Introduction to EELS

Transcription

1 Introduction to EELS Nestor J. Zaluzec Brief Review of Energy Loss Processes Instrumentation: Detector Systems Instrumentation: AEM Systems Data Analysis and Quantification: Advanced Topics 1

2 Brief Review of Energy Loss Processes Electron Excitation of Inner Shell & Continuum Processes Spectral Shapes Notation of Edges Electron Scattering Angular Distributions The Emission Process: 1-Excitation, 2-Relaxation, 3-Emission 2

3 Electron Energy Loss Spectroscopy Measure the changes in the energy distribution of an electron beam transmitted through a thin specimen. Each type of interaction between the electron beam and the specimen produces a characteristic change in the energy and angular distribution of scattered electrons. The energy loss process is the primary interaction event. All other sources of analytical information ( i.e. X-rays, Auger electrons, etc.) are secondary products of the initial inelastic event. Thus, EELS has the highest potential yield of information/inelastic event E V Electron Emission 3 Photon Emission 3 h E F M L K a b c 3

4 DOS L shell M shell d-band s-band } XEDS/XRF AES XPS/UPS EELS/XAS Incident Electrons Incident Photons Schematic Diagram Illustrating Sources of Inelastic Scattering Signals Experimental XEDS, XPS, and EELS data from the Copper L shell. Note the differences in energy resolution, and spectral features. Intensity (Arb Units) XEDS XPS EELS Energy (ev) 4

5 5

6 Comparision Light Element Spectroscopy Resolution XEDS vs EELS Comparision of WL XEDS Detector and EELS spectra taken from the same NiO specimen O K Relative Intensity (Arb. Units) XEDS NiL EELS Energy (ev) Note the enhanced spectral information in the EELS data. Vertical scale is arbitrary and chozen for clarity of presentation. 6

7 7

8 8

9 Instrumentation: Detector Systems Energy Loss Spectrometers Basic Principles Electrostatic/Electromagnetic Serial/Parallel Detector Systems Spectral Artifacts Multichannel Analyzers 9

10 10

11 11

12 12

13 13

14 14

15 Data Analysis and Quantification: Spectral Processing Thin Film Quantification Methods Specimen Thickness Effects 15

16 16

17 M 23 C 6 in Steel : Spectral Overlap 17

18 Spectral OverLap Problems also exist in EELS 18

19 Two Related Methods are sometimes used: Second Difference Filtering (Shuman( etal MAS,1983) Record 3 Energy Loss Spectra which are displaced in energy by de Mathematically combine in computer to form the Second Difference (SD) Spectrum SD(E) = I (E-dE) - 2 I 1 2 (E) + I 3 (E+dE) Digital Filtering Spectrum has the appearance of a derivitive, removes channel to channel gain variation in parallel EELS and slowly varying backgrounds. Sharp features are enhanced in visibility. This is related to a simple numerical differentiation of a single spectrum 19

20 I k = P k * I o I k = Number of electron having excited a kth inner shell P k = Probability of excitation of the kth shell I o = Incident electron current P k = N k N k = Number of atoms of the element analyzed = Ionization cross-section for the kth shell Alternatively; Consider the ratio of Intensities of any two Edges in the same spectrum I and I A B Invoke the Ratio Method and obtain the exact equation: Note the similiarity of this equation with that of Thin Film XEDS 20

21 But in the real world the assumptions used above simple arguments are never realized: in the Measure all scattered electrons ( = 180 0) Integration over all energy Losses Because we must measure over a finite energy window ( E) we modify the expression to: e - N A= I A( E) A( E)*I 0 we also measure over a finite angular window ( ) and therefore, we modify the expression to: Incident Beam Convergence α N A= I A( E, ) A( E, )*I 0 and the ratio equation becomes: N A IA( E, ) = kab N B I B( E, ) Specimen β Inelastic Scattering Angle with k AB = B( E, ) A( E, ) 21

22 Problems in EELS Quantification Cross-section Calculations 22

23 Problems In EELS Quantification Collection Angle Errors 23

24 Effects of Specimen Thickness Multiple Scattering Low Loss Core Loss -Visiblity Quantification Effects Effects of Specimen Thickness Low Loss Spectra Intensity τ/λ = 0.9 τ/λ =1.5 τ/λ = 1.9 τ/λ = 2.5 I P I o = t ln I T Io = t Energy Loss (ev) 24

