`
`Enhanced Durability of a Cu/Zeolite Based SCR Catalyst
`
`Giovanni Cavataio~ Hung-Wen Jen, James R Warner, James W. Girard,
`Jeong Y. Kim and Christine K. Lambert
`Ford Motor Company
`
`Copyright© 2008 SAE International
`
`2008-01-1025
`
`ABSTRACT
`
`Passenger and light duty diesel vehicles will require up to
`90°/o NOx conversion over the Federal Test Procedure
`(FTP) to meet future Tier 2 Bin 5 standards. This
`accomplishment is especially challenging for low exhaust
`temperature applications that mostly operate in the 200 -
`350°C temperature regime. Selective catalytic reduction
`(SCR) catalysts formulated with Cu/zeolites have shown
`the potential to deliver this level of performance fresh,
`but their periormance can easily deteriorate over time as
`a result of higt1 temperature thermal deactivation. These
`high
`temperature SCR deactivation modes are
`unavoidable due to
`the requirements necessary to
`actively regenerate diesel particulate fiiters and purge
`SCRs
`irom sulfur and hydrocarbon contamination.
`Careful vehicle temperature control of these events is
`necessary to prevent unintentional thermal damage but
`not always possible. As a result, there is a need to
`develop thermally robust SCR catalysts.
`Fe/zeolite
`formulations are known to exhibit superior hydrothermal
`stability over Cu/zeolite formulations. However, current
`Fe/zeolite forrnulations are not very active for NOx
`conversion in the desired 200 -- 350°C temperature
`regime under conditions having low N02/NOx ratios.
`From previous studies, Cu/zeolite formulations have
`demonstrated never-to-exceed
`temperatures up
`to
`In this work, a laboratory flow reactor was
`775°C.
`utilized to hydrothermally age and evaluate tl1e latest
`state-of-the-art Cu/zeolite formulations. Results confirm
`remarkable high temperature hydrothermal stability up to
`950°C while maintaining stable low temperature NOx
`activity.
`A broad
`range of
`time-at-temperature
`hydrothermal aging was carried out to clearly define the
`full durability range. n1e aging time was varied from 1
`hour to 256 hours while the aging temperature was
`varied from 670 °C to i 100 °C. The catalyst performance
`was evaluated under a synthetic exhaust gas mixture
`commonly known as the "Standard" SCR reaction.
`
`INTRODUCTION
`
`the Federal Test
`The operating conditions over
`Procedure (FTP) results in high NOx emissions in the
`200 --· 350°C temperature range. From current ligl1t·duty
`
`diesel applications, the future Tier :?. Bin 5 em1smon
`standards will require up to 90"/o reduction in the tailpipe
`NOx emissions.
`
`Implementation of zeolite based components has been
`extensively studied for application in gasoline and diesel
`aftertreatment devices.
`However,
`the harsh high
`exhaust
`temperatures observed
`in
`typical gasoline
`vehicles have limited their widespread use. On the other
`hand, the relatively milder diesel exhaust temperatures
`have encouraged continued development of zeolites as a
`major component in aftertreatment devices.
`
`tecl1nology
`aftertreatrnent
`diesel
`prom1smg
`One
`containing zeolite is the Selective Catalytic Reduction
`(SCR) of NOx with an ammonia-based reductant such as
`aqueous urea. As stated in equation ( i), NOx reduction
`is possible due to the high selectivity of the ammonia
`(NHa) and nitrogen oxide (NO) reaction to form elemental
`In the absence of nitrogen dioxide (NO;:), this
`N2•
`reaction is referred to as the ''Standard" SCR reaction
`[i ]. Additionally, the SCR reaction containing 50% NO
`and 50% N02 is referred to as the "Fast" SCR reaction
`(equation 2).
`
`"Standard"
`
`( 1)
`
`"Fast"
`
`(2)
`
`Vanadium, Fe/zeolite, and Cu/zeolite based SCR
`formulations are very active for the "Standard" SCR
`reactions. However, vanadium based formulations have
`been shown to easily deactivate when exposed to
`temperatures necessary to actively regenerate Diesel
`particulate filters (DPFs) with oxygen [2]. This cannot be
`avoided since a DPF is currently required to meet Tier 2
`Bin 5 particulate matter (PM} emission standards.
`Fe/zeolites have been shown to be much more durable
`to high temperature exposure. However, in the absence
`of N02, Fe/zeolites lack the low temperature (200 ·
`35ocq NOx activity necessary for high FTP efficiency.
`In this critical temperature range, Cu/zeolite formulations
`have been reported to have much lower s;::nsitlvity to the
`N02/NOx ratio. As a result, Cu/zeolite formulations have
`been shown to achieve hiqh NOx conversion at the
`desired
`low operating tem-peratures.
