Cu/Zeolite SCR on High Porosity Filters:
`Laboratory and Engine Performance Evaluations
`
`2009-01-0897
`
`Copyright © 2009 SAE International
`
`Giovanni Cavataio, James W. Girard and Christine K. Lambert
`Ford Motor Company
`
`ABSTRACT
`
`Selective catalytic reduction (SCR) is expected to be
`used extensively in the U.S. for diesel vehicle NOx
`control. Much progress has been made on improving
`performance and reducing complexity of SCR systems
`for vehicles in the past several years. SCR system
`complexity can be reduced further by implementation of
`SCR-coated diesel particulate filters (SCRFs). In this
`system, a high porosity (> 50%) filter substrate is coated
`with an SCR formulation, ideally in the pores of the filter
`walls, so that the DPF and SCR functions can be
`combined into a single catalyst.
`
`Two state-of-the-art Cu/zeolite SCR formulations and
`three types of high porosity filter substrates were
`included
`in
`this study.
` Laboratory and engine-
`dynamometer tests were performed to measure NOx
`conversion under a variety of conditions to assess the
`impact of ammonia oxidation, inlet NO2/NOx ratio,
`ammonia/NOx ratio, oxygen level, space velocity, soot
`loading, and ammonia loading level. In general, SCRF
`technology was found to perform similarly to traditional
`channel-flow catalysts with the same type of active
`materials.
`
`
`INTRODUCTION
`
`Urea-based selective catalytic reduction (SCR) is one
`method for controlling oxides of nitrogen (nitric oxide,
`NO, and nitrogen dioxide, NO2, commonly referred to
`together as NOx) in lean exhaust gas. SCR may be used
`in light duty diesel vehicle applications to help meet Tier
`2 Bin 5 emissions requirements [1]. In urea-based SCR,
`
`aqueous urea is injected into the hot exhaust gas, where
`it vaporizes and converts to ammonia (NH3). NH3 reacts
`with NOx on the SCR catalyst to form primarily nitrogen
`(N2) and water (H2O).
`
`One of the challenges faced by SCR technology for
`diesel vehicles is the complexity of the system. In order
`to fully meet Tier 2 Bin 5 emissions requirements,
`additional aftertreatment components are
`typically
`needed in addition to the SCR system. A common
`arrangement of a diesel aftertreatment system including
`NOx control is shown in Figure 1. A diesel oxidation
`catalyst (DOC) controls carbon monoxide (CO) and
`hydrocarbon (HC) emissions. A porous wall-flow diesel
`particulate filter (DPF) is used for meeting particulate
`matter
`(PM)
`requirements.
` The DOC
`in such
`applications can have the added function of providing
`heat
`for active DPF
`regenerations by oxidizing
`hydrocarbons introduced into the exhaust system, either
`by late in-cylinder injection or by hydrocarbon injection
`into the exhaust system.
`
`
`
`From
`Engine
`
`DOC
`
`SCR
`
`DPF
`
`Urea
`
`To
`Tailpipe
`
`
`
`FIGURE 1: Example of a typical diesel aftertreatment
`system including DOC, SCR and DPF.
`
`
`
`
`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.
`ISSN 0148-7191
`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.
`877-606-7323 (inside USA and Canada)
`SAE Customer Service: Tel:
`724-776-4970 (outside USA)
` Tel:
`724-776-0790
` Fax:
` Email: CustomerService@sae.org
`SAE Web Address: http://www.sae.org
`
`*9-2009-01-0897*
`
`Printed in USA
`
`BASF-2032.001
`
`

`
`There is great potential for system improvement through
`reduced complexity of the diesel aftertreatment system.
`One potential simplification of the system, combining
`SCR and DPF into a single substrate, is the subject of
`this paper. A sketch of this concept is given in Figure 2.
`
`ammonia/NOx ratio, oxygen level, space velocity, soot
`loading, and ammonia loading level. In general, SCRF
`technology was found to perform similarly to traditional
`channel-flow catalysts with the same type of active
`materials.
`
`
`
`From
`Engine
`
`DOC
`
`SCRF
`
`Urea
`
`To
`Tailpipe
`
`
`
`FIGURE 2: Sketch of a diesel aftertreatment system
`including DOC and combined SCR-on-filter (SCRF).
`
`
`
`In the SCRF concept, the filter substrate has an
`increased porosity in comparison to the standard DPF of
`Figure 1. The SCR washcoat is ideally coated inside of
`the pores of the DPF, rather than on the substrate
`surface, allowing room for adequate SCR washcoat and
`open DPF channels for gas flow and soot storage. The
`performance of SCRF systems on vehicles is reported in
`Dobson, et al. [2] and Lee, et al. [3].
