Deactivation of Cu/Zeolite SCR Catalyst Due To Reductive
`Hydrothermal Aging
`
` 2008-01-1021
`
`Yinyan Huang, Yisun Cheng and Christine Lambert
`Ford Motor Company, Research and Advanced Engineering
`
`Copyright © 2008 SAE International
`
`ABSTRACT
`
`Temperature programmed reduction by CO, H2, and
`propylene (C3H6), as well as hydrothermal aging in the
`presence of mixture of NO, HC, CO, H2 and O2 were
`used
`to study
`the deactivation of Cu/zeolite SCR
`catalysts. The presence of CO had no detrimental effect
`on catalyst activity. Carbonaceous deposit on the catalyst
`surface from propylene (C3H6) reduction suppressed the
`catalyst activity and burn off of carbonaceous deposit
`recovered activity, the presence of O2 suppressed
`carbonaceous deposit formation. The presence of H2
`under lean conditions had much less effect on catalyst
`activity than H2 presence under rich conditions. Rich
`conditions with O2 presence represented the most
`detrimental effect on catalyst activity.
`
`INTRODUCTION
`
`Diesel powered vehicles have a bright future since the
`diesel engine provides higher
`fuel efficiency
`than
`gasoline powered vehicles. However, diesel engine
`exhaust streams have to meet future stringent emission
`regulations, which are especially challenging for nitrogen
`oxides (NOx) and particulate matter (PM). Selective
`catalytic reduction (SCR) of NOx with aqueous urea is
`considered as one of the primary approaches for NOx
`removal. Diesel particulate filters (DPFs) have been used
`for PM control. For urea SCR, aqueous urea is injected
`into the exhaust stream. It undergoes vaporization and
`hydrolysis to form ammonia for NOx reduction. For PM
`removal, a
`loaded particulate
`filter needs
`to be
`regenerated when it has a high soot loading. Different
`system configurations containing SCR and DPF for
`diesel exhaust aftertreatment have been reported. One
`system that contains a DOC (diesel oxidation catalyst),
`urea SCR, and DPF has been described (1-3). Another
`system containing the same catalyst devices, but having
`the DPF in front of the SCR (DOC-DPF-SCR) is also
`reported (4, 5). In addition, a combined SCR-DPF single
`brick system has been disclosed (6, 7). With the single
`brick system, some of the packaging constraints are
`relieved and the system is simplified.
`
`Under normal operating conditions, a diesel engine
`levels of O2 (lean
`exhaust stream contains high
`
`conditions). However, there are several situations when
`O2 can be low and even depleted, resulting in rich gas
`conditions. For example, hydrocarbons adsorbed on a
`catalyst during
`idling are
`released during hard
`accelerations. If the released hydrocarbons are ignited,
`O2 is consumed, resulting in low O2 levels in the gas
`stream or even rich conditions. This can change the
`chemistry of the catalyst and cause catalyst deactivation.
`During the DPF regeneration process, soot combustion
`results in O2 depletion and a high exotherm which can
`potentially cause deactivation of a downstream SCR
`catalyst or SCR coating on a DPF. In addition, there can
`also be a small amount of hydrogen in the exhaust
`stream. Under typical lean conditions CO/H2 ratio in the
`diesel engine exhaust stream is 40/1, much lower than
`the ratio of 3/1 in typical gasoline exhaust. However,
`under rich conditions, besides hydrocarbons (HC) and
`carbon monoxide (CO), hydrogen (H2) is a main reducing
`agent in engine exhaust stream.
`
`In the present paper, the performance of a typical
`Cu/zeolite SCR catalyst designed for light-duty vehicle
`use is studied after aging under reducing conditions. The
`reducing agents include propylene (C3H6), CO, H2, and
`their mixtures. The impacts of O2 and H2 concentrations
`are studied. The results elucidate the impacts of various
`reducing aging conditions on Cu/zeolite SCR catalyst
`durability.