`Their lack of
`hydrothermal stability above T75 °C has drawn questions
`
`SAE Int. J. Fueis Lu!Jr. I Voiume 1 I Issue 1
`
`4"77
`
`Exhibit 2002.001
`
`
`
`EXHIBITS TO DECLARATION OF STANLEY ROTH, PH,D, UNDER 37 C.F.R. § U32
`
`14%0i
`5"/a H20
`5%1C02
`Balance N~
`€L44 nters/min
`
`4 i:::::::
`
`Cordierite
`
`DIAGRAM 1. Sample configuration during hydrotherrnai aging.
`
`A wide range of time-at-temperature hydrothermal aging
`was carried out to clearly define tile full durability range
`of a promising Cu/zeolite SCR
`formulation.
`The
`hydrothermal aging duration was varied from 1 hour to
`256 hours while the aging temperature was varied from
`670"C to 1100"C. Totaling ii 16 aging hours, Table i
`defines the 24 different aging conditions utilized in this
`study. Special attention was considered to determine
`the shoti-term never-to-exceed (NTE) temperature and
`the long··term SCR durability necessary to withstand the
`temperature resulting frorn DPF regeneration events.
`For a given aging duration, the NTE is defined as the
`temperature at which
`the NOx conversion decay
`accelerates significantly.
`
`about their long-term in-use durability and robustness to
`occasional oveHemperature events.
`
`Improvements in the thermal durability of Cu/zeo!ite
`based SCR formulations has been higt1!y desirable and
`pursued by many research
`institutes and catalyst
`suppliers.
`
`This paper discusses the performance and hydrothermal
`durability of an enhanced Cu/zeolite based SCR
`low temperature NOx
`formulation exhibiting durable
`activity under a wide matrix of time-at-temperature aging
`conditions.
`On key aged samples, surface area
`measurements and Cu reduction measurements are
`performed to investigate changes in the zeolite and Cu
`state, respectively.
`
`EXPERIMENTAL
`
`SAMPLE PREPARATION
`
`A full size monolith washcoated with a state-of-the-mi
`Cu/zeolite based SCR formulation was obtained from a
`catalyst supplier in 2007. The cordierite-based monolith
`measured 20.3cm diameter x 15.2cm length with 400
`cells per square inch (CPSI) and 4.5 mil wall thickness.
`The SCR monolith was completely cored and cut into
`i60 round samples measuring c:54cm diameter x
`2.54cm length. From this, a normal distribution was
`observed where the 95'% confidence interval around the
`mean mass was determined to be ± 0.4%. Older
`formulations mentioned in this paper did not necessarily
`exhibit the same distribution in mass.
`
`HYDROTHERMAL AGING
`
`i, sample cores were
`in Diagram
`As configured
`hydrothermally aged in flowing gas from an automated
`fiow reactor system. The total flow rate utilized was 6A4
`liters/min. The synthetic gas composition consisted ol
`14% 0 2 , 5% H~,o, 5% C02, and balance N2. For each
`aging,
`three samples rneasuring 2.54cm diameter x
`2.54cm length were placed in a quartz reactor tube and
`labeled A, 8, and C. The three SCR samples were
`separated by 30mm to ensure weli distributed gas flow in
`all channels. An uncoated cordierite monolith was
`placed upstream to serve as a gas heat exchanger. The
`an
`isothermal gas
`uncoa!ed monolith ensured
`temperature across each sample. Samples positioned in
`location "A" were used for surface area measurements.
`Samples positioned
`in
`location "B" were used
`for
`temperature-programmed
`reduction measurements
`(TPR). Samples positioned in location "C" were used for
`the NOx conversion evaluation tests.
`
`478 SAE Int. J. Fuels Lubr. i Volume 1 I Issue 1
`
`Mydrotherma! Aging Duration (hours}
`i 4
`8 ! 16
`··---~~~---- -······ ·······--·t-·-_ ....... J..
`
`Temp.
`
`("C)
`
`2
`
`7001X
`
`32
`
`64
`
`140 256
`
`t--x--+---+----1
`
`jX XX
`
`/X i
`
`_;t~-~~.~=L~~~
`
`...... 950jx···· .... x x ! .. x.
`,
`i
`x
`x
`
`1000
`1100
`'----!.........-.1.........-..<........-.J.---'----'--------- ~- -~-------.. -~~ .....
`
`I
`i
`I
`,
`~-----~
`I
`:
`
`TABLE 1. Time-at-temperature hydrothermal aging matrix.
`
`Exhibit 2002.002
`
`
`
`EXHIBITS TO DECLARATION OF STANLEY ROTH, PH,D, UNDER 37 C.F.R. § U32
`
`LABORATORY CATALYST EVALUATION
`
`Fundamental catalyst activity data were obtained using
`an automated
`laboratory-scale flow reactor system.
`Custom .. written labVIEW based software with National
`Instruments data acquisition hardware controlled MKS
`mass flow controllers and Lindberg Mini··Mite tubular
`furnaces. A computer controlled evaluation protocol was
`developed and run ior each sample to decrease the test(cid:173)
`to-test variations commonly observed by manual
`operation. Table 2 shows the simulated diesel exhaust
`gas composition flowed through each sample core to
`study the "Standard" SCR reaction.
