O. The samples
Fig. 3. Images of (a) P2O5/SiC aged sample under air condition, (b) P2O5/SiC aged sample under air with H2O, (c) P2O5/Pt–SiC aged sample under air with H2O condition and (d) P2O5 fresh sample. The results are magnified 100 times by microscope.
710
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
condition, and extensive pitting surface of the P2O5/SiC sample was observed under hydrothermal aging condition at 1100 8C in Fig. 4(c). However, the CaSO4/SiC sample showed no change at low temperatures until the temperature reached 1100 8C, while different phenomena were observed under the hydrothermal aging condition because of the H2O. This result suggests that very little partial reaction occurred because of the H2O; this partial reaction in air at 1100 8C with or without H2O was proven by SEM micrographs (see Fig. 5). As shown in Fig. 5(a), the particle grains had a relatively uniform distribution on the sample surface, but a lot of non-uniform grains developed on the surface, as shown in Fig. 5(b). Fig. 5(a) shows the open pores, and Fig. 5(b) shows the blocked pores covered with melted particles at 1100 8C. Therefore, more extensive damages were achieved with air and H2O than with only air at high temperature. In the case of the mixture of ash materials composite and SiC powders, the chemical reaction was weak not only under the air condition but also under the air with H2O condition up to 900 8C, but more partial damages occurred under the hydrothermal condition than under only the air condition at 1100 8C (see Fig. 7). As shown in Fig. 6(c), few bigger melted holes on the sample surface were observed at 900 8C, and these holes were due to the presence of a precious metal catalyst on the sample surface under the air with H2O condition. Fig. 7 shows the SEM images of the sample surfaces of mixed ash materials with SiC at 1100 8C. As shown by Fig. 7(b), much more partial grains were developed
Fig. 4. SEM micrograph of (a) P2O5/SiC sample and (b) P2O5/Pt–SiC sample aged at 900 V under H2O/air condition, (c) P2O5/SiC sample aged at 1100 V under H2O/air condition. The results are magnified 200 times by microscope.
have shown the same characteristic of slight staining on their surfaces as a result of aging. The P2O5/SiC sample showed slight staining at 900 8C in each condition, and the surface expansion observed at 1100 8C under the hydrothermal aging condition was more extensive than that under the thermal aging condition, in which air was supplied without H2O (see Fig. 3(a) and (b)). Moreover, the adherence of the P2O5/SiC sample started at 800 8C under the hydrothermal aging condition but started at 1100 8C under the thermal aging condition, whereas the other ash materials (CeO2 and MgO) showed almost no change at any high temperature under both hydrothermal and thermal aging conditions. From Fig. 3(c), partial slight pitting was observed on the surface of the P2O5/Pt-SiC sample (Pt catalyzed SiC with P2O5) at 900 8C under the hydrothermal aging condition, in which staining and moderate pitting between the P2O5/SiC and the P2O5/Pt–SiC sample was shown in Fig. 4(a) and (b) under the same
Fig. 5. SEM micrograph of CaSO4/SiC sample produced at 1100 V under (a) air condition, (b) H2O/air condition. The results are magnified 200 times by SEM.
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
711
Fig. 6. Images of (a) AM/SiC fresh sample (b) AM/SiC aged sample under H2O/air condition, (c) AM/Pt–SiC aged sample under H2O/air condition. The results are magnified 100 times by microscope.
around the bigger particles on the sample surface than on the sample surface shown in Fig. 7(a). This result suggests that H2O may have been an influential factor of SiC-DPF durability in atmosphere under the hydrothermal condition.
Fig. 7. SEM micrograph of AM/SiC sample produced at 1100 V under (a) air condition (b) H2O/air condition. The results are magnified 1000 times by SEM.
Finally, in the case of a single material in ash, no significant damage like the pinhole was obtained, except for CaSO4 and P2O5, which were pitted at 1100 8C under the air with or without H2O condition. However, the sample surface under the hydrothermal condition had more severe damages than under the thermal aging condition with only air because of the thermal reaction of the ash mixture at 1100 8C. This result indicates that the single material in ash did not affect the SiC significantly; then, the damage by the mixed ash materials at 1100 8C was caused by the H2O. As shown above, the P2O5/Pt–SiC sample showed slightly partial pitting at 900 8C under the air with H2O condition. In fact, the accumulated ash in the SiC-DPF was a composite of the ash materials that came from engine oil components, engine cylinder wear and erosion of the exhaust system, and the SiC-DPF after-treatment system was always exposed to H2O. Therefore, the H2O in the after-treatment system can be regard as an influential factor of SiC-DPF durability even though the SiC itself has a very high melting point. As a result, considering the fact that large quantities of ash materials and H2O components are included in
Fig. 8. XRD pattern of the P2O5/Pt–SiC sample by hydrothermal aging.
