Effects of Alkali and Alkaline Earth Cocations on the Activity and Hydrothermal Stability of Cu/SSZ-13 NH 3 -SCR Catalysts

Transcription

1 Effects of Alkali and Alkaline Earth Cocations on the Activity and Hydrothermal Stability of Cu/SSZ-13 NH 3 -SCR Catalysts Feng Gao, Márton Kollár, Yilin Wang, Nancy M. Washton, János Szanyi, Charles H.F. Peden InsAtute for Integrated Catalysis, Pacific Northwest NaAonal Laboratory, P.. Box 999, Richland, WA 99352, United States CLEERS Workshop, 4/29/215

2 Cu Loading Dilemma in Cu/CHA SCR Catalysts N x Conversion (%) Cu-SSZ-13, /Al = 6, GHSV = 4, h -1.65/.6.95/.8.198/ / / Reaction Temperature ( o C) Cu (wt%)/(cu/al Ratio) N x Conversion (%) Cu-SSZ-13, /Al = 6, GHSV = 4, h -1 Cu (wt%)/(cu/al Ratio) 1.31/ / / / / Reaction Temperature ( o C) For good low- temperature ac1vity, high Cu loadings (exchange levels) are required 2

3 Cu Loading Dilemma in Cu/CHA SCR Catalysts However, high Cu loading is detrimental to high- temperature hydrothermal stability Kim, et al., Journal of Catalysis 311 (214)

4 rigin on Cu Loading Effect to Hydrothermal Stability Kwak, et al., Chem. Commun., 48 (212) XRD/Maximum Entropy Andersen, et al., IUCrJ, 1 (214) DFT 4

5 rigin on Cu Loading Effect to Hydrothermal Stability Ø Cu(H) + species in CHA cage agglomerate to form Cu x clusters more readily than Cu MR during aging. Ø Cu x clusters promote destruc@on of the CHA framework at high temperatures! Ø An easy solu1on to the Cu loading dilemma: synthesize and u1lize Cu/CHA catalysts with intermediate Cu/Al ra1os. Ø Can we do even be[er? Can we manipulate Cu ion posi1oning at a constant Cu loading? Can we op1mize both Cu loading and Brønsted acid site density? Can we avoid dealumina1on during aging? 5

6 ne Possible Solution Use coca1ons for: Ø Space filling. Ø Neutralize Brønsted acidity. Ø Prevent dealumina1on. Ø Lower redox barriers. Ø 6

7 utline 1. Catalyst Synthesis. 2. Characteriza1ons: H 2 - TPR, NH 3 - TPD, 27 Al NMR, DRIFTS, XRD, etc. 3. NH 3 - SCR reac1ons. 4. Summary. 7

8 Catalyst Synthesis 1. Na/SSZ- 13 forma1on (/Al = 6) hydrothermally. 2. Solu1on ion- exchange to NH 4 /SSZ Solu1on ion- exchange to M/SSZ- 13 (M = Li +, Na +, K +, Cs +, Mg 2+, Ca 2+ ). 4. Solu1on ion- exchange to Cu,M/SSZ- 13 (~1. wt% Cu if full exchange). Ø Wash, dry, calcine in air at 55 C (Fresh); Ø Hydrothermal aging, 1% H 2 in zero air, 75 C for 16 h (HTA). Na/SSZ- 13 NH 4 /SSZ- 13 M/SSZ- 13 Cu,M/SSZ- 13 8

9 Catalyst Composition Elemental analysis via ICP Sample Cu Content (wt%) Coca1on Content (wt %) Ideal Stoichiometry Cu,H/SSZ Cu.3 H 4.8 Al P 72 Cu,Li/SSZ Cu.33 Li 1.25 H 3.5 Al P 72 Cu,Na/SSZ Cu.33 Na 1.67 H 3.1 Al P 72 Cu,K/SSZ Cu.32 K 2.33 H 2.4 Al P 72 Cu,Cs/SSZ Cu.25 Cs 2.86 H 2. Al P 72 Cu,Mg/SSZ Cu.24 Mg 1. H 2.9 Al P 72 Cu,Ca/SSZ Cu.33 Ca 1.23 H 2.3 Al P 72 9

10 NH 3 -TPD on M/SSZ-13 NH 4 /SSZ- 13 M/SSZ- 13 NH 3 Desorption gnal (ppm) Temperature ( o C) H/SSZ-13 Li/SSZ-13 Na/SSZ-13 K/SSZ-13 Cs/SSZ-13 Mg/SSZ-13 Ca/SSZ-13 6 o K + and Cs + completely neutralize Brønsted acidity. o Mg 2+ and Ca 2+ are the least efficient in removing Brønsted acid sites. 1

