TAP Temporal Analysis of Products

Transcription

1 Lecture Series at Fritz-Haber-Institute Berlin Modern Methods in Heterogeneous Catalysis TAP Temporal Analysis of Products Cornelia Breitkopf Institut für Technische Chemie Universität Leipzig

2 Transient methods are a powerful tool for gaining insights into the mechanisms of complex catalytic reactions.

3 Dimensions in Heterogeneous Catalysis

4 Macrokinetics and Microkinetics

5 Kinetic measurements Complexity of heterogeneously catalyzed reactions Macrokinetics and Microkinetics O. Deutschmann, XXXIV. Jahrestreffen Deutscher Katalytiker/ Fachtreffen Reaktionstechnik, Weimar

6 Kinetic measurements in practise Tasks of lab investigations Laboratory reactors o catalyst preparation o catalyst screening o activity o selectivity o stability o scale-up o process optimization o microcatalytic pulse reactor o gaschromatographic reactor o single pellet diffusion reactor o catalytic fixed bed reactor o recycle reactor o TAP

7 Questions What distinguishes TAP from other lab reactors? Which information do we need to understand TAP? How can TAP applied in heterogeneous chemistry?

8 Outline Steady state and unsteady state experiments Diffusion and mathematics TAP-Method and modeling Applications

9 Recommended literature and books Gleaves JT et al. Appl. Cat. A 160 (1997) Phanawadee P. PhD thesis Washington Chen S. PhD thesis Washington Yablonsky GS et al. J. Catal. 216 (2003) Gleaves JT et al. Cat. Rev.-Sci. Eng. 30 (1988) Dumesic JA et al (Ed) The Microkinetics of Heterogeneous Catalysis ACS, Washington Stoltze P. Progress in Surface Science 65 (2000) Kobayashi H et al.. Cat. Rev.-Sci. Eng. 10 (1974) Engel T, Ert G. Adv. Catal. 28 (1979) Ertl G. The dynamics of interactions between molecules and surfaces. Berichte Bunsengesellschaft für Physikalische Chemie 99 (1995) Creten G et al. J. Catal. 154 (1995) Haber J Concepts in catalysis by transition metal oxides. Bonelle JP et al. (Ed.) D. Reidel Publishing Company. Surface Properties and Catalysis by Nonmetals. 45, Hinrichsen O. DECHEMA-Kurs Angewandte Heterogene Katalyse, Bochum Dewaele O, Froment GF. J. Catal. 184 (1999) Christoffel EG Laboratory Studies of Heterogeneous Catalytic Processes Stud.Surf.Sci.Catal. 42. Müller-Erlwein E. Chemische Reaktionstechnik, Teubner Stuttgart-Leipzig, 1998, 237.

10 continued Jens Hagen - Emig G., Klemm E.. Technische Chemie. Springer Baerns M., Behr A., Brehm A., Gmehling J., Hofmann H., Onken U., Renken A.. Technische Chemie, Viley VCH 2006.

11 Steady state and unsteady state experiments

12 Kinetic investigations Evaluation of the reaction mechanism Constructing of a mathematical model Parameter estimation How to do? Steady-state or unsteady-state experiment? Quantitative evaluation of kinetic data?

13 Steady-state experiments Most common reaction technique used in heterogeneous catalysis Achieved by operation such that temperature, pressure, concentration, and flow rate at any point in the reactor is time invariant Access to activity, selectivity, reaction order, activation energy

14 Steady-state experiments Advantages: - Easy to build and operate - Results can be described with mathematical models based on algebraic equations - Most industrial processes are operated under steady-state conditions Disadvantages: - Provide global kinetic parameters, limited information on individual reaction steps - Interpretations often based on simple assumptions

15 Unsteady-state experiments Transient techniques provide information on - Reaction intermediates (pulse response) (Gleaves 1988) - Reaction sequence in a multistep reaction (Kobayashi 1975) - Rate constants of elementary steps (Ertl 1979, Creten 1995)

16 Unsteady-state experiments Transient techniques provide information on - Investigation of complex kinetic phenomena (oscillating chemical reactions, hysteresis) that are not observable under steady-state conditions - Probe of catalyst surfaces that are not easily observed under steady-state conditions (oxidation catalysis) (Haber 1983)

17 Unsteady-state experiments Disadvantages - Not easy to build up, expensive - Main problem: theory is very complex

18 Steady-state and transient methods Steady-state methods measure overall performance give integrated picture of reaction system have minimum reactor residence time of 1 s Transient methods give information on individual steps operate in millisecond time regime; resolution increase

