Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
Infrared spectroscopyDavide Ferri :: Paul Scherrer Institut
Molecular aspects of catalysts and surfaces :: ETHZ
source: Andor.com
INFRARED
1 mm-1 m
1012-1015 Hz
The electromagnetic spectrum
-300 200 700 1200
EPR
UV-Vis
XAFS
NMR
Raman
IR
Number of publications
Number of publications containing in situ, catalysis, and respective method
Source: ISI Web of Knowledge (Sept. 2008)
• pros
- economic
- non-invasive
- versatile (e.g. solid, liquid, gas, interfaces)
- very sensitive (concentration)
- fast acquisition (down to ns!)
• cons
- no atomic resolution
Importance of IR spectroscopy in catalysis
Infrared spectroscopy
Far
Mid
Near
[mm] [cm‐1] E [kJ/mol]
25‐1000
2.5‐25
0.78‐2.5
visible
microwave
12800‐4000
4000‐400
400‐10
150‐50
50‐5
5‐0.1
freq
uency
high
low
overtone, combinationvibrations
fundamentalvibrations
fundamental vibrationsof heavy elements,lattice vibrations
wavelength wavenumber energy
• ‘quality control’: identification of compounds according to their fingerprintspectrum also inorganic materials, e.g. metal oxides ex situ, but also after degassing in cell (vacuum)
• Identification of surface sites│Detailed characterization of surface use of molecular probes in situ experiments, controlled dosage of probe
• Identification of surface sites under reaction conditions in situ/operando experiments to obtain molecular reaction mechanism, exposure
to reaction conditions
Why IR spectroscopy
• Interaction with matter energy causes vibration of molecular bonds energy is absorbed in correspondence of vibrational modes an absorption band is generated
Vibrational spectroscopy
symmetric stretchingasymmetric stretching bending
energy
Vibrations
wavenumber (cm-1)4000 4003000 2000 1000
bend
ing
stre
tchi
ngC-H
O-H
C-H
O-H
C=O C-O
C-CC=CC≡C
alkenesaromatics
• Why do vibrations appear in the IR spectrum?
Vibrations
selection rule
0Q
molecular dipole moment must change due to vibration or rotation along itscoordinate (so called, normal mode or normal coordinate, Q)
ArHClO2
3756 cm-1
3657 cm-1
N=3, non-linear, 3 fundamental modes
asymmetric stretching symmetric stretching scissoring (bending)1595 cm-1
All modes IR active
H2O
4000 3500 3000 2500 2000 1500 1000 500
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Abso
rban
ce
Wavenumber / cm-1
bendingstretching
4000 3500 3000 2500 2000 1500 1000 5000.0
0.5
1.0
1.5
2.0
2.5
3.0
Abso
rban
ce
Wavenumber / cm-1
roto-vibrational bands
(gases)
Gas and liquid phase H2O
• Harmonic oscillator The stretching frequency of a bond can be approximated by Hooke’s law. Two
atoms and the connecting bond are treated as a harmonic oscillator composed of two masses (atoms) joined by a spring.
