`
`SAE TECHNICAL
`PAPER SERIES
`
`2005 -01 -0583
`
`Aluminum Titanate Compositions for Diesel
`Particulate Filters
`
`S. B. Ogunwumi, P. D. Tepesch, T. Chapman, C. J. Warren,
`I. M. Melscoet- Chauvel and D. L. Tennent
`Corning Incorporated
`
`Reprinted From: Diesel Exhaust Emission Control 2005
`(SP -1942)
`
`ISBN 0- 7680 -1635-
`
`1111
`9 780768 016352
`
`ZInternationa[rc
`
`$
`
`EXHIBIT
`
`(20,2 Lf
`
`2005 SAE World Congress
`Detroit, Michigan
`April 11 -14, 2005
`
`400 Commonwealth Drive, Warrendale, PA 15096 -0001 U.S.A. Tel: (724) 776 -4841 Fax: (724) 776 -5760 Web: www.sae.org
`
`BASF-2024.001
`
`
`
`The Engineering Meetings Board has approved this paper for publication. It has successfully completed
`SAE's peer review process under the supervision of the session organizer. This process requires a
`minimum of three (3) reviews by industry experts.
`
`All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
`transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise,
`without the prior written permission of SAE.
`
`For permission and licensing requests contact:
`
`SAE Permissions
`400 Commonwealth Drive
`Warrendale, PA 15096- 0001 -USA
`Email: permissions @sae.org
`724 -772 -4028
`Tel:
`724 -772 -4891
`Fax:
`
`t+wlff
`
`e
`Global Mobility Database®
`All SAE papers, standards, and selected
`books are abstracted and indexed in the
`Global Mobility Database.
`
`For multiple print copies contact:
`
`SAE Customer Service
`877 -606 -7323 (inside USA and Canada)
`Tel:
`724 -776 -4970 (outside USA)
`Tel:
`724 -776 -1615
`Fax:
`Email: CustomerService @sae.org
`
`ISSN 0148 -7191
`Copyright © 2005 SAE International
`
`Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE.
`The author is solely responsible for the content of the paper. A process is available by which discussions
`will be printed with the paper if it is published in SAE Transactions.
`
`Persons wishing to submit papers to be considered for presentation or publication by SAE should send the
`manuscript or a 300 word abstract to Secretary, Engineering Meetings Board, SAE.
`
`Printed in USA
`
`BASF-2024.002
`
`
`
`Aluminum Titanate Compositions for Diesel Particulate Filters
`
`2005 -01 -0583
`
`S. B. Ogunwumi, P. D. Tepesch, T. Chapman, C. J. Warren,
`I. M. Melscoet- Chauve) and D. L. Tennent
`Corning Incorporated
`
`Copyright © 2005 SAE International
`
`ABSTRACT
`
`the mixed strontium /calcium feldspar
`Compositions in
`([Sr /Ca]OA12032Si02) - aluminum titanate (A1203TIO2)
`system have been investigated as alternative materials
`for the diesel particulate filter (DPF) application. A key
`attribute of these compositions is their low coefficient of
`thermal expansion (CTE). Samples have been prepared
`with porosities of >50% having average pore sizes of
`between 12 and 16pm. The superior thermal shock
`resistance, increased resistance to ash attack, and high
`volumetric heat capacity of these materials, coupled with
`monolithic fabrication, provide certain advantages over
`currently available silicon carbide products. In addition,
`based on testing done so far aluminum titanate -based
`filters have demonstrated chemical durability and
`comparable pressure drop (both bare and catalyzed) to
`current, commercially available, silicon carbide products.
`
`INTRODUCTION
`
`Diesel particulate matter (PM) emissions pose serious
`health concerns and are under environmental regulation.
`Diesel filter after -treatment technology is currently used
`to remediate PM emissions. SiC and cordierite filters are
`two of the most viable solutions available for use today.
`Cordierite has a low coefficient of thermal expansion and
`this application, but
`thermal shock
`can survive
`in
`low heat capacity. It
`limited by its
`is also
`cordierite is
`reaction during exceedingly high
`susceptible
`to ash
`temperature applications [1]. SiC, on the other hand,
`has a lower thermal shock resistance and thus needs to
`segmentation
`segmented.