`
`There are several potential advantages of an SCRF
`system. There is one less canned catalyst, reducing
`manufacturing complexity. Also, since both the DPF and
`the SCR are relatively large catalysts, eliminating one of
`them reduces both the amount of volume under the
`vehicle required for the aftertreatment system and the
`weight of the aftertreatment system. Reduced weight
`means less stringent exhaust support requirements and
`potential fuel economy improvements. In addition, the
`SCRF system (Figure 2) enables the soot-loaded filter to
`be placed closer to the DOC and thus reducing the heat
`loss upstream of the filter during active regenerations.
`
`Since the SCRF concept involves the placement of the
`SCR washcoat within a high porosity DPF substrate, the
`SCR washcoat should be able to tolerate extreme
`temperature swings due to typical and non-typical active
`DPF regenerations. For this reason, the SCRF catalyst
`formulation was aged to 800°C for 64 hours to represent
`120,000 miles of equivalent
`time-at-temperature
`exposure during soot regeneration events. In addition to
`thermal
`robustness, high SCRF NOx
`reduction
`performance will be required in the 200°C to 350°C
`temperature range for city driving and potentially 300°C
`to 450°C temperature range for highway driving. Finally,
`the SCRF must also be able to handle uncontrolled NOx
`emissions during the active DPF regeneration events
`(450°C – 700°C).
`
`the
`in
`tested
`this study, SCRF samples were
`In
`laboratory and on an engine dynamometer to understand
`the impact of ammonia oxidation, inlet NO2/NOx ratio,
`
`
`
`
`EXPERIMENTAL
`
`LABORATORY EVALUATIONS
`
`All SCRF samples were aged on a flow reactor for 64hr
`at 800°C in flowing gas that contained 14% oxygen (O2),
`5% carbon dioxide (CO2), 5% H2O and balance N2. The
`gas flow rate through the sample during aging was 6.44
`standard liters per minute (slm). All samples were
`loaded with the same amount of washcoat to within 5%
`of the nominal loading. The samples were 2.54cm in
`diameter and 7.62cm long, with the plugs at each end of
`the filter being 0.95 cm long, such that the active catalyst
`length was 5.72 cm.
`
`The laboratory reactor system was made up of quartz
`tubing and a pre-heat tube furnace to ensure a relatively
`uniform catalyst sample temperature followed by a
`second tube furnace with the catalyst sample. Ammonia
`gas diluted in nitrogen was injected between the two
`furnaces to reduce the effect of oxidation on the tube
`walls at elevated temperatures. Gas measurements
`were taken with a Midac FTIR equipped with a heated
`sample cell. The standard space velocity was 30,000/hr
`and was varied from 15,000/hr up to 60,000/hr.
`
`Testing was done on two promising copper-zeolite SCR
`catalyst washcoat formulations coated on three types of
`high porosity filters (two types of cordierite and one
`silicon carbide).
` Table 1 summarizes
`the SCR
`washcoat/substrate combinations tested.
`
`TABLE 1: SCR Catalyst and Substrate Properties
`
`
`Sample
`
`Description
`
`Testing
`
`A-C1
`
`A-C2
`
`A-SiC
`
`B-C1
`
`B-SiC
`
`Cu-Zeolite SCR A on
`Cordierite 1
`
`Cu-Zeolite SCR A on
`Cordierite 2
`
`Cu-Zeolite SCR A on
`Silicon Carbide
`
`Laboratory
`
`Laboratory
`
`Laboratory
`
`Cu-Zeolite SCR B on
`Cordierite 1
`
`Laboratory and
`Engine
`Dynamometer
`
`Cu-Zeolite SCR B on
`Silicon Carbide
`
`Laboratory
`
`
`
`
`
`2
`
`BASF-2032.002
`
`

`
`Three main reactions were studied:
`
`STANDARD SCR REACTION:
`
`TABLE 3: Inlet gas conditions for the standard SCR
`reaction at varying NH3/NOx ratios
`
`
`4NH3 + 4NO + O2  4N2 + 6H2O (1)
`
`
`
`Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7
`
`NO
`
`NO2
`
`NH3
`
`350
`ppm
`
`350
`ppm
`
`350
`ppm
`
`245
`ppm
`
`315
`ppm
`
`350
`ppm
`
`350
`ppm
`
`0
`
`385
`ppm
`
`350
`ppm
`
`350
`ppm
`
`350
`ppm
`
`420
`ppm
`
`525
`ppm
`
`700
`ppm
`
`Alpha
`
`0.7
`
`0.9
`
`1.0
`
`1.1
`
`1.2
`
`1.5
`
`2.0
`
`O2
`
`H2O
`
`CO2
`
`N2
`
`14%
`
`5%
`
`5%
`
`Balance
`
`
`
`For the ammonia oxidation test, NOx was removed
`completely from the inlet gas and replaced by N2 (Table
`4). The other gas concentrations remained the same.