712
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
Fig. 9. Images of (a) Na/SiC fresh sample, (b) Na/SiC aged under air condition, (c) Na/SiC aged under H2O/air condition and (d) Na/Pt–SiC aged at 900 8C under H2O/air condition. The results are magnified 7 and 100 times by microscope.
the exhaust gas, the regeneration temperature has to be controlled under a critical value to ensure the durability of SiC-DPF in the after-treatment system. Fig. 8 represents the XRD patterns of only the SiC substrate and P2O5/SiC sample. These patterns are formed by hydrothermal aging
at 1100 8C with or without Pt/g-Al2O3. As the ash material P2O5 and the catalyst Pt deposited on the SiC substrate, the SiC peaks decreased rapidly, whereas, the amount of SiO2 increased for the P2O5/SiC sample. For the P2O5/Pt–SiC sample, SiC and SiO2 peaks became weaker than those of the SiC-only and P2O5/SiC samples,
Fig. 10. Images of (a) Cs/SiC fresh sample, (b) Cs/SiC aged under air condition, (c) Cs/SiC aged under H2O/air condition and (d) Cs/Pt–SiC aged at 900 V under H2O/air condition. The results are magnified 7 and 100 times by microscope.
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
713
Fig. 11. Images of ALM/SiC aged sample at 900 V under (a) air condition and (b) H2O/air condition, (c) ALM/Pt–SiC sample aged under H2O/air condition. The results are magnified 7 and 100 times by microscope.
and the Pt composite peaks emerged from the coexistence of SiC and Al2O3 at high temperature. Therefore, these results suggested that the Pt catalyst in the P2O5/SiC promoted the inter-reaction of P2O5/SiC with other substrates in hydrothermal aging.
3.2. Inter-reaction of alkali metal with SiC and Pt catalyzed SiC In the case of an alkali metal with SiC powder, the temperature of inter-reaction in the aging conditions was lower than that in the
Fig. 12. Images of (a) AALM/SiC fresh sample, (b) AALM/SiC sample aged under air condition, (c) AALM/SiC sample aged under H2O/air at 1100 V and (d) AALM/Pt–SiC sample aged under H2O/air condition. The results are magnified 7 and 100 times by microscope.
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
714
case of the ash material, generally. The inter-reaction of Na with SiC was stronger than anything else in our experimental conditions, and same results of glazing, melting and pinholes were shown at different temperatures in hydrothermal or thermal aging conditions (see Fig. 9). At 900 8C under hydrothermal aging, the reaction of the Na/Pt–SiC sample gave a different result from slight pitting (see Fig. 9(d)) and extensive pitting (see Fig. 9(c)). On the other hand, the Cs/SiC sample showed a higher reaction temperature than the Na/SiC sample, and the result of the reaction showed no difference between the aging conditions. However, different results between the Cs/SiC sample and Cs/Pt–SiC sample under the hydrothermal aging condition were observed at 900 8C by moderate pitting and staining, respectively (see Fig. 10). In Fig. 9(a)–(c) shows that Na/SiC sample damaged more extensively as the temperature increased under different conditions, showing pitting at 900 8C and glazing at 1100 8C. In the comparison of (c) with (d) at 900 8C for the hydrothermal aging, the surface of the Na/SiC sample expands much more than the Na/Pt– SiC sample, and glazing also occurs. As shown in Fig. 10, the inter-reaction in the Cs/SiC sample was more severe as the temperature increased under the hydrothermal or thermal aging condition. It shows the same reaction: slight pitting at 900 8C and moderate glazing at 1100 8C. Thus, there was slight adherence at 900 8C and extensive adherence at 1100 8C in the presence of H2O, whereas there was no adherence under the thermal aging condition. However, in the comparison of (c) with (d) at 900 8C for the hydrothermal aging condition, the surface of the Cs/Pt–SiC sample shows only staining whereas the Cs/SiC sample shows pitting. The reaction of the ALM (the composite of alkali metals)/SiC sample shows pitting from 800 8C. The sample was damaged seriously as the temperature increased to 1100 8C under different conditions because of the extensive interaction of the ALM/SiC sample, which prevented the observation of any differences between aging conditions. In the comparison of the ALM/SiC sample with the ALM/Pt–SiC sample at 900 8C under hydrothermal aging, the surface of ALM/SiC sample expanded more extensively than the ALM/Pt–SiC sample, and in this case, glazing and pinholes occurred on the ALM/SiC sample, but only slight pitting occurred on the ALM/Pt–SiC sample (see Fig. 11). The interaction temperature of the ALM/SiC sample was much lower than that of the ash materials, and the adherence became more severe with the increase of temperature. Fig. 12 shows the results of the interaction of AALM (ash materials and alkali metals mixture)/SiC in the thermal and hydrothermal aging conditions. The pitting phenomenon happened at 800 8C. With the increase of temperature, partly melting and pinholes were presented at 900 8C, and finally, the sample was
Fig. 13. SEM micrograph of AALM/SiC sample produced at 1100 V under (a) air condition (b) H2O/air condition. The surfaces are magnified 200 times by SEM.