`increases
`The
`be
`is a concern because of
`manufacturing costs and
`potential mechanical integrity issues. Other issues with
`SiC have been reported recently and solutions have
`been implemented [2 -3].
`
`is a novel Aluminum
`An alternate DPF candidate
`Titanate (AT) ceramic oxide composite. The composition
`is highly refractory with a melting temperature exceeding
`1500 °C. The high heat capacity of the composition is an
`attribute that is beneficial for thermal management and
`allows the filter regeneration temperature to be
`low.
`Although the intrinsic coefficient of thermal expansion of
`aluminum titanate is quite high (CTE = >9 *10- 6 / °C), the
`
`composite AT composition discussed herein has
`excellent thermal shock resistance due to
`the lower
`the secondary phases and
`thermal expansion of
`the AT phase. Microcracking
`microcracking of
`is a
`feature that has been used successfully for cordierite in
`catalytic converters. In addition, Sakai and Bradt have
`that microcracking can be used
`demonstrated
`to
`improve the toughness and increase the thermal shock
`resistance of a ceramic [4].
`
`this study, an overview of
`the compositions,
`In
`resulting material properties of this
`processing, and
`novel AT composition are presented. Two generations of
`subsequent
`compositions are discussed
`showing
`for
`improvement
`porosities. Strategies
`the
`of
`optimization of the porosity are presented resulting in
`considerably lower backpressures. Finally, the durability
`investigated under various conditions. The material
`is
`has met internal durability requirements for the DPF
`application.
`
`COMPOSITION
`
`The strength of pure aluminum titanate ceramics
`is
`limited due to extensive microcracking [5], which results
`from the crystallographic thermal expansion anisotropy.
`In order to simultaneously optimize CTE, strength, and
`durability, an additional phase or phases are required.
`The first generation AT composition (AT -Gen A) was
`based on the pseudo phase diagram shown in Figure 1.
`
`Mullite
`3A1 0, 2Si0.
`
`SAS
`SrOAI.O,2SiO;
`
`Al2TiO5
`
`Figure 1. AT -Gen A Phase assembly diagram
`
`BASF-2024.003
`
`
`
`The composition was batched to include 7.5% mullite
`(3A1203 2SiO2), 22.5% strontium feldspar (SrO A1203
`2SiO2) and 70% aluminum titanate [6]. Less than 0.25%
`iron oxide (Fe203) was added to the batch. The Fe203
`forms a solid solution with the aluminum titanate phase,
`and improves the thermal decomposition resistance of
`the AT phase
`[7]. The processing
`other
`and
`components, including the resulting properties will be
`two sections. Subsequent
`discussed
`the next
`in
`generations of the AT composition aimed at increasing
`improving strength and
`reducing CTE, and
`porosity,
`durability resulted in the current composition (AT -Gen B)
`which incorporates pore formers and includes calcium.
`The calcium is added to form a solid solution of a mixed
`Feldspar (CaO,SrO A1203 2Si02) [8]. Approximately
`4% of the current batch is Ca- Feldspar (CaO A1203
`2SiO2), leaving a reduced strontium feldspar level of
`18.5 %, and the same levels of mullite and aluminum
`titanate as described previously. In addition, the AT -Gen
`B composition has no Fe203 additive.
`
`MATERIALS AND PROCESSING
`
`The ease of processing and the choice of standard raw
`materials associated with this oxide composite make it
`attractive. A desirable feature of aluminum titanate filters
`is that the firing temperatures can be as low as 1400 °C,
`[9]. Another desirable processing
`is done
`in air
`and
`advantage of the AT DPF
`is extruded as a
`that it
`is
`monolith. According to the current information available to
`us this does not seem to be the case with SiC filters
`which require assembly of extruded SiC segments and
`higher firing temperatures under controlled atmosphere.
`In addition, segmentation of SiC might pose potential
`mechanical integrity issues and decreases open frontal
`area.