Vibrations
12
k
1 21 2
m mm m
m: reduced massk: force constant
Vibrations
C-C C=C C≡C
C-H C-D C-C C-O C-Cl
1200 cm-1 1650 cm-1 2200 cm-1
3000 cm-1 2100 cm-1 1200 cm-1 1100 cm-1 800 cm-1
larger k
larger µ
12
k
The spectrometer
sample
beamsplitter
moving mirror
fixed mirror
x
SiC
interferogram
Dispersive vs. FT
FT-IR spectrometer has significant advantages over dispersive one
Multiplex (Fellgett) advantageAll source wavelengths are measured simultaneously
Throughput (Jacquinot) advantageFor the same resolution, the energy throughput in an interferometer can be higher → the same S/N as a dispersive-IR in a much shorter time
Precision (Connes) advantageThe wavenumber scale of an interferometer is derived from a HeNe laser that acts as an internal reference for each scan
The IR spectrum
‘no sample’
‘sample’
dete
ctor
inte
nsity
inte
nsity
FT
FT
I
I0II0
T =
4000 3500 3000 2500 2000 1500 1000
0.6
0.7
0.8
0.9
1.00.00
0.05
4000 3500 3000 2500 2000 1500 10000.00
0.05
wavenumber (cm-1)
4000 3500 3000 2500 2000 1500 1000wavenumber (cm-1)
wavenumber (cm-1)
15600 15800 16000 16200
-0.05
0.00
0.05
0.10
relative mirror position
15600 15800 16000 16200
-0.05
0.00
0.05
0.10
relative mirror position
‘background’
CO adsorbed on Pd/Al2O3
interferogram
The IR spectrum
4000 3500 3000 2500 2000 1500 1000 5000.0
0.5
1.0
1.5
2.0
2.5
3.0
abso
rban
ce
wavenumber (cm-1)
The IR spectrum
4000 3500 3000 2500 2000 1500 1000 5000.0
0.5
1.0
1.5
2.0
2.5
3.0
abso
rban
ce
wavenumber (cm-1)-Al2O3
180°C150°C
200°C
(O-H)
-Al2O3
-Al2O3
-Al2O3
Morterra, Catal. Today 27 (1996) 497
(Al-O)
• DRIFT spectra of V-W-TiO2 catalysts after adsorption of NH3 aspect of spectra changes with background
The background
4000 3500 3000 2500 2000 1500 1000
1419; NH4+*
1V5WTi(I)5WTi(I)
1V5WTi(C)1VTi(C)
5WTi(C)
Abs
orba
nce
(a.u
.)
Wavenumbers (cm-1)
1603; NH3*
1180; NH3*
3161; NH3*
3257; NH3*3346; NH3*
3382; NH4+*
wavenumber (cm-1)4000 3500 3000 2500 2000 1500 1000
0.1
dry KBr or air sample after treatment
400°C│O2 200°C│O2 200°C│NH3+O2 200°C│O2dehydration adsorption desorptionload sampleno cell
positive+negative signalspositive signals
• The spectrum contains information on terminal O-H bonds│3800-3600 cm-1
bridge hydroxyls│Brønsted acidity
H-bonded hydroxyls
M-O and M=O bonds, bulk and surface fundamental (n) and overtone (2×n) modes
other groups, e.g. C-H, carbonates, carboxylates…
Information on materials
M
OH
MO
H
M
Mm+O
H
Mn+
• Perturbation of hydroxyls adsorption of probe molecule
Information on materials
terminal
bridged
Busca, in Metal Oxide Catalysis, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim (2009) 95│Uslamin, Zeolite-based catalysis for sustainable aromatics production, Technische Universiteit Eindhoven (2019)
• Perturbation of hydroxyls adsorption of probe molecule
H-bonded hydroxyls
Information on materials
after CO adsorption
terminal
bridged
Busca, in Metal Oxide Catalysis, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim (2009) 95
Information on materials
Moros et al., Anal. Bioanal. Chem. 385 (2006) 708
sugar-added yogurt
plain yogurt
water
bkg: waternutritional parameters of yogurt samples
• Also other disciplines…
• The carbonate ion
Adsorbates by FTIR
1900 1500 1100wavenumber (cm-1)
free ion
monodentate
bidentate
organic-like
bicarbonates
as(CO)
s(CO) s(OH)
1700 1300
Morterra, Catal. Today 27 (1996) 497
D3hC2vC2v
as(CO)
simmetry
Adsorbates by FTIR
2200 2000 1800 1600 1400 1200
0.02abs.
units
(a.u.
)
wavenumber (cm-1)2200 2000 1800 1600 1400 1200
0.02
wavenumber (cm-1)
0.02
2200 2000 1800 1600 1400 1200
wavenumber (cm-1)
time time time
CO2H2
CH4CH4
as(HCO3-) s(HCO3-)
as(CO32-)
(CO)
as(OCO-)as(OCO-)
CO2H2
Techniques, sample form, environment
transmission powder gas/vacuum
reflection
internal reflection powder liquid
external reflection single crystal vacuum
diffuse reflection powder gas/vacuum
ATR-IRS
IRRASDRIFTS
• Lambert-Beer law
Transmission
0
log( ) log( )IA T cdI
I0 I
sample
d
c
T: transmittance, A: absorbance, : molar absorption (extinction) coefficient, c: concentration, d: path length
Transmission
Solid samplesLarge solid particles generally absorb too much IR light, therefore particles should be small and also special preparations are often necessary.