`
`EXPERIMENTS
`
`1" diameter by 1" long cores were taken from a
`Cu/zeolite catalyst monolith. The catalyst was fully-
`formulated and designed for diesel vehicle use by a
`major catalyst supplier. The cores were aged and tested
`in a quartz flow reactor. The total flow for aging and
`testing was 6.44L/min, corresponding to a space velocity
`of 30,000 hr-1. Temperature was measured at the point
`of about 3/8" away from the inlet face of the catalyst.
`
`Catalyst degreening was carried out in the flow stream of
`5% CO2, 4.5% H2O, 14% O2, and balance N2 at 650oC
`for one hour. Hydrothermal aging of catalyst was carried
`out in the flow of 5% CO2, 4.5% H2O, 14% O2, and
`balance N2 at 650oC for 16 hours.
`
` SAE Int. J. Fuels Lubr. | Volume 1 | Issue 1
`
`466
`
`BASF-2036.001
`
`

`
`Degreened
`
`As-reduced-Run1
`
`Repeat-Run2
`
`250
`
`350
`Temperature (°C)
`
`450
`
`550
`
`100
`
`80
`
`60
`
`40
`
`20
`
`NOx Conversion (%)
`
`0
`150
`
`Fig. 1. SCR activities of fresh catalyst and catalyst
`reduced by 3000ppm CO
`
`2. Catalyst deactivation due to carbonaceous coke
`formation
`
`Figure 2 shows the catalytic activities of the degreened
`catalyst and catalyst reduced by 1000ppm C3H6 at 650oC
`for 1 hour. The as-reduced catalyst sample was dark in
`color and had lower activity than the degreened catalyst
`(Run1). After Run1 test, the color of the catalyst became
`lighter and turned to grey. As shown in Figure 2, the
`second run (Run2) catalyst activity was almost identical
`to the degreened catalyst.
`
`Degreened
`
`As-Reduced-Run1
`
`Repeat-Run2
`
`250
`
`350
`Temperature (°C)
`
`450
`
`550
`
`100
`
`80
`
`60
`
`40
`
`20
`
`NOx Conversion (%)
`
`0
`150
`
`Fig. 2. SCR activities of fresh catalyst and catalyst
`reduced by 1000ppm C3H6
`
`The results indicated that the reduction of Cu/zeolite
`catalyst by C3H6 results in the formation of carbonaceous
`coke formation on the catalyst surface. The surface
`carbonaceous deposit blocked the active sites and
`suppresses the catalyst activity. These results were
`consistent with literature report (10). During the Run1
`activity testing, 14% O2 was present and the highest
`to 600oC. As a result,
`the
`temperature was up
`carbonaceous deposit was burned off and activity was
`recovered. Therefore, the catalyst deactivation due to
`carbonaceous coke formation is recoverable once the
`deposit is burned off.
`
`3.
`
`Impact of O2 presence on catalyst deactivation
`
`Catalysts were aged under reaction conditions in the
`presence of NO, C3H6, CO and O2. Without the presence
`
`To study the impact of CO and hydrocarbon, catalysts
`were aged in the reduction gas stream. The temperature
`was ramped up at 5oC/min to 650oC, held at 650oC for
`one hour, followed by natural cooling. The reduction gas
`stream contained 5% CO2, 4.5% H2O, balance N2, and
`either 3000ppm CO or 1000ppm C3H6 respectively.
`
`To study the impact of O2 level, catalysts were aged in
`different reductive gas streams at 650oC for 16 hours.
`The temperature was ramped up at 5oC/min to 650oC,
`held at 650oC for 16 hours, followed by natural cooling.
`The gas stream contained 5% CO2, 4.5% H2O and
`specified level of O2. Reducing agents were added
`during the time period of 16 hours at 650oC. The
`reductive gas stream contained 5% CO2, 4.5% H2O,
`350ppm NO, 1000ppm C3H6, 2% CO and O2, at levels 0,
`1% and 3%.