`
`-----~-:~--~~~-~~~;~~---------- ------------------;~~~~~~;~~~~----------------1
`------------------------------------------------------ ---------------------------------------------------i
`-I
`NO (ppm)
`350
`N02 (ppm)
`0
`·
`·--------------NH~-(p·j;;)·--------------- --------------------------···35()·-------------·-··----------1
`
`!
`02 (%)
`14
`-----------------------------------------------------------------····------------------------j
`!
`5
`C02 (%)
`I
`s
`r----------------H~------------.
`[---------------~-~!~_n_c_~ _______________ J ___ -_---_______________
`-N_~_---_---_----------=---------_---~--J
`
`TABLE 2. Simulated gas composition used to study performance for
`the "Standard"' SCR reaction.
`
`For all evaluations, the total gas flow rate was held
`constant at 6.44 liters/min while the sample size was
`held constant at 2.54cm diameter x 2.54cm length. As a
`result, a space velocity equal to 30,000ihr was used in
`this study.
`For the typical light-duty diesel vehicle
`operating over the FTP drive cycle, this space velocity
`corresponds to a SCR monolith size between 100% -
`150% ol the engine swept volume.
`
`The SCR inlet gas temperature was maintained with one
`preheat tubular furnace followed by a second tubular
`furnace. SCR samples were loaded in quartz tubing and
`placed in the second tubular furnace. A Tl1ermo Electron
`Antaris IGS FT!R Oas Analyzer with a heated sample
`cell was used at the outlet of the reactor to measure NO,
`N02, N20, NH:3, C02, and H,.o levels. To cover the fuli
`exhaust ternperatures expected on diesel vehicles, data
`were
`taken at SCR
`inlet gas
`temperatures
`from
`approximately i50'C to 700°C in 25-50°C steps. Tl1e
`computer-controlled evaluation protocol stepped
`the
`reaction temperature setting from a high temperature to
`low temperature for a prescribed duration. At each
`temperature setting, the duration was chosen so that
`post SCR gas composition was allowed enough time to
`completely equilibrate.
`
`The fiow reactor used in this study was examined to
`determine
`the
`test-to-test variability of
`the entire
`measurement system. Among nurnerous variables, the
`
`FTIR measurement, the thermocouple measurement,
`and precision of
`the mass
`flow controllers may
`collectively contribute large discrepancies in the data.
`This may make it difficuit to conclude with confidence
`that one ·result is statistically diiferent than another result.
`Repeated evaluation runs were made on a single pre(cid:173)
`aged (64hr/670°C) SCR sample to determine the 95%
`confidence
`intervaL
`Five evaluation
`runs were
`pe1iormed over the entire temperature range.
`
`SURFACE AREA MEASUREMENTS
`
`A Micromeritics ASAP 2400 instrument in conjunction
`witri the well-known Brunauer, Emmet, and Teller (BET)
`equation was employed to determine the surface area of
`each SCR sample. The BET equation determines the
`surface area by establishing the relationship between the
`volurne adsorbed at a given partial pressure and the
`volume adsorbed at monolayer coverage [3].
`
`TEMPERATURE PROGRAMMED REDUCTION (TPR)
`
`The Cu state within zeolite lorrnulations changes during
`the SCR reaction and after tiydrothermal aging. These
`physical-chemical changes yield different
`reduction
`temperatures. The TPR results reveal direct evidence of
`changing Cu-species
`in
`the catalyst and rnay be
`correlated to the deactivation of SCR activity after aging.
`
`was
`(TPR)
`Temperature-Programmed-Reduction
`Ii 2920
`conducted on a Microrneritics AutoChem
`instrument. Part of a catalyst sample (location "B") was
`sliced off and cut into small pieces, about 3 mm long.
`These small pieces were then loaded into a quartz
`reactor for TPR. Typically, 0.5 ~iram oi sample was used
`in the experiment. The temperature was measured with
`a thermocouple in the catalyst bed. Prior to TPR, the
`catalyst sample was pretreated in 10%02/He at 600°C
`for 30 minutes and
`then cooled down
`to
`room
`temperature in 10°/o02/He. After the pretreatment, the
`gas fiow was changed to 9%H2iAr at :?.O ml/min. During
`TPR, the catalyst bed was heated to 600CC at a linearly
`rate of 1 o °C/min. The change
`increasing
`in H2
`concentration was monitored
`using
`Thermal(cid:173)
`Conductivity·Detector (TCD). The consumption of H2
`indicated the reduction of oxidized Cu.
`
`RESULTS AND DISCUSSION
`
`FLOW REACTOR VARIABILITY
`
`The variability of
`flow reactor was
`the evaluation
`determined by running a pre-aged SCR sample five
`times. The steady state "Standard" SCR reaction results
`for each of the five runs are overlaid in Figure 1.