19 Pulse methods response to pertubations describe mathematically the transient behaviour evaluation of rate parameters from response measurements, such as mass transfer coefficients, diffusivities, and chemical kinetic constants use of: - fixed-bed (column) chromatography - isotope technique - slurry adsorber - single-pellet

20 Pulse methods

21 Diffusion and mathematics

22 Diffusion in porous solids Characteristic value to characterize the influence of internal transport phenomena of heterogeneous reactions on the surface between a fluid phase and a porous solid is Da II (see textbooks) Da II = f (k, c external, l characteristic, D eff ) if fluid is gaseous: diffusion in pores depends on dimensions of pore system

23 Diffusion mechanisms Transport mechanisms in porous solids Pore diffusion depending on pore diameter

24 Atkins Molecular diffusion

25 Molecular diffusion mixture of two components A and B, concentration gradient (in one dimension y): under steady-state conditions the diffusional flow of one component is described by 1. Fick law J A = D AB dc dy A D AB...binary molecular diffusion coefficient of component A diffusing through B D AB = f( molecular properties of A and B, T, c or p)

26 General view to transport equations for gases Transport Mass Heat Energy Flux of mass density Flux of heat density Flux of momentum j n = D i dc dz dt du z j Q = λ jp = η dz dy Fick s Law Fourier s Law Newton s Law (viscosity)

27 Atkins General view to transport equations for gases

28 Molecular diffusion coefficient for gases Hirschfelder/Curtiss/Bird oder Chapman-Enskog relation for gases at low T, p (± 10% accuracy; extrapolation possible) D gs AB T 3 M r, A 2 M + M r, B r, A r, B 2 1 = , m s 2 pσ AB ΩD M T p M σ Ω D r,a 2 AB, M r,b Temperature [K] Pressure [bar] Relative molecular masses of molecules Aand B Average collision factor of gases Aand B Collision integral with respect to force constants Ω = f(k B *T / ε AB ) k B Boltzmann constant ε AB..force constant (see tables in textbooks)

29 Molecular diffusion coefficient for gases Ω = f( k B *T / ε AB )

30 Knudsen diffusion under low pressure conditions and/or for small pores: collision of gas with pore wall > collision of gas with gas mean free path length of molecule > pore diameter Λ = πσ V N A Λ mean free path length σ 2 molecular cross-section V gas molar volume at p N A /V at 298 K : c ges *p (molecules/cm 3 ) (dimension of p 10 5 Pa) mean free path length with typical σ (9-20*10-16 cm 2 ) 10 2 Λ ( nm) p

31 Knudsen diffusion 10 2 Λ ( nm) p Conditions for Knudsen diffusion d pore [nm] <1000 <100 <10 <2 p [bar]

32 Knudsen diffusion Knudsen flow through one zylindrical pore D K 2 d p 8 RT 6 m, i = 10 ( at 293 K,0.01MPa) 3 π M s for porous solids, the relative pore volume ε b and the tortuousity factor τ K have to be considered D eff ε d P p 8 RT τ 3 π M K, i = effective diffusion coefficient K estimationof τ is a complex procedure (see literature)

33 Effective diffusion coefficients diffusional flow in the pores may be described by an effective diffusion coefficient relation to surface: surface of pore mouths is representing only a part of the outer surface of a particle ε pores are not ideally cylindrical pores are connected by a network τ

34 Characterization of porous structures Porosity pore volume total volume,- mass m m m, kg Pore radius, pore diameter Pore radius distribution, -density H ( r Specific surface area Tortuousity factor p r r p ) = h ( r p,min p ) dr p ; h( r p ) = d H ( r dr p p )

35 Transition region of diffusion for heterogeneous reactions in a porous solid, the conditions of pressure or pore diameter may be such that the system is between Knudsen and molecular diffusion mean free path length pore diameter both equations for D M and D Kn apply 1 D * = 1 D M + 1 D Kn

36 Emig, Klemm Knudsen and molecular diffusion for porous solids experimental evaluation of diffusion coefficients

37 Background - Mathematics Delta function f(x) = δ (x-a) - functional which acts on elements of a function - integral y 1 x < a < y δ ( t a) dt = a < x y < a x 0, for x < y - multiplication with a function f(x) + f ( x) δ ( x a) dx = f ( a)

