X-ray absorption spectroscopy Jagdeep Singh Jeroen A. van Bokhoven
X-ray absorption spectroscopy
Jagdeep Singh
Jeroen A. van Bokhoven
Absorption as function of energy of the x-ray
8800 9000 9200 9400 9600 9800-0.5
0.0
0.5
1.0
1.5
2.0
Ab
sorp
tion
(a.u
.)
Energy (eV)
0 200 400 600 8000.0
0.2
0.4
0.6
0.8
1.0
Abso
rptio
n (a
.u.)
Energy (eV)
8950 8975 9000 9025
0.0
0.5
1.0
1.5
2.0
Abso
rptio
n (a
.u.)
Energy (eV)
2 4 6 8 10 12 14-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
χ(k)
k (Å-1)
a.
c. d.
b.
8800 9000 9200 9400 9600 9800-0.5
0.0
0.5
1.0
1.5
2.0
Ab
sorp
tion
(a.u
.)
Energy (eV)
0 200 400 600 8000.0
0.2
0.4
0.6
0.8
1.0
Abso
rptio
n (a
.u.)
Energy (eV)
8950 8975 9000 9025
0.0
0.5
1.0
1.5
2.0
Abso
rptio
n (a
.u.)
Energy (eV)
2 4 6 8 10 12 14-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
χ(k)
k (Å-1)
8800 9000 9200 9400 9600 9800-0.5
0.0
0.5
1.0
1.5
2.0
Ab
sorp
tion
(a.u
.)
Energy (eV)
0 200 400 600 8000.0
0.2
0.4
0.6
0.8
1.0
Abso
rptio
n (a
.u.)
Energy (eV)
8950 8975 9000 9025
0.0
0.5
1.0
1.5
2.0
Abso
rptio
n (a
.u.)
Energy (eV)
2 4 6 8 10 12 14-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
χ(k)
k (Å-1)
a.
c. d.
b.
Pre-edge subtraction Edge energy determination
Background and Normalization EXAFS Function
Data-analysis
4 6 8 10 12 14-60
-40
-20
0
20
40
60
χ(k)
k(Å-1)
1 2 3 4 5 6
-4
-2
0
2
4
Four
ier t
rans
form
(k2 *χ
(k))
R (Å)
Fourier transformation
1st neighbor
2nd neighbor
EXAFS formula
radial distribution function
exp
Scatter power
Damping Disorder
(−2σi k ) ( )( )2
2 202
2R(k) i
i i
SkR
χλ
−∑= NiFi(k) exp sin 2kRi + ϕj k
Scatter power
Damping Disorder
( ) ( )( )2
2 202
2( ) ( ) exp exp 2 sin 2ii i i i j
i i
S Rk N F k k kR kkR
χ σ ϕλ
−⎛ ⎞= − +⎜ ⎟⎝ ⎠
∑
0 5 10 15 200.00
0.05
0.10
0.15
0.20
0.25
Bac
ksca
tterin
gsAm
plitu
de
k (Å)
Pt - O
Pt - Pt
Temperature effect on EXAFS / XANES
Temperature effectScatter power
Damping Disorder
( ) ( )( )2 2exp 2 sin 2i i jk kR kϕ+20
2
2( ) ( ) exp ii i
i i
S Rk N F kkR
χ σλ
−⎛ ⎞= −⎜ ⎟⎝ ⎠
∑
exp(-2σ2k2)
0 5 10 15 200.0
0.2
0.4
0.6
0.8
1.0
Low Tσ2 = 0.001
High Tσ2 = 0.01
K (A-1)
Fj, ϕj, and So2 from reference compound or theory
( ) ( )( )2
2 202
2( ) ( ) exp exp 2 sin 2ii i i i j
i i
S Rk N F k k kR kkR
χ σ ϕλ
−⎛ ⎞= − +⎜ ⎟⎝ ⎠
∑
0 1 2 3 4 5 6
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
Four
ier t
rans
form
(k3 )
R (Å)
Au/Al2O3
3.04 < K < 11.5 A-1
0
Coordination number 6.8Au-Au distance 2.76 ÅΔDWF 0.0058C3 9 E-6C4 3E-6
Getting structural information from EXAFS
Added parameter: ΔE0
Abstract
• EXAFS gives local structure
• XANES gives geometry and oxidation state (empty density of states)
exp
Scatter power
Damping Disorder
(−2σi k ) ( )( )2
2 202
2R(k) i
i i
SkR
χλ
−∑= NiFi(k) exp sin 2kRi + ϕj k
EXAFS (χ) Extended X-ray Absorption Fine-StructureSingle scattering
XANES X-ray Absorption Near-Edge StructureMultiple scattering
Single Scattering
Multiple Scattering
A
Pre edge
XANES versus EXAFS
EXAFS- Single scattering dominates
XANES- Multiple scattering- Electronic transitions- Multiple electron transitions