`In
`addition, the 95% confidence interval around the mean
`
`SAE Int. J Fuels Lubr. i Volume 1 I Issue ·t
`
`479
`
`Exhibit 2002.003
`
`
`
`EXHIBITS TO DECLARATION OF STANLEY ROTH, PH,D, UNDER 37 C.F.R. § U32
`
`RECENT SCR DURABILITY IMPROVEMENTS
`
`Since current and future diesei aftertreatment systems
`formulations are
`required
`to
`contain DPFs, SCR
`withstand the high temperature process of regenerating
`soot-loaded particulate filters. A robust engine controi
`strategy
`that
`lessens
`the variability of
`the actual
`regeneration temperature is critical to the durability of the
`SCR. For this study, the target active DPF temperature
`has been determined to be 670"C. Also, the total
`cumulative duration for the full vel1icle useful life has
`been determined to be 64 hours. Therefore, the long(cid:173)
`term hydrothermal stabHiiy of base rnetal-zeolite SCR
`catalysts for typical light-duty diesel applications must be
`able to endure, at minimum, 670"C for 64 hours.
`
`Figure 3 shows the recent progress that has been made
`in Cu/zeo!ite SCR development. Many Cu/zeo!ite
`iorrnulations have been aged and evaluated betvveen
`Steady
`improvements of NOx
`2005 and 2007.
`conversion have been made in the low temperature
`range (200 - 350°C). For example, at 200°C, the NOx
`!n
`conversion has been enhanced from 70"/,, to 90%.
`addition, the 200"7 state·of-the··art SCR maintained 90%+
`NOx conversion over a much larger temperature range.
`However, note
`that
`the enhanced
`iow
`temperature
`activity came with a trade-off in the performance above
`400"C.
`
`1ll0
`
`SD
`
`80
`
`-~ ?()
`i:
`60
`0
`·~
`"' 50
`> c:
`0 u 40.
`x
`0
`z
`
`30
`
`20
`
`10
`
`NOx conversion is shown in Figure 2. From Figure 1,
`the NOx conversion traces are virtually line on line for
`operating temperatures beiow 600"C. Above 600"C, the
`NOx conversion drops slightly after each subsequent
`evaluation run. For this particular SCR formulation, the
`explanation for this slight deactivation has to do with the
`additional aging the sample experiences during high
`temperature performance evaluations.
`This
`trend
`becomes more apparent with data presented later in this
`paper. Figure 2 more clearly defines the variability in the
`overall now reactor system. The data from Figure 1 was
`manipulated in Minitab to yield the 95% confidence
`interval at each evaluation temperature. For evaluation
`temperatures below 600°C, the 95% confidence interval
`around the mean NOx conversion was better than ±2'?/c.
`Due to catalyst deactivation with testing, the higher
`temperature points showed variability up to ±6%.
`
`100
`
`90
`so
`7t1
`~
`60
`ao
`40
`
`" 0 '° '" >
`
`i::
`
`0 u
`"' 0 z
`
`30
`w
`1C ·
`O·
`
`-10
`
`·20
`
`-30
`100
`
`1 :SO WO
`
`250
`
`500
`450
`400
`350
`300
`Inlet Gas Temperature FC)
`
`550
`
`SOO
`
`SSO
`
`JOO
`
`FIGURE 1. NOx conversion results for the STANDARD SCR
`REACTION. Five consecutive evaluation runs on a single sample
`aged 64 hours at 6 70 "C.
`
`100
`
`150
`
`200
`
`350
`300
`250
`450
`~00
`ln!el Gas T empernlme (•C)
`
`500
`
`S50
`
`GOO
`
`Mean
`
`FIGURE 3. NOx conversion results for the STANDARD SCR
`REACTION. Best in class SCR ca1alys! formulations fmm 2005 --
`2007 after hydrothermal aging for 54 hours at 670"G.
`
`95% Confidence Lower Umil
`
`;7 ·!-. __________________ __ ,
`
`100
`
`150
`
`200
`
`250
`300
`350
`4'}0
`450
`500
`550
`Ave•age !rile! Gas Temperature (•C)
`
`600
`
`650
`
`?{}0
`
`FIGURE 2. Calculated from Figure ·1, the differential NOx conversion
`variability around the mean as determined by !he 95% confidence
`interval.
`
`480 SAE Int. J. Fuels Lubr. j Volume 1 I Issue 1
`
`formulations
`As shown in Figure 4, the three SCR
`generate measurable levels of N20 as a by-product.
`The N20 formation has a bi-modal profile as a function of
`low temperature N20
`temperature.
`The
`formation
`is a result of NH a oxidation by NO
`around 200 "C
`whereas the high temperature N20
`formation around
`525'C is mainly from the oxidation of NH3 by 0 2 . The
`latest SCR formulation generates much less N20. At
`200"C, the 200-7 SCR formulation yielded up to 3 fanes
`less N20 compared to the tvvo older iorrnulations.
