Modern Methods in Heterogeneous Catalysis Research

Transcription

1 Modern Methods in Heterogeneous Catalysis Research Axel Knop-Gericke, January 09, 2004 In situ X-ray photoelectron spectroscopy (XPS) In situ near edge X-ray absorption fine structure (NEXAFS) in the soft energy range

2 Outline In situ X-ray photoelectron spectroscopy (XPS) Basics (you heard an hour ago) Experimental Set up Examples: methanol oxidation on copper pentene hydrogenation on Pd In situ near edge X-ray absorption fine structure (NEXAFS) in the soft energy range Basics Experimental Set up Examples: Methanol oxidation on copper n-butane oxidation on vanadium phosphorous oxide

3 Motivation Example: catalysis hν H C O e - ex situ analysis does not give correct results! catalyst Many other processes cannot be studied in UHV: environmental chemistry, corrosion, electrochemistry Very few methods can investigate the solid-gas interface at high pressures Pressure Gap (SFG, scanning probe microscopies, X-ray diffraction) Photoemission methods are very powerful

4 Transmission of 20 cm air Transmission Energy (ev)

5 In situ XPS: obstacles Fundamental limit: elastic and inelastic scattering of electrons in the gas phase Mean Free Path [Torr-mm] Elastic Mean Free Path in Oxygen Gas Electron Kinetic Energy [ev] Technical issues: - Differential pumping to keep analyzer in high vacuum - Sample preparation and control in a flow reactor

6 In situ XPS : basic concept Photons enter through a window Electrons and a gas jet escape through an aperture to vacuum

7 In situ XPS instruments: previous designs K. Siegbahn et al. (1968) K. Siegbahn et al. (1973) M.W. Roberts et al. (1979) M. Grunze et al. (1988) H. Siegbahn et al., J. Electron Spectrosc. Relat. Phenom. 2, 319 (1973) H. Siegbahn et al., J. Electron Spectrosc. Relat. Phenom. 24, 205 (1981)

8 In situ XPS : differential pumping + electrostatic lens system conventional X-ray source previous designs X-rays from synchrotron our design

9 In situ XPS system X-rays enter the cell at 55 incidence through an SiN x window (thickness ~ 1000 Å) Hemisphercal electron analyzer (10-9 p 0 ) mass spectrometer and additional pumping Analyzer input lens Focal point of analyzer input lens Third differential pumping stage (10-8 p 0 ) Experimental cell supplied by gas lines (p 0 ) First differential pumping stage (10-4 p 0 ) Second differential pumping stage (10-6 p 0 )

10 Gas phase analysis using XPS spot size 200 µm x (exit slit width) exit slit µm hν e - gas gas phase signal: 1 mbar mm ~ a few monolayers

11 BerlinerElektronenspeicherringgesellschaft für Synchrotronstrahlung Photon scource

12 Application of in situ XPS to catalysis Catalytic oxidation of methanol Cu Partial oxidation: CH 3 OH + ½O 2 CH 2 O + H 2 O 3 Cu Total oxidation: CH 3 OH + O 2 CO 2 + 2H 2 O 2 What is the state of the surface under reaction conditions? surface properties catalytic activity In situ measurements are necessary

13 Partial oxidation of methanol UHV XPS I.E. Wachs & R.J. Madix, Surf. Sci. 76, 531 (1978); A. F. Carley et al., Catal. Lett. 37, 79 (1996). O - (a) + 2CH 3 OH(g) CH 3 O - (a) 2CH 3 O - (a) + H 2 O(g) CH 2 O(g) + H + (a) In situ NEXAFS A. Knop-Gericke et al., Topics Catal. 15, 27 (2001). CO+H 2 CH O+H 2 2 CH OH 3 CO +H O 2 2 Ox surf CH OH 3 CH OH 3 Subox CH 3 OH + O 2 ~ 0.5 mbar suboxide phase: - only present in situ C u 0 O xb u lk O+ O CH O+H 2 2O O vol O 2 Questions for in situ XPS: - Quantitative analysis of surface species - Carbon species on the surface - Depth-dependent analysis