Information- Number & kind of neighbor- Distance- Disorder
- Geometry / subtle distortions- Oxidation state- Electronic information- DOS in final state
Theoretical description of XANES- Detailed electronic information- Aids interpretation of spectra of unknown compounds- Time-consuming - Needs an expert
A
GoalLocal structure of catalysts under well-defined conditions
precursor state
during / after activation
during reaction
deactivation
time-resolution (few msec)
space-resolution (few μm)
… … .
XAS in Catalysis
3160 3170 3180 3190 3200-0.5
0.0
0.5
1.0
1.5
Difference
HydrideNaked paticle
2%Pd/SiO2
Energy (eV)
Nor
mal
ized
Abs
orpt
ion
10 nm
3160 3170 3180 3190 3200-0.5
0.0
0.5
1.0
1.5
Palladium hydride formation
Difference
Hydride
Naked paticle
Nor
mal
ized
Abs
orpt
ion
Energy (eV)
2%Pd/Al2O3
<2 nm
Metal d-band reactant level
Fermi level
εd
-
+
Anti bonding state
3160 3170 3180 3190 3200-0.5
0.0
0.5
1.0
1.5
Difference
HydrideNaked paticle
2%Pd/SiO2
Energy (eV)
Nor
mal
ized
Abs
orpt
ion
-15 -10 -5 0 5 10 150.0
0.5
1.0
1.5
2.0
2.5
0.0
0.5
1.0
1.5
2.0
s p d s (H) p (H)
DOS vs XANES for PdH0.64
(93 atoms) M
uX
eV
-15 -10 -5 0 5 10 150.0
0.5
0.0
0.5
1.0
1.5
2.0
s p d s (H) p (H)
DOS vs XANES for PdH0.64 (93 atoms)
MuX
eV
XAS
Partial / complete oxidation of hydrocarbons methane, alkenes, methanol
Hydrogenation / dehydrogenation reactions alkenes, alkynes, alkadienes, (un)saturated ketones
Methanol synthesisCO + 2H2 CH3OHCO2 + 3 H2 CH3OH + H2O
WGSH2O + CO CO2 + H2
Nitric Oxide reduction (with CO, olefins, or H2)
CO oxidation1925: Active in CO oxidationhighly active in presence of H2: Haruta Catal. Today 36 (1997) 153Selective CO removal, air purification, high-purity N2 and O2
Catalysis by Gold
Physical properties• bulk metallic gold is thermodynamically stable • melting point and metallicity of the particle is function of particle size
Catalysis by Gold
Buffat Phys. Rev. A 13 (1976) 2287
CO oxidation: particle-size effect
Physical properties• bulk metallic gold is thermodynamically stable • melting point and metallicity of the particle is function of particle size
Catalysis by Gold
Goodman Science 281 (1998) 1648 Haruta Cattech 3 (2002) 102
CO oxidation: particle-size effect
Large support effects: SiO2: hardly activeAl2O3, MgO: moderately active(TiO2) Fe2O3, CeO2, other reducible supports: very active & less dependent on particle sizes; No clear relation to reducibility of support
Physical properties:• bulk metallic gold is thermodynamically stable • melting point and metallicity of the particle is function of particle size
Active species in gold oxidation catalysis? • Carbonate-mechanism excluded• Small particles become active as soon as they are non-metallic (Goodman)• Oxidic gold (I or III) is active species (Gates)• Theory supports both gold-only and support-aided mechanism• Support supplies oxygen via molecularly (activated) adsorbed oxygen
via Mars van Krevelen
Catalysis by Gold
Geometry / coordination
Density of states
Oxidation state
Time resolved
In situ
How is oxygen activated on the catalyst?
