Diffuse Reflectance IR and UV-vis Spectroscopy

Diffuse Reflectance IR and UV-vis Spectroscopy

Modern Methods in Heterogeneous CatalysisFriederike Jentoft

December 10, 2004

ACFHI

ACACFHI

© 2004 F.C. Jentoft, Fritz-Haber-Institut der Max-Planck-Gesellschaft

Outline

I. Background

1. UV-vis, NIR and IR spectroscopy

2. Transmission

3. Specular & Diffuse Reflectance4. Mie and Rayleigh Scattering Theories5. Kubelka-Munk Theory

II. Experimental & Examples1. White Standards2. Integrating Spheres3. Mirror Optics4. Fiber 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)


Interaction of Light with Sample

v how to extract absorbing 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


How to Deal with Reflection (1)

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

incident light


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


transmitted light


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


Transmittance

if we manage to eliminate reflection:1 = α + τ

in 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 + ρ

α 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−=

I0 I0II

=τ

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


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


4000 3500 3000 2500 2000 1500 1000

0

10

20

30

40

50

161216

32

1625

2119

2040

1379

1257

2118

2036

1934

H4PVMo11O40

Cs2H2PVMo11O40

Cs3HPVMo11O40

Cs4PVMo11O40

Tran

smis

sion

(%)

Wellenzahl (cm-1)

Limitations of Transmission Spectroscopy

CaF2-Window-

Absorption

self-supporting wafers


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 possible


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 & Diffuse Reflection

Reflection of radiant energy at boundary

surfacesmirror-type

(polished) surfacesmat (dull, scattering)

surfaces

mirror-type reflectionmirror reflectionsurface reflectionspecular reflectionreguläre Reflexion

gerichtete Reflexion

reflecting power called“reflectivity”

reflecting power called“reflectance”

multiple reflections at surfaces of small

particles


Specular Reflection (Non-Absorbing Media)

v fraction of reflected light increases with α

v β depends on α and ratio of the refractive indices (Snell law)

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°


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!


Light Reflection from Powders: Why is the Distribution Isotropic?

particle diameter d > λ

v light partially reflected on elementary mirrors (crystal facets) inclined statistically at all possible angles to the macroscopic surface

v light penetrates into sample, undergoes numerous reflections, refractions and diffraction and emerges finally diffusely at the surface

particle diameter d ≤ λ

v scattering occurs

v Mie’s theory for single scattering says distribution is not isotropic

v multiple scattering should produce isotropic distribution


The Rayleigh Scattering Theory

v according to Rayleigh (Gans, Born)

v electromagnetic wave induces vibrating dipole in molecule (i.e. d << λ), electrons assumed to be in phase, molecular antenna

v dipole emits electromagnetic waves in all directions, i.e. the primary wave is being scattered

40

1λ

∝scatI


The Mie Scattering Theory

v according to the Mie formula, there is no principle distinction between reflection, refraction, and diffraction on one hand and scattering on the other hand

v rigorous theory for single scattering of a planar wave at spherical particles, both dielectric and absorbing, of any size

v in the limit of infinite diameter, the Mie theory approaches a scattering distribution as can be obtained from the law of geometrical optics by dealing separately with reflection, refraction, and diffraction


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

sk

RRRF =

−=

∞

∞∞ 2

)1()(2

2 constants are needed to describe the reflectance:absorption coefficient kscattering coefficient s

Kubelka-Munk functionremission function


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 spectroscopy

v in case of a dilute species is proportional to the concentration of the species (similar to the Lambert-Beer law):

scRF ε

∝′∞ )(


Deviations from & Limitations of the Kubelka-Munk Theory

incident radiation

v should be diffuse and not directed, however, it is assumed that after a few scattering events the distribution is isotropic

v directed incident radiation can be a problem on very rough surfaces: shadowing effects

any regular reflection

v will cause deviation

low reflectance values

v applicability of Kubelka-Munk theory questionable

v smallest error between 0.2 < R∞ < 0.6


Transmission vs. Reflection Spectroscopy

TA ln−=RRRF

2)1()(

2−=

0.0 0.2 0.4 0.6 0.8 1.00

1

2

3

4

5

6

Kubelka-Munk-function

Absorbance

Abs

orba

nce

/ Kub

elka

-Mun

k

Transmission / Reflectance

v for quantification and to be able to calculate difference spectra:calculate absorbance / Kubelka-Munk function