11 NH 3 -TPD and DRIFTS on Cu,M/SSZ-13 5 M/SSZ- 13 Cu,M/SSZ , Cu H 36/3575, (H) Al NH 3 Desorption gnal (ppm) 5 Fresh Samples Temperature ( C) Cu,H Cu,Li Cu,Na Cu,K Cu,Cs Cu,Mg Cu,Ca 6 Absorbance (a.u.) , H 3645, Al H Cu,H Cu,Li Cu,Na Cu,K Cu,Cs Cu,Mg Cu,Ca Frequency (cm 1 ) o Residual Brønsted acid site density: consistency between NH 3 - TPD and ν H. o NH 3 desorbs from Cu sites at ~32 C. o ν H in Cu- H at ~366 cm

12 H 2 -TPR on Cu,M/SSZ-13 H 2 Consumption (a.u.) Hydrated Dehydrated Temperature ( C) Cu,Ca Cu,Mg Cu,Cs Cu,K Cu,Na Cu,Li Cu,H Ø Hydrated samples: samples exposed to air and saturated with moisture. Ø Dehydrated samples: samples treated in an 2 /He flow at 55 C and cooled to RT in the same flow. Ø During dehydra1on, Cu 2+ compete favorably against H +, Li + and Na + for 6MR. Ø However, K +, Cs +, Mg 2+ and Ca 2+ prevent all Cu 2+ ions in finding the most favorable 6MR sites. Compete for the same sites and/or pore blocking? Ø Cu 2+ ions autoreduce if they cannot find the favorable sites. Cu /2H 2 = Cu + + H + Godiksen, et al., J. Phys. Chem. C 118 (214)

13 best Effects from hydrothermal aging: 27 Al NMR and XRD Cu, Cs Cu, K Cu, Na Cu, Mg Cu, Li Cu, Ca 58 ppm Intensity (Counts) Cu,H/SSZ-13 HTA, 75 o C, 16h Fresh Sample worst Cu, H Al Frequency (ppm from AlCl 3 (aq)) θ (Deg.) Ø All coca1ons prevent hydrolysis and dealumina1on of the Cu,M/SSZ- 13 catalysts be[er than Cu,H/SSZ- 13. Ø Even for Cu,H/SSZ- 13, no apparent degrada1on of the CHA framework in aging. 13

14 Effects from hydrothermal aging: NH 3 -TPD NH 3 Desorption gnal (ppm) HTA Samples Cu,H Cu,Li Cu,Na Cu,K Cu,Cs Cu,Mg Cu,Ca NH 3 Desorption gnal (ppm) 5 Fresh Samples Temperature ( C) Cu,H Cu,Li Cu,Na Cu,K Cu,Cs Cu,Mg Cu,Ca Temperature ( C) 6 For all of the catalysts, loss of Brønsted acidity occurs during hydrothermal aging. Loss of Brønsted acidity without apparent dealumina@on! 14

15 Effects from hydrothermal aging: hydrolysis pathways H H 2 H H 2 Al Al H + H H 2 Al H 2 H + H Al H H 2 H 2 Al(H) Al Frequency (ppm from AlCl 3 (aq)) 15

16 Effects from hydrothermal aging: H 2 -TPR.1 H 2 -TPR, HTA Samples H 2 Consumption (a.u.) Cu,Ca Cu,Mg Cu,Cs Cu,K Cu,Na Sample total Cu Content (wt%) Cu 2+ reduced to Cu + before 65 C (wt%) Percen tage (%) Cu,H/SSZ Cu,Li/SSZ Cu,Na/SSZ Cu,K/SSZ Cu,Cs/SSZ Cu,Li Cu,Mg/SSZ Cu,H Cu,Ca/SSZ Temperature ( C) Cu /2H 2 = Cu + + H + Cu 2+ + H 2 = Cu + 2H + H 2 - TPR consistent with 27 Al NMR. 16

17 Standard NH 3 -SCR Fresh Catalysts HTA Catalysts 1 Fresh Samples 1 HTA Samples N Conversion (%) Cu,H Cu,Li Cu,Na Cu,K Cu,Cs Cu,Mg Cu,Ca N Conversion (%) Cu,H Cu,Li Cu,Na Cu,K Cu,Cs Cu,Mg Cu,Ca Reaction Temperature ( C) Reaction Temperature ( C) Ø NH 3 light- off curves mimic the N light- off curves (i.e., excellent SCR selec1vi1es). T 5 ( C) Sample Cu,H Cu,Li Cu,Na Cu,K Cu,Cs Cu,Mg Cu,Ca Fresh HTA