How can the most inert metal be so active?
Small gold particles adsorb oxygen(and react)
Nørskov et al. Angew. Chem. 44 (2005) 1824
• Deposition precipitation HAuCl4 adjusted pH• Washing with a base to remove chlorine• Reduction in hydrogen
SupportsAl2O3, SiO2, CeO2, TiO2, ZrO2, Nb2O5
Sample Preparation
Full EXAFS & XANES analyses
Structure of gold catalysts
EXAFS Fitting Results of Reduced Catalysts
2.70
2.72
2.74
2.76
2.78
2.80
2.82
2.84
2.86
2.88
2 4 6 8 10 12
Au-Au CN
Au
Bon
d D
ista
nce,
ÅBulk value
2.88 Å
± 10-20%
± 0.02
0 10 20 30 40 50 60 70 80
2.72
2.76
2.80
2.84
2.88
Al2O3 TiO2 SiO2 Nb2O5 ZrO2 CeO2
Au-A
u D
ista
nce
Particle Size (Å)
EXAFS Fitting Results of Reduced Catalysts
Strong reduction in Au-Au distance with particle sizeNo visible influence of support
30%50%
100%
Electronic structure from LIII XANES
Whitelines reflect number of holes in the d-bandGold whiteline: spd-rehybridization results in 5d10–x6sp1+x
11875 11900 11925 11950 11975 120000.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Au(0)Au(III)
Nor
mal
ized
Abs
orpt
ion
Energy (eV)
Mattheiss, L.F. et al. Phys. Rev. B 1980, 22, 1663
11.91 11.92 11.93 11.94 11.950.0
0.2
0.4
0.6
0.8
1.0
Bulk Gold 32 Å 16 Å 11 Å
Nor
mal
ized
Abs
orpt
ion
Energy (keV)11.91 11.92 11.93 11.94 11.95
0.0
0.2
0.4
0.6
0.8
1.0
Bulk Gold 32 nm 12 nm 5 nm
Nor
mal
ized
Abs
orpt
ion
Energy (eV)
11.92 11.930.4
0.6
0.8
Au/ZrO2 Au/Al2O3 Au/TiO2 Au/Al2O3 Bulk Gold
Nor
mal
ized
Abs
orpt
ion
Energy (keV)
Au/Al2O3 Au/TiO2
1.1 nm Au
Whiteline is particle-size dependent
Whitelines reflect number of holes in the d-band
3 4 5 6 7 8 9 10 11 12
0.00
0.15
0.30
0.45
0.60
Au/Al2O3 Bulk Au/ZrO2 CeO2 SiO2 Nb2O5 TiO2
Diff
eren
ce in
WL
Inte
nsity
Au-Au Coordination Number2.70 2.72 2.74 2.76 2.78 2.80 2.82 2.84 2.86 2.88 2.90
0.00
0.15
0.30
0.45
0.60
Au/Al2O3 Bulk Au/ZrO2 CeO2 SiO2 Nb2O5 TiO2
Diff
eren
ce in
WL
Inte
nsity
Au-Au Distance (Angstron)
Whiteline intensity versus particle sizeDifference intensity with bulk
Six supports, one trendLarger particles fewer d-electrons
CN R(Å) %Au(III)Reduced 3.6 2.72 0Reox. RT 3.6 / 0.3 2.72 / 2.04 10Reox. 225C 2.7 / 0.5 2.71 / 2.04 15
11900 11920 11940 119600.0
0.2
0.4
0.6
0.8
1.0
Reduced (H2 250C)Reoxidized (20% O2 RT)Reoxidized (20% O2 225C)
Nor
mal
izd
Abs
orpt
ion
Energy (eV)
