Global kinetic modelling of the reaction of soot with NO x and O 2 on Fe 2 O 3 catalyst

Global kinetic modelling of the reaction of soot with NO x and O 2 on Fe 2 O 3 catalyst Dirk Reichert, Henning Bockhorn, Sven Kureti Germany 1

Techniques for catalytic removal of NO x from diesel exhaust Selective Catalytic Reduction (SCR) Catalyst: V 2 O 5 /WO 3 /TiO 2 or Fe based zeolites 4 NO + 4 NH 3 + O 2 4 N 2 + 6 H 2 O NO x Storage Reduction (NSR) Catalyst Basic storage components (BaCO 3, Al 2 O 3, CeO 2 ), noble metals (Pt/Rh) Lean phase: NO + 0,5 O 2 NO 2 NO 2 + BaCO 3 Nitrate Rich phase: Nitrate + HC, CO, H 2 BaCO 3 + N 2, CO 2, H 2 O 2

Removal of soot from diesel exhaust Separation of soot by Diesel Particulate Filters (DPF) Pt catalyst DPF DPF regeneration Fast regeneration by NO 2 (T = 200 450 C) C + 2 NO 2 2 NO + CO 2 Induced regeneration by O 2 C + O 2 CO 2 - engine management (T > 600 C) - fuel additives (Fe, Ce; T > ca. 300 C) - catalytic coating (Fe, Ce; T > ca. 450 C) 3

Alternative concept: Simultaneous conversion of soot and NO x 2 NO + C N 2 + CO 2 Catalysts: oxide systems (Fe, Cu) Y. Teraoka, S. Kagawa, W.F. Shangguan, Appl. Catal. B 5 (1995) L181 K. Hizbullah, W. Weisweiler, S. Kureti, Appl. Catal. B 43 (2003) 281 Catalytic soot/no x conversion might be used as additional deno x procedure X(soot) / % X(NO x ) / % bare DPF bare DPF c(no) = 500 ppm n(c)/n(no x ) = 0.4 c(o 2 ) = 7.0 vol.% c(co 2 ) = 7.5 vol.% c(h 2 O) = 7.5 vol.% S.V. = 20.000/h P. Balle, H. Bockhorn, B. Geiger, N. Jan, S. Kureti, D. Reichert, T. Schröder, Chem. Eng. Process. 45 (2006) 1065 4

Laboratory bench for production of soot and evaluation of catalytic DPF systems Characterisation of the soot 2.6 wt.% adsorbed species 98.8 wt.% C 0.7 wt.% O 0.5 wt.% H 0 wt.% N S BET = 91 m 2 /g most frequent diameter: ca. 50 nm Model catalyst: α-fe 2 O 3 P. Balle, H. Bockhorn, B. Geiger, N. Jan, S. Kureti, D. Reichert, T. Schröder, Chem. Eng. Process. 45 (2006) 1065 5

Effect of the Fe 2 O 3 catalyst on soot/no x /O 2 reaction without catalyst with Fe 2 O 3 catalyst 28 μmol N 2 23 μmol N 2 O 83 μmol N 2 14 μmol N 2 O n(co x ) n(n 2 )+n(n 2 O) =320 n(co x ) n(n 2 )+n(n 2 O) =108 c(no)=500 ppm total flow: 500 ml/min W/F=0.39 gs/ml c(o 2 )=6.0 vol.% n(fe 2 O 3 )=20 mmol ΔT/Δt= 90 K/h Ar balance n(c)=10 mmol 6

Temporal separation of NO/O 2 /soot reaction without Fe 2 O 3 T c(co),c(co 2 ) / vol.% I O 2 II III NO IV V CO 2 CO N2 I: C + O 2 C(O y ) (X(C) = 50 %) II, III: C(O y ) C f + CO x IV: C* + NO N 2 + C(N) inhibition of NO reduction, N accumulation V: C + O 2 CO x C(N) + O 2 N 2 + NO n(c) = 10 mmol (0,12 g), Ar balance, F = 500 ml/min I: c(o 2 ) = 6 Vol.-%; II: Argon; III: Argon, β = 10 K/min; IV: c(no) = 500 ppm; V: c(o 2 ) = 6 vol.%, β = 10 K/min. 7

Temporal separation of NO/O 2 /soot reaction on Fe 2 O 3 catalyst I II III IV I: C + O 2 C(O y ) (X(C) = 50 %) II: C(O y ) C f + CO x III: C* + NO N 2 + C(N) inhibition of NO reduction, N accumulation Conditions: n 0 (C) = 10 mmol (0,12 g), Ar balance, F = 500 ml/min; I: c(o 2 ) = 6 Vol.-%; II: Argon; III: c(no) = 500 ppm; IV: c(o 2 ) = 6 vol.%, β = 10 K/min. D. Reichert, H. Bockhorn, S. Kureti, Appl. Catal. B 80 (2008) 248 IV: C + O 2 CO x C(N) + O 2 N 2 + NO 8

Model for N 2 formation in soot/no reaction Model bases on transient experiments and quantum-mechanical calculations Also possible: T. Kyotani, A. Tomita, J. Phys. Chem. B 103 (1999) 3434 M. Sander, A. Raj, O.R. Inderwildi, M. Kraft, S. Kureti, H. Bockhorn, Carbon 47 (2009) 866 appropriate orientation of molecular orbitals of soot surface is crucial oxygen regenerates active surface configuration 9