completely damaged at 1100 8C in the air with or without H2O condition. More melting and pinholes were observed in the SEM micrograph (see Fig. 12(b)) than in Fig. 13(a). Therefore, more damages occurred in the hydrothermal aging than in the thermal aging. As shown in Fig. 12(c) and (d) at 1100 8C, the surface of the AALM/Pt–SiC sample did not expand under the hydrothermal aging, and slight melting and pinholes occurred. Table 2 shows the results of the inter-reaction between ash materials/alkali metals and the SiC sample according to temperature. Fig. 14 represents the XRD pattern of the Na/SiC samples with or without Pt catalyst by hydrothermal aging. The result shows composite chemical elements, such as sodium aluminium silicates. These composite chemicals were produced by the chemical interreactions among Na, Al2O3, SiC and Pt. Therefore, the catalyzed SiC
Table 2 Inter-reaction of sample on air (500 cm3/min) with or without 10% H2O. Subject
Ash material
Alkali metal
CeO2
Fe2O3
CaSO4
CaO
P2O5
MgO
AM
Na
Cs
ALM
AALM
Air 800 8C 900 8C 1100 8C
^
^*
^~ ~~ &~
^ ~
^~ ~~ &~
^ ~* * &~
Air + H2O 800 8C 900 8C 1000 8C
^
* & ^~
^*
^~ ~~ &~
^* ~~
^~ ~~ &~
^ ~* *&~
Pt, Air + H2O 900 8C 1000 8C
^
*
^
^ ~
^* ^&
^~
AM: ash materials mixture with SiC; ALM: alkali metals mixture with SiC; AALM: ash materials and alkali metals mixture with SiC. () No reaction, ( (~) Glazing, (*) Pinhole, (&) Bulk melting (*) Slight adherence, (&) Moderate adherence, (~) Extensive adherence.
) Staining, (^) Pitting,
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
715
of composites. Moreover, the mixture of ash materials and alkali metals with Pt–SiC restrained surface expansion. In fact, the accumulated ash materials and alkali metals in the SiC-DPF were an admixture, and the SiC-DPF after-treatment system always contained H2O. According to the results, H2O in the after-treatment system can be regard as an influential factor of SiCDPF durability even though the SiC itself has a very high melting point. As a result, the regeneration temperature has to be controlled under a critical value to ensure the durability of SiCDPF in the after-treatment system, considering the fact that large quantities of ash materials, alkali metals and H2O components are included in the exhaust gas. Acknowledgement This works was supported by the 2nd stage BK (Brain Korea) 21 Project in 2008. Fig. 14. XRD pattern of the Na/Pt–SiC sample by hydrothermal aging.
accumulated with alkali metals, such as Na, etc., was easily exposed to damage by hydrothermal aging. 4. Conclusions The results of the hydrothermal aging (under air with H2O condition) for ash material samples showed that the inevitable H2O in the after-treatment system of a diesel engine can be an influential factor of SiC-DPF durability even though the SiC itself has a very high melting point. The results were observed by microscopy and SEM-EDS. However, the hydrothermal aging durability of the Pt–SiC-CDPF with ash materials, such as the P2O5 sample and the ash material mixture sample, was changed by high temperatures. Alkali metal samples (alkali metal/SiC) did not provide sufficient evidence to prove that H2O is a possible influential factor of SiC-DPF durability in the after-treatment system. On the other hand, the results of XRD analysis showed a high degree of inter-reaction between Pt–SiC with alkali metals for the formation
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]
O. Tomokazu, SAE paper No. 2005-01-0579, 2005. J.W. Jeong, B.C. Choi, M.T. Lim, J. Ind. Eng. Chem. 14 (2008) 830. G.A. Merkel, W.A. Cutler, C.J. Warren, SAE paper No. 2001-01-0190, 2001. M. Harada, S. Ichikawa, A. Otsuka, M. Hattori, S. Hashimoto, SAE paper No. 200501-0582, 2005. M. Matti Maricq, Aerosol. Sci. 38 (2007) 1079. I. Atribak, A. Bueno-Lopez, A. Garcia-Garcia, Catal. Commun. 9 (2008) 250. T.V. Johnson, SAE paper No. 2006-01-0030, 2006. J.C Tan, C.N. Opris, K.J. Baumgard, J.H. Johnson, SAE paper No. 960136, 1996. H. Okazoe, K. Shimizu, Y. Watanabe, E. Santiago, P. Kugland, W. Ruth, SAE paper 960130, 1996. B.C. Choi, H.Y. Kang, G.S. Son, C.W. Hwang, SAE paper No. 2007-01-1939, 2007. Y. Uchida, S. Ichikawa, T. Harada, T. Hamanaka, SAE paper No. 2003-01-0384, 2003. S. Ichikawa, Y. Uchida, A. Kaneda, T. Hamanaka, SAE paper No. 2004-01-0951, 2004. K. Yuuki, T. Ito, H. Sakamoto, R. Matsubara, S. Hashimoto, T. Hamanaka, SAE paper No. 2003-01-0383, 2003. C.R. Scherik, J. McDonald, B. Olson, SAE paper No. 2001-01-1351, 2001. M.J. Murtagh, D.L. Sherwood, L.S. Socha, SAE paper No. 940235, 1994. A. Gubernat, J. Eur. Ceram. Soc. 27 (2007) 781. T. Mizutani, Y. Watanabe, K. Yuuki, S. Hashimoto, T. Hamanaka, SAE paper No. 2004-01-0159, 2004. P.L. Yaroshchyk, R.J.S. Morrison, D. Body, B.L. Chadwick, Spectrochim. Acta B 60 (2005) 986.