`
`To study the impact of H2, catalysts were aged in
`reducing gas streams containing hydrogen. The
`temperature was ramped up at 5oC/min to 650oC, held at
`650oC for 16 hours, followed by natural cooling. The gas
`stream contained 5% CO2, 4.5% H2O and a specified
`level of O2. Reducing agents were added during the time
`period of 16 hours at 650oC. The reductive gas stream
`contained 5% CO2, 4.5% H2O, 350ppm NO, 1000ppm
`C3H6, 2% CO, H2 and O2. Two O2 levels, i.e. 1% and 3%
`were studied. At each O2 level, three H2 levels, i.e. 0,
`500ppm and 0.67% were included. 1% O2 is equivalent
`to a rich condition of Ȝ=0.98. 3% O2 is equivalent to a
`lean condition of Ȝ=1.1.
`
`After aging, all catalysts were evaluated for steady state
`activities. The testing gas stream contained 350ppm NO,
`350ppm NH3, 5% CO2, 4.5% H2O, and balance N2. The
`testing was performed from low temperature to high
`temperature (up to 600oC). After the first test (Run1), the
`catalyst was cooled down in a gas flow with 14% O2.
`Then, the activity testing was repeated and called Run2.
`No special cleaning procedure was used between two
`runs.
`
`RESULTS AND DISCUSSIONS
`
`1.
`
`Impact of CO reduction on catalyst deactivation
`
`Figure 1 shows the activities of a degreened catalyst,
`catalyst as-reduced by 3000ppm CO at 650oC for 1 hour
`(Run1) and the repeat steady state testing (Run2). The
`results show
`that CO as-reduced catalyst had
`comparable activity to the degreened catalyst. The
`consecutive repeat testing shows no apparent activity
`change.
`
`Reduction of Cu/ZSM5 catalyst by CO at high
`temperature has been reported in literature (8, 9).
`Though Cu2+ can be reduced to Cu1+ by CO in inert
`atmosphere, Cu1+ is readily oxidized into Cu2+ once O2 is
`available. This process is reversible and it has no
`negative impact on Cu2+ dispersion. As a result, CO
`reduction has no impact on catalyst deactivation.
`
`SAE Int. J. Fuels Lubr. | Volume 1 | Issue 1
`
`467
`
`BASF-2036.002
`
`

`
`cages and agglomerated to cause more deactivation,
`same as reported in literature (11, 12).
`
`HT Aged
`
`0%O2-Run2
`
`1%O2-Run2
`
`3%O2-Run2
`
`250
`
`350
`Temperature (°C)
`
`450
`
`550
`
`100
`
`80
`
`60
`
`40
`
`20
`
`NOx Conversion (%)
`
`0
`150
`
`Fig. 4. SCR activities of hydrothermally aged catalyst and
`second run of catalyst aged with 350ppm NO,
`1000ppm C3H6, 2% CO at different O2 levels at
`650oC for 16 hours
`
`4.
`
`Impact of H2 on catalyst deactivation
`
`Impact of H2 presence on catalyst activities was
`conducted at 1% O2 (rich) and 3% O2 (lean) levels
`respectively. The aging gas stream contained NO, C3H6,
`CO and H2. It was found that H2 presence under rich
`conditions had different effect on catalyst activity from H2
`present under lean conditions.
`
`4.1. H2 presence under rich conditions
`
`Catalysts were aged under reaction conditions with
`350ppm NO, 1000ppm C3H6, 2% CO, H2 and 1% O2. H2
`levels were 0, 500ppm and 0.67%. Under
`these
`conditions, the aging stream was rich and equivalent to
`Ȝ=0.98. During
`the aging process, a significant
`temperature increase across the catalyst was generated.
`lead
`to significantly higher
`H2 presence did not
`temperature increase, compared to aging with C3H6 and
`CO present.
`
`All as-aged catalysts were grey in color, indicating no
`carbonaceous deposit
`formation during
`the aging
`process. The presence of O2 suppressed carbonaceous
`deposit formation.
`
`Figure 5 shows the activities of catalysts as-reducing-
`aged with 1% O2 present (Run1). Comparing with
`hydrothermally aged catalyst, all catalysts as-reduced at
`1% O2 showed obvious deactivation. Zero H2 and
`500ppm H2 resulted in similar catalyst deactivation effect.