1.3wt% Au/Al2O3
11875 11900 11925 11950 11975 120000.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Au(0)Au(III)
Nor
mal
ized
Abs
orpt
ion
Energy (eV)
Exposure to 20% O2
11900 11910 11920 11930 11940 11950 119600.0
0.2
0.4
0.6
0.8
1.0
1.2
Nor
mal
ized
Abs
orpt
ion
Energy (eV)
Au/Al2O3
Au whiteline
CN(Au) R(Au)He 6.5 2.771:1 5.3 2.732:1 5.7 2.731:2 5.2 2.77
Small oxygen contribution
11930 11940 11950
0.6
0.8
1.0
XAS during CO Oxidation
CO/O2 2/1CO/O2 1/1CO/O2 1/2HeN
orm
aliz
ed A
bsor
ptio
nEnergy (eV)
119200.4
More intense with more CO:holes in the d-band (anti-bonding states)
11910 11920 11930 11940 119500.0
0.2
0.4
0.6
0.8
1.0
Norm
aliz
ed A
bsor
ptio
n
Energy/eV11910 11920 11930 11940 11950
0.0
0.2
0.4
0.6
0.8
1.0
Norm
aliz
ed A
bsor
ptio
n
Energy/eV11910 11920 11930 11940 11950
0.0
0.2
0.4
0.6
0.8
1.0
Norm
aliz
ed A
bsor
ptio
n
Energy/eV
Gold catalysts and activation of oxygen
• Under (diluted) O2: surface oxidation (Au/Al2O3 & Au/TiO2)
• Switch to CO/O2: CO keeps gold reduced
Au/Al2O3Switch Oxygen to CO2s repetition
Reduced gold is active phase
Gold participates in oxygen activation
Au(0) Au(0)
Aun+
O
OC
O
CO + O2
Au(0)Au(0) Au(0)
Aun+
O
OC
O
CO + O2
Au(0) Au(0)
Aun+
O
OC
O
CO + O2
Au(0)
Au(0) Au(0)
Aun+
OO
O2
Au(0) Au(0)
Aun+
OO
CO + O2
Au(0)
OC
Au(0)
OC
Fast
Van Bokhoven Angew. Chem. (2006) VIP
Slow
• Rehybridization of spd-orbitals (5d10–x6sp1+x)
• Smaller particles have fewer holes in the d-band
• Particle size dominates support-effect
• Oxygen is activated on gold particle
0 10 20 30 40 50 60 70 80
2.72
2.76
2.80
2.84
2.88
Al2O3 TiO2 SiO2 Nb2O5 ZrO2 CeO2
Au
-Au
Dis
tanc
e
Particle Size (Å)
Boiling point Gold-Gold distance d electrons
2.70 2.72 2.74 2.76 2.78 2.80 2.82 2.84 2.86 2.88 2.90
0.00
0.15
0.30
0.45
0.60
Au/Al2O3 Bulk Au/ZrO2 CeO2 SiO2 Nb2O5 TiO2
Diff
eren
ce in
WL
Inte
nsity
Au-Au Distance (Angstron)
Fermi level
d-LDOS
He Traces O21%CO(He)
Pt/Al2O3
Adsorption sites from XAS
bare atop
FEFF8 simulation
HeCO
Experimental
platinum catalystStructure of the active phase?
fuel cell (PEM) catalytic converter
surface rougheningsurface patternssurface reaction
Single Crystals
Langmuir-Hinshelwood
UHV conditions
Single Crystals
Langmuir-Hinshelwood
UHV conditions high pressure
Two reaction regimes
• Low activity – CO poisoning
• High activity ???