`
`
`
`FAST SCR REACTION:
`
`2NH3 + NO + NO2  2N2 + 3H2O (2)
`
`
`
`AMMONIA OXIDATION REACTION:
`
`4NH3 + 3O2  2N2 + 6H2O (3)
`
`
`
`The inlet gas conditions for the standard SCR reaction
`are described in Table 2.
`
`
`
`
`
`TABLE 4: Inlet gas conditions (ammonia oxidation)
`
`Feed Gas Composition
`
`AMMONIA OXIDATION REACTION
`
`NO (ppm)
`
`NO2 (ppm)
`
`NH3 (ppm)
`
`O2 (%)
`
`CO2 (%)
`
`H2O (%)
`
`N2
`
`0
`
`0
`
`350
`
`14.0
`
`5.0
`
`5.0
`
`balance
`
`
`
`The Ammonia Storage Capacity measurements were
`carried in the laboratory flow reactor system on just one
`of
`the aged sample cores. The NH3 adsorption
`measurement was
`performed at 14
`separate
`temperatures ranging from 125°C to 500°C. Before each
`measurement, the SCRF was conditioned at 600°C for
`ten minutes in flowing nitrogen containing 5% water, 5%
`CO2, and 14% oxygen. The catalyst was cooled to the
`evaluation temperature and then a 350ppm NH3 step
`function was introduced for 3600s. The 3600s of NH3
`exposure allowed the sample to completely saturate.
`
`For low temperature evaluations, Figure 3 shows the
`typical pre-SCR NH3 profile compared to the post-SCR
`NH3 profile in the absence of NH3 oxidation activity. The
`these profiles were used
`to
`differences between
`calculate the NH3 storage capacity per liter of catalyst
` For each higher
`temperature
`geometric volume.
`
`3
`
`
`
`
`
`TABLE 2: Inlet gas conditions (standard SCR reaction)
`
`
`Feed Gas Composition
`
`STANDARD SCR REACTION
`
`NO (ppm)
`
`NO2 (ppm)
`
`NH3 (ppm)
`
`O2 (%)
`
`CO2 (%)
`
`H2O (%)
`
`N2
`
`350
`
`0
`
`350
`
`14.0
`
`5.0
`
`5.0
`
`balance
`
`
`
`For more detailed studies, the inlet NO2/NOx ratio was
`varied
`from 0
`to 1 while holding
`the
`total NOx
`concentration constant. To understand the impact of
`oxygen, the O2 concentration was varied from 1 to 20%.
`The ammonia/NOx ratio was also varied from 0.7 to 2.0
`at the conditions shown in Table 3 under standard SCR
`conditions (14% O2, 5% H2O, 5% CO2, balance N2).
`
`
`
`
`
`BASF-2032.003
`
`

`
`100
`90
`80
`70
`60
`50
`40
`30
`20
`10
`0
`150
`
`
`
`NOx Conversion (%)
`
`evaluation where the SCRF sample was active for NH3
`oxidation,
`the pre-SCR 350ppm
`step
`function
`assumption could not be used. A significant portion of
`NH3 can be oxidized rather than stored. For these higher
`temperatures,
`the
`equilibrated
`post-SCR NH3
`concentration at the end of 3600s was used to represent
`the corrected pre-SCR NH3 concentration step function
`level.
`
`Pre - SCR Profile
`
`Post - SCR Profile
`
`500
`
`450
`
`400
`
`350
`
`300
`
`250
`
`200
`
`150
`
`100
`
`50
`
`0
`
`NH3 Concentration (ppm)
`
`
`
`0
`
`600
`
`1200
`
`1800
`Elapsed Time (s)
`
`2400
`
`3000
`
`3600
`
`
`
`FIGURE 3: Example of typical NH3 concentration
`profiles in the pre-SCR location and post-SCR location.
`Difference in the two profiles was used to determine the
`NH3 storage capacity for each SCRF sample.
`
`
`
`ENGINE DYNAMOMETER TESTING
`
`In addition to the detailed laboratory testing, engine
`dynamometer testing was carried out. The catalyst size
`for engine testing was 19 cm in diameter and 20.3 cm
`long and was degreened for 2 hours at 750oC on the
`engine dynamometer prior to catalyst evaluations. A Pd-
`containing DOC was used upstream of the SCRF and
`was degreened on the engine for two hours at 800oC.
`Very little NO2 was present in the gas entering the SCRF.
`The layout of the exhaust system for the engine testing
`resembled that shown in Figure 2.