14 Methanol oxidation on Cu: O1s spectra 400 C 160 Normalized count rate (cps/ma) O 2 (g) CO 2 (g) H 2 O (g) CH 2 O (g) CH 3 OH (g) Cu 2 O?? O 1s CH 3 OH:O 2 1:2 3:1 VB ~ 2 ev : Binding energy (ev) Binding energy (ev)

15 Methanol oxidation on Cu: O1s spectra 400 C Normalized count rate (cps/ma) O 2 (g) CO 2 (g) H 2 O (g) CH 2 O (g) CH 3 OH (g) Surface oxygen Cu 2 O O 1s Subsurface oxygen CH 3 OH:O 2 1:2 3:1 CO 2 (g) CH 2 O (g) CH 3 OH (g) C 1s coke : Binding energy (ev) Binding energy (ev)

16 O 1s depth profile CH 3 OH : O 2 = 3:1 sub-surface/ surface peak area ratio b Sub-surface O / Cu peak area ratio a 3:1 6: Inelastic mean free path (Å) Inelastic mean free path in Cu (Å) from: Tanuma at al., SIA 17, 911 (1991) Kinetic energy (ev) Inelastic mean free path (Å) I /I =n /n exp[-(z z )/λ] z = 4 Ǻ, n /n = 1.6

17 Correlation of catalytic activity and surface species CH 2 O yield vs sub-surface oxygen peak area CH 2 O partial pressure (mbar) C C 200 C 250 C C 1:2 normalized sub-surface oxygen peak area 3 3:1 450 C 4 6:1 400 C 5 mixing ratio series (T = 400 C) temperature series (CH 3 OH:O 2 =3:1) Open questions: What is the nature of the sub-surface oxygen species? What is its role in the catalytic reaction?

18 Basics of X-ray absorption spectroscopy Prinzip der Röntgenabsorptionsspektroskopie (XAS) Auger-Elektron Anregung von von Rumpfniveauelektronen mittels mittels Röntgenstrahlung Relaxation mittels mittels Fluoreszenzstrahlung oder oder durch durch Aussenden eines eines Auger- Auger- Elektrons (TEY, (TEY, PEY, PEY, AEY) AEY) Untersuchung der der unbesetzten Zustände I f r p i 2

19 Experimental set up for in situ XAS investigations manipulator mass spektrometer properties of the set-up 150 mm butterfly-valve heating heatingup up to to K pressure pressureup up to to mbar mbar sample 150 mm process pump gas inlet by MFC batch- batch-and and flow-through-mode flow-through-mode angular angular dependent dependent measurements measurements UHV-valve Plattenventil turbo pump φ 100 mm

20 In situ XAS detector system Simultaneous Simultaneous detection detection of of gas gas phase- phase-and and sample sample signal signal

21 O K-edge Gas phase substraction Analysis of the Near Edge X-ray Absorption Fine Structure (NEXAFS) Intensity (a. u.) π* ev O 2 σ* ev ev ev ev CH 3 OH + O 2 CH 3 OH I gas I det I gas ( I det - ) * 3 NEXAFS of the O K-edge Total Total electron electron yield yield of of the the gas gas phase phase dominates dominates all all signals, signals, therefore therefore only only small small differences differencesin in the the detector detector signals signals Substraction Substraction allows allowsto to separate separate the theabsorption absorptionsignal signal of ofthe thesurface surfaceof ofthe thecatalyst 0 Cu 2 O ev Photon Energy / ev

22 Cu L 3 -NEXAFS Catalytic Activity NEXAFS at the Cu L 3 -edge 100 Conversion CH 3 OH (%) Flow ratio O 2 / CH 3 OH Increased Increased activity activity for for gas gas flow flow ratios: ratios: O 2 2 // CH CH 3 OH 3 OH < Transition Transition from froman an oxidic oxidiccopper-phase copper-phaseto to the themetallic state state