38 Background - Mathematics Delta function - graphical representation as Sprungfunktion (limes) - graphical representation as Pulsfunktion x u(x-a) 1 a ) ( sin ) ( 1 lim ) ( a x n a x a x n = π δ ) ( ) ( a x dx a x du = δ 1 x u(t-a) a = 0 1 ) ( dt a t δ

39 Background - Mathematics Laplace transformation - aim: transformation of differential and integral expressions into algebraic expressions - definition - examples f ( s) = 0 e st F( t) dt f f ( s) = L [] 1 = 0 t [ ] e ( s) = L e = e 1 dt = e s st st 1 α st α t 1 0 e 0 = s dt = α s

40 TAP method and modeling

41 Transient experiments Transients are introduced into a system by varying one or more state variables (p, T, c, flow) Examples: - Molecular beam experiments pressure change - Temperature-programmed experiments temperature change - Step change experiments concentration change

42 TAP A transient technique The key feature which distinguishes it from other pulse experiments is that no carrier gas is used and, gas transport is the result of a pressure gradient. At low pulse intensities the total gas pressure is very small, and gas transport occurs via Knudsen diffusion only. Pulse residence time under vacuum conditions is much shorter than in conventional pulse experiments. Thus a high time resolution is achievable.

43 TAP Features Extraction of kinetic parameters differs compared to steady-state and surface science experiments - Steady-state: kinetic information is extracted from the transport-kinetics data by experimentally eliminating effects of transport - Surface science: gas phase is eliminated - In TAP pulse experiments gas transport is not eliminated. The pulse response data provides information on the transport and kinetic parameters.

44 TAP Features TAP pulse experiments are state-defining - Typical pulse contains molecules or moles - Example: - Sulfated zirconia: 100 m 2 /g with 3 wt-% S, 5*10 18 S/m 2-1 pulse of n-butane: 1*10 14 molecules - per pulse 1/50000 of surface addressed

45 TAP System hardware Simplified schematic of a TAP pulse response experiment Gas pulse Reaction zone Detector Injection of a narrow gas pulse into an evacuated microreactor Gas pulses travel through the reactor Gas molecules (reactant and product) are monitored as a function of time and produce a transient response at the MS

46 TAP System hardware

47 TAP System hardware

48 TAP System hardware 1) High speed pulse valve 2) Pulse valve manifold 3) Microreactor 4) Mass spectrometer 5) Vacuum valve 6) Manual flow valve 7) Mosfet switch

49 TAP Response curves Typical pulse response experimental outputs

50 TAP Multipulse experiment Key features of input: - typical pulse intensity range from to molecules/pulse - pulse width µs - pulse rates 1-50 s -1 Key features of output: - different products have different responses - individual product response can change with pulse number

51 TAP Pump probe experiment Key features of input: - different reaction mixtures are introduced sequentially from separate pulse valves Key features of output: - output transient response spectrum coincides with both valve inputs

52 TAP Transient response data In contrast to traditional kinetic methods, that measure concentrations, the observable quantity in TAP pulse response experiments is the time dependent gas flow escaping from the outlet of the microreactor. The outlet flow is measured with a MS (QMS) that detects individual components of the flow with great sensitivity. The composition of the flow provides information on the types of chemical transformation in the microreactor. The time dependence of the flow contains information on gas transport and kinetics.

53 TAP Theory Goal Interpretation of pulse response data - determine typical processes - find parameters for these processes - develop a model Analyzation of experiments that provide parameters of diffusion, irreversible adsorption or reaction and reversible reaction

54 TAP General models Currently there are three basic models in application based on partial-differential equations. - One-zone-model - Three-zone-model - Thin-zone-model The mathematical framework for the one-zone-model was first published in 1988 (Gleaves). Basic assumptions of one-zone-model: - catalyst and inert particle bed is uniform - no radial gradient of concentration in the bed - no temperature gradient (axial or radial) - diffusivity of each gas is constant

55 TAP Gas transport model The gas transport is the result of Knudsen diffusion. An important characteristic of this tranport process is that the diffusivities of the individual components of a gas mixture are independent of the presssure or the composition of the mixture. D M 1 e 1 = T 1 D e2 M T 2 2 D e,i effective Knudsen diffusivity M i molecular weight T i temperature 1,2 gas 1, gas 2

56 TAP Transport model (1) Diffusion only case Mass balance for a non-reacting gas A transported by Knudsen diffusion C ε b t A = D ea 2 C z A 2 C A concentration of gas A (mol/cm 3 ) D ea effective Knudsen diffusivity of gas A (cm 2 /s) t time (s) z axial coordinate (cm) ε b fractional voidage of the packed bed in the reactor