Transmission vs. Reflection Spectroscopy

200 400 600 8000.00.10.20.30.40.50.60.70.80.91.0

TC+A

TC+A

Tran

smitt

ance

or R

efle

ctan

ce

Wavelength / nm

TA ln−=RRRF

2)1()(

2−=

200 400 600 800

0.00.10.20.30.40.50.60.70.80.91.0

higher reflectance

lower reflectance

Kub

elka

Mun

k un

its

Wavelength / nm200 400 600 800

0.00.10.20.30.40.50.60.70.80.91.0

higher transmission

lower transmission

Abs

orba

nce

Wavelength / nm


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 light

v need to avoid specularly reflected light

v need reference standard (white standard)


“White” Standards

v KBr: IR (43500-400 cm-1)

v BaSO4: UV-vis

v MgO: UV-vis

v Spectralon: UV-vis-NIR


White Standard MgO

v ages rapidly


White Standard BaSO4

0 500 1000 1500 2000 2500

40

50

60

70

80

90

100%

Ref

lect

ance

Wavelength / nm

v reflects well in range 335 – 1320 nm

v humid!


White Standard Spectralon®

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


„Gray“ Standards

0 500 1000 1500 2000 25000

20

40

60

80

100

"20%="20Sp vs. Sp.

"50%="50Sp vs. Sp.

Ref

lect

ance

%

Wavelength / nm

100%=Sp vs. Sp.


The Collection Elements

v integrating spheres (UV-vis-NIR)

v fiber optics (UV-vis-NIR)

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


Integrating Spheres

v also called: photometer spheres, Ulbricht spheres

KK

MmPWP F

ffIIρ

ρρρ−

=1

120

F total internal surface of spherefm,,fM relative fractions of measuring surface and opening apertureρW absolute reflectance of wallρP absolute reflectance of sampleρK absolute reflectance of sphereS screen blocks regular reflection

PP

WPEM

K Ff

Ffff

ρρρ +

++

−= 1


Integrating Spheres: Coatings

v BaSO4: UV-vis

v MgO: UV-vis

v Spectralon: UV-vis-NIR

v Au: 800 nm to MIR


In Situ Cell & Integrating Sphere

v beam path equivalent for sample and reference beam

v spectrometer background correction can be used

v window (Suprasil or Infrasil): absorbs and reflects light

top view


sulfated zircona, 378 K5 vol% n-butane

Catalyst Deactivation: in situ UV-vis Spectroscopy

300 400 500 600 700 8000.0

0.2

0.4

0.6

0.8

1.0

difference

Ref

lect

ance

Wavelength / nm

after activation after reaction

v formation of additional bands during n-butane isomerization

v band at 310 nm, allylic cations? Analyzed using apparent absorption A=1-R

v spectroscopic and catalytic data can be correlated

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0 60 120 180 240 300 360

0

50

100

150

200

250

300 isobutane propane

Rat

e of

form

atio

n / µ

mol

g-1 h

-1Time on stream / min

Ban

d ar

ea a

t 310

nm

/ a.

u.

sulfated zirconia, 358 K5 vol% n-butane


Mirror Optics

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

v good in IR

v some disadvantages in UV-vis


How to Avoid Specular Reflection (1)

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°


The Harrick Praying Mantis

v first ellipsoidal mirror focuses beam on sample

v second ellipsoidal mirror collects reflected light

v 20% of the diffusely reflected light is collected

alignment can be tested

v align with flat mirror (= only specular reflection): minimize throughput

v align with tilted mirror: maximize throughput


Mirror Optics: Throughput in MIR Range

v throughput measured with tilted mirror

v mirror optics work well in the IR range….