`
`For extrusion of the AT oxide composite, fine powders of
`together, mulled and
`the raw materials are batched
`extruded as illustrated in Figure 2 below. The resulting
`monolith after extrusion
`is dried and fired, and finally
`plugged to make a DPF.
`
`Mix A1203 + TiO2 + (Fe203) + Si02+ Sr-0 +
`CaO + pore former + binder
`
`Extrude, Dry
`
`Fire
`
`Figure 2. Processing Steps for AT Filters
`
`METHOD FOR POROSITY IMPROVEMENT
`
`For the first generation of AT, no calcium or pore former
`the composition. The porosity was
`were added
`to
`
`achieved simply by reactive sintering of known precursor
`the
`formed during
`materials, with
`the pores being
`evolution of the reaction. The particle size of the starting
`improve
`the
`resulting
`materials were optimized
`to
`typically 40 -42% for the first generation. An
`porosity;
`the
`example of the resulting microstructure showing
`highly microcracked nature of the material is shown in
`Figure 3. The microcracks come from the anisotropy in
`CTE of the aluminum titanate and the mismatch of the
`CTE of the individual phases, resulting in a composite
`with an overall low CTE and elastic modulus. The result
`is a highly thermal shock resistant material.
`
`Figure 3. SEM Image of the Polished Cross Section of
`AT -Gen A.
`
`The white phase in the upper SEM is the Sr- feldspar
`phase, while the gray phase is the aluminum titanate
`phase which is mixed with some glass, trace alumina
`and titania.
`
`The next generation of AT incorporated changes in the
`raw materials and
`the
`particle sizes of the starting
`addition of pore formers to increase the porosity to 51%
`and create a well connected microstructure. In addition,
`the batch after durability
`Fe203 was removed
`from
`indicated that compositions without it passed
`studies
`internal durability requirements. Finally, calcium was
`added in order to improve property uniformity and ease
`processing.
`
`ALUMINUM TITANATE DPF PROPERTIES
`
`lists the properties of the two generations of AT
`Table 1
`discussed herein.
`
`Table 1. Physical Properties of AT Compositions
`
`% porosity
`Pore Size (pm)
`CTE (x10-' / °C; 1000 °C)
`MOR (PSI [MPa])
`eMod (PSI x105 [GPa])
`Firing Temperature ( °C)
`
`AT -Gen A
`41
`
`17
`
`5
`206 [1.42]
`2.65 [1.83]
`1500
`
`AT -Gen B
`
`51
`
`15
`
`9
`213 [1.47]
`2.10 [1.45]
`1450
`
`BASF-2024.004
`
`
`
`The progression of properties from a product with 41%
`porosity to one having 51% porosity is clearly evident in
`the SEM images in Figure 4. The polished cross sections
`clearly show more porosity with a more evenly distributed
`pore size in the Gen B composition. By achieving such a
`distribution without the large pores observed in the Gen A
`composition, we have been able to maintain nearly the
`same strength with 10% higher porosity.
`
`PRESSURE DROP PERFORMANCE
`
`Clean and soot -loaded pressure drop were measured on
`a series of 4 different AT and SiC samples in 144 mm x
`150 mm geometry (5.66" x 6 "). The filters were loaded
`with artificial soot (Printex U - Degussa) by aerating the
`fine powder into an air stream at a flow rate of -76
`m3 /hr. After a specified amount of artificial soot was
`loaded, each filter was removed, weighed, and placed
`on Corning's internal pressure drop set -up where the
`pressure drop was measured as a function of flow rate
`according to published procedures [8]. The sample was
`then loaded with more soot and the process repeated
`to -5 g/L was obtained. The
`full curve out
`until a
`loading curves for these
`pressure drop versus soot
`samples obtained at the highest measured flow rate
`the Corning setup
`obtainable
`(of 356 m3 /hr) are
`in
`illustrated in Figure 5.
`
`Figure 4. Polished Cross Section SEM Images of AT -Gen
`A (left) and AT -Gen B (right), both at 100X.