18 Standard NH 3 -SCR Apparent Activation Energies (kj/mol) Cu/H Cu/Li Fresh Samples HTA Samples Cu/Na Cu/K Cu/Cs Cu/Mg Cu/Ca Cu, Cs Cu, K Cu, Na Cu, Mg Cu, Li Cu, Ca Cu, H 58 ppm Al Frequency (ppm from AlCl 3 (aq)) Ø For HTA samples, reac1on results indicate redistribu1on of K + and Cs + ca1ons in faces of 8MR to bring intracrystalline mass transfer limita1ons. 18

19 Intracrystalline mass transfer limitations E ob = E a + E D /2 Apparent Activation Energies (kj/mol) Cu/H Cu/Li Fresh Samples HTA Samples Cu/Na Cu/K Cu/Cs Cu/Mg Cu/Ca C. G. Hill, Jr., An Introduc1on to Chemical Engineering Kine1cs and Reactor Design, Wiley, NY, K + Cs + 19

20 Intracrystalline mass transfer limitations N x Conversion (%) Cu-SSZ-13, /Al = 6, GHSV = 4, h -1 Cu (wt%)/(cu/al Ratio) 1.31/ / / / / Reaction Temperature ( o C) E ob = E a + E D /2 E D silent at high Cu loadings! Gao, et al., Journal of Catalysis, 3 (213) 2 29; 319 (214) Cu(H) + [Cu- - Cu] 2+ 2

21 NH 3 storage effects N Conversion (%) HTA Samples Cu,H Cu,Li Cu,Na Cu,K Cu,Cs Cu,Mg Cu,Ca Reaction Temperature ( C) K Cs NH 3 Desorption gnal (ppm) 5 HTA Samples Temperature ( C) Cu,H Cu,Li Cu,Na Cu,K Cu,Cs Cu,Mg Cu,Ca 6 NH 3 storage important to high- temperature SCR rates, less significant to low- temperature SCR rates. 21

22 Summary (1) weaken between Cu- ions and the CHA framework. This allows Cu- ions to be posi1oned deeper in the CHA cages and become more readily reducible. During dehydra1on of Cu,M/SSZ- 13 samples, Li + and Na + do not prevent Cu 2+ ions from migra1ng to energe1cally most favorable sites while other coca1ons, to various extents, do prevent all Cu 2+ ions from residing in such sites. A por1on of Cu 2+ ions thus hydrolyze and even autoreduce to form Cu +. (2) Coca@ons help to improve hydrothermal stability of Cu/SSZ- 13. This is mainly due to removal of Brønsted acid sites that are most vulnerable for hydrolysis during hydrothermal aging. (3) Li + and Na + coca@ons promote low- temperature SCR ac@vity. This is assigned to modifica1ons to the redox and/or loca1ons of Cu- ion ac1ve centers by the coca1ons. Large sized K + and Cs + coca1ons appear to decrease accessibility of Cu- ion ac1ve centers by blocking pore openings of the catalysts as evidenced from decrease in apparent ac1va1on energies, due likely to intracrystalline mass transfer limita1ons. At elevated reac1on temperatures, catalysts with K + and Cs + coca1ons display decrease in N conversion due to the lack of high- temperature NH 3 storage in these catalysts. (4) From an applica1on point of view, this study demonstrates that introduc@on of a coca@on is a useful strategy in manipula@ng Cu/CHA catalysts for higher stability, higher selec1vity and wider opera1onal temperature window. F. Gao, M. Kollár, Y. Wang, N.M. Washton, J. Szanyi, C.H.F. Peden, ACS Catalysis, submifed. 22

23 Acknowledgements Ø DeNx research team at PNNL: Márton Kollár, Yilin Wang, János Szanyi, Chuck Peden, George Muntean, Mark Stewart. Ø DE user facility, EMSL, located at PNNL; and EMSL scien1sts Ravi Kukkadapu, Eric Walter and Nancy Washton. Ø DE ffice of Energy Efficiency and Renewable Energy/Vehicle Technologies ffice for the support of this work. Ø CLEERS for financial support for the fundamental studies. Ø Industrial partners from Cummins and Johnson- Ma[hey: Alex Yezerets, Krish Kamasamudram, Neal Currier, Jennifer Li, John Luo, Haiying Chen and H. Hess. 23