Most popular sample preparation methods (for mid-IR):
Alkali halide disk method Typically solid samples are diluted in KBr and ground Then pressurized to form a disk
Mull method Most common one is Nujol (liquid paraffin) Samples are ground and suspended in one or two drops of a mulling agent Followed by further grinding until a smooth paste is obtained
Film method By solvent casting or melt casting
NOT FOR IN SITU/OPERANDO EXPERIMENTS
Transmission
Busca, in Metal Oxide Catalysis, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim (2009) 95
KBr
MgO
pressed powder
• Sample preparation self supporting wafers (few mg)
• Controlled exp. conditions vacuum and controlled dosages
• Baseline slope increases at high frequency
(beam scattering increases with increasing frequency)
slope depends on particle size (very steep for powders with large particles, ca. 1 μm)
T @ 4000 cm-1 is ca. 0 for large particle size oxides
• Quantification│Molar absorption coefficient
Transmission
SiO2
■ 1605 cm‐1
□ 1585 cm‐1
Onfroy et al., Micropor. Mesopor. Mater. 82 (2005) 99
ε SAn=
ε, integrated molar absorption coefficient
ℓ, disc thickness (optical path)
n, amount of adsorbed molecule
S, disc area
ASℓ
εℓ= n
A = Sεn
Diffuse reflection
• Combination of transmission and scattering/reflection
Chalmers and Dent, Industrial analysis with vibrational spectroscopy, 1997, 153
normal specular component
incident beam
i r
Reflectance increases by increasing scattering → spectrum baseline is flat or even decreasing
transmission
diffuse reflection
• Qualitative analysis very sensitive to surface species due to its diffuse reflective nature the detected light is reflected multiple times at powder surfaces
• Quantitative analysis can be very complicated the spectra are largely influenced by various experimental parameters, e.g.
particles shape and size, refractive index of particles, absorption characteristics of particles, and porosity of the powder bed
a popular method is to use the Kubelka-Munk (K-M) function to transform reflectance to a sort of absorbance (K-M) unit
solid (approximated) theory applicability and accuracy for highly absorbing and non-absorbing samples is
questionable
Diffuse reflection
• Infinitely thick medium
• Adsorbate on infinitely thick medium
Kubelka-Munk function
R
R∞
0
J = R∞ ‐ RR∞
K/S = (1-R∞)2/2R∞K, absorption coeff.S, scattering coeff.
(K+C)/S = (1-R)2/2R
Matyshak et al., Catal. Today 25 (1995) 1
F(R) = J(1/R-R∞) = 2C/S optical length (d in L-B Law) much larger than in transmission → more sensitivity
Diffuse reflection
Sirita et al., Anal. Chem. 79 (2007) 3912
Pt/CeO2
R’= Icat+ads/Icat
• Reflectance or Absorbance?