Contents lists available at ScienceDirect
Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec
Effects of hydrothermal aging on SiC-DPF with metal oxide ash and alkali metals Byungchul Choi a,b,*, Bao Liu b, Jong-Woo Jeong a a b
School of Mechanical Systems Engineering, Chonnam National University, Gwangju 500-757, Republic of Korea Graduate School of Mechanical Engineering, Chonnam National University, Gwangju 500-757, Republic of Korea
A R T I C L E I N F O
A B S T R A C T
Article history: Received 31 December 2008 Accepted 7 February 2009
The silicon carbide used to make diesel particulate filters (SiC-DPF) has a maximum temperature of use, which is not the melting point of the filter material itself but rather the eutectic melting points of the ash materials and alkali metals deposited on the filter wall. Chemical reactions between the SiC filter and the other materials, i.e. ash materials and/or alkali metals, decrease the filtration efficiency and catalytic reactivity of engine out emission. The objective of this study is to understand the effect of hydrothermal aging on the SiC-DPF, and on the SiC-CDPF (catalyzed diesel particulate filter) deposited with ash materials and/or alkali metals. Hydrothermal aging simulated for the extreme condition of uncontrolled regeneration in DPF is carried out by using H2O at high temperature. The surface change of the SiC filter was characterized in terms of the geometric microstructure and metal composites of the filter by using the SEM-EDS, BET and XRD. The accumulated ash materials and alkali metals in the SiC-DPF were an admixture, and the SiC-DPF after-treatment system always contained H2O. According to the results, H2O in the after-treatment system can be regarded as an influential factor of SiC-DPF durability even though the SiC itself has a very high melting point. The regeneration temperature has to be controlled under a critical value to ensure the durability of SiC-DPF in the after-treatment system, considering the fact that large quantities of ash materials, alkali metals and H2O components are included in the exhaust gas. ß 2009 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Keywords: Diesel engine Silicon carbide (SiC) Diesel particulate filter (DPF) Hydrothermal aging
1. Introduction Diesel particulate matter (PM) discharged from diesel vehicles is harmful to both human health and the environment and has been an obstacle to the development of diesel vehicles. An effective technology against PM is the diesel particulate filter (DPF) in the exhaust after-treatment system [1,5]. The materials of the DPF have to have considerable endurance because the filter has to trap the PM from the diesel exhausts during severe operating conditions. At present, silicon carbide (SiC) is one of the most promising candidates as the base material of DPF or catalyzed diesel particulate filter (CDPF) because it has good mechanical performance, excellent heat-resistant properties, and wide applicability in diesel emission control. SiC also has high thermal conductivity, which allows the higher regeneration of the filter by both thermal oxidation and catalytic PM oxidation at the SiC-DPF or CDPF system. With these advantages, SiC-DPF or CDPF has been developed and used on diesel vehicles [2,4,6]. The PM accumulated in the filter system needs to be oxidized or removed physically from the system. In the oxidation method, inorganic oxide materials are added into the diesel fuel as additives
* Corresponding author. Tel.: +82 62 530 1681; fax: +82 62 530 1689. E-mail address: [email protected] (B.C. Choi).
admixture to remove the accumulated PM. On the other hand, the physical removal method is hardly used because of its ineffectiveness [3,7]. Ash materials of sulfates, phosphates, or other oxides of calcium and zinc are formed from the burned additives of the engine lubricant in the engine’s combustion chamber. Some oxides of metals such as iron, manganese, and cerium, which are presented as additives of catalyst to CDPF to reduce the activation temperature of the catalytic reaction, are required to initiate the catalytic reaction of the soot in the CDPF. In addition, iron-containing oxides may result from the diesel engine friction, and from the erosion of the exhaust system [3,8]. However, the oxidation method may still have some disadvantages because it uses additives as admixtures (Fe, Ce, etc.) in the diesel fuel to decrease the temperature of PM oxidation. For example, the filter will become clogged with ash materials that result from the burning of inorganic materials. Furthermore, a second contamination or slip by the admixture of additives may occur. The accumulated carbon ingredients of the PM are removed during DPF’s regeneration process, but the inorganic materials are not removed and can accumulate continuously [3,4,9]. As a result, the pressure drop can increase, gradually, and in turn the temperature of the DPF can increase greatly, and this would ultimately result in the reaction of materials with DPF. Finally, DPF can experience partial thermal aging or can malfunction [4,11,12]. Alkali metals such as sodium and cesium are used as a type of