`
`The engine used for testing was a 2.7 liter turbocharged
`direct injection Ford Lion V6 engine designed to meet
`Euro 4 emissions standards. All testing was done using
`ultra-low sulfur diesel fuel. Space velocity was varied by
`changing the engine speed in a range from 1500 rpm to
`4000 rpm. The speed and load were varied in order to
`evaluate various catalyst temperatures while the space
`velocity was held constant. This was repeated at space
`velocities of 15,000, 30,000 and 40,000/hr.
`
`
`
`
`
`4
`
`RESULTS AND DISCUSSION
`
`PERFORMANCE OF SCRF UNDER STANDARD SCR
`CONDITIONS
`
`The laboratory performance of Washcoat A for the
`standard SCR reaction on three different substrate types
`(A-C1, A-C2, A-SiC) was found to be similar (Figure 4).
`A potential cause for slight differences in the results
`might be washcoat distribution in the various substrates.
`Since the washcoat is coated in the filter wall, the
`washcoat distribution may depend on the pore structure.
`Due to the low quantity of available samples with this
`new technology, it was not possible to determine the
`part-to-part variability in performance. Overall, the
`performance of Washcoat A on a filter substrate was
`found to be similar to that of a conventional channel-flow
`SCR catalyst [4].
`
`
`
`Sample A-C2Sample A-C2
`
`Ceramic T yp e 2
`Si licon Carbi de
`Ceramic T yp e 1
`
`
`
`Sample A-SiCSample A-SiC
`
`
`
`Sample A-C1Sample A-C1
`
`200
`
`250
`
`300
`
`500
`450
`400
`350
`Catalyst T em perature (ºC)
`
`550
`
`600
`
`650
`
`700
`
`
`
`FIGURE 4: SCRF performance under standard SCR
`conditions for Washcoat A with varied substrates.
`
`
`
`The standard SCR reaction was also performed in the
`laboratory flow reactor on samples B-C1 and B-SiC
`(Figure 5). Again the results are similar to each other. It
`can be seen from Figure 5 that this catalyst has poorer
`performance when compared with Washcoat A (Figure
`4) for temperatures below 500°C. Washcoat B does not
`reach NOx conversion efficiencies of 90% or greater until
`250°C, a shift of approximately 50°C versus Washcoat
`A.
` However, washcoat B performed better
`than
`washcoat A for temperatures above 500°C.
`
`BASF-2032.004
`
`

`
`AMMONIA OXIDATION REACTION
`
`In order to understand the steep decline in NOx activity
`at higher temperatures (Figure 6), the gross oxidation of
`ammonia in the absence of NOx was performed (Figure
`7). It was found that the NH3 oxidation reaction
`generated NO, NO2, and N2O by-products at
`temperatures greater than 400°C. By difference, the
`remaining nitrogen species unaccounted
`for was
`assumed to be N2. If true, the NH3 oxidation reaction
`below 400°C would be as high as 100% selective to N2.
`Direct measurement of N2 was not possible due to the
`use of N2 as the carrier gas in the reaction mixture.
`Although no evidence of NO, NO2, and N2O byproducts
`were observed between 225°C and 400°C, we can not
`rule out the possibility of NH3 oxidation by O2 to form
`NOx (NOx remake) where the generated NOx may still
`further react with available NH3 to yield N2 (no undesired
`by-products).
`
`NO Formation (ppm)
`
`200
`
`180
`
`160
`
`140
`
`120
`
`100
`
`80
`
`60
`
`40
`
`02
`
`0
`
`NH3 Conversion (%)
`
`NO2 Formation (ppm)
`
`N2O Formation (ppm)
`
`NO Formation (ppm)
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`NO2 and N2O Formation (ppm)
`
`NH3 Conversion (%)
`
`
`
`100 150 200 250 300 350 400 450 500 550 600 650 700 750
`
`Inlet Gas Temperature (ºC)
`
`
`FIGURE 7: Ammonia oxidation activity of SCRF A-C1 in
`the absence of NOx.
`
`
`
`IMPACT OF INLET NO2/NOx RATIO
`
`To study the NO2 dependency, the inlet NO2/NOx ratio
`was varied in the laboratory from 0 to 1 (i.e. 0% - 100%)
`(Figure 8). The low temperature NOx performance of
`SCRF A-C1 was improved by 5-10% with the addition of
`NO2 when the percentage was varied from 0% NO2/NOx
`(i.e. Standard SCR Reaction) to 50% NO2/NOx (i.e. Fast
`SCR Reaction). Further increasing the NO2/NOx level
`from 50% to 100% worsened the low temperature NOx
`performance significantly and dramatically increased the
`formation of undesirable N2O (Figure 9).