57 TAP Transport model The equation can be solved using different sets of initial and boundary conditions that correspond to different physical situations. initial conditions: 0 z L, t = 0, C A =δ z N ε A b pa boundary conditions: z = 0, z C A =0 z = L, C A = 0 N pa number of moles of gas A in one pulse A cross-sectional area of the reactor (cm 2 ) L length of the reactor (cm)

58 TAP Transport model The gas flow at the reactor exit F A (mol/s) is described by C z A F A = A DeA z= L and the gas flux (mol/cm 2 s) by Flux A = F A A To solve for the gas flow it is useful to express initial and boundary conditions in terms of dimensionless parameters.

59 TAP - Transport model Use of dimensionless parameters ξ = z L dimensionless axial coordinate C A = N PA CA ε A L b dimensionless concentration t D τ = ε L b ea 2 dimensionless time

60 TAP Transport model Dimensionless form of mass balance τ ξ 2 C A CA = 2 initial conditions: 0 ξ 1, t = 0, C A =δξ boundary conditions: ξ = 0, ξ = 1, C A =0 ξ C A = 0 Solution: - analytical a) method of separation of variables b) Laplace transformation - numerical for more complex problems

61 TAP Transport model Method of separation of variables: solution for dimensionless concentration C F F A A A 2 2 (( n + 0.5) πξ ) exp( ( n 0.5) π τ ) ( ξ, τ ) = 2 cos + n= 0 solution for dimensionless flow rate C ( ξ, τ ) ( ξ, τ ) = = π A (2n + 1)sin + ξ n= (( n + 0.5) πξ ) exp( ( n 0.5) π τ ) dimensionless flow rate at the exit (ξ=1) = π n= 0 n ( 1) (2n + 1) exp ( ( n 0.5) π τ )

62 TAP Transport model F A = π = n 0 n ( 1) (2n + 1) exp ( ( n 0.5) π τ ) Expression for the dimensionless exit flow rate as a function of dimensionless time Standard diffusion curve For any TAP pulse response experiment that involves only gas transport, the plot of the dimensionless exit flow rate versus dimensionless time will give the same curve regardless the gas, reactor length, particle size, or reactor temperature.

63 TAP Transport model Due to initial condition the surface area under SDC is equal to unity flow rate in dimensional form F N A pa D = ea π ε L n= b ( 1) (2n + 1) exp ( n n 2 ) π t D ε L b ea 2 An important property of this dimensional dependence is that its shape is independent on the pulse intensity if the process occurs in the Knudsen regime.

64 TAP Transport model F N A pa D = ea π ε L = b ( 1) (2n + 1) exp ( n n 2 ) 2 n 0 2 π t D ε L b ea 2 Characteristics of standard diffusion curve F 1 b L τ p = 1, t p = ε 6 6 D A ea D, p = 1.85, H p = ε 2 ea 2 b L peak maximum corresponding height F A, p τ p = H p t p 0.31 fingerprint for Knudsen regime

65 TAP Transport model a) Standard diffusion curve showing key time characteristics and the criterion for Knudsen diffusion b) Comparison of standard curve with experimental inert gas curve over inert packed bed

66 TAP Transport + adsorption model (2) Diffusion + irreversible adsorption (adsorption is first order in gas concentration) C t C z 2 A A ε b = DeA as Sv (1 ε 2 b ) k a C A θ A t = k a C A a s k a S v θ A surface concentration of active sites (mol/cm 2 of catalyst) adsorption rate constant (cm 3 of gas/mol s) surface area of catalyst per volume of catalyst (cm -1 ) fractional surface coverage of A

67 TAP - Transport + adsorption model (2) Diffusion + irreversible adsorption Flow rate F A at reactor exit F N A pa D = ea π n 2 2 exp( k t) ( 1) (2n + 1) exp ( n + 0.5) 2 a π ε b L n= 0 t D ε L b ea 2 with k as Sv (1 a = ε b ε ) k b a Exit flow curve for diffusion+irreversible adsorption -issmaller than SDC by a factor of exp(-k a t) - is always placed inside the SDC (fingerprint)

68 TAP - Transport + adsorption model Exit flow curve for the diffusion+irreversible adsorption case Comparison of irreversible adsorption curves with standard diffusion curve (A) k a =0 (SDC), (B) k a =3 (C) k a =10