25 Effects of Specimen Thickness Core Loss Spectra C o u n t s Energy (ev) Experimental EELS Edge/Background Ratio as a Function of Specimen Thickness at Constant Accelerating Voltage Variation in EELS Edge P/B with Thickness EELS Intensity Normalized to Pre-Edge Bgnd t/ = 0.23 t/ = 0.62 O K Edge in NiO t/ = Energy Loss (ev) 25

26 Mean Free Path Calculated Variation In Mean Free Path as a Function of Accelerating Voltage 2.5 T = m ov 2 2 E m 1 n 2 = m oc a o T ( ) ln 2 T c E m 1 + E o 2m = Eo o c E 2 o m o c 2 Normalized Mean Free Path Accelerating Voltage (kv) Experimental Measurement of MFP 26

27 Experimental EELS Edge/Background Ratio as a Function of Accelerating Voltage at Constant Specimen Thickness O K Edge in NiO Variation in EELS Edge P/B with kv EELS Intensity Normalized to Pre-Edge Background 300 kv 200 kv 100 kv Energy Loss (ev) Specimen Thickness Effects on EELS 27

28 Effects of Thickness and Voltage on EELS P/B Deconvolution of Multiple Scattering using Leapman/Swyt Method 28

29 This defines a "Thin Film Approximation" for EELS i.e. / < 1.0 Advanced Topics Low Loss Spectroscopy Plasmon Losses Studies Dielectric Properties Core Loss Spectroscopy Near Edge Structure Extended Fine Structure Radiation Damage 29

30 EELS Measurements of Valence Electron Densities Table 2 Experimental and Calculated Plasmon Shifts in Selected Metal-Hydrogen Systems Material Expt. E p (ev) Calculated FEM Ep/H/M Expt. Ep/H/M FEM Calc. Group II Mg MgH Group III Sc ev ScH Y YH Group IV Ti TiH Zr ZrH Group V V VH PseudoGroup VI FeTi /21.9 * FeTiH( β) /21.8 * /-0.3 * FeTiH 2 (γ ) /21.5 * * Calculated using modified FEM with m s =m e and m d =1.9 m e 30

31 Information in Core-Loss Profiles L3 L2 L1 M5 M4 M3 M2 M1 N 45.. O23.. K L M N & O 31

32 Oxygen K Shell in superconducting and non-superconducting specimens in the Bi 2 Sr 2 Ca n-1 Cu n O x system. Note the disappearance of the pre-edge peak in the non-superconducting specimen. Intensity CaSrBiCuO Non-SuperConducting 525 Energy Loss (ev) 575 O K Shell NES in Y-Ba-Cu-O Perpendicular to C Axis Parallel to C Axis Orientation dependance of the O K shell in Y 1 Ba 2 Cu 3 O 7- δ. The intensity variation of the pre-edge peak indicates that the hole state just above the Fermi level has p x,y symmetry. Parallel and perpendicular here refer to the momentum transfer directions which are by conservation perpendicular to the incident beam directions. 32

33 33

34 Copper L-shell Core Loss Spectroscopy in Metallic, Oxide and High Tc Superconductor Phases 34

35 35

36 Amorphous Fe x Ge 1-x Magnetic Materials x=0.2 x=0.3 x=0.4 x=0.5 x=0.6 x=1.0 Fe x Ge 1-x Energy Loss (ev) 36

37 Silicon 111 ZAP- Gamma Corrected Unfiltered Elastic 2nd Plasmon 1st Plasmon 37

38 Unscattered Elastic Plasmon I( )/Io Inner-Shell Phonon Scattering Angle [mr] Elastic Scattering : Crystalline Solids I( ) ~ F(hkl) 2 j F(hkl ) = f j ( ) exp (-2 ig r) V(r) = - g? V g exp (2πig.r) ; V g = h 2 2πmV e f(θ) Energy Filtered Convergent Beam Electron Diffraction Filtering Removes Inelastic Scattering Dynamical Calculations Used to Compare Experimental Data to Models Low Order Structure Factors Determined (Semi-Conductors, Metals) Applied to Ti-Al Intermetallics and Charge Density (Difference) Maps 38