`
`Sample B-C1
`
`Washcoat B Ceramic 1
`Washcoat B Silicon Car bid e
`
`Sample B-SiC
`
`200
`
`250
`
`300
`
`500
`450
`400
`350
`Catalyst Tem perature (ºC)
`
`550
`
`600
`
`650
`
`700
`
`100
`90
`80
`70
`60
`50
`40
`30
`20
`10
`0
`150
`
`
`
`NOx Conversion (%)
`
`FIGURE 5: Comparison of SCR-on-filter performance at
`alpha = 1 for samples BC1 and B-SiC.
`
`
`
`SCRF sample A-C1 was used for a more detailed
`laboratory study on the standard SCR reaction. The
`ammonia conversion and N2O formation during the
`standard SCR reaction were measured (Figure 6). At
`200°C, the NOx conversion was around 80% for the
`Standard SCR Reaction. The peak NOx conversion for
`both SCR
`reactions
`reached 99%
`for a narrow
`temperature
`range.
` However, as
`the evaluation
`temperature was
`increased,
`the NOx conversion
`declined rapidly.
`
`The same SCRF formulation (SCRF A-C1) but two
`different samples and flow reactor systems were used to
`obtain the data in Figures 4 and 6. As a result, some
`slight variability in the NOx conversion was realized.
`
`NH3 Conver sio n (%)
`
`NOx Conversion (%)
`
`N2O Formation (ppm )
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`-10
`
`-20
`
`N2O Formation (ppm)
`
`NOx and NH3 Conversion (%)
`
`
`
`100
`
`150 200 250 300
`
`350 400 450
`
`500 550 600 650
`
`700 750
`
`Inlet Gas Temperature (ºC)
`
`
`FIGURE 6: NOx and NH3 conversion efficiency of SCRF
`A-C1 at standard SCR conditions. N2O formation is also
`included.
`
`5
`
`
`
`
`
`BASF-2032.005
`
`

`
`Alp ha = 0.7
`Alp ha = 0.9
`Alp ha = 1.0
`Alp ha = 1.1
`Alp ha = 1.2
`Alp ha = 1.5
`Alp ha = 2.0
`Alp ha=1, 4% O2
`
`150
`
`200
`
`250
`
`500
`450
`400
`350
`300
`C atalyst Temperature (ºC)
`
`550
`
`600
`
`650
`
`700
`
`100
`90
`80
`70
`60
`50
`40
`30
`20
`10
`0
`100
`-10
`
`
`
`NOx Conversion (%)
`
`FIGURE 10: Performance with varied ammonia/NOx
`ratio (alpha) for SCRF A-C1. Also shown is the effect of
`oxygen level on NOx conversion at high temperature.
`
`
`
`IMPACT OF OXYGEN LEVEL
`
`In contrast to the relatively stable flue gas conditions
`from a typical stationary source, the transient operation
`of diesel vehicles requires the aftertreatment system to
`function properly under dynamic levels of O2 content.
`For example, during hard accelerations or full speed/load
`conditions, the O2 content can drop to as low as 1%.
`Due to the integration of the SCR catalyst within a DPF
`substrate, the SCRF catalyst must also function during
`active soot-regeneration events where the O2 content
`can also reach as low as 1%. On the other hand, under
`idle-type conditions, the O2 content can reach up to 20%.
`
`The effect of oxygen level was studied over the SCRF for
`the standard SCR reaction only, as O2 must be present
`for that reaction to occur. For temperatures below
`450°C, results indicated a substantial catalytic decline in
`the NOx conversion as the oxygen level was decreased
`from 20% to 1% (Figure 11). Low temperature oxidation
`functions within catalysts are often kinetically-limited.
`For temperatures above 450°C, results indicated an
`improvement in NOx conversion as the oxygen level was
`decreased from 20% to 1%. This trend at high
`temperature was
`favorable
`for
`improving NOx
`conversion during high temperature events (i.e. such as
`active DPF regeneration events) and was probably due
`to the corresponding decrease in the undesired ammonia
`oxidation reaction with O2 (Figure 12).
`
`
`
`0% NO2/NOx
`
`20% NO2/NOx
`
`50% NO2/NOx
`
`80% NO2/NOx
`
`100% NO2/NOx
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`-10
`
`-20
`
`-30
`
`NOx Conversion (%)
`
`100 150 200 250 300 350 400 450 500 550 600 650 700 750
`
`Inlet Gas Temperature (ºC)
`
`
`FIGURE 8: Impact of feedgas NO2/NOx ratio on the
`activity of SCRF A-C1.