Single Crystals
UHV conditions
Real Catalysts
Real Conditions
1.8 nm
1 nm
1.3 nm
pressure gap
material gap
Conversion data
350 400 450 500 5500.00
0.02
0.04
rate
of C
O c
onve
rsio
n (c
m3 /m
in/g
cat)
Temperature (K)
Low activity region
O2/CO = 521
High activity region
2wt%Pt/Al2O3
Conversion dataignition or extinction
temperature (K)O2/CO ratio
heating cooling
1 472 456 yes 340
2 445 440 yes 338
5 433 421 yes 329
hysteresis temperature for
onset ofconversion
(K)
XAS
11560 11570 11580 115900
1
2
3
heating
Energy (eV)
heating
308 K340 K386 K
455 K
422 K
439 K
472 K
O2/CO = 2
350 400 450 500 550 600
Temperature (K)
11560 11570 11580 115900
1
2
3
Energy (eV)
O2
CO
XAS
11560 11570 11580 115900
1
2
3
cooling
Energy (eV)
cooling
308 K340 K386 K
455 K
422 K
439 K
472 K
O2/CO = 2
11560 11570 11580 115900
1
2
3
Energy (eV)
O2
CO
350 400 450 500 550 6000.00
0.02
0.04
Temperature (K)
ratio O2 / CO = 1
11560 11570 11580 115900
1
2
3
heating
(a)
Energy (eV)
heating
ratio O2 / CO = 2
11560 11570 11580 115900
1
2
3
heating
(a)
Energy (eV)
heating
ratio O2 / CO = 5
11560 11570 11580 115900
1
2
3
heating
(a)
Energy (eV)
heating
oxidation parallels ignition
350 400 450 500 5500.00
0.02
0.04
rate
of C
O c
onve
rsio
n (c
m3 /m
in/g
cat)
Temperature (K)
125
Kinetics and XAS
350 400 450 500 5500.00
0.02
0.04
rate
of C
O c
onve
rsio
n (c
m3 /m
in/g
cat)
Temperature (K)
Low activity region
O2/CO = 521
High activity region
adsorbed CO
partially oxidicsurface
Kinetics and XAS
350 400 450 500 5500.00
0.02
0.04
rate
of C
O c
onve
rsio
n (c
m3 /m
in/g
cat)
Temperature (K)
Low activity region
O2/CO = 521
High activity region
adsorbed CO
partially oxidicsurface
Qexafs signalsPt foilL3, L2, L1 edges
XAS (QEXAFS)
470 471 472 473 474 47540
50
60
70
80
90
100
CO
Con
vers
ion
%
Temperature/ K
11580 11600 11620
0.9
1.2
1.5
Energy (eV)
O2/CO = 1
Kinetics and XAS
350 400 450 500 5500.00
0.02
0.04
rate
of C
O c
onve
rsio
n (c
m3 /m
in/g
cat)
Temperature (K)
heating
11560 11570 11580 115900
1
2
3
heating
Energy (eV)
heating
308 K340 K386 K
455 K
422 K439 K
472 KCO poisoning
partially oxidicsurface
J. Singh, et al. Angew. Chem. Int. Ed., accepted
incr
easi
ngly
oxid
ized
Conclusions
• a highly active state of the catalyst is discovered
• the catalyst shows different structure in low- and high-
activity regime; low-activity region : CO adsorbed on
platinum, high-activity region: partially oxidized platinum
• the catalyst increasingly oxidizes during the ignition
• high temperature and a high oxygen concentration
benefit the formation of the more active partially oxidic
catalyst.