`In contrast, 0.67% H2 presence during aging resulted in
`the most significant catalyst deactivation.
`
`Figure 6 shows the results of repeat testing (Run2) which
`had no impact on activity recovery. Zero H2 and 500ppm
`H2 had almost the same activities while 0.67% H2 still
`had much lower activity. The results indicated 500ppm
`H2 presence did not have an obvious impact on catalyst
`deactivation. However, a high concentration of H2
`(0.67%) generated significant catalyst deactivation.
`
`of NO, C3H6 and CO during the aging process, there was
`no exotherm generation. The temperature difference
`between the catalyst inlet and outlet was less than 10oC.
`When NO, C3H6 and CO were present and O2 was
`absent during the aging process there also was no
`exotherm. Catalyst had a dark color after aging. When
`NO, C3H6, CO and O2 were present during the aging
`process, a significant exotherm was generated. The
`temperature increase across the catalyst was in the
`range of 60-80oC. Catalysts were grey in color after
`aging.
`
`Figure 3 shows the activity of catalysts as-aged with NO-
`(Run1).
`In comparison with a
`CO-C3H6 present
`hydrothermally aged catalyst,
`there were obvious
`deactivations with reducing-aged catalysts. Aging with
`C3H6 and CO present without O2 resulted in the most
`significant deactivation. In contrast, aging with NO-CO-
`C3H6 and O2 present resulted in less deactivation. Rich
`aging with 1% O2 presence (Ȝ=0.98) generated more
`significant deactivation than lean aging when 3% O2 was
`present (Ȝ=1.1).
`
`HT Aged
`
`0%O2-As-aged
`
`1%O2-As-aged
`
`3%O2-As-aged
`
`250
`
`350
`Temperature (°C)
`
`450
`
`550
`
`100
`
`80
`
`60
`
`40
`
`20
`
`NOx Conversion (%)
`
`0
`150
`
`Fig. 3. SCR activities of hydrothermally aged catalyst and
`catalyst aged with 350ppm NO, 1000ppm C3H6,
`2% CO at different O2 levels at 650oC for 16
`hours
`
`After the first activity testing, the catalysts were grey in
`color and were tested again (Run2). Figure 4 shows the
`activities of the second run (Run2). As with the catalyst
`aged with C3H6 and CO present and O2 absent, repeated
`testing (Run2) showed complete activity recovery. As
`described in the previous section, catalyst deactivation in
`this case was caused by carbonaceous coke formation
`which suppressed the catalyst activity. First activity
`testing (Run1) with 14% O2 presence up to 600oC
`burned off carbonaceous deposit and cleaned catalyst
`surface. As results, the activity was fully recovered.
`
`When O2 was present during the aging, hydrocarbon
`(C3H6) and CO were ignited on the catalyst. The
`temperature increase across the catalyst indicated that
`the outlet section of the catalyst was exposed to high
`temperature, which
`caused
`agglomeration
`and
`deactivation. Compared to lean aging at 3% O2 level, rich
`aging at 1% O2 level resulted in a reduction of Cu2+ to
`Cu0. The resulting Cu0 species migrated out of zeolite
`
` SAE Int. J. Fuels Lubr. | Volume 1 | Issue 1
`
`468
`
`BASF-2036.003
`
`

`
`HT Aged
`
`0%H2-As-aged
`
`500ppmH2-As-aged
`
`0.67%H2-As-aged
`
`250
`
`350
`Temperature (°C)
`
`450
`
`550
`
`100
`
`80
`
`60
`
`40
`
`20
`
`NOx Conversion (%)
`
`0
`150
`
`Fig. 7. SCR activities of hydrothermally aged catalyst and
`catalysts aged at Ȝ=1.1 with 3% O2, 350ppm NO,
`1000ppm C3H6, 2% CO and various H2 levels
`
`HT Aged
`
`0%H2-Run2
`
`500ppmH2-Run2
`
`0.67%H2-Run2
`
`250
`
`350
`Temperature (°C)
`
`450
`
`550
`
`100
`
`80
`
`60
`
`40
`
`20
`
`NOx Conversion (%)
`
`0
`150
`
`Fig. 8. SCR activities of hydrothermally aged catalyst and
`second run activities of catalysts aged at Ȝ=1.1
`with 3% O2, 350ppm NO, 1000ppm C3H6, 2% CO
`and various H2 levels
`
`Figure 7 shows the activities of catalysts as-aged at 3%
`level. All as-aged catalysts showed catalyst
`O2
`deactivation. Zero H2 and 500ppm H2 resulted in little
`difference while 0.67% H2 resulted in slightly more
`catalyst deactivation. Slight activity
`recovery was
`observed for the zero and 500ppm H2 aged samples,
`and no activity recovery was observed for the 0.67% H2
`aged one, as shown in Figure 8 during the second run
`catalyst activities. Again zero H2 and 500ppm H2 showed
`little difference while the presence of high concentration
`H2 (0.67%) during the aging generated higher level of
`catalyst deactivation. However, compared to rich aging,
`the
`lean aging
`results
`indicated
`that
`lean aging
`generated less catalyst deactivation.