69 TAP Transport and adsorption model (3) Diffusion + reversible adsorption Mass balances - for component A in gas phase ) )( (1 2 2 A d A a b v s A ea A b k C k S a z C D t C θ ε ε = - for component A on the catalyst surface A d A a A k C k t θ θ = k d desorption rate constant (s -1 )

70 TAP - Transport + adsorption model Exit flow curve for the diffusion+reversible adsorption case Comparison of reversible adsorption curves with standard diffusion curve (A) k a =0 (SDC), (B) k a =20, k d =20 (C) k a =20, k d =5

71 TAP Transport, adsorption, reaction (3) Diffusion + reversible adsorption + irreversible reaction A(gas) B(gas) A(ads) B(ads) k r ) )( (1 2 2 A da A aa b v s A ea A b k C k S a z C D t C θ ε ε = A r A da A aa A k k C k t θ θ θ = ) )( (1 2 2 B db B ab b v s B eb B b k C k S a z C D t C θ ε ε = B db B db A r B k C k k t θ θ θ + =

72 TAP Transport, adsorption, reaction (3) Diffusion + reversible adsorption + irreversible reaction The initial and boundary conditions used in the transport-only case can be applied to the transport+reaction case as well. Initial condition t = 0, C A = 0 C A,1 Inlet boundary condition z = 0, DeA, 1 = δ ( t) z

73 TAP Transport, adsorption, reaction (3) Diffusion + reversible adsorption + irreversible reaction D ea C A,1, 1 = δ ( t) z The inlet flux is respresented by a delta function. With this combination, analytical solutions in the Laplace-domain can be easily derived when the set of differential equations that describe the model is linear.

74 TAP Characteristic values Use of dimensionless parameters k a L = k dimensionless apparent adsorption 2 ε b a DeA rate constant with k as Sv (1 a = ε b ε ) k b a k a contains: -thekinetic characteristics of the active site (k a ) -thestructural characteristics of the whole catalytic system a s S v ( 1 ε b) ε b

75 TAP Extended models Three-zone-model Additional boundary conditions between the different zones have to be applied pulse inert catalyst inert outcome to QMS zone 1 zone 2 zone 3 Evaluation of curve shapes and fingerprints are basicly the same as for the one-zone-model

76 TAP Theory Moment-based quantitative description of TAP-experiments idea behind: observed TAP data consist of a set of exit flow rates versus time dependencies analytical solutions in integral form can be usually obtained analysis of some integral characteristics (moments) of the exit flow rate moments reflect important primary features of the observations

77 TAP Theory - Moments momentm n of the exit flow rate (i.e. not of concentration) t = n M n t F( t) dt 0 n order of moment representation in dimensionless form m n n t FA dτ with τ 2 ( ε / 0 bl DeA = τ = )

78 TAP Theory - Moments Application to TAP experiments: - for irreversible adsorption / reaction 1 m0 = 1 X = with cosh Da I Da I 2 ka ε b L = D ea Da I Damköhler number I - mean dimensionless residence time τ res m = m 1 0 = n= 0 n= 0 A( n) [ B( n) + Da ] A( n) B( n) + Da I I τ Dif n n A( n) = ( 1) (2 + 1) B 2 2 ( n) = ( n + 0.5) π ε bl τ Dif = D 2 e, A

79 Applications

80 TAP - Examples Mechanism and Kinetics of the Adsorption and Combustion of Methane on Ni/Al 2 O 3 and NiO/Al 2 O 3 (Dewaele et al. 1999) - Supported Ni catalysts for steam reforming of natural gas to synthesis gas - Combination of TPR with single pulse experiments

81 TAP - Examples Selective oxidation of n-butane to maleic anhydride (MA) over vanadium phosphorous oxide (VPO) (Gleaves et al. 1997) important heterogeneous selective oxidation complex reaction (electron transfer, fission of C-H-bonds, addition of oxygen, ring closure) literature: (VO) 2 P 2 O 7 -lattice as active-selective phase, but it is not clear how the lattice supplies the oxygen - possible involvation of oxygen adspecies adsorbed on vanadium surface sites - possible involvation of bulk oxygen - possibly other crystal phases are active

82 TAP - Examples Selective oxidation of n-butane to maleic anhydride (MA) over vanadium phosphorous oxide (VPO) Steady-state kinetic studies yielded different rate expressions and kinetic parameters this may be related to differences in c. composition, c. structure, c. oxydation state, etc. no unique information Aim of TAP investigations: Relate changes in kinetic dependencies to changes in the catalyst How to do? Systematic alteration of the state of the catalyst and measuring kinetic dependencies