Development of the contact between Fe 2 O 3 catalyst and soot (HRTEM study) 20 nm 10 nm 10 10 nm 10 nm X(soot) = 0% X(soot) = 90% H. Bockhorn, S. Kureti, D. Reichert, Top. Catal. 42-43 (2007) 283

TPO with isotope labelled oxygen ( 18 O 2 ) Conditions: c(o 2 )=0.19 vol.% N 2 balance n(fe 2 O 3 )=3.8 mmol n(c)=1.9 mmol ΔT/Δt=600 K/h D. Reichert, H. Bockhorn, S. Kureti, Appl. Catal. B 80 (2008) 248 11

Role of the Fe 2 O 3 catalyst Fe 2 O 3 catalyst acts as oxygen pump transfer of O from the contact points of Fe 2 O 3 and soot contact maintains up to high conversion levels of the soot soot/no reaction is not directly affected by the catalyst D. Reichert, H. Bockhorn, S. Kureti, Appl. Catal. B 80 (2008) 248 G. Mul, F. Kapteijn, C. Doornkamp, J.A. Moulijn, J. Catal. 170 (1998) 258 12

Global kinetic modelling Potential law: r(co x ) = k COx n(c f ) c(o 2 ) n O 2 k COx = r(no) = k NO g(x) n(c f ) c(no) n NO k NO = A CO x e A E A,COx RT NO x e E A,NO RT Determination of n(c f ) : From N 2 physisorption (BET): S(X) = S 0 m 0 (1-X) (1+300 X) 1/3 From TPD with X(C) = 0, 25, 50, 75%: X(soot) / % n(co x) des S(X) = 8,7 µmol m 2 n(c f ) = λ S(X) = λ λ: Surface concentration of C f (λ λ(x)) 13

Observation in transient studies: Only a part (C*) of the C f sites is involved in NO reduction n(c*) = g(x) n(c f ) g(x) = 0,13 e 1,4 X X(soot) / % 14

Global kinetic modelling of soot/no x /O 2 reaction (without catalyst) E A, COx = 115 kj/mol (gradient free loop reactor) A COx = 7 10 6 m 3 /(mol s) (transient experiments) n O2 =1 J.P.A. Neeft et al., Fuel 76 (1997) 1129 n NO =1 P. Chambrion et al., Energy Fuel 12 (1998) 416 From fitting calculations: E A,NO = 56 kj/mol A NO = 4 10 3 m 3 /(mol s) c(co x ) r(no) X(C) experimental data are described well by kinetic model c(co x ) model r(no) model X(C) model n(c*) 15

Global kinetic modelling of soot/no x /O 2 reaction on Fe 2 O 3 Taken from non-catalytic conversion: = 56 kj/mol E A,NO From fitting calculations: A NO E A, COx = 85 kj/mol A COx = 5.0 10 3 m 3 /(mol s) = 1.8 10 6 m 3 /(mol s) c(co x ) r(no) X(C) c(co x ) model r(no) model X(C) model n(c*) experimental and modeled data are in fair agreement D. Reichert, T. Finke, N. Atanassova, H. Bockhorn, S. Kureti, Appl. Catal. B 84 (2008) 803 16

Model validation by simulating catalytic TPO profiles Variation of O 2 (10 vol.%) and soot (0.01g) c(no)=500 ppm total flow: 500 ml/min W/F=0.39 gs/ml c(o 2 )=10 vol.% n(fe 2 O 3 )=20 mmol ΔT/Δt= 90 K/h Ar balance 17

Model validation by simulating catalytic TPO profiles Variation of O 2 (10 vol.%) and soot (0.01g) c(no)=500 ppm total flow: 500 ml/min W/F=0.39 gs/ml c(o 2 )=10 vol.% n(fe 2 O 3 )=20 mmol ΔT/Δt= 90 K/h Ar balance 18

Variation of O 2 (14 vol.% O 2 ) c(no)=500 ppm total flow: 500 ml/min W/F=0.39 gs/ml c(o 2 )=14 vol.% n(fe 2 O 3 )=20 mmol ΔT/Δt= 90 K/h Ar balance 19

Variation of O 2 (14 vol.% O 2 ) c(no)=500 ppm total flow: 500 ml/min W/F=0.39 gs/ml c(o 2 )=14 vol.% n(fe 2 O 3 )=20 mmol ΔT/Δt= 90 K/h Ar balance 20

Summary and conlusions Fe 2 O 3 catalyst enhances the soot oxidation as well as the NO x reduction Catalyst transfers gas-phase oxygen to the soot surface ( oxygen pump ) Contact between catalyst and soot maintains, even up to high conversion levels Soot/NO reaction takes place on active carbon sites formed by decomposition of surface oxygen compounds (no direct effect of the catalyst) Based on mechanistic and kinetic studies a global kinetic model is constructed which describes the experimental data well Validation of the model on the level of catalyst/soot powder mixtures Model is used for the targeted development of advanced Fe 2 O 3 catalysts 21

Acknowledgement Prof. Dr.-Ing. Henning Bockhorn Dipl.-Chem. Dirk Reichert Dipl.-Ing. Nadya Atanassova Dipl.-Ing. Thomas Schöne Dipl.-Ing. Sandra Nobst Prof. Dr. Hizbullah Khan Dr. Thomas Schröder M.Sc. Naveed Jan M.Sc. Orkun Nalcaci Dr. Markus Kraft, University of Cambridge (UK) Dipl.-Phys. Markus Sander, University of Cambridge (UK) 22