39 Fig. 1a - (200) CBED dark field disk Experimental Calculated Difference Angular Dispersion in mrad

40 Radiation Damage in Materials Damage Creation Event/Process Mobile Vacancy & Interstitials Created Solid State Order/DisOrder/Amorphization/ Studies Lindemann Melting Criterion Perfect Crystals Melt: µ 2 vib = 9 h2 T 2 M k B θ c > µ 2 crit T m = M k B θ2 c µ 2 9 h 2 crit Defective Crystal Melt : µ 2 total = µ 2 vib + µ 2 static > µ 2 crit T d = M k 2 B θ d 2 µ 9 h 2 crit Materials Research Society Bulletin : 19 #7 (1994) - Lam & Okamoto P.C. Liu, P. R. Okamoto, 40

41 Solid State Order/DisOrder/Amorphization/ Studies We can experimentally measure < I( ) ~ F(hkl ) 2 1-e -2M M = 8 u2 sin ( ) using Energy Filtered Electron Diffraction P.C. Liu, P. R. Okamoto, Energy Filtered Electron Diffraction 41

42 Solid State Order/DisOrder/Amorphization/ Studies A A B B P.C. Liu, P. R. Okamoto, Solid State Order/DisOrder/Amorphization/ Studies P.C. Liu, P. R. Okamoto, 42

43 EELS / NES Measurements in Order/DisOrder/Amorphization Studies in TM Alloys Irradition Effects in TM alloys <---> Structure/Properties Mechanical Properties <---> Electronic Structure TM's have tightly bound valence d band which overlaps and hybridizes with broader nearly Free-electron sp band In TM's with partially filled d-bands the NES is readily measured. P.C. Liu, P. R. Okamoto, Phase Stablitiy of Ion Irradiated of Minerals for HLW Storage Materials Zirconolite ~ Ca ZrTi 2 O 7 Perovskite ~ Ca Ti O 3 Pyrochlore ~ VIII A 2 VI B 2 IV X 6 Y Candidate Phases for Immobilization of REE, & ACT in HLW forms Over Geologic Time Scales alpha decay causes the minerals to become metamict. In-situ Kr Ion Irradiation used in HVEM-Tandem to simulate long-term α-decay damage What is the Polytype, Dose, and Composition dependance? Are there other methods of measuring the damage and can they tell us about the structure? K.L. Smith 43

44 Zirconolite - 2M Illustrating the loss of crystallinity with irradiation dose (figure 1) Selected area diffraction patterns (SADs) of individual grains of various zirconolites were recorded as a function of dose to establish the critical dose for amorphisation (D ). c Phase Stablitiy of Ion Irradiated of Minerals for HLW Storage Materials Material Dc (10 14 ions/cm 2) Zirconolite - 2M Nd + Zirconolite- 2M Nd + Zirconolite - 3T Nd + Zirconolite - 4M U + Zirconolite - 2M U - Pyrochlore Th - Zirconolite 3T Perovskite Dc = f ( complexity/polytype) = f (dopant) but only by a factor of ~ 2 Additional studies planned Vary Ion Energy Fabrication Methodology(Sintering/Hot Pressing/Temperature) Dopant Partitioning K.L. Smith, 44

45 EELS Measurements 11 Figure 2. The Ti L edges of undoped zirconolite before and after ion irradiation, and two Ti standards (TiO 2 and Ti 2 O 3 ) EELS Results Comparison of the Zirconolite and Ti-oxide spectra suggests that Ti predominantly exhibits a valence of 4 + in both unirradiated and irradiated Zirconolite and that the bond lengths in zirconolite are similar to those in TiO 2. (figure 2) The electron energy loss near edge structure (ELNES) of the Ti L shell is consistent with the Ti changing from octahedral coordination to tetrahedral coordination (figure 3). 45

46 HR EELS of Zirconolite Ti L shell Figure 3. HR EELS of Ti L Shell in Zirconolite. EELS Results Specimen L2 /L 3- Spin L* 2 /L * 3- Energy Shift Orbit Molecular L3- L* 3 Ti 2O TiO CaTiO (CaxGdy)TiO (CaxGdy)(TinAl m)o CaZrTi 2O7(Zirconolite) UnIrrad Irradiated Table 2 46

47 EELS Results To quantitatively measure the changes due to irradiation, we fitted the zirconolite Ti-L spectra with four Lorentzians and calculated L 2 /L 3 and L* 2 / L* 3 and L 3 - L* 3 * (Table 2). L 2 /L 3 is the same before and after irradiation which suggests that radiation damage does not significantly affect the number of holes in the d-band (and hence valence). L* 2 / L* 3 increases and L 3 - L* 3 that radiation damage in zirconolite causes a distortion of the octahedral field around Ti atoms toward a tetrahedral configuration 47