Pt/SiO2
• Advantages of diffuse reflection easier sampling applicability to powders that scatter too much in transmission, assuming the
surface area is sufficiently high to detect surface vibrations with a sufficiently high signal-to-noise ratio slightly lower sensitivity to bulk conduction phenomena, because of a higher
surface-to-bulk sensitivity ratio ideally suited for in situ/operando studies
• Disadvantages of diffuse reflection less obvious optical setup work in flow rather than in vacuum more difficult sample activation (i.e. water removal from highly porous materials)
Diffuse reflection vs Transmission
• Comparison between techniques with different sensitivity (bulk/surface) should be careful
• DRIFTS more sensitive than TIRS• Band assignment depends on
surface sensitivity of the technique
Diffuse reflection vs Transmission
Rödel et al., PCCP 10 (2008) 6190
• Typical mirror unit, sample holder and in situ cell
Diffuse reflection
Harrick Scientific│Praying Mantis
in situ cell
mirror unit
sample holder
Attenuated total reflection
dp
IRE
n1=nIRE
n2
evanescent wavez
12 2
212 sinpd
n
11n
dp: penetration depththe distance frominterface where theelectric field hasdecayed to 1/e of itsvalue E0 at the interface
221
1
nnn
IRE, internal reflection element; high refractive index material
: angle of incidence
• Sample preparation sample deposition on IRE from aqueous slurry dry in air
Attenuated total reflection
2000
µm
85 µm
ZnSe
Pd/Al2O3
4 μm
0 1 2 3 4 5 6 7
10
20
numbe
r of p
articles (%)
particle size (nm)
drying
suspension
usedrying
suspension
useuse
suspensionIR TEM
SEM
Attenuated total reflection
drying
suspension
usedrying
suspension
use
microbalance
metals and metal oxides
IRE
e‐ beam
crucible
shutter
10‐6 mbar
100
0
50
Pt(1 nm)/Al2O3
Particulate filmModel film
4 μm
Pd/Al2O3
Attenuated total reflection
• Stable films needed for in situ investigations
Materials for internal reflection elements
Material Useful range / cm-1 n dp Properties
ZnSe 20000-700 2.43 long soluble in strong acid; usable up to ca. 300°C
Ge 5000-900 4.02 short good chemical resistance; hard and brittle; becomes
opaque at 250°C
Si 9400-1500; 350-FIR 3.42 short excellent chemical resistance; hard; usable up
to ca. 300°C
KRS-5(Thallium bromoiodide)
14000-330 2.45 long toxic; slightly soluble in water and soluble in base; usable
up to ca. 200°C
4000 3500 3000 2500 2000 1500 1000
wavenumber (cm-1)
1000 μm500 μm200 μm100 μm50 μm
250 μm
(0.3 to 1.4 μm !!!)
• The meaning of dp and nIRE
Attenuated total reflection
abs.
uni
ts
TIR
ATR-IRS
same spectrum as in transmission but withmuch smaller sample thickness
Attenuated total reflection
ZnSe
Si
• The meaning of dp and nIRE
cell section
cell section
powder layer
low nIRE↓
high dp
high nIRE
↓
low dp
• Solid (powders, crystals, foils, plates, seeds…) and liquid samples
• Ex situ structure assignment identification quality control
Attenuated total reflection
IRE
• Solid (powders, crystals, foils, plates, seeds…) and liquid samples
• Ex situ
• In situ/operando only liquids suspensions solid in contact with liquid
Attenuated total reflection
• Cell mounting
cell
coated IRE
mask for IRE
to/from thermostat
to pump
from liquid supply
gasket
gasket
samplesolution
Attenuated total reflection
• Sample compartment
Attenuated total reflection
liquid in to/from thermostat
cell holder
manual mirrors
liquid out
• γ-aminopropyl modified SiO2 (APS-SiO2) 1.5 mmol/g NH2 202 m2/g deposited on ZnSe from CCl4 slurry toluene/PE slurry prep. 80ºC
dried in vacuum
• toluene, 60ºC
Knoevenagel condensation
inletoutlet
IRZnSeinlet
HO
CN
COOEt+
ethyl cyanoacetatebenzaldehyde ethyl ‐cyanocinnamate
COOEt
CNbase
HO
CN
COOEt+
Knoevenagel condensation
toluene, 60°C, 20 mM
1800 1700 1600 1500 1400 1300 1200
abs.
uni
ts(a
.u.)
wavenumber (cm-1)
0.01
flow-through mode
COOEt
CN
17551701
Knoevenagel condensation
toluene, 60°C, 20 mM
1800 1700 1600 1500 1400 1300 1200
abs.
uni
ts(a
.u.)
wavenumber (cm-1)
0.01 1645Si N
flow-through mode
HO
+
CN
COOEt
• Consecutive dosage of reactants• Time dependence
abs.