83 TAP - Examples Selective oxidation of n-butane to maleic anhydride (MA) over vanadium phosphorous oxide (VPO) Kinetics of n-butane conversion over VPO in the absence of gas phase oxygen - n-butane curve lies completely within the generated diffusion curve irreversible process - initial step is irreversible (fissure of CC or CH bond)

84 TAP - Examples Selective oxidation of n-butane to maleic anhydride (MA) over vanadium phosphorous oxide (VPO) Kinetics of n-butane conversion on oxygen-treated VPO - values of apparent activation energy is strongly dependent on the VPO oxidation state - decreasing number of active centers Arrhenius plots for pulsed n-butane conversion over an oxygen-treated VPO catalyst sample as a function of oxidation state -TAP + Raman: V 4+ V 5+

85 TAP sulfated zirconia (SZ) TAP investiagtions on sulfated zirconias C. Breitkopf. J. Mol. Catal. A: Chem. 226 (2005) 269. X. Li, K. Nagaoka, L. Simon, J.A. Lercher, S. Wrabetz, F.C. Jentoft, C. Breitkopf, S. Matysik, H. Papp. J. Catal. 230 (2005) C. Breitkopf, S. Matysik, H. Papp. Appl. Catal. A: Gen. 301 (2006) 1. C. Breitkopf, H. Papp. Proceedings of The 1st International Conference on Diffusion in Solids and Liquids DSL-2005, July 6-8, 2005, Aveiro, Portugal, 67. (oral presentation)

86 TAP Investigation of poisoned samples Comparison of active sample Kat1 with samples of Na-modified surfaces 3 % of active centres poisoned 40 % of active centres poisoned all centres poisoned Sulfation Step: 5.7x10-3 mol (NH 4 ) 2 SO 4 Kat1 5x10-5 mol Na 2 SO x10-3 mol (NH 4 ) 2 SO 4 SZ-Na1 5x10-4 mol Na 2 SO x10-3 mol (NH 4 ) 2 SO 4 SZ-Na2 5x10-3 mol Na 2 SO x10-3 mol (NH 4 ) 2 SO 4 SZ-Na3 873 K for 3 h

87 Characterization Sample Kat1 SZ-Na1 SZ-Na2 SZ-Na3 BET [m 2 /g] Pore diameter [Å] Pore volume [cm 3 /g] SO 4 content w.% S atoms/m 2 5.3* * * *10 19 Na + ions/m 2-2.2* * *10 20 S/Na

88 Catalysis and TPD Activation: 2 h at 600 C in air, 30 ml/min Reaction: 5 % n-butane (99.95 % purity) in N 2, 20 ml/min yield i-butane / % C SZ-Na1 SZ-Na2 SZ-Na3 Kat TPD Kat1 Na time on stream / min Temperatur [ C] Poisoning of 3 % of active Brønsted centres led to significant decrease in activity, loss of Brønsted centres led to inactive samples

89 DRIFTS Sample L/B Ratio (by pyr ads.) at 323 K Surface Content Na [%] S=O mode [cm -1 ] at 673 K Kat SZ-Na1 1.1 not detectable 1397 SZ-Na SZ-Na L/B ratio determined according to Platon A, Thomas WJ. Ind. Engin. Chem. Res. 24 (2003) 5988.

90 TAP n-butane single pulses Kat1 SZ-Na1 SZ-Na2 Höhennorm. Int m/e 43 [w.e.] 1,0 0,8 0,6 0,4 0,2 0,0 448 K 423 K 398 K 373 K 348 K Zeit [s] Höhennorm. Int. m/e 43 [w.e.] 1,0 0,8 0,6 0,4 0,2 0,0 448 K 423 K 398 K 373 K 348 K 0,0 0,2 0,4 0,6 0,8 1,0 Zeit [s] Höhennorm. Int. m/e 43 [w.e.] 1,0 0,8 0,6 0,4 0,2 0,0 448 K 423 K 398 K 373 K 348 K 0,0 0,2 0,4 0,6 0,8 1,0 Zeit [s] τ (348 K) = 1.58 s τ (423 K) = 0.49 s τ residence time τ (348 K) = 0.47 s τ (423 K) = 0.16 s τ (348 K) = 0.20 s τ (423 K) = 0.11 s Interaction of n-butane single pulse with surface decreases with increasing amount of sodium

91 Kinetic modeling

92 Thank you for your attention!