Investigation of Sulfur Deactivation on Cu/Zeolite SCR Catalysts in Diesel Application
Yisun Cheng, Christine Lambert
Ford Motor Company
Do Heui Kim, Ja Hun Kwak, Charles H.F. Peden
Pacific Northwest National Laboratory
August 4, 2009
2009 DEER Conference
Urea SCR Catalysts in Diesel Application
• Cu, Fe Zeolite catalyst
• Leading candidate for treatment of NOx emission for North America Diesel applications.
• Better options for Heavy-duty Diesel Engines
• Already commercialized in Europe
• Durability issues: – Dealumination and zeolite structure collapse due to
hydrothermal aging at high temperature
– Cu sintering due to the reducing agents
– Poisonings: such as S, P, Zn
2
SOx on Cu/zeolite SCR catalysts
• Sulfur poisoning is still a durability issue for base metal/zeolite SCR catalysts, especially for Cu/Zeolite SCR catalysts.
• Most studies have been based on SO2.
• As DOCs are employed upstream of the SCR catalysts, it is likely that a portion of the SO2 are oxidized into SO3.
• Investigation of the impact of SO3 on Cu/zeolite SCR catalysts is important.
3
Experimental
• Catalysts: – Fully formulated monolith Cu/zeolite.
– Six (6) 1x1 samples.
• Procedure: – Hydrothermal aging at 670C for 20 hr. – S poisoning:
• 40ppm for 1.5hr with SO2 or SO3 at
• 200°C, 300°C, 400°C
• Total S throughput equivalent to 500 miles with 350ppm sulfur fuel
– DeSOx: 170°C to 770°C at 5°C/min.
• Characterization: – XPS, Cu XAFS, Cu XANES
4
SCR Activity Testing Sequence
Hydrothermal Aging
Test #1: Baseline
Sulfation with SO2 or SO3
Test #2: Sulfation effect 770 oC
De-Sulfation with O2, H2O and CO2 up to 770C: TPD SOx
desorption
Test #3: DeSOx effect 200 oC
5
NOx Activity After Thermal Aging
100
90 Almost 100% conversion after 200C
80
70 Test1, Sample 1
60 Test1, Sample 2
50 Test1, Sample 3
40 Test1, Sample 4
30 Test1, Sample 5
20 Test1, Sample 6 10
0
100 125 150 175 200 225 250 275 300 325 350
Inlet Gas Temperature (ºC) 6
NO
x C
on
ve
rsio
n (
%)
300 325 350
SO2 vs. SO3 – The impacts on NOx conversion N
Ox C
on
ve
rsio
n (
%)
200C SO2, T2
300C SO2, T2
400C SO2, T2
200C SO3, T2
300C SO3, T2
400C SO3, T2
T1
Significant deactivation by SO3!
100
90
80
70
60
50
40
30
20
10
0
100 125 150
NOx Conversion SV = 30,000/hr - NO only
175 200 225 250 275
Inlet Gas Temperature (ºC) 7
275 300 325 350
NOx Activity Recovered after DeSOx
0
10
20
30
40
50
60
70
80
90
100
NO
x C
on
vers
ion
(%
)
Test1, after thermal aging
Test2, after SO3 aging at 400C
Test3, after deSOx
100 125 150 175 200 225 250
Inlet Gas Temperature (ºC) 8
SO2 vs. SO3 – Sulfur released during deSOx
SO2 Measurement During DeSOx
45 More SO2 desorption for the sample sulfated with SO3
40 SO2/200C/ afterT2
35 SO2/300C/ afterT2
SO2/400C/ afterT2 30
SO3/200C/ afterT2
SO3/300C/ afterT2 25 SO3/400C/ afterT2
20
15
10
5
0
150 250 350 450 550 650 7509
Inlet Gas T, C
SO
2, P
PM
10
Sulfated with SO3: XPS S 2p region
3500
3000
2500
2000
1500
1000
500
0
165 170 175 180
c/s
Hydrothermal aging
Sulfated with SO3 at 200C
Sulfated with SO3 at 300C
No observable sulfur for the
sample sulfated with SO2
Binding Energy (eV)
Only sulfates exists, only if sulfated with SO3. The sample sulfated with
SO2 does not contain sulfur on the catalyst.
Sulfates formed during reaction at 200C is larger than those at 300C.