23 NEXAFS at the O-K edge 10 O K-edge Temperature ev ev Flow ratio O 2 / CH 3 OH 0.6 less active NEXAFS NEXAFS of of the the active active state state is is completly completly different different from from the the NEXAFS NEXAFS of of the the known known copper-oxides copper-oxides 2 2 oxidic- oxidic- and and 1 1 suboxidic suboxidic species species can can be be distinguished distinguished Total Electron Yield (a. u.) K 670 K 570 K Reference Cu 2 O very active K ev Photon Energy / ev

24 Correlation between the suboxide species and CH 2 O yield Variation of temperature at O 2 / CH 3 OH = 0.2 Intensity Intensity of of the the suboxide suboxide species species increases increases with with increasing increasing temperature temperature Intensity Intensity of of the the suboxide suboxide species species is is positively positively correlated correlated to to the the yield yield of of CH CH 2 2 O and and CO CO

25 Correlation between oxidic species and CO 2 Intensity Intensityof ofthe theoxidic oxidicspecies speciesox Ox surf surf decreases decreases with with increasing increasingco CO 2 -yield 2 -yield 2 2 areas areas of of activity activity can can be be distinguished distinguished

26 Model Proposed model of the copper surface under reaction conditions for methanol oxidation CO+H 2 CH OH 3 CO +H O 2 2 CH OH 3 CH OH 3 CH O+H 2 2 Ox surf Subox Cu 0 Ox bulk O+ O CH O+H 2 2 O O vol O 2

27 n-butane Oxidation to MA by Vanadium Phosphorus Catalysts VPO + 3,5 O H 2 O 400 C, 1 bar O O O 1,5 Vol% air Maleic Anhydride (MA) C 4 H ,5 O 2 4 CO H 2 O C 4 H ,5 O 2 4 CO + 5 H 2 O Active phase: highly ordered vanadyl pyrophosphate (VO) 2 P 2 O 7?

28 The VPO V L 3 -NEXAFS Analysis of spectral shape by unconstrained least squares fit Total Electron Yield (norm. u.) V L 3 -edge V1 V2 V3 V4 V5 V6 V Photon Energy / ev V valence Details of of the the local local chemical bonding Local geometric structure

29 Interpretation of V L 3 NEXAFS Experimental finding: NEXAFS resonances appear in in a sequence of of V-O V-O bond bond lengths Relative Resonance Position / ev Relative Resonance Position / ev V 4 V 3 V 2 V 1 V 2 O 5 V Bond Length / Å Bond Length / Å

30 Changes of NEXAFS while heating Decrease while active Proportion of int. intensity of V5 0,55 0,50 0,45 0,40 0,35 0,30 Relative spectral Intensity of V 5 at V L 3 -edge RT 400 C RT 400 C RT Spectra number Total Electron Yield (norm. u.) V L 3 -edge Photon Energy / ev V 5

31 Interpretation of V L 3 NEXAFS Identification of resonances (V5, V6): V 2 O 5 as model substance for VPO DFT DFT calculation of of DOS DOS (V (V 2 O 2 5!)*: 5!)*: O(3) O(1a) O(2) V 2 O 2 5 : 5 : Close relationship between geometric and and O(3) electronic structure at at V L 3 -absorption 3 edge edge O(1b) main main contributions to to NEXAFS resonances appear V6: V6: O(1a) in in a sequence of of V-O V5: V5:? (estimated value of of bond bond length V-O bond bond length between O(2) O(2) and and O(1a): Å) Å) O(2)

32 Summary In situ XPS: Estimation of stoichiometry, valence, depth profiles (using a tuneable photon scource), information depth can be reduced to 0.5 nm (depends on the kinetic energy) In situ XAS Information depth of about 50 nm in TEY, details of chemical bonding, local geometric structure, calculations or refrence samples are required Methanol oxidation: a non oxidic subsurface oxygen species correlates with the yield of formaldehyde Pentene hydrogenation: different carbon species were detected under reaction conditions, C-H correlates with activity n-butane oxidation: indication that (VO) 2 P 2 O 7 is not the active phase