`
`The properties of the latest version of AT are compared
`with those of a commercially available SiC product in
`Table 2. The most striking difference is, of course, the
`CTE which requires the SiC product to be fabricated in
`segments. The mechanical strength of SiC is over 5
`times higher than it
`in AT. On the other hand the
`is
`thermal shock parameter (TSP), which is a calculated
`value based on mechanical strength, CTE and e-
`modulus values [10],
`is nearly an order of magnitude
`this material's greater
`for AT,
`higher
`indicating
`resistance to thermal shock and ability to fabricate it as a
`In addition, the strain tolerance (stress divided
`monolith.
`by elastic modulus) is much greater in AT, indicating that
`AT can elongate much more than SiC before breaking.
`
`Table 2. Comparison of AT and SiC Properties
`
`% Porosity
`Pore Size (um)
`MOR (PSI [MPa])
`eMod (PSI x 105 [GPa])
`CTE (x1071°C)
`Strain Tolerance
`(MOR /eMod)
`Cell Density (Cells /in2)
`Wall Thickness (mil)
`Bulk Density (g /cm3)
`TSP (MOR /CTE x eMod)
`
`AT -Gen B
`51
`15
`213 [1.47]
`2.10 [1.45]
`9
`
`SiC
`58
`17
`1185 [8.17]
`18.9 [13.0]
`44
`
`1.01e-3
`
`6.27e -4
`
`320
`12
`0.74
`1127
`
`320
`13
`0.71
`142
`
`5
`
`4
`
`3
`
`2
`
`1
`
`0
`
`Gen A; 200/17; 40% porosity
`-s- Gen A: 300/15; 40% porosity
`
`- sic; 320/13; 58% porosity
`
`Gen B; 320/12; 51% porosity
`
`0
`
`1
`
`4
`2
`3
`Soot Loading (g /L)
`
`Figure 5. Soot Loaded Pressure Drop Curves for AT
`and SiC Filters.
`
`Note that according to these test results pressure drop is
`highly dependent on cell density and web thickness
`combinations. For example, at the same 40% porosity,
`the nominal 200/17 AT -Gen A DPF has a higher
`pressure drop than the nominal 300/15 AT -Gen A filter.
`Increasing the porosity to 51% while decreasing the wall
`thickness to 12 mils results in a significant pressure drop
`decrease for the nominal 320/12 AT -Gen B filter, which
`lower backpressure
`the nominal
`than
`has an even
`filter with 58% porosity. While
`lower
`320/13 SiC
`is certainly a goal
`for better engine
`backpressure
`filter's ability
`performance and
`fuel economy, a
`to
`survive under the extreme conditions of regeneration will
`dictate what cell geometry /porosity combinations will be
`chosen for a given application.
`
`BASF-2024.005
`
`
`
`DURABILITY OF ALUMINUM TITANATE
`FILTERS
`
`In order to verify fitness for use in diesel applications, a
`series of durability evaluations was conducted on AT-
`Gen A. These tests included: 1. exposure to engine ash
`(which contains Fe, P, Ca, Zn, etc.) that can react with
`form a eutectic melt or
`some substrate materials to
`cause densification [1]; 2. exposure to iron and
`iron
`oxide (which are common forms of debris in the exhaust
`react similar
`high
`system and can
`to ash);
`3.
`temperature oxidizing and reducing conditions (which
`vary as the engine cycles from rich to lean and as soot
`accumulates on the filter); and 4. acidic solutions (which
`the
`reaction of SOX with H2O
`may
`result
`from
`in
`presence of a catalyst).
`
`temperature exposure conditions were
`Extreme high
`in order to
`to cause failure in
`the samples
`chosen
`It should be stressed, however,
`establish thresholds.
`temperature exposure under service
`that expected
`less severe.
`Since
`the peak
`is much
`conditions
`temperatures observed during regeneration
`for
`last
`periods of seconds, the cumulative time in service at the
`(assuming -500 regenerations
`highest temperatures
`where only 10% are uncontrolled) is on the order of only
`minutes. Because there are time /temperature exposure
`limitations on both catalyst activity and on the sintering
`of engine ash (temperatures > 1050 °C cause permanent
`changes in surface permeability due to ash consolidation
`[1]), the conditions of the tests performed in this study
`are well beyond conditions expected during practical
`service conditions.