`
`0% NO2/NOx
`
`20% NO2/NOx
`
`50% NO2/NOx
`
`80% NO2/NOx
`
`100% NO2/NOx
`
`50
`
`45
`
`40
`
`35
`
`30
`
`25
`
`20
`
`15
`
`10
`
`05
`
`N2O Formation (ppm)
`
`
`
`100 150 200 250 300 350 400 450 500 550 600 650 700 750
`
`Inlet Gas Temperature (ºC)
`
`
`
`
`FIGURE 9: N2O formed at varying NO2/NOx inlet ratios
`with SCRF A-C1.
`
`
`
`AMMONIA/NOx RATIO EFFECT AT STEADY-STATE
`
`Figure 10 shows the results for SCRF A-C1 under
`standard SCR conditions (NO2/NOx = 0, 14% O2) with
`ammonia/NOx ratios from 0.7 (underdosing) to 2.0
`(overdosing). In general, NOx conversion improved with
`increasing ammonia. Depending on the overdosing
`ratio, ammonia slip was evident until high temperatures,
`at which point the SCR formulation was able to oxidize
`the excess ammonia (as in Figure 7). The efficiency for
`a ratio of 1 in 4% oxygen rather than 14% was measured
`at high
`temperature
`to better simulate a
`filter
`regeneration condition, resulting in an improvement in
`NOx conversion. The oxygen impact was studied in
`detail in the next section.
`
`
`
`6
`
`BASF-2032.006
`
`

`
`SV = 15,000/hr
`
`SV = 30,000/hr
`
`SV = 45,000/hr
`
`SV = 60,000/hr
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`-10
`
`-20
`
`NOx Conversion (%)
`
`100 150 200 250 300 350 400 450 500 550 600 650 700 750
`
`Inlet Gas Temperature (ºC)
`
`
`
`FIGURE 13: NOx conversion over SCRF A-C1 as a
`function of temperature and space velocity (Standard
`SCR Reaction).
`
` A
`
`the engine
`in
` similar study was performed
`dynamometer using SCRF B-C1 formulation (Figure 14).
`Using steady-state engine operating points, the catalyst
`performance was evaluated at each point by allowing
`the exhaust gas composition to stabilize for 30 minutes
`to ensure full ammonia storage levels on the catalyst.
`The space velocities tested were 16,000/hr, 30,000/hr,
`and 40,000/hr. The moderate decline in NOx conversion
`between 16,000/hr and 30,000/hr seemed reasonable
`based on
`the data collected on the SCRF A-C1
`formulation (Figure 13). Further comparison of the
`SCRF B-C1 catalyst results from 30,000/hr to 40,000/hr
`revealed a more negative impact than expected. At this
`point, it is unclear why the SV = 40,000/hr showed such
`a large impact.
`
`
`
`
`
`30k h-1 SV
` SV
`40k h-1 SV
`
`-1
`16k h
`
`250
`
`450
`350
`
`SC R Bed T em perature ( o C)
`
`550
`
`100%
`90%
`80%
`70%
`60%
`50%
`40%
`30%
`20%
`10%
`0%
`150
`
`
`
`NOx Conversion (%)
`
`FIGURE 14: Engine dynamometer performance of
`SCRF B-C1 at varied space velocities.
`
`20%
`
`14%
`
`10%
`
`4%
`
`2%
`
`1%
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`-10
`
`-20
`
`NOx Conversion (%)
`
`100 150 200
`
`250 300
`350
`400 450
`500
`550 600 650
`Inlet Gas Temperature (ºC)
`
`700 750
`
`FIGURE 11: NOx conversion over SCRF A-C1 as a
`function of temperature and O2 level (Standard SCR
`Reaction).
`
`14%
`
`10%
`
`4%
`
`2%
`
`1%
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`NH3 Oxidation (%)
`
`
`
`100 150 200
`
`250 300
`350
`400 450
`500
`550 600 650
`Inlet Gas Temperature (ºC)
`
`700 750
`
`FIGURE 12: Ammonia oxidation over SCRF A-C1 as a
`function of temperature and O2 level in the absence of
`NOx.
`
`
`
`EFFECT OF SPACE VELOCITY
`
`In the laboratory, space velocity using the standard SCR
`reaction conditions was increased from 15,000/hr to
`60,000/hr, decreasing the residence time by a factor of
`four. As a result, the NOx conversion of SCRF A-C1
`dropped for the entire temperature range tested (Figure
`13).
`
`
`
`
`
`7
`
`BASF-2032.007
`
`

`
`(Figure 16). During the NH3 adsorption mode, the
`threshold storage capacity (TSC) was the amount of
`adsorbed NH3 that can be stored at greater than 97%
`capture efficiency (no more than 10ppm slip). Because
`of this, the TSC NH3 amount was considered to be
`strongly held. It was interesting that the TSC was
`approximately 50% of the total NH3 stored on the SCRF
`catalyst. As a result, the NH3 storage capacity was
`considered to be composed of 50% strongly held NH3
`and 50% weakly held NH3. The NH3 storage capacity
`results were obtained in the presence of oxygen. As a
`result, the oxidation of NH3 needed to be considered in
`the calculations.