`
`During the lean aging process with 3% O2, C3H6, CO and
`H2 were ignited on the catalysts. Excess O2 led to
`complete combustion of H2, CO and C3H6. Thus, the
`deactivation due to Cu2+ reduction had no apparent
`impact on catalyst deactivation and the exposure to high
`temperature during the aging becomes the major reason
`for catalyst deactivation.
`
`Most diesel exhaust aftertreatment systems contain a
`DOC to promote the formation of NO2. It has been
`known that presence of NO2 has enhanced SCR activity
`
`Catalyst deactivation during rich aging could not be
`recovered.
`
`HT Aged
`
`0%H2-As-aged
`
`500ppmH2-As-aged
`
`0.67%H2-As-aged
`
`250
`
`350
`Temperature (°C)
`
`450
`
`550
`
`100
`
`80
`
`60
`
`40
`
`20
`
`NOx Conversion (%)
`
`0
`150
`
`Fig. 5. SCR activities of hydrothermally aged catalyst and
`catalysts as-aged at Ȝ=0.98 with 1% O2, 350ppm
`NO, 1000ppm C3H6, 2% CO and various H2 levels
`
`HT Aged
`
`0%H2-Run2
`
`500ppmH2-Run2
`
`0.67%H2-Run2
`
`250
`
`350
`Temperature (°C)
`
`450
`
`550
`
`100
`
`80
`
`60
`
`40
`
`20
`
`NOx Conversion (%)
`
`0
`150
`
`Fig. 6. SCR activities of hydrothermally aged catalyst and
`second run activities of catalysts aged at Ȝ=0.98
`with 1% O2, 350ppm NO, 1000ppm C3H6, 2% CO
`and various H2 levels
`
`The exposure to high temperature and reduction of Cu2+
`were responsible for catalyst deactivation during the rich
`aging process. H2 presence during aging provided
`low H2 concentration
`additional deactivation. At
`(500ppm), H2 was ignited and consumed quickly in the
`catalyst inlet face resulting in less impact on catalyst
`deactivation. With high concentration H2 (0.67%), a
`larger catalyst section was reduced, resulting in more
`catalyst deactivation.
`
`4.2. H2 presence under lean conditions
`
`In this section, catalysts were aged under reaction
`conditions with the presence of 350ppm NO, 1000ppm
`C3H6, 2% CO, H2 and 3% O2. H2 levels were 0, 500ppm
`and 0.67%. Under these conditions, the aging stream
`was lean, equivalent to Ȝ=1.1. During the aging process,
`significant
`temperature across
`the catalyst was
`generated due to hydrocarbons, CO and H2 ignition. H2
`presence did not lead to significantly higher temperature
`increase. All as-aged catalysts at 3% O2 level were grey
`in color, indicating no carbonaceous deposit formation.