`
`Ash Stability
`
`testing was performed under extreme
`Ash stability
`conditions of 8 hours at 1100 °C, 1200 °C, and 1300 °C
`with 10% humidity. For this test, actual ash (collected
`on- engine) is sieved onto several modulus of rupture
`bars which are broken with the ash contact surface in
`treatment according
`following heat
`the
`tension
`to
`process previously described [1]. No change in strength
`failure mode was observed versus
`or preferential
`untreated parts. As a matter of fact, no reaction was
`observed with the substrate under any of the above
`conditions. Cross sections of the samples showed
`negligible penetration of ash constituents into the bulk of
`the sample up to 1200 °C. After 1300 °C, Ca and P had
`penetrated the top cell wall and some discoloration
`occurred.
`
`Iron Oxide
`
`iron oxide was conducted in a
`Exposure to
`iron and
`testing.
`similar fashion to the ash
`iron oxide
`Fine
`(hematite) powder and coarse black iron filings were
`
`applied to the substrate in separate tests. Test coupons
`in air and soaked for 1 hour at 100 °C
`were heated
`increments between 900 °C and 1500 °C. No detrimental
`reactions with the substrate were observed for any of the
`test conditions. Even after exposure to 1500 °C, the iron
`A cross
`material did not penetrate the cell wall.
`sectional view of an AT cell wall after this exposure is
`shown in Figure 6.
`
`Figure 6. Optical Microscope Cross Sectional View of
`AT -Gen A Substrate Indicating no Fe Penetration or
`Reaction Following Exposure at 1500 °C for 1 hour.
`
`High Temperature Oxidizing and Reducing Conditions
`
`temperature
`Exposures of the AT substrate to high
`under oxidizing and reducing conditions were performed
`in order to determine the breakdown conditions for the
`AT -Gen A composition. The aluminum titanate phase,
`thermal expansion anisotropy
`which provides
`the
`for micro -cracking,
`is metastable. Given
`necessary
`this phase can decompose (Al2TiO5
`sufficient time,
`becomes Al2O3 + TiO2) at temperatures below 1250 °C,
`and this decomposition results in a higher CTE. The
`reaction are
`thermodynamics of
`this
`kinetics and
`the presence or absence of
`influenced by
`strongly
`stabilizing additives [11].
`
`In order to test the AT -Gen A samples under extreme
`reducing conditions, the filters were loaded with 10g /I of
`for various
`soaked
`in argon
`carbon
`soot and
`temperatures and times to verify fitness for use. These
`conditions are many orders of magnitude more reducing
`than the exhaust stream. Samples were also soaked in
`conditions.
`Results are
`study oxidizing
`air
`to
`summarized in Table 2 which also includes expected
`service conditions: 500 regenerations, 10% uncontrolled.
`
`BASF-2024.006
`
`
`
`Table 2. Results of Exposure of AT -Gen A to Oxidizing
`and Reducing Conditions
`
`Atmosphere
`
`p02
`
`T ( °C)
`
`Reducing
`
`Oxidizing
`
`Argon +Soot
`Argon +Soot
`Air
`Air
`
`10e -20
`10e -20
`0.2
`0.2
`
`1200
`1000
`1200
`1000
`
`Time
`(Min)
`
`25
`600
`3000
`6000
`
`Thermal
`Expansion
`(CTE)
`
`Unchanged
`Unchanged
`Unchanged
`Unchanged
`
`Acid Stability
`
`Exposure of the AT substrate to low pH solutions was
`for 5 hours.
`conducted using sulfuric acid at 80 °C
`Several solutions at pH's between 7 and 1 were used to
`determine acceptable conditions for service on engine,
`in catalyst coater operations.
`for processing
`and
`Commercially available SiC and cordierite filters were
`also included in this test. Previous studies [12] have
`shown that pure SiC is unaffected by extreme acidic
`conditions. However, the product consists of segments
`which are held together by cement. The authors were
`not aware of any public disclosure of the performance of
`these cements in low pH solutions. AT -Gen A showed
`no degradation in strength under any of these conditions
`as shown in Figure 7.