` Significant oxidation activity was
`observed at temperatures above 250°C for sample A-C1.
`For each evaluation
`temperature,
`the gross NH3
`oxidation percentage was calculated by comparing the
`concentration of NH3 at the end of the 3600s (post-SCR)
`exposure to the 350ppm NH3 inlet concentration (pre-
`SCR). Although a step function of 350ppm NH3 was
`introduced into the SCR catalyst for 3600s, the post-SCR
`steady state concentration was less than 350ppm NH3
`for evaluations temperatures above 250°C. This result
`indicated a reaction between NH3 and O2 as described in
`Equation 3. As expected, the NH3 oxidation profile in
`Figure 16 matched well with the previous evaluation
`performed in Figure 7. Recall from Figure 7 that some
`by-products (N2O, NO, and NO2) were detected below
`500°C while the oxidation activity reached as high as
`90% conversion. By performing a material balance of all
`the known nitrogen species, the authors infer that the
`NH3 + O2 reaction (Figures 7 and 16) was greater than
`90% selective toward N2 formation (for T < 500°C).
`
`NH3 Oxidation (%)
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`0
`
`01
`
`NH3 Adsorption
`
`Threshold Storage Capacity (TSC)
`
`NH3 Oxidation
`
`2000
`
`1800
`
`1600
`
`1400
`
`1200
`
`1000
`
`800
`
`600
`
`400
`
`200
`
`0
`
`100
`
`150
`
`400
`350
`300
`250
`200
`Inlet Gas Temperature (°C)
`
`450
`
`500
`
`
`
`
`
`NH3 Stored (mg/L)
`
`
`
`FIGURE 16: Summary of the NH3 storage capacity
`(adsorption), TSC, and ammonia oxidation activity as a
`function of temperature for SCRF A-C1.
`
`
`For SCRF A-C1, Figure 17 summarizes the steady-state
`NOx conversion, the TSC NOx conversion, the initial
`NOx conversion as a function of catalyst temperature.
`The
`initial NOx conversion represented the SCRF
`catalyst performance after the cumulative NH3 exposure
`
`EFFECT OF SOOT LOADING
`
`A unique aspect of the engine dynamometer testing as
`opposed to the laboratory testing was that soot was
`generated. The impact of soot on NOx conversion was
`measured over a narrow temperature range at steady
`state points from 250°C to 425°C (Figure 15). Soot
`loading on
`the
`filter appeared
`to
`influence NOx
`conversion at temperatures roughly between 300°C and
`350°C.
` At both
`lower
`temperatures and higher
`temperatures the soot loading did not impact NOx
`conversion. The most likely cause for the result seen in
`Figure 15 would be some type of coking reaction of the
`soot or hydrocarbons with the SCR washcoat. This
`phenomenon has been seen previously with zeolite SCR
`washcoats [5-6] in approximately the same temperature
`range.
` Cu/zeolite SCR catalysts are poor as
`hydrocarbon and soot oxidation catalysts when
`compared to Pt-based DOC formulations. Cu/zeolite
`SCR catalysts partially oxidize hydrocarbons to yield
`significant levels of carbon monoxide (CO). The partial
`oxidation of hydrocarbons to CO can lead to the
`formation of undesirable coke byproducts that happen to
`be prominent in the 300°C – 350°C temperature range
`[6].
`
`The results in Figure 15 showing a drop of up to 10%
`NOx conversion with soot loaded on the SCRF was
`determined
`to be reproducible
`from repeated
`test
`evaluations (not shown).
`
`
`
`N Ox Conv. Clean N Ox Conv. Clean
`
`
`NOx Conv with SootNOx Conv with Soot
`
`N Ox Conv. wit h SootN Ox Conv. wit h Soot
`
`
`
`250 250
`
`
`
`350 350
`450 450
`
`
`
`SCR Bed Temp er ature ( o SCR Bed Temp er ature ( o C) C)
`
`
`
`550 550
`
`
`100% 100%
`
`90% 90%
`
`80% 80%
`
`70% 70%
`
`60% 60%
`
`50% 50%
`
`40% 40%
`
`30% 30%
`
`20% 20%
`
`10% 10%
`
`0% 0%
`
`150 150
`
`
`
`
`
`NOx Conversion (%)
`NOx Conversion (%)
`
`FIGURE 15: Engine dynamometer performance of
`SCRF B-C1 showing the effect of soot loading on NOx
`conversion.