`
`SAE Int. J. Fuels Lubr. | Volume 1 | Issue 1
`
`469
`
`BASF-2036.004
`
`

`
`12. J.Y. Yan, G.D. Lei, W.M.H. Sachtler and H.H. Kung,
`"Deactivation of Cu/ZSM-5 Catalysts for Lean NOx
`Reduction: Characterization of Changes of Cu State
`and Zeolite Support", J. Catal., 1996-161-43
`13. M. Koebel, M. Elsener, G. Madia, "Recent Advances
`in the Development of Urea-SCR for Automotive
`Applications", SAE 2001-01-3625
`14. M. Koebel, M. Elsener, T. Marti, "NOx-Reduction in
`Diesel Exhaust Gas with Urea and Selective
`Catalytic Reduction", Combust. Sci. Tech. 1996-
`121- 85
`
`compared with NO-only activity, specifically in the low
`temperature region [13, 14]. The current work used the
`NO-only reaction for activity testing to represent the
`highest impact of catalyst aging. If NO2 was included and
`a mixture of NO/NO2 reduction was tested, some of the
`effects of aging could be masked.
`
`CONCLUSIONS
`
`The reduction of a fully-formulated Cu/zeolite SCR
`catalyst by CO had no impact on catalyst performance.
`Reduction by propylene formed a carbonaceous deposit
`that suppressed NOx activity, but the activity was
`recovered after burn-off of the deposit. Presence of O2
`suppressed the carbonaceous deposit formation. H2
`presence under lean conditions had little or no impact on
`catalyst deactivation. High concentration of H2 under rich
`conditions generated extra catalyst deactivation
`in
`addition to thermal deactivation. Rich conditions with O2
`present
`resulted
`in
`the most significant catalyst
`deactivation in this study. Therefore, exposure of
`Cu/zeolite SCR catalysts to low O2 levels and rich
`conditions should be avoided.
`
`REFERENCE
`
`1. P. Tennison, P. Laing, C. Lambert, R. Hammerle, W.
`Ruona, "Exhaust gas aftertreatment", US Patent
`6,928,806
`2. Paul Tennison, Christine Lambert, Michael Levin,
`"NOx control development with urea SCR on a diesel
`passenger car", SAE 2004-01-1291
`3. G. Cavataio, J. Girard, J. Patterson, C. Montreuil, Y.
`Cheng, C. Lambert, "Laboratory Testing of Urea-
`SCR Formulations to Meet Tier 2 Bin 5 Emissions",
`SAE 2007-01-1575
`4. R. Kakwani, K. Voss, J. Patchett, K. Grimston,
`"Exhaust system for enhanced reduction of nitrogen
`oxides and particulates from diesel engines", US
`Patent 7,143,578
`5. P. Blakeman, M. Twigg, "Exhaust system for lean
`burn IC engines", US Patent 7,178,331
`6. J. Patchett, J. Dettling, E. Przybylski, "Catalyzed
`SCR filter and emission treatment system", US
`Patent 7,229,597
`7. R. Hammerle, C. Lambert, "Diesel engine after
`treatment device for conversion of nitrogen oxide
`and particulate matter", US Patent 7,225,613
`8. V.B. Kazansky and E.A. Pidko, "A new insight in the
`unusual adsorption properties of Cu+ cations in Cu-
`ZSM-5 zeolite", Catal. Today, 2005-110-281
`9. J. Sárkány, "IR study of partial reduction of over-
`exchanged, O2-calcined CuII-ZSM-5 using flowing
`CO/He at stepwise increasing temperatures, Part 1",
`Surf. Sci., 2004-566-723
`10. J.L. d'Itri and W.M.H. Sachtler, "Selective reduction
`of nitric oxide over Cu/ZSM-5: The role of oxygen in
`suppressing catalyst deactivation by carbonaceous
`deposits", Appl. Catal. B: Environ., 1993-2-L7.
`11. G.D. Lei, B.J. Adelman, J. Sárkány and W.M.H.
`Sachtler, "Identification of copper(II) and copper(I)
`and
`their interconversion in Cu/ZSM-5 De-NOx
`catalysts", Appl. Catal. B: Environ., 1995-5-245
`
` SAE Int. J. Fuels Lubr. | Volume 1 | Issue 1
`
`470
`
`BASF-2036.005