`
`--AT Gen A
`
`800
`
`700
`
`600
`
`500
`
`400
`
`300
`
`200
`
`100
`
`0
`
`a
`
`0
`2
`
`5.52
`
`4.52
`
`3.52
`
`2.52
`
`1.52
`
`0.52
`
`-0.48
`
`Untreated
`
`7
`
`4
`
`3
`
`2
`
`1
`
`pH
`
`Figure 7. Strength vs. Acid Exposure for AT -Gen A
`
`Durability evaluations of AT -Gen B are ongoing.
`In fact,
`as of this publication, AT -Gen B exceeds requirements
`in the areas discussed above. More detailed results will
`be published separately.
`
`CONCLUSION
`
`A novel mixed oxide system has been developed for
`Diesel Particulate Filters. Based on the testing described
`above, due to the highly microcracked structure and
`thermal expansion,
`this material has higher
`lower
`
`resistance
`than silicon carbide.
`thermal shock
`In
`addition, internal studies have shown that AT is resistant
`to chemical attack even when subjected to ash and iron
`in excess of 1200 °C
`for
`long exposures.
`oxides
`Furthermore, the material seems able to withstand acidic
`environments that would be expected during the life of
`the filter. Finally the unique microstructure and 51%
`porosity of the AT -Gen B composition allows for a better
`porosity SiC while
`pressure drop
`than
`58 -60%
`maintaining a higher bulk heat capacity.
`
`ACKNOWLEDGMENTS
`
`The authors wish to thank Mr. J. Schermerhorn, Mr. A.
`Gorges, Ms. Barbara Oyer and Ms. Elizabeth Wheeler
`the samples, Ms. K.
`for preparing and processing
`Robbins and Ms. Elaine McDonald for pressure drop
`and regeneration testing, and Mr. Jeff Payne for help
`to thank
`with the durability studies. We also wish
`members of Corning Characterization Services
`for
`properties measurements.
`
`REFERENCES
`
`3.
`
`1. Merkel, G.A., Cutler, W.A. and Warren, C.J.,
`"Thermal Durability of wall -Flow Ceramic Diesel
`Particulate Filter ", SAE 2001 -01 -0190.
`"Further Durability
`2. Kazushige
`al.,
`O.,
`et.
`Enhancement of Re- crystallized SiC -DPF ", SAE
`2001 -01 -0190.
`Ishikawa. S, et. al., "Durability Study on Si -SiC for
`DPF ", SAE 2004 -01 -0951.
`4. Sakai, M., Bradt, R.C., "Fracture Toughness Testing
`of Brittle Materials ", Int. Mat. Rev. (1993) 38, 2, 53-
`78
`5. D. Hasselman and J. Singh, Ceram. Bull., 1979, 58,
`856.
`6. Tepesch, P. and Ogunwumi, S., "Aluminum -Titanate
`for Diesel Particulate Filters ",
`Based Materials
`presentation at the American Ceramics Society 28th
`on Advanced Ceramics
`International Conference
`and Composites, January 2004.
`7. G. Tilloca, Journal of Material Science, 26 (1991)
`2809 -2814.
`8. Tennent, D, Tepesch, P., Ogunwumi, S., Warren, C.,
`Chapman, T. and Melscoet -Chauvel, I., "Aluminum
`Titanate Compositions for Diesel Particulate Filters ",
`the American
`for presentation
`Submitted
`at
`Ceramics Society 29th International Conference on
`Advanced Ceramics and Composites, January
`2005.
`9. Locker, R.J., Gunasekaran, N. and Sawyer, C.,
`"Diesel Particulate Filter Test Methods ", SAE 2002-
`01 -1009.
`10. Gulati, S., Hampton, L. and Lambert, D., SAE 2002-
`02 -0738.
`11. Freudenberg et al, US Patent #5,153,153, Oct. 6th
`1992.
`12. Uchida, Y., et al., 2003. "Durability Study on Si -SiC
`Material for DPF ", SAE 2003 -01 -0384.
`
`BASF-2024.007