1226-086X/$ – see front matter ß 2009 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jiec.2009.09.050
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
708
Table 1 Prepared sample matrix for inter-reaction between ash/alkali materials and SiC powder. Material
Components (wt.%)
Ash material Ash material mixture Alkali metal Alkali metal mixture Ash and Alkali metal mixture
CeO2
Fe2O3
SO3
CaO
P2O5
MgO
Na
Cs
100 58 – – 29
100 12 – – 6
100 12 – – 6
100 8 – – 4
100 8 – – 4
100 2 – – 1
– – 100 50 25
– – 100 50 25
Specification of SiC substrate Melting point (8C)
Porosity (%)
Mean pore diameter (fan)
Purity (%)
Thermal conductivity (W/mK)
2400
55
12
99
60
catalytic washcoat for CDPF [5,13–15,17]. Therefore, when these ash materials or alkali metals join with the DPF, the SiC material of the DPF interacts with the ash materials or/and alkali metals. The effect of this interaction on the temperature at which sintering and adherence to the filter wall occur is studied because this effect can decrease the filtration efficiency and catalytic reactivity of the DPF and CDPF, respectively [1,3,10,16]. The present study aims to understand the behavior and the effect of ash material and alkali metal depositions on the damage of SiC-CDPF monolith during hydrothermal aging upon simulated uncontrolled regeneration. Moreover, hydrothermal aging behavior of catalyzed SiC-DPF with Pt/g-Al2O3 was observed. 2. Experimental 2.1. Sample preparation The SiC substrate used in this experiment is a commercial product from IBIDEN Co. Engine lubrication oil (Hyundai 5W30) includes compositions of Zn (950 ppm), P (768 ppm), Ca (1850 ppm), S (1700 ppm) and Mg (17 ppm). Al (600 ppm), Ca (400 ppm), Fe (600 ppm), Mg and Mn (50 ppm), Na (150 ppm), Zn (300 ppm) were detected by measurements of wear metals in used engine oil [18]. SiC-DPF system uses fuel-borne catalysts as CeO2, Fe2O3 [4]. The composition of an actual ash of DPF
consisted of CaO (29.6 wt.%), ZnO (9.9 wt%), MgO (5.5 wt.%), P2O5 (15.8 wt%) and Fe2O3 (0.41%) [3]. The mixing portion of ash materials and alkali metals considered this composition. The properties of a SiC substrate are shown in Table 1 [3,10]. An SiC sample extracted from SiC-DPF was fine-crushed into powder, and the catalyst (3 wt.% Pt/g-Al2O3) was added to the SiC-DPF powder to produce a catalyzed DPF. A CDPF was obtained by gently shaking the both the catalyst and the DPF powder in a vessel and by sonification using an ultrasonic cleaner. Subsequently, to initiate the inter-reaction between the SiC substrate and ash materials and/or alkali metals, which was accumulated and deposited in performing the roles of DPF, the samples were pressed into a pellet of 2 mm in thickness and 13 mm in diameter at a special weight ratio (1:10) of ash material and/or alkali metal to SiC powder. The simplified structure of the experiment sample is showed in Fig. 1(d). Fig. 1 shows the powder sample, the equipment used for pressing the sample (pressure is 400 kg/cm2), the pellet sample, and the experiment sample. CeO2, Fe2O3, CaSO4, CaO, P2O5 and MgO powders were mixed with SiC powder at the weight ratio of 58:12:12:8:8:2. Also, the alkali metals, Cs and Na, were mixed at the weight rate of 1:1. The inter-reaction between SiC substrate and ash material mixture (or alkali metal mixture) was observed, and each mixture was prepared at 1:1 weight ratio. The sample components used in the experiment are shown in Table 1.
Fig. 1. Images of (a) mixture of Fe2O3/SiC powder, (b) equipment used for pressing sample, (c) the pellet shape sample and (d) experiment sample.
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
709
Fig. 2. Images of Fe2O3/CaO sample at 1100 V: (a) fresh sample (b) aged under air condition (c) aged under H2O/air. The results are magnified 100 times by microscope.
2.2. Hydrothermal aging The experiment was carried out at a high temperature in atmosphere of air (500 cm3/min) with (thermal aging) or without 10% H2O content (hydrothermal aging) for 10 h. The samples were heated up to 800–1100 8C at the heating rate of 10 8C/min by using an electric furnace. The experimental temperature was set higher than that of real applications to confirm the stability of the DPF against high temperature during regeneration. The results of the chemical reaction in during the hydrothermal aging process were classified as ‘‘staining–pitting–surface glazin– pinhole–bulk melting’’ to represent the inter-reaction [1,9]. The
concept of adherence declares the degree of binding force derived from the sample and SiC-base, which was cut from SiC-DPF. The aged samples were investigated microscopically by using SEM-EDS, and the chemical composition was inspected by physicochemical (BET isotherm) analysis and phase analysis (XRD). 3. Results and discussion 3.1. Inter-reaction of ash materials with SiC and Pt catalyzed SiC Fig. 2 shows the samples of Fe2O3/SiC and CaO/SiC aged at 1100 8C in atmosphere of air with or without H2O. The samples
Fig. 3. Images of (a) P2O5/SiC aged sample under air condition, (b) P2O5/SiC aged sample under air with H2O, (c) P2O5/Pt–SiC aged sample under air with H2O condition and (d) P2O5 fresh sample. The results are magnified 100 times by microscope.