`
`Exhibit 2002.004
`
`
`
`EXHIBITS TO DECLARATION OF STANLEY ROTH, PH,D, UNDER 37 C.F.R. § U32
`
`45
`
`40
`E' 35.
`"' "' -; 3!)
`" ~ 25
`E
`~ 20
`~ .. 15
`10.
`
`100
`
`1 SI)
`
`200
`
`250
`300
`'.J.50
`4.?0
`¢00
`Inlet G;;s Tempera!ure (•C)
`
`500
`
`55-0
`
`Q(U'J
`
`100
`
`90
`
`80
`
`'IS'
`"" 70
`s
`" 60
`" ~ 5(.1
`§
`" 40.
`I.<.
`" 0 z 30.
`
`20
`
`10
`
`0 ·.
`100
`
`"!5(1
`
`:!00
`
`3fiV
`350
`400
`450
`250
`Inlet Gas Tempem!um ("C)
`
`500
`
`550
`
`GOV
`
`FIGURE 4. N20 iormalion resu!ls for the STANDARD SCR
`REACTION in Figure 1. Best in class SCR catalyst formulations from
`200E) - 2007 after hydrothermal aging for 64 hours at 670 "C.
`
`FIGURE 6. NOx formation (ppm) results for the ammonia oxidation
`reaction in the absence of NOx (FIGURE 5). Best in class SCR
`catalyst 1ormuiations from 2005 - 2007 alter hydrothermal aging for 64
`hours at 670"C.
`
`Durable low temperature NOx performance ls desirable
`ior
`light-duty diesel applications.
`However, a
`considerable amount of NOx
`is emitted at high
`temperature during the time when the vehicle undergoes
`an active DPF regeneration. This added NOx emission
`must be compensated by additional NOx conversion
`during
`low
`temperature operation.
`As mentioned
`previously, the high temperature NOx performance of the
`2007 SCR catalyst drops sharply as the temperature
`increases beyond 400"C (Figure 3). Figure 5 plots the
`NHs oxidation of the three catalysts in the absence of
`NOx. Ammonia is more strongly oxidized by the c~007
`SCR catalyst.
`In addition, a clear inflection point at
`400"C is observed which corresponds to the formation of
`NOx (Figure 6). As a result, the NOx performance in
`Figure 3 declines rapidly due to, in part, the remake of
`NOx from NH3 oxidation (Figure 6}.
`
`100.
`
`~~KKK~K~KK
`
`90
`
`so
`
`70
`
`f~
`" £!)
`.S!
`'iii
`·o
`·;;:
`0
`£ z
`
`50
`
`40
`
`30
`
`Unrefined engine exhaust temperature control during
`DPF
`regeneration events
`coupled with
`inexact
`temperature measurement may expose SCR catalysts to
`an occasional unexpected oveHernperature.
`As a
`result,
`the SCR
`formulations are screened with a
`robustness
`test protocol consisting of hydrothermal
`exposure at 900 °t~ for i hour. These types of data are
`used to define the short-term never-to-exceed (NTE)
`temperature. The NTE testinhl provides a higher degree
`of discrimination among similar performing formulations
`compared to the less severe 64hri670 "C standard aging.
`
`Figure 7 iilustrates the remarkable progress that has
`been made in the past year with tl1e durability of the 2007
`Cu/zeolite based SGR formulation. Among the dozens
`of Cuizeolite formulations tested
`in past years, no
`formulation has been able to withstanding exposure up
`to 900"C while maintaining stable NOx performance at
`200"C. Under the 1 hour/900"C aging condition, the
`2007 SCR catalyst retained 90°/,, NOx conversion at
`200 "C. Ali older SCR formulations have achieved no
`better than 20% NOx conversion.
`The enhanced
`durabHity of the 2007 SCR formulations has been mainly
`to advances
`in
`the zeolite
`type and
`attributed
`composition.
`
`...
`
`(-
`
`:I
`
`20
`
`10
`
`FIGURE 5. NH3 conversion results ior the ammonia oxidation
`reaction in the absence of NOx. Best in class SCR cataiyst
`formulalions from 2005 - 2007 alter hydrothermal aging for 64 hours
`al 670"C.
`
`SAE In/. J Fuels Lubr. I Volume 1 I Issue 1
`
`481
`
`Exhibit 2002.005
`
`
`
`EXHIBITS TO DECLARATION OF STANLEY ROTH, PH.D. UNDER 37 C.F.R. § 1.132
`
`100
`
`so
`eo
`c 70
`" 0
`~
`" 50
`> " 0
`" 0
`<:
`
`(.) 40
`aa
`
`60
`
`20
`
`10
`
`0 -
`100
`
`1.?t'l
`
`WO
`
`3{.V
`4SO
`250
`&00
`3SC
`inlet Gas Temperature !'C)
`
`500
`
`.?50
`
`500
`
`FIGURE 7. NOx conversion of bes! in class SCR catalyst
`formulations from 2005 - 2007 after hydrothermal aging for 1 hour at
`900"C.