uni
ts
elapsed time (s)
0.020
0.025
0.030
0.035
0.040
0.045
0 100 200 3000.000
0.005
0.010
0.015
0.020
0.025
0.030product
1645 cm-1
1701 cm-1
1755 cm-1
Wirz et al., Langmuir 22 (2006) 3698
Knoevenagel condensation
O O
O
CN
NH2 NH2
Si Si
H2O
O
O
CN
HO
CN
COOEt
• Stopped-flow mode
Knoevenagel condensation
toluene, 60°C, 20 mM
0.01 30 min
1800 1700 1600 1500 1400 1300 1200wavenumber (cm-1)
abs.
uni
ts(a
.u.)
stopped-flow mode
flow-through mode
Knoevenagel condensation
Si NH2
Si N
HO
CN
COOEt
COOEt
CN
H2O
H O+
1645 cm-1
1733 cm-1
2225 cm-1
CN
COOH
OHO HHOH
+
2
2300 22202260
2264
2225
Wirz et al., Langmuir 22 (2006) 3698
CO adsorption
C
O
4
1 13
22
22 5
CO donates electrons from the s orbital to the metal
metal donates electrons back to the anti-bonding orbital of CO
Donation
Back-donation (BD)
• Low CO coverage: CO depends on thegeometry of adsorption site (face order:terrace – corner – edge) – BD is strong
• High CO coverage: CO depends on dipole-dipole interactions – BD is weak
metal
Lewis acid
CO adsorption
2300 2200 2100 2000 1900 1800 1700
abs.
uni
ts
wavenumber (cm-1)
gas phase
CO in organic solvent
CO@Pt
(red-) shift
4 cm-1 resolution
ro-vi spectrum0.5 cm-1 resolution
effect of bond order and condensed phase
Adsorbateassignments on powdersby comparison withreference exps. in UHV (single crystals)
CO adsorption
XPS
FTIR
detector
manipulator
LEED
MS
Empa
• Model studies – Surface science stainless steel UHV setup with flanges, pumps,
pressure gauges, etc. 10-10 to 10-11 mbar base pressure tools and components for preparation,
characterization, sample manipulation, resistiveheating
50 nm
• Model studies – Surface science
CO adsorption
2 nmnanoparticles
model systems
single crystals NiAl(110)
Al2O3
• Powders
CO adsorption
Pd/SiO2
Pd(111)Pd(100)
fcc(100)
fcc(111)
nanoparticle
Szanyi et al., J. Vac. Sci. Technol. A 11 (1993) 1969
• Powders
the larger the particles, the less CO adsorbs (intensity) the larger the particles, the less the available defects (nr. of signals)
CO adsorption
50 nm
800°C‐2h‐air
fresh
particle size
Pt disp
ersio
n
Pt/Al2O3
2100 2000 1900 1800 17000.0
0.2
absorbance (a.u.)
wavenumber (cm‐1)
0.1fresh
agedCOB
COL
intensity
site distribution
Pt
PtPt
COTEM
How does the CO stretching frequency shift when a Pt surface is covered with hydrogen or oxygen?
CO
H H H H H H
Pt
CO
O O O O O O
Pt
CO adsorption
CO
Pt
Lear et al., J. Chem. Phys. 123 82005) 174706
size confirmed by TEM
CO adsorption
CO adsorption
Carenco, Chem. Eur. J. 20 (2014) 10616
M-M bond distance increasein EXAFS
time dependent COL/COB ratio
COL
COB
COL1COL2
time
COgas
Y. Soma-Noto et al., J. Catal. 32 (1974) 315
0.01 Torr CO0.5 Torr CO Pd
Ag
CO adsorption
Pd/Ag alloy on SiO2
Pearce and Sheppard, Surf. Sci. 59 (1976) 205
Total dipole = 0
Total dipole = 2
The surface selection rule
AS(OCO) S(OCO) 1500 cm-1
O O
R
x
yz
O O
Rmetal metal
The surface selection rule
• Carboxylate groups
Greenler et al. Surf. Sci. 118 (1982) 415
Note that the selection rule can break down for particles smaller than ca. 2 nm
L (Langmuir)= exposure of 10-6 Torr gas for 1 s
wavenumber (cm-1)
trans
mitta
nce (
%)
Haq et al., J. Phys. Chem. 100 (1996) 16957; Preuss et al., Phys. Rev. B 73 (2006) 155413
Pt(111)/90 K
Reflection-absorption (IRRAS)
• Also RAIRS; specular/external reflection method
higher T
Reflection-absorption (IRRAS)
Haq et al., J. Phys. Chem. 100 (1996) 16957
Chesters et al., Surf. Sci. 187 (1987) L639
94 K
5 L C2H4
Reflection-absorption (IRRAS)
• Adsorption of ethylene
Perpendicular polarizationParallel polarization
RsRp ‐ = RDifference
surface + gas gas surface
The surface spectra are often shown in R/R (R=Rs+Rp)
Perpendicular (s-) polarization (y-axis)
Parallel (p-) polarization (x, z-axis)
yx
z
CaF2 window
IR light
Gas inlet
Gas outlet
Sample
Heating element
Urakawa et al., J. Chem. Phys. 124 (2006) 054717
Phase-modulation IRRAS (PM-IRRAS)
• Generation of 2 polarizations (photoelastic modulator) excellent gas-phase compensation non-UHV experiments pssible highly sensitive, time-resolved studies possible
• Quality and quantity of acid sites│Criteria unequivocal analysis of intermolecular interaction selective interaction with acidic or basic sites sufficient accuracy in frequency shift determination high (and available) extinction coefficients of adsorbed probe appropriate acid (base) strength to induce interaction high specificity (allow discrimination between sites with different strength) - Use
different molecules ! small molecular size - Use different molecules ! pyridine (smaller channels) and picoline (larger channels or surface only)
low reactivity under exp. conditions
• Examples acidity of zeolites with different channel sizes acid sites located in all channels
Acid sites
• Adsorption of NH3
Probe molecules
15002000250030003500wavenumber (cm-1)
abs.
uni
ts
1000
NH3(g)
TiO2
WO3-TiO2
• Adsorption of NH3
Probe molecules
15002000250030003500wavenumber (cm-1)
20002050
V=O W=O
abs.
uni
ts
1000
NH3
TiO2
L-NH3
B-NH4+
WO3-TiO2
V2O5-TiO2
V2O5-WO3-TiO2
Acid sites
BASBrønsted sites(protic)
LASLewis sites(aprotic)
OH
base
base
M
base
M
OH
base+
OH
base
HBhydrogen bond
• Adsorption of NH3 V2O5/WO3-TiO2
Acid sites
wavenumber (cm-1)
abs.
uni
ts (a
.u.)
500 ppm NH3 + 5 vol% O2
4000 3500 3000 2500 2000 1500 1000
0.1
0 s
LNH3
BNH3
NH4+: 1440 cm-1NH3: 1605 cm-1
250°C
LNH3
V
Ti
BNH3BAS
V
Ti
LAS
Vittadini et al., J. Phys. Chem. B Lett. 109 (2005) 1652
Arnarsson et al., PCCP 18 (2016) 17071
Acid sites
• Probe molecules for the study of localization of active sites in microporous materials
acetonitrile methanolamine
benzonitrilepivalonitrile
pyridine 2,6-dimethyloyridine
dimethyl amineo-xylene
p-xylene
m-xylene
trimethylamine
2,2 - diphenylpropionitrile
di-tert-butyl-pyridine
o-toluonitrile
isobutyronitrilepropionitrile
Busca, in Metal Oxide Catalysis, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim (2009) 95
• Pyridine Transmission
Acid sites
+
3800 3700 3600 3500
abso
rban
ce (a
.u.)
wavenumber (cm-1)
BAS
1700 1600 1500 1400 1300wavenumber (cm-1)
BAS
LAS
Liu et al., J. Phys. Chem. C 121 (2017) 23520
solution