710
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
condition, and extensive pitting surface of the P2O5/SiC sample was observed under hydrothermal aging condition at 1100 8C in Fig. 4(c). However, the CaSO4/SiC sample showed no change at low temperatures until the temperature reached 1100 8C, while different phenomena were observed under the hydrothermal aging condition because of the H2O. This result suggests that very little partial reaction occurred because of the H2O; this partial reaction in air at 1100 8C with or without H2O was proven by SEM micrographs (see Fig. 5). As shown in Fig. 5(a), the particle grains had a relatively uniform distribution on the sample surface, but a lot of non-uniform grains developed on the surface, as shown in Fig. 5(b). Fig. 5(a) shows the open pores, and Fig. 5(b) shows the blocked pores covered with melted particles at 1100 8C. Therefore, more extensive damages were achieved with air and H2O than with only air at high temperature. In the case of the mixture of ash materials composite and SiC powders, the chemical reaction was weak not only under the air condition but also under the air with H2O condition up to 900 8C, but more partial damages occurred under the hydrothermal condition than under only the air condition at 1100 8C (see Fig. 7). As shown in Fig. 6(c), few bigger melted holes on the sample surface were observed at 900 8C, and these holes were due to the presence of a precious metal catalyst on the sample surface under the air with H2O condition. Fig. 7 shows the SEM images of the sample surfaces of mixed ash materials with SiC at 1100 8C. As shown by Fig. 7(b), much more partial grains were developed
Fig. 4. SEM micrograph of (a) P2O5/SiC sample and (b) P2O5/Pt–SiC sample aged at 900 V under H2O/air condition, (c) P2O5/SiC sample aged at 1100 V under H2O/air condition. The results are magnified 200 times by microscope.
have shown the same characteristic of slight staining on their surfaces as a result of aging. The P2O5/SiC sample showed slight staining at 900 8C in each condition, and the surface expansion observed at 1100 8C under the hydrothermal aging condition was more extensive than that under the thermal aging condition, in which air was supplied without H2O (see Fig. 3(a) and (b)). Moreover, the adherence of the P2O5/SiC sample started at 800 8C under the hydrothermal aging condition but started at 1100 8C under the thermal aging condition, whereas the other ash materials (CeO2 and MgO) showed almost no change at any high temperature under both hydrothermal and thermal aging conditions. From Fig. 3(c), partial slight pitting was observed on the surface of the P2O5/Pt-SiC sample (Pt catalyzed SiC with P2O5) at 900 8C under the hydrothermal aging condition, in which staining and moderate pitting between the P2O5/SiC and the P2O5/Pt–SiC sample was shown in Fig. 4(a) and (b) under the same
Fig. 5. SEM micrograph of CaSO4/SiC sample produced at 1100 V under (a) air condition, (b) H2O/air condition. The results are magnified 200 times by SEM.
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
711
Fig. 6. Images of (a) AM/SiC fresh sample (b) AM/SiC aged sample under H2O/air condition, (c) AM/Pt–SiC aged sample under H2O/air condition. The results are magnified 100 times by microscope.
around the bigger particles on the sample surface than on the sample surface shown in Fig. 7(a). This result suggests that H2O may have been an influential factor of SiC-DPF durability in atmosphere under the hydrothermal condition.
Fig. 7. SEM micrograph of AM/SiC sample produced at 1100 V under (a) air condition (b) H2O/air condition. The results are magnified 1000 times by SEM.
Finally, in the case of a single material in ash, no significant damage like the pinhole was obtained, except for CaSO4 and P2O5, which were pitted at 1100 8C under the air with or without H2O condition. However, the sample surface under the hydrothermal condition had more severe damages than under the thermal aging condition with only air because of the thermal reaction of the ash mixture at 1100 8C. This result indicates that the single material in ash did not affect the SiC significantly; then, the damage by the mixed ash materials at 1100 8C was caused by the H2O. As shown above, the P2O5/Pt–SiC sample showed slightly partial pitting at 900 8C under the air with H2O condition. In fact, the accumulated ash in the SiC-DPF was a composite of the ash materials that came from engine oil components, engine cylinder wear and erosion of the exhaust system, and the SiC-DPF after-treatment system was always exposed to H2O. Therefore, the H2O in the after-treatment system can be regard as an influential factor of SiC-DPF durability even though the SiC itself has a very high melting point. As a result, considering the fact that large quantities of ash materials and H2O components are included in
Fig. 8. XRD pattern of the P2O5/Pt–SiC sample by hydrothermal aging.
712
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
Fig. 9. Images of (a) Na/SiC fresh sample, (b) Na/SiC aged under air condition, (c) Na/SiC aged under H2O/air condition and (d) Na/Pt–SiC aged at 900 8C under H2O/air condition. The results are magnified 7 and 100 times by microscope.
the exhaust gas, the regeneration temperature has to be controlled under a critical value to ensure the durability of SiC-DPF in the after-treatment system. Fig. 8 represents the XRD patterns of only the SiC substrate and P2O5/SiC sample. These patterns are formed by hydrothermal aging
at 1100 8C with or without Pt/g-Al2O3. As the ash material P2O5 and the catalyst Pt deposited on the SiC substrate, the SiC peaks decreased rapidly, whereas, the amount of SiO2 increased for the P2O5/SiC sample. For the P2O5/Pt–SiC sample, SiC and SiO2 peaks became weaker than those of the SiC-only and P2O5/SiC samples,
Fig. 10. Images of (a) Cs/SiC fresh sample, (b) Cs/SiC aged under air condition, (c) Cs/SiC aged under H2O/air condition and (d) Cs/Pt–SiC aged at 900 V under H2O/air condition. The results are magnified 7 and 100 times by microscope.