`
`
`EFFECT OF AMMONIA LOADING
`
`The data discussed thus far were collected at steady-
`state conditions when the SCRF was saturated with NH3.
`in
`the
`The NH3 storage capacity was measured
`laboratory for sample A-C1 as a function of temperature
`
`
`
`8
`
`BASF-2032.008
`
`

`
`aligned with the decrease in the reaction of ammonia
`and NOx at temperatures greater than 450°C.
`
`• The addition of NO2 in the inlet gas to the SCRF was
`beneficial up to 50% and then became detrimental to
`NOx performance. Undesirable N2O emissions
`increased with increasing NO2 levels.
`
`lower NOx
`in
`resulted
`• Underdosing ammonia
`conversion; overdosing ammonia resulted in higher
`NOx conversion and ammonia slip. Above 550°C
`the ammonia slip was oxidized by the SCRF.
`
`• Reduced oxygen levels led to lower NOx conversion
`at low to mid-range temperatures (150-450°C) and
`improved NOx conversion at higher temperatures.
`For temperatures above 450°C, lower O2 levels were
`to suppress
`the undesired ammonia
`thought
`oxidation reaction with O2 and therefore benefit the
`desired NH3 + NOx reaction.
`
`•
`
`to
`from 15,000/hr
`Increasing space velocity
`60,000/hr reduced the NOx conversion in both the
`lab and the engine dynamometer.
`
`• Soot loading on the SCRF B-C1 (Washcoat B on
`cordierite
`type 1)
`in
`the engine dynamometer
`reduced NOx conversion by as much as 10% at
`temperatures between 250°C and 425°C, most likely
`due to coking of the SCRF.
`
`adsorption/desorption
`ammonia
`• Laboratory
`measurements of SCRF A-C1 revealed that about
`50% of the ammonia was strongly held and 50% was
`weakly held. A threshold storage capacity of strongly
`held ammonia was defined as the amount of
`ammonia needed to reach the near peak NOx
`conversion with little risk of ammonia slip.
`
`This study focused on NOx conversion with ammonia
`using several combinations of SCR catalysts and filter
`materials. While this is an extremely important aspect of
`SCRF technology, the loading of SCR material into the
`filter substrate presents new challenges for diesel vehicle
`applications. The backpressure of SCRF compared to
`today's precious-metal or bare filter technology can be
`much higher and needs to be improved. For example,
`some of our unpublished results
`indicated similar
`backpressure levels of a washcoated conventional cDPF
`device with 5g/L soot loading to a washcoated SCRF
`device with no soot loading [7]. The filter material
`properties should be well defined for coating to optimize
`the accessibility of the SCR catalyst by the exhaust gas.
`The physical and thermal properties of the coated filter
`will play a major role in the filter regeneration frequency
`and ultimately the fuel economy of the vehicle.
`
`
`
`9
`
`of only 100mg/L. The TSC NOx conversion represented
`the peak performance of the SCRF catalyst with no more
`than 10ppm NH3 slip. The steady-state NOx conversion
`was determined after waiting one hour at each
`temperature for conditions to stabilize and represented
`the best possible NOx conversion with unrestricted NH3
`slip. Additional cumulative exposure up to the TSC value
`changed the NOx performance dramatically without
`much NH3 slip risk. Recall from the previous section that
`the TSC level was approximately 50% of the total NH3
`storage capacity. Therefore, operation of the SCRF
`catalyst should be operated within the 0% to 50% NH3
`storage capacity range to avoid excessive NH3 slip.
`Performance at temperatures >450°C was independent
`of the cumulative NH3 exposure level.
`
`SS NOx Conversion
`
`TSC NOx Conversion
`
`Initial NOx Conversion
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`NOx Conversion (%)
`
`
`
`100
`
`150
`
`200
`
`400
`350
`300
`250
`Inlet Gas Temperature (°C)
`
`450
`
`500
`
`550
`
`
`
` Summary of steady-state NOx
`
`
`FIGURE 17:
`conversion, TSC NOx conversion, and
`initial NOx
`conversion (100mg/L NH3 exposure level) as a function
`of inlet gas temperature for sample A-C1.
`
`
`
`CONCLUSION
`
`The following conclusions emerge from this study:
`
`• Wall-flow SCRF technology yielded similar NOx
`performance
`to channel-flow SCR catalysts
`in
`laboratory testing after extended hydrothermal aging
`at 800oC, representing full useful life (120k mi).
`
`• Two Cu/zeolite washcoats were successfully coated
`on high porosity cordierite and SiC filters. Washcoat
`A performed slightly better than Washcoat B at
`T<500°C. However, Washcoat B performed better
`at T>500°C.
`
`laboratory study of SCRF A-C1
`• A detailed
`(Washcoat A on cordierite type 1) under standard
`SCR condi