NOx Conversion: sulfated with SO3 at different T
0
10
20
30
40
50
60
70
80
90
100
NO
x C
onvers
ion (
%)
Test2, SO3 aged at 200C
Test2, SO3 aged at 300C
Test2, SO3 aged at 400C
Test1, Before Sulfur Aging
100 125 150 175 200 225 250 275 300 325 350
Inlet Gas Temperature (ºC)
Not much difference in NOx conversion in spite of higher amount of sulfates over the sample aged at 200C than that at 300C.
11
After DeSOx: XPS S 2p region
Inte
nsity (
A. U
.)
Sulfated at 300 oC
Desulfated (300 oC)
Desulfated (200 oC)
175 170 165
Binding Energy (eV)
•Sulfur is completely removed after desulfation for both samples.
•Full recovery of NOx conversion after deSOx can be explained by
the complete removal of sulfate after deSOx.
12
13
ZCuSO4 Most stable adsorbed SOx species
better than ZCuSO3 or ZCuSO2
Can we identify the species
(structure) with XAFS?
XAFS (Xray absorption fine structure)
XAFS = XANES (Xray absorption nearedge spectroscopy) +
EXAFS (extended Xray absorption fine structure)
• EXAFS can give information about bond lengths, element and coordination number surrounding the atom.
• XANES yields information about the electronic structure of the absorbing atom, including valence and oxidation state.
• XAFS works for a wide variety of samples: amorphous and crystalline; solid, liquid, and gas; magnetic and nonmagnetic, etc..
14
15
How about the state of Cu?: Cu XANES study
Normalized spectra Derivative spectra
Not much change in Cu after DeSOx
Not like bulk CuSO4, but highly dispersed CuO
Cu XAFS: aged, sulfated with SO3, DeSOx
16
Informative, but the 2nd shell structure
is not obtained due to the noise signal.
Distance (nm)
0.0 0.1 0.2 0.3 0.4 0.5 0.6
FT
in
ten
sit
y (
a. u
.) Distance (nm) 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44
FT
in
ten
sit
y (a
. u
.)
Summary of XAFS curve fitting for Cu sample
Sample Pair CN1 r (nm)2 σσσσ2(pm2)3 ΔΔΔΔE (eV)
Cu aged Cu-O 3.5±±±±0.5 0.195±±±±0.001 57±±±±11 -3.4±±±±2.1
Cu DeSOx Cu-O 3.4±±±±0.3 0.195±±±±0.001 54±±±±8 -3.4±±±±1.4
Cu sulfated Cu-O 3.5±±±±0.3 0.195±±±±0.001 46±±±±6 -3.5±±±±1.2
1Coordination number. 2Coordination distance. 3The Debye-Waller factor accounting thermal and statistical
vibration. *The many body reduction factor was fixed to 0.9. The fitting ranges were 20 - 139 nm-1 for Δk and 0.100 – 0.300 nm for Δr, respectively. The restraint was applied to the Debye-Waller factor for the multiple scattering.
Only 1st shell information, which is the same among the samples, is
available due to the noisy signal.
17
Summary
• SCR activity was significantly reduced for samples poisoned
by SO3 compared with those by SO2, indicating that sulfur poisoning by SO2 and SO3 are not equivalent, with different poisoning mechanisms and impacts.
• Upon the sulfation with SO3, Sulfur exists as sulfate forms (not bulk CuSO4 form, but highly dispersed CuSO4), but maintain its highly dispersed Cu-O species during SOx and DeSOx, which can explain the reversible recovery of activity after desorption as SO2 at elevated temp.
• This study raises an important sulfur poisoning concern for the systems with DOCs in front of on Cu/zeolite SCR catalysts in diesel engine applications.
18
Acknowledgement • Clifford Montreuil and Giovanni Cavataio (Ford) for some testing.
• Prof. Sung June Cho (ChonNam Nat. Unv. of Korea) for analyzing XAFS.
• Funding from the U.S. Department of Energy (DOE), Freedom Car and Vehicle Technologies Program.
• NSLS for the beam time at X19A.
• Studies at PNNL were performed in the Environmental Molecular
Sciences Laboratory (EMSL), a National Scientific User Facility funded by U.S. DOE, Office of Science/Biological and Environmental Research.
• PNNL is a multi-program national laboratory operated for the U.S. Department of Energy by Battelle Memorial Institute under contract number DE-AC06-76RLO 1830.
19