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
713
Fig. 11. Images of ALM/SiC aged sample at 900 V under (a) air condition and (b) H2O/air condition, (c) ALM/Pt–SiC sample aged under H2O/air condition. The results are magnified 7 and 100 times by microscope.
and the Pt composite peaks emerged from the coexistence of SiC and Al2O3 at high temperature. Therefore, these results suggested that the Pt catalyst in the P2O5/SiC promoted the inter-reaction of P2O5/SiC with other substrates in hydrothermal aging.
3.2. Inter-reaction of alkali metal with SiC and Pt catalyzed SiC In the case of an alkali metal with SiC powder, the temperature of inter-reaction in the aging conditions was lower than that in the
Fig. 12. Images of (a) AALM/SiC fresh sample, (b) AALM/SiC sample aged under air condition, (c) AALM/SiC sample aged under H2O/air at 1100 V and (d) AALM/Pt–SiC sample aged under H2O/air condition. The results are magnified 7 and 100 times by microscope.
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
714
case of the ash material, generally. The inter-reaction of Na with SiC was stronger than anything else in our experimental conditions, and same results of glazing, melting and pinholes were shown at different temperatures in hydrothermal or thermal aging conditions (see Fig. 9). At 900 8C under hydrothermal aging, the reaction of the Na/Pt–SiC sample gave a different result from slight pitting (see Fig. 9(d)) and extensive pitting (see Fig. 9(c)). On the other hand, the Cs/SiC sample showed a higher reaction temperature than the Na/SiC sample, and the result of the reaction showed no difference between the aging conditions. However, different results between the Cs/SiC sample and Cs/Pt–SiC sample under the hydrothermal aging condition were observed at 900 8C by moderate pitting and staining, respectively (see Fig. 10). In Fig. 9(a)–(c) shows that Na/SiC sample damaged more extensively as the temperature increased under different conditions, showing pitting at 900 8C and glazing at 1100 8C. In the comparison of (c) with (d) at 900 8C for the hydrothermal aging, the surface of the Na/SiC sample expands much more than the Na/Pt– SiC sample, and glazing also occurs. As shown in Fig. 10, the inter-reaction in the Cs/SiC sample was more severe as the temperature increased under the hydrothermal or thermal aging condition. It shows the same reaction: slight pitting at 900 8C and moderate glazing at 1100 8C. Thus, there was slight adherence at 900 8C and extensive adherence at 1100 8C in the presence of H2O, whereas there was no adherence under the thermal aging condition. However, in the comparison of (c) with (d) at 900 8C for the hydrothermal aging condition, the surface of the Cs/Pt–SiC sample shows only staining whereas the Cs/SiC sample shows pitting. The reaction of the ALM (the composite of alkali metals)/SiC sample shows pitting from 800 8C. The sample was damaged seriously as the temperature increased to 1100 8C under different conditions because of the extensive interaction of the ALM/SiC sample, which prevented the observation of any differences between aging conditions. In the comparison of the ALM/SiC sample with the ALM/Pt–SiC sample at 900 8C under hydrothermal aging, the surface of ALM/SiC sample expanded more extensively than the ALM/Pt–SiC sample, and in this case, glazing and pinholes occurred on the ALM/SiC sample, but only slight pitting occurred on the ALM/Pt–SiC sample (see Fig. 11). The interaction temperature of the ALM/SiC sample was much lower than that of the ash materials, and the adherence became more severe with the increase of temperature. Fig. 12 shows the results of the interaction of AALM (ash materials and alkali metals mixture)/SiC in the thermal and hydrothermal aging conditions. The pitting phenomenon happened at 800 8C. With the increase of temperature, partly melting and pinholes were presented at 900 8C, and finally, the sample was
Fig. 13. SEM micrograph of AALM/SiC sample produced at 1100 V under (a) air condition (b) H2O/air condition. The surfaces are magnified 200 times by SEM.
completely damaged at 1100 8C in the air with or without H2O condition. More melting and pinholes were observed in the SEM micrograph (see Fig. 12(b)) than in Fig. 13(a). Therefore, more damages occurred in the hydrothermal aging than in the thermal aging. As shown in Fig. 12(c) and (d) at 1100 8C, the surface of the AALM/Pt–SiC sample did not expand under the hydrothermal aging, and slight melting and pinholes occurred. Table 2 shows the results of the inter-reaction between ash materials/alkali metals and the SiC sample according to temperature. Fig. 14 represents the XRD pattern of the Na/SiC samples with or without Pt catalyst by hydrothermal aging. The result shows composite chemical elements, such as sodium aluminium silicates. These composite chemicals were produced by the chemical interreactions among Na, Al2O3, SiC and Pt. Therefore, the catalyzed SiC
Table 2 Inter-reaction of sample on air (500 cm3/min) with or without 10% H2O. Subject
Ash material
Alkali metal
CeO2
Fe2O3
CaSO4
CaO
P2O5
MgO
AM
Na
Cs
ALM
AALM
Air 800 8C 900 8C 1100 8C
^
^*
^~ ~~ &~
^ ~
^~ ~~ &~
^ ~* * &~
Air + H2O 800 8C 900 8C 1000 8C
^
* & ^~
^*
^~ ~~ &~
^* ~~
^~ ~~ &~
^ ~* *&~
Pt, Air + H2O 900 8C 1000 8C
^
*
^
^ ~
^* ^&
^~
AM: ash materials mixture with SiC; ALM: alkali metals mixture with SiC; AALM: ash materials and alkali metals mixture with SiC. () No reaction, ( (~) Glazing, (*) Pinhole, (&) Bulk melting (*) Slight adherence, (&) Moderate adherence, (~) Extensive adherence.
) Staining, (^) Pitting,
B.C. Choi et al. / Journal of Industrial and Engineering Chemistry 15 (2009) 707–715
715
of composites. Moreover, the mixture of ash materials and alkali metals with Pt–SiC restrained surface expansion. In fact, the accumulated ash materials and alkali metals in the SiC-DPF were an admixture, and the SiC-DPF after-treatment system always contained H2O. According to the results, H2O in the after-treatment system can be regard as an influential factor of SiCDPF durability even though the SiC itself has a very high melting point. As a result, the regeneration temperature has to be controlled under a critical value to ensure the durability of SiCDPF in the after-treatment system, considering the fact that large quantities of ash materials, alkali metals and H2O components are included in the exhaust gas. Acknowledgement This works was supported by the 2nd stage BK (Brain Korea) 21 Project in 2008. Fig. 14. XRD pattern of the Na/Pt–SiC sample by hydrothermal aging.
accumulated with alkali metals, such as Na, etc., was easily exposed to damage by hydrothermal aging. 4. Conclusions The results of the hydrothermal aging (under air with H2O condition) for ash material samples showed that the inevitable H2O in the after-treatment system of a diesel engine can be an influential factor of SiC-DPF durability even though the SiC itself has a very high melting point. The results were observed by microscopy and SEM-EDS. However, the hydrothermal aging durability of the Pt–SiC-CDPF with ash materials, such as the P2O5 sample and the ash material mixture sample, was changed by high temperatures. Alkali metal samples (alkali metal/SiC) did not provide sufficient evidence to prove that H2O is a possible influential factor of SiC-DPF durability in the after-treatment system. On the other hand, the results of XRD analysis showed a high degree of inter-reaction between Pt–SiC with alkali metals for the formation
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]
O. Tomokazu, SAE paper No. 2005-01-0579, 2005. J.W. Jeong, B.C. Choi, M.T. Lim, J. Ind. Eng. Chem. 14 (2008) 830. G.A. Merkel, W.A. Cutler, C.J. Warren, SAE paper No. 2001-01-0190, 2001. M. Harada, S. Ichikawa, A. Otsuka, M. Hattori, S. Hashimoto, SAE paper No. 200501-0582, 2005. M. Matti Maricq, Aerosol. Sci. 38 (2007) 1079. I. Atribak, A. Bueno-Lopez, A. Garcia-Garcia, Catal. Commun. 9 (2008) 250. T.V. Johnson, SAE paper No. 2006-01-0030, 2006. J.C Tan, C.N. Opris, K.J. Baumgard, J.H. Johnson, SAE paper No. 960136, 1996. H. Okazoe, K. Shimizu, Y. Watanabe, E. Santiago, P. Kugland, W. Ruth, SAE paper 960130, 1996. B.C. Choi, H.Y. Kang, G.S. Son, C.W. Hwang, SAE paper No. 2007-01-1939, 2007. Y. Uchida, S. Ichikawa, T. Harada, T. Hamanaka, SAE paper No. 2003-01-0384, 2003. S. Ichikawa, Y. Uchida, A. Kaneda, T. Hamanaka, SAE paper No. 2004-01-0951, 2004. K. Yuuki, T. Ito, H. Sakamoto, R. Matsubara, S. Hashimoto, T. Hamanaka, SAE paper No. 2003-01-0383, 2003. C.R. Scherik, J. McDonald, B. Olson, SAE paper No. 2001-01-1351, 2001. M.J. Murtagh, D.L. Sherwood, L.S. Socha, SAE paper No. 940235, 1994. A. Gubernat, J. Eur. Ceram. Soc. 27 (2007) 781. T. Mizutani, Y. Watanabe, K. Yuuki, S. Hashimoto, T. Hamanaka, SAE paper No. 2004-01-0159, 2004. P.L. Yaroshchyk, R.J.S. Morrison, D. Body, B.L. Chadwick, Spectrochim. Acta B 60 (2005) 986.