`
`Based on these encouraging results, a more severe
`time-at-temperature aging study was undertaken with the
`2007 SCR iorrnulation defined in Figure 7. The aging
`and evaluation helped determine the full robustness map
`of this promising Cuizeoiite based SCR formulation.
`
`TIME-AT-TEMPERATURE PERFORMANCE
`
`Current diesel engines require a DPF to meet the Tier 2
`particulate matter standards. As a result, the SCR must
`be abie to tolerate extreme temperature swings due to
`typica! and non-typical active DPF regenerations. Aiso,
`the durability requirement of the SCR formulation will
`hinge largely on the location of the SCR relative to the
`DPF. SCR formulations placed directly upstream must
`withstand high exhaust temperatures generated from the
`engine or over a DOC. SCR formulations placed directly
`downstream of the DPF wi!i need to withstand extended
`temperatures coming from soot regeneration. However,
`the most severe conditions will likely arise from future
`combination systems where the DPF filter is coated with
`a SCR
`formulation.
`Clearly,
`for promising SCR
`lorrnulatioris in close relationship to the DPF, there is a
`need to determine the full temperature based durability
`map to better assess the thermal robustness.
`
`time-at-temperature hydrothermal
`A broad range of
`aging was carried out to clearly define the full durability
`range: The aging time was varied frorn i hour to 256
`hours while the aging temperature was varied from
`670"C to i 100'C. The catalyst NOx performance was
`based on the "Standard" SCR reaction, the "Fast" SCR
`reaction
`(in
`the
`reaction, and ammonia oxidation
`absence of NOx).
`
`the
`For
`the
`light-duty driving conditions,
`typical
`"Standard" SCR reaction is considered to be the most
`
`48.2 SAE Int. J. Fuels Lul1r. I Volume 1 I Issue 1
`
`chalienging since little or no N02
`250"C.
`
`is expected below
`
`The effect of long-term hydrothermal aging at 700'C
`versus duration was carried out in the laboratory. The
`duration was varied from 1 hour to 256 hours while the
`aging temperature was held constant at 700'C, NOx
`conversion as a function of temperature was measured
`on each individual sample. Results show that the 2007
`Cu/zeoiite SCR catalyst demonstrated outstanding
`stability (Figure 8). Recall that tile 64 hour aging
`duration has been calculated to be an equivalent of
`120,000 miles for a typical ligtlt-duty diesel. As long as
`the SCR catalyst temperature does not exceed 700"C,
`these results heighten the industry's confidence for
`successfully implementing a Cu/zeolite SCR formulation
`into production.
`
`10!i
`so
`BO
`
`70 -
`~ eo -
`c
`so
`0
`"(;)
`40 -
`~
`,,
`" "'
`;;o
`u w
`"' 0
`10
`z
`
`l)
`
`-1!i-
`
`n21)
`
`~30
`100
`
`r+-;;;;--100-c----1
`i .~ ... tGh-; ?fJOC i
`! -llB-64h,.7uOC i
`b-2%h~_?OOC !
`
`""i 50
`
`200
`
`25!)
`
`400
`350
`300
`fHH}
`~sn
`Inlet Gas Tempern!ure (2C)
`
`!iStl
`
`SfJO
`
`650
`
`700
`
`FIGURE 8. NOx conversion results tor the ST AND ARD SCR
`REACTION. SCR samples hydrothermally aged at ?OO''C tor-; hour -
`256 hours.
`
`For configurations where the SCR catalyst is piaced
`upstream ol the DPF, temperatures greater than 700"C
`are not expected normally.
`However, higher
`temperatures may be apparent for SCR catalysts placed
`for SCR
`immediately downstream of a DPF and
`formulations coated on the DPF itself. Therefore, the
`effect of long-term hydrothermal aging for up to 256
`The results
`hours at 800"C was also conducted.
`showing
`the NOx conversion as a
`function of
`temperature are plotted in Figure 9. The Cu/zeoiite SCR
`catalyst proved to demonstrate durable NOx conversion
`up to 64 hours. Further aging out to 140 hours and then
`out to 256 hours resulted in a continual decline in the low
`temperature NOx conversion. The high temperature
`f\JOx conversion declines steadily from i hour to 64
`hours but then mildly improves from 64 !lours to 256
`the higll NOx conversion becomes
`hours. Recall,
`negative due to aggressive oxidation of NH3 with 0 2 to
`yield NOx.
`
`Exhibit 2002.006
`
`
`
`EXHIBITS TO DECLARATION OF STANLEY ROTH, PH,D, UNDER 37 C.F.R. § U32
`
`Given the observed hydrothermal aging !imitation oi 64
`hours at 800'C, these data provide encouragement for
`the potential development of a single combined
`SCR/DPF substrate where
`the SCR
`formulation
`is
`coated within the DPF substrate. This consolidation
`would provide smaller vehicle packaging and lower cost
`possibilities.