Mirror Optics: Throughput in UV-vis-NIR Range

v aluminum coating of mirrors absorbs light! Here: 7 mirrorsv spectra will be more noisy in region of low throughput

0 500 1000 1500 2000 25000

10

20

30

40

50

60

70

80

tilted mirror in sampling position

BC (100%): empty versus emptyCBM=90integration time NIR=0.24, UVvis=0.2 data points 1 per 1 nm (ca. 240 nm/min)slit 2 nm, gain (NIR) 1

8 June, 2004, Lambda 950, PERKIN ELMERAlignment Harrick Praying Mantis

% R

efle

ctan

ce

Wavelength / nm


The Harrick Praying Mantis Reaction Chamber

v quartz windows (high pressure version available)

v up to 873 K (lot T version with Dewar available)

v powder bed rest on stainless steel grid: flow through bed

v low dead volume


Mirror Optics: UV-vis-NIR Range with Quartz Dome and Spectralon

v very low light yield

v need spectrometer that works well under these conditions

0 500 1000 1500 2000 25000

10

20

30

40

50

60

70

80

tilted mirror in sampling position reaction chamber with spectralon and dome

BC (100%): empty versus emptyCBM=90integration time NIR=0.24, UVvis=0.2 data points 1 per 1 nm (ca. 240 nm/min)slit 2 nm, gain (NIR) 1

8 June, 2004, Lambda 950, PERKIN ELMERAlignment Harrick Praying Mantis%

Ref

lect

ance

Wavelength / nm


v an attenuator may be necessary in the reference beam

v artifacts from change of optical components (gratings, filters, detectors, lamps) may become visible

0 500 1000 1500 2000 25000123456789

10

tilted mirror in sampling position reaction chamber with spectralon and dome

% R

efle

ctan

ce

Wavelength / nm

Mirror Optics: UV-vis-NIR Range with Quartz Dome and Spectralon


Example: Catalyst Activation

v UV-vis NIR gives information on valences and water content

0 600 1200 1800 24000

40

80

120

160

Temperature / °C 64 66 100 125 148 170 197 222 244 267 292 314 344 366 393 416 441 465 488 498 500 500

BC with different parameters!

BC (100%): spectralon + dome versus 10%TCBM=90integration time NIR=0.08, UVvis=0.08 data points 1 per 1 nm (ca. 570 nm/min)slit 2 nm, gain (NIR) 2

9 June, 2004, Lambda 950, PERKIN ELMERcat.: AH222, heating in N

2 from 50-500°C (5 K/min)

% R

efle

ctan

ce

Wavelength / nm


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

1.2

Tran

smitt

ance

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 appearance

v 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


n-Butane Isomerization over MnSZ: In Situ DRIFTS

v bands at 1600 and 1630 cm-1 increase

v also range of C=C stretching vibrations, but corresponding CH vibrations not ob served

v water bending vibration

1700 1650 1600 1550

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Kub

elka

-Mun

k un

its

Wavenumber / cm-1

16301600

Time on stream

1 kPa n-C4 323 K

0

10

20

30

40

50

0 10 20 30 40 50 600

50

100

150

200

250

Isob

utan

e fo

rmed

/ µm

ol g-1

h-1

Band area (1600+1630 cm-1) / cm-1

2% Mn-promoted sulfated zirconia, 323 K

sulfated zirconia, 358 K

v Rate of isomerization proportional to the amount of water formed (induction period)


Fiber Optics

v light conducted through total reflectance

v 6 around 1 configuration: illumination through 6 (45°), signal through 1avoids collection of specularly reflected light

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


Fiber Optics

v fibers need to be coupled into spectrometer beam bath

v problems: losses at couplers

picture: Hellma (http://www.hellma-worldwide.de)example shows transmission fiber


Summary


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

Documents

Diffuse Reflectance IR and UV-vis Spectