MAX-PLANCK-GESELLSCHAFT

ACFHI

From Surfaces to Three Dimensions:Experimental Methods in Catalysis

Friederike C. Jentoft

Department of Inorganic ChemistryFritz-Haber-Institut der Max-Planck-Gesellschaft

Faradayweg 4-6, 14195 Berlin

IMPRS “Complex Surfaces in Materials Science”

Block Course WS 05/06

October 13, 2005

Outline Part I

1. IntroductionComplexity of Heterogeneous Catalysts

2. Role of Surface and Bulk Structure

3. Active Sites

4. Physisorption, Porosity

5. ChemisorptionIsothermsIR SpectroscopyThermal Desorption SpectroscopyAdsorption Calorimetry

Questions:Type? Number / density?„Strength“ (interaction with a certain molecule)?

Examples for Surface Sites

Metal sites

Brønsted-sites

Lewis-sites

Oxide Surfaces

v can be terminated by O or Mx+ or mixture (may depend on environment)

v can be terminated by foreign ions

v most frequently surfaces feature OH groups:

HO

Zr

HO

Zr….Zr

HO

Zr..Zr..Zr

type I type II type III

v zeolite OH groups inside the poresHO

Si Al

OH groups can beBrønsted acids = proton donorsBrønsted bases = proton acceptors

Detection of OH-Groups

v IR spectroscopy, NMR spectroscopy

v band position or chemical shift delivers information

4000 3500 3000 2500 2000 1500 1000

0

10

20

30

40

50

161216

32

1625

2119

2040

1379

1257

2118

2036

1934

H4PVMo

11O

40Cs

2H

2PVMo

11O

40Cs

3HPVMo

11O

40Cs

4PVMo

11O

40

Tran

smitt

ance

(%)

Wavenumbers / cm-1

CaF2window

absorption

Lewis-sites coordinatively unsaturated (cus) metal cations (acidic, electron pair acceptors)oxygen anions (basic, electron pair donors)

Lewis Sites

v note: on a surface, acidic and basic sites (Brønsted and/or Lewis) can coexist (would neutralize each other in solution)

v only detectable in an adsorption experiment

OH O

H NHH

HIR, XPS, NMR, UV/Vis...shift of bands(probe/surface)desorption T

TDS/TPD

adsorption ∆ Hcalorimetry

+ NH3

Probing Sites by Chemisorption

Different methods deliver different informationv isotherms: number of sitesv spectroscopies: type, strength (but not number unless extinction

coefficient known)v calorimetry, TDS: number and strength (but not type)

Metal Surface Sites

v specific adsorption of probe on only one type of site (e.g. on metal and not on support): depends on strength of interaction of probe with metal/support sites, conditions (T, p) can be optimized

Metal particles

Support (e.g. oxide) surface

CO molecules

Measuring Metal Surface Area with H2 or CO

v if adsorption is specific, number of sites can be derived from isotherm(no site heterogeneity, no spill-over)

IR Spectra of CO Adsorption on SAPO at 77K

10-3 < p < 10-2 torr

Onida et al. J. Phys. Chem. B 45 (1997) 9244-9249

Hydroxyl vibrations Carbonyl vibrations

OH

OH C O

v different OH groups will experience different shifts, shift is a measure of the acid strength of the OH group

Adsorption of CO on Pt/Mordenite

2300 2250 2200 2150 2100 2050 2000 19500.000

0.001

0.002

0.003

2131

p(CO): ca. 0.40 mbar dj6norm 250°C unter Vakuumdl7norm 450°C unter Vakuumdl71norm 375°C unter H2

2195

2171

2110

2088

2050

2228

Abs

orba

nce

(nor

mal

ized

)

Wavenumber / cm-1

v multiple sites!

v need reference data (zeolite without Pt, Pt single crystal data)

NH3 TPD

100 200 300 400 500 600 700 800

sulfated ZrO2M

S-In

tens

ity [a

.u.]

T [°C]

v NH3 is a strong base and a probe for acidic sites (Brønsted and Lewis)

v presence of sulfate on the zirconia surface generates strongly acidic sites

NH H

H

CO2 TPD

v CO2 is a probe for basic sites (oxygen anions)

v ZrO2 is a basic oxide, but the basic sites can be completely coveredwith sulfate

100 200 300 400 500 600 700

sulfated ZrO2

MS-

Inte

nsity

[a.u

.]

T [°C]

Integral Heat of Adsorption

v A probe may chemisorb on different sites under production of different heats of adsorption

v If all sites are covered at once, the evolved heat will be an integral heatv if the number of adsorbed molecules is known, an average / mean heat

of adsorption can be calculated

Differential Heats of Adsorption

v A general concept: example dissolution“first”, “last” heat of dissolution = differential heats

v Sites can be covered step by step, e.g. 1.2. 3.

v Differential heats of adsorption as a function of coverage can be determined

ATdiff nQq ,int )/( δδ=

Adsorption Calorimetry

v the sorptive must be introduced stepwise, i.e. at constant temperature, the pressure is increased slowly

v for each adsorption step, the adsorbed amount must be determined(isotherm)

v for each adsorption step, the evolved heat must be determined

v the differential heat can then be determined by division of evolved heat through number of molecules adsorbed in a particular step

20 21 22 23 24 25 26 27 28 29 300.2

0.4

0.6

0.8

1.0

1.2

1.4

0.0

0.5

1.0

1.5

2.0

2.5

3.0Th

erm

osig

nal /

V

Time / h

Equi

libriu

m p

ress

ure

/ hPa

vgeneration of adsorption isotherm

vdifferential heats of adsorption

Adsorption Calorimetry Raw Data: Equilibrium Pressure and Thermosignal

0 1 2 3 4 5 6 70.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

Act. 473 KAct. 573 KAct. 723 KAct. 723 KA

dsor

bed

amou

nt /

mm

ol g-1

Isobutane partial pressure / hPa

Surface Sites: Effect of Activation

v with increasing activation temperature, the differential heats of adsorption increase

Isotherms313 K

0.00 0.01 0.02 0.03 0.04 0.050

20

40

60

80

100Act. 473 KAct. 573 KAct. 723 KAct. 723 K

Diff

. hea

t of a

dsor

ptio

n / k

J m

ol-1

Adsorbed amount / mmol g-1

Ammonia Adsorption on Heteropolyacids

v H3PW12O40 * x H2O reaction with ammonia

Diffuse Reflectance IR and UV-vis Spectroscopy

Outline Part II

1. Interaction of Light with a Sample2. Transmission & Lambert-Beer Law3. Scattering and Reflection4. Schuster-Kubelka-Munk Theory5. Experimental & Examples

White StandardsIntegrating SpheresMirror OpticsFiber Optics

UV- vis- NIR – MIR

v electronic transitions, vibrations (rotations)

v diffuse reflectance UV-vis spectroscopy (DR-UV-vis spectroscopy or DRS)

v diffuse reflectance Fourier-transform infrared spectroscopy (DRIFTS)

Type of transition

Spectral range

Molecular rotation Electronic excitation

X-ray radiation

InfraredRadio waves Micro waves

F M Nvisvis UV

6.2 to 1.5Energy /eV

200 bis 800(700-1000) bis 30003000 bis(25000-40000)

Wavelength / nm

50000 bis 1250012500 bis 33003300 bis 250Wavenumber/ cm-1

UV-visNear IR (NIR)Mid IR (MIR)

Molecular vibrations

Spectroscopy with Powder Samples

v how to measure spectra of catalyst (+ adsorbate)?v absorption as a function of wavelength, qualitatively and

quantitatively

ZrO2

Interaction of Light with Sample

v how to extract absorption properties from transmitted light?v how to deal with reflection and scattering?

incident light

reflection at phase boundaries

transmitted light

scattered light

absorption of light

(luminescence)

How to Deal with Reflection (1)

v fraction of reflected light can be eliminated through reference measurement with same materials (cuvette+ solvent)

incident light

reflection at phase boundaries

transmitted light

How to Deal with Reflection (2)

v fraction of reflected light can be eliminated through variation of sample thickness: number of phase boundaries remains the same

incident light

reflection at phase boundaries

transmitted light

Transmittance

if we manage to eliminate reflection:1 = α + τin this case, the absorption properties of the sample can be calculated from the transmitted light

I0 I

0II

=τ

Sample

if luminescence and scattering are negligible:1 = α + τ + ρα absorptance, absorption factor, Absorptionsgradτ(T) transmittance, transmission factor, Transmissionsgradρ reflectance, reflection factor, Reflexionsgrad

Transmitted Light and Sample Absorption Properties

separation of variables and integration

sample thickness: l ∫∫ −=lI

I

dlcI

dI

00

κ

decrease of I in an infinitesimally thin layer

lcII

κ−=0

ln

dlcIdI κ−=c: molar concentration of absorbing species [mol/m-3]κ: the molar napierian extinction coefficient [m2/mol]

I0 I

0II

=τ

Transmitted Light and Sample Absorption Properties

ατ κ −=== − 10

lceII

)1ln( ακ −−=== lcBAe

)1log(10 αε −−== lcA

extinction E (means absorbed + scattered light)absorbance A (A10 or Ae)

optical density O.D.

all these quantities are DIMENSIONLESS !!!!

Lambert-Beer Law

standard spectroscopy software uses A10!

napierian absorbanceNapier-Absorbanz

(decadic) absorbancedekadische Absorbanz

Scattering

v scattering is negligible in molecular disperse media (solutions)

v scattering is considerable for colloids and solids when the wavelength is in the order of magnitude of the particle size

Wavenumber WavelengthMid-IR (MIR) 3300 to 250 cm-1 3 to (25-40) µmNear-IR (NIR) 12500 to 3300 cm-1 (700-1000) to 3000 nmUV-vis 50000 to 12500 cm-1 200 to 800 nm

v scattering is reduced through embedding of the particles in media with similar refractive index: KBr wafer (clear!) technique, immersion in Nujol

1100 1000 900 800 7000

20

40

60

Wavenumbers (cm-1)

1083

10731059

983960

896862 788

(NaCl)(KBr)(CsCl)

Tran

smis

sion

(%)

But…….Reaction with Material Used for Embedding

H4PVMo11O40

v reaction with diluent possiblev dilution usually not suitable for experiments at high T/ with reactive gases

Limitations of Transmission Spectroscopy

v transmission can become very low! even in IR regime!

v catalysts are fine powders (high surface area)

v embedding (KBr) not desirable for in situ studies (heating, reaction)

5000 4000 3000 2000 10000.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07sulfated ZrO2 after activiton at 723 K

Tran

smitt

ance

Wavenumber / cm-1

Can We Use the Reflected Light?

v instead of measuring the transmitted light, we could measure thereflected light

v can we extract the absorption properties of our sample from the reflected light?

Detector

Specular Reflection (Non-Absorbing Media)

v fraction of reflected light increases with αv β depends on α and ratio of the refractive indices (Snell law)v insignificant for non-absorbing media, for air/glass about 4%

incident beam reflected beam

refracted beam

n0

n1

n1>n0

α

β

surface SIDE VIEW

Specular Reflection: Angular Distribution

v the intensity of the specularly reflected light is largest at an azimuth of 180°

incident beam reflected beam

surfaceTOP VIEW

azimuth180°

azimuth<180°

Light scatteringdeflection of electromagnetic or corpuscular

radiation from its original direction

Raman-scattering

Compton-scattering

Brillouin-scattering

elastic

Scattering

Rayleigh-scatteringλ > d

wavelength dependent: ∼1/ λ4

no preferred direction

Mie-scatteringλ ≤ d

wavelength independentpreferentially in forward (and

backward) direction

inelastic

Rayleigh- and Mie-Scattering

Quelle: RÖMPP on line

d >> λ: Mie-Theory approaches laws of geometric optics

h·υ

d < λ: Rayleigh-Scatteringisotropic distribution

d = λ: Mie-Scatteringin forward and backward directions

d > λ: Mie-Scatteringpredominantly in forward direction

Diffuse Reflection

v intensity of diffusely reflected light independent of angle of incidence (isotropic)

v Due to multiple reflection, refraction, and diffraction inside the sample

Reflection of Radiation

mirror-type (polished) surfaces mat (dull, scattering)

surfaces

mirror-type reflectionmirror reflectionsurface reflection

equally intense in all directions (isotropic)

images: Wolfgang Böhme Sphere Optics Hoffman GmbH

Typical Catalyst Particles

v need theory that treats light transfer in an absorbing and scattering medium

v want to extract absorption properties!

ZrO2

ca. 20 µm

scanning electron microscopy image

The Schuster-Kubelka-Munk Theory

Assumptions

v incident light diffuse

v isotropic distribution (Lambert cosine law valid) , i.e. no regular reflection

v particles randomly distributed

v particles much smaller than thickness of layer

A Simplified Derivation of the SKM Function

I

J

IJR =

dlSJdlSIdlKIdI +−−=

dlSIdlSJdlKJdJ +−−=with K, S: absorption and scattering coefficient [cm-1]

Divide equations by I or J, respectively, separate variables, introduce R=J/I

Integrate via expansion into partial fractions

∫ ∫∞

=++−

R

R

ldlS

SKRR

dR

0 02 1)1(2

2

A Simplified Derivation of the SKM Function

SK

RRRF =

−=

∞

∞∞ 2

)1()(2

2 constants are needed to describe the reflectance:absorption coefficient Kscattering coefficient S

Kubelka-Munk functionremission function

assume black background, so that R0 = 0make sample infinitely thick, i.e. no transmitted light (typical sample thickness in experiment ca. 3 mm)

)2( SKKSKSR

+++=∞

for K→0 (no absorption) R∞→1, i.e. all light reflectedfor S→0 (no scattering) R∞→0, i.e. all light transmitted or absorbed

Kubelka-Munk Function

v mostly we measure R’∞, i.e. not the absolute reflectance R∞but the reflectance relative to a standard

v depends on wavelength F(R’∞)λ

v corresponds to extinction in transmission spectroscopyv in case of a dilute species is proportional to the concentration of the

species (similar to the Lambert-Beer law):

scRF ε

∝′∞ )(

Spectroscopy in Transmission

reference „nothing“= void, empty cell, cuvette with solvent

Catalyst

LightSource Detector

spectrum: transmission of catalyst vs. transmission of reference

v in IR transmission is widely applied for solids v in UV-vis transmission is rarely used for solids

Reflectance Spectroscopy

reference„white standard“

Catalyst

LightSource

Detector

spectrum: reflectance of sample (catalyst) vs. reflectance of standard

v need element that collects diffusely reflected lightv need to avoid specularly reflected lightv need reference standard (white standard)

White Standards

Spectralon® thermoplastic resin, excellent reflectance in UV-vis region

v KBr: IR (43500-400 cm-1)v BaSO4: UV-visv MgO: UV-visv Spectralon: UV-vis-NIR

The Collection Elements

v integrating spheres (UV-vis-NIR)

v fiber optics (UV-vis-NIR)

v mirror optics (IR and UV-vis-NIR)

Diffuse Reflection & Macroscopic Surface Properties

v randomly oriented crystals in a powder: light diffusely reflected

v flattening of the surface or pressing of a pellet can cause orientation of the crystals, which are “elementary mirrors”

v causes “glossy peaks” if angle of observation corresponds to angle of incidence

v solution: roughen surface with (sand)paper or press between rough paper, or use different observation angle!

Integrating Sphere

Image: Römpp on-line

Sample

Source

Detector

Gloss trap

Screen

v the larger the sphere the smaller errors from the ports

v the larger the sphere the lower the intensity onto the detector

v typically 60-150 mm diameter

v coatings: BaSO4, Spectralon

400 500 600 700 8000.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5K

ubel

ka-M

unk

units

Wavelength / nm

Diffuse Reflectance Spectrum: Green Sheet of Paper

How to Avoid Specular Reflection

v specular reflection is strongest in forward direction

v collect light in off-axis configuration

sample surface

SIDE VIEW TOP VIEW

sample surface

90°

120°

Praying Mantis Optical Accessory (Harrick)

v first ellipsoidal mirror focuses beam on sample

v second ellipsoidal mirror collects reflected light

v 20% of the diffusely reflected light is collected

vcan be placed into the normal sample chamber (in line with beam), no rearrangement necessary

Source

Detector

Diffuse Reflectance IR (DRIFTS): Graseby Specac Selector Optics and Environmental Chamber

v ZnSe window (chemically resistant)

v up to 773 K

v no flow through be (rest in cup)

v large dead volume (est. 100 ml)

Comparison Transmission - Diffuse Reflectance

5000 4000 3000 2000 10000.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.0

0.2

0.4

0.6

0.8

1.0

1.2Tr

ansm

ittan

ce

Wavenumber / cm -1

Ref

lect

ance

sulfated ZrO2 after activation at 723 K Transmission: self-supporting waferReflectance: powder bed

v spectra can have completely different appearancev transmission decreases, reflectance increases with increasing wavenumber

Comparison Transmission - Diffuse Reflectance

v vibrations of surface species may be more evident in DR spectra

5000 4000 3000 20003.03.23.43.63.84.04.24.44.64.85.0

0.0

0.2

0.4

0.6

0.8

1.0

Abs

orba

nce

A Nap

ier

Wavenumber / cm -1

Kub

elka

-Mun

k Fu

nctio

n

sulfated ZrO2 after activation at 723 K

Transmission: self-supporting waferReflectance: powder bed

OH vibrations

SO vibrations

Fiber Optics

v light conducted through total reflectancev fiber bundle with 6 around 1 configuration: illumination through 6 (45°), signal

through 1v avoids collection of specularly reflected light

Bilder: Hellma (http://www.hellma-worldwide.de) and CICP (http://www.cicp.com/home.html)

Literature

v W. WM. Wendlandt, H.G. Hecht, “Reflectance Spectroscopy”, Interscience Publishers/John Wiley 1966, FHI: 22 E 13

v G. Kortüm, “Reflectance Spectroscopy” / “Reflexionsspektroskopie” Springer 1969, FHI: missing