`
`Figure 11 shows the NOx conversion results for samples
`hydrothermally aged at 950"C for 1, 2, 4, and 8 hours.
`The Cu/zeolite formulation can only tolerate i hour
`exposure to 950'C. Tl1ere is a significant drop in
`performance af1er 2 hours and complete deactivation
`after only 4 hours of exposure.
`
`00
`
`lGQ r---==:a=~!§::'.~-------;-.:===::::;i
`-11>-1 h~ __ 950C i
`go .
`-~~!
`-111-4hr ___ S50C i
`~~~_r.,.~5_0('.: __ j
`
`70
`
`100
`
`90
`
`BO
`
`70
`
`60.
`
`50
`
`4Q
`
`30
`
`20
`10.
`
`0.
`
`l
`" " -~
`"' "' §
`" 0 z
`
`(.)
`
`·10
`
`.2()
`
`-30
`100
`
`r+·.;;;;-·e;;r;c;··1
`1-$-iBh~ __ EOOC
`·:
`I ·0· 32h•· BOOC !
`-lll-o4hr~eooc i
`
`1 ,50
`
`200
`
`:.:sv
`
`500
`450
`400
`350
`:JOO
`iniel Gas Temperature (~C)
`
`5SV
`
`tiM
`
`650
`
`700
`
`-30 -~---------------------!
`
`soo
`300
`350
`400
`450
`lnlet Gas Temperature ('C)
`
`550
`
`GOO
`
`fiStl
`
`700
`
`10-fJ
`
`~so
`
`200
`
`250
`
`FIGURE 9. NOx conversion results for the STANDARD SCR
`REACTION. SCR samples hydrothermal!y aged at 80Q~ for i hour -
`256 hours.
`
`FIGURE 11. NOx conversion resuits for the STANDARD SCR
`REACTION. SCR samples hydroti1ermaliy aged at -~~Q_".Q. for i hour -
`8 hours.
`
`For a i hour exposure, the sllorHerm never-to-exceed
`(NTE)
`temperature
`is defined as
`the maximum
`temperature the SCR formulation can tolerate without
`showing signs of significant deactivation. This
`is
`particularly helpful for use by engine control calibration
`engineers., Figure ·12 sllows the NOx conversion results
`for samples hydrothermally aged
`ior
`i hour wHh
`temperatures ranging from 700 °C to 1100 °C. The results
`indicated that the NTE temperature was 950°C but
`without much margin for error. For example, at 250"C,
`increasing the aging temperature from 950 ''C to 1000 "C
`decreased the NOx conversion from 95% to 18%. At
`1 "lOO'C, the NOx conversion was further reduced to 0"1o.
`It was clear tl1at structural damage occurred and further
`work to understand the deactivation will be carried out in
`the near future.
`
`Figure 10 shows the NOx conversion results for samples
`hydrothermally aged at 900'C for durations from 1 hour
`to 64 hours. Durable NOx conversion is observed up to 4
`hours. The Cu/zeolite cannot tolerate 8 hours and the
`performance is completely destroyed with the individual
`samples aged out to 16 hours and beyond.
`
`H}{}
`so.
`so
`70
`
`GO
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`-IO
`
`-20
`
`! -$-16!·n_9fJ>JC i
`• -lli-S4hr__B00~j
`
`l
`c
`0
`·~
`
`"' > "' ,,
`"' 0 :z
`
`0
`
`.,30
`IC>ll
`
`153
`
`2fi0
`
`250
`
`300
`lSO
`400
`453
`500
`lnlel Gas Tempernlure (•C)
`
`550
`
`GOO
`
`fiSO
`
`700
`
`FIGURE 10. NOx conversion results for the STANDARD SCR
`REACTION. SCR samples hydrothermally aged at 900"C ior i hour-·
`64 hours.
`
`SAE Int. J. Fuels Lubr. I Volume 1 I issue 1
`
`Exhibit 2002.007
`
`
`
`EXHIBITS TO DECLARATION OF STANLEY ROTH, PH.D. UNDER 37 C.F.R. § 1.132
`
`"Standard'' SCR Reaction but with slightly better activity
`at the low to moderate temperatures.
`
`1C!l.
`
`SO
`
`83
`
`70
`~
`" 0
`·~
`" ,.
`g
`"' 0 z
`
`40
`
`30
`w
`10
`
`(.)
`
`-~04hr_700C
`
`...,64hr_800C
`
`.....,64hr __ 850C
`
`15-0
`
`2M
`
`2SO
`
`.500
`400
`400
`350
`300
`inlet Gas Temperature (2C)
`
`?50
`
`GGn
`
`f).>0
`
`100
`
`t:
`
`mu
`so.
`eo ·
`::o.
`SQ
`~
`50.
`" ]?
`40.
`:;o.
`"' >
`20
`i:; "' u
`10
`"' 0 :z
`0
`-10
`-20
`-30
`-40
`-50 .