Other supports(normal XAS)
Pt/Al2O3L, 2 nm
11.55 11.56 11.57 11.58 11.59 11.600.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
norm
aliz
ed a
bsor
ptio
n
Energy, keV
125oC, 2% conversion 194oC, 9% conversion 210oC, 20% conversion 218oC, 32% conversion 231oC, 100% conversion
SiO2, 1.5-3 nm
11.55 11.56 11.57 11.58 11.59 11.600.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Energy, keV
40oC, 0% conversion 140oC, 1% conversion 240oC, 6% conversion 294oC, 45% conversion 306oC, 100% conversion
norm
aliz
ed a
bsor
ptio
n
TiO2, 1.3 nm
11.55 11.56 11.57 11.58 11.59 11.600.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
norm
aliz
ed a
bsor
ptio
n
Energy, keV
68oC, 4% conversion 134oC, 24% conversion 180oC, 67% conversion 190oC, 100% conversion
Extinction plots
100 150 200 250 3000
20
40
60
80
100
perc
ent C
O c
onve
rsio
n
temperature, oC
Pt/TiO2 Pt/Al2O3Pt/SiO2
3 4 5 6 7 8 9 10 11 12
-6
-4
-2
0
2
4
6
o
Chi
x k
3
k (A-1)
Chi FitA
1.5 2.0 2.5 3.0 3.5 4.0
-4
-2
0
2
4
oFT
(A3 )
R (A)
Fourier Transform Fit
o
B
K3 weighted CHI and Fourier transform2.5 < k < 12 Å-1
Pt/Al2O3L reduced, hydrogen removed; He (RT)
EXAFS analysisBelow and above temperature of ignition
Conditions atom N DW R(Å) Eo
RT, He Pt 5.7 0.0042 2.72 1.68
below ignition Pt 6.2 0.0054 2.77 0.58
O 2.4 0.0021 1.99 4.36above ignition
Pt 3.0 0.0065 2.61 7.19
RT, O2/CO Pt 5.6 0.0039 2.76 1.87
EXAFS analysis
350 400 450 500 550 6000.00
0.02
0.04
rate
of C
O c
onve
rsio
n (c
m3 /m
in/g
cat)
Temperature (K)
0.25
0.5125
Pt /Al2O3 small particles
Kinetic oscillations – Single Crystals
T = 470KpCO= 3×10-5mbarpO2= 2.0 2.7×10-4mbar
M. Eiswirth and G. Ertl, Surface Sci. 177 (1986), 90
Pt (110)
Kinetic oscillations – Single Crystals
bulk crystal plane
Pt (110)
Kinetic oscillations – Single Crystals
T. Gritsch et al. Phys. Rev. Lett. 63 (1989) 1086N. Freyer et al. Surface Sci. 166 (1986) 206
higher adsorption energy of CO
lower adsorption energy of CO
sticking coefficient of O on 1x1 is about 1.5 times than on 1x2
Kinetic oscillations - 2wt% Pt/Al2O3
385 380 375 370 365-0.001
0.000
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0.009
IR P
eak
Inte
nsity
(a
.u.)
Mas
s S
pec.
Sig
nal
(a.u
.)
Temp. (K)
0
5
10
15
20
25
30
35
40
45
50
55
CO2
CO
OCPt
OC
Pt Pt
• storage of CO that is released in sudden spike of CO2
• CO poisoning; activityinversely proportional to adsorbed CO
Kinetic Oscillations and QEXAFS
cata
lyst
par
ticle
s
CO/O2
To exhaust / mass spec
O2 : CO = 192 wt% Pt/Al2O320 mg catalyst30 ml/min
• activity loss due to reduction of active (oxidized) surface
• sudden increase in activity parallels oxidation of surface
kinetic oscillations originate from
the reduction and re-oxidation
of the surface
Best of luck for your exams ☺
EXAFS error margins:N ± 0.5 20%R ± 0.01 – 0.03σ2 ± 20%