`10V
`
`FlGUF!E 12. Short-term Never-To-Exceed (NTE) Temperature: NOx
`conversion results for the STANDARD SCR REACTION. SCR
`samples hydroihermally aged at i hour from 700"C -110occ.
`
`100
`
`150
`
`200
`
`253
`
`500
`450
`4t>J
`350
`300
`Inlet Gas Temperature !'Ci
`
`550
`
`500
`
`GSiJ
`
`700
`
`FIGURE ·14. NOx conversion results for the FAST SCR REACTION.
`SCR samples hydrothermally aged at 64 hours from fJ70"C -- 900"C.
`
`Figure 13 shows the NOx conversion results for samples
`hydrothermally aged for 64 hours and temperature
`exoosure from 670"C to 900"C. For the baselirn~. 64
`ho'urs at 670"C aging (120k mi equivalent), the Cu/zeolite
`activity data clearly shows > 90°/o NOx conversion in the
`200"C - 350"C temperature window. However, the
`maximum temperature for tilis extended duration is
`800"C. An additional 50"C, corresponding to 850"C, had
`a severe impact on the catalyst durability.
`
`100
`su ·
`80
`·m
`
`-1>-S'iiir _70GC
`
`-111-64hr_800C
`
`~u.
`
`~ 60.
`c
`SC
`0 -ra
`lo
`>
`"' ()
`3l)
`() 20.
`x
`0
`z
`
`10
`
`-10
`
`-20.
`
`Inlet Gas Temperature ("C)
`
`FlGURE 13. NOx conversion results tor the STANDARD SCR
`REACTION. SCR samples hydrothermally aged at 64 hours from
`6?0"C --- 900'C.
`
`The same identical samples presented in Figures 8 - 13
`were further evaluated under two other SCR reactions.
`These reactions were the "Fast" SCR Reaction and
`"Ammonia Oxidation" Reaction in the absence of NOx.
`Comparison of Figure 13 and Figure 14, the data
`resulting from the "Fast" SCR Reaction experiments
`yielded the same trends in NOx performance as the
`
`484 SAE Int. J. Fuels Lubr. I Volume 1 l Issue 1
`
`illustrated in Figure 15, the ammonia oxidation
`As
`evaluation in tt1e absence of NOx yielded curves showing
`the deactivation of the catalyst with respect to time-at(cid:173)
`temperature. The results show a similar deactivation
`trend as the NOx performance.
`
`~~l;i{f:(,. 6'iOC
`
`#+~fNhr _ _7cnc
`
`-11!-64hr_SGOC
`
`"*"S4hr __ 850C
`
`.....64hr_900C
`
`mo
`
`~c
`
`80
`
`fi3 -
`
`~ ..
`"' 70
`i:;
`"'
`'§
`"' SC
`> " 0
`£
`z
`
`(.)
`
`40
`
`30
`
`20
`
`rn.
`o .L. ~lE1!iia:~~--~=-;:d::::::".:......,_~--1
`HiO
`HiO
`24!0
`250
`300
`3S!l
`400
`4511
`Sfl3
`StiO
`600
`f.50
`700
`Inlet Gas Temperature (2C)
`
`FlGURE 15. NH3 conversion results !or the AMMONIA OXiDATION
`REACTION in the absence of NCk SCR samples hydrothermally
`aged a1 64 hours from 670 "C - 900 "C.
`
`The results just described were for samples aged for 64
`hours but at varying temperatures, Furthermore, the
`corresponding samples aged at the various other aging
`conditions showed a similar deactivation trend as the
`corresponding NOx perrormance. As a result, these
`data sets of NOx conversion and NH3 oxidation have
`been excluded from this paper for brevity.
`
`Exhibit 2002.008
`
`
`
`~ 0.7 ---"lc---·+--·+--++-H1~,•H-·--------+----+·+--l--+L\'-l+-----·---1------~----'--'--1--U-i.I
`\
`~
`:;a O.G
`\
`1\\900· ....
`w
`1- o.5 ---·\t-----r--Tt:t-t-H--'1:--t--t--f-C-H-H-\--,;e~.-+--+++H+l
`w
`\
`\<IUUC
`"' .. ..
`\
`MD'•----------f+-f--+-t-+-H-l+----~+---+---i--++-H+l~------f----•!---+·-l--f-4-4-1-!
`~
`, .. L--4--4--H+L-\~------H!l!l-----1----i---'--L-1-Hl
`ill
`950'i
`""'
`~ 0.3
`---
`_____ ,,,,0
`. I
`0.1 !---t----+---~!-'i><H-i--H---+---+--+4+H1Hi·+-----i-l--+--!-l~
`ii I\ 1111
`ii
`~ii
`
`\
`
`\
`
`'
`
`•
`
`0 0.2
`
`; ; : :
`
`! \ .
`
`BET SURFACE AREA MEASUREMENTS
`
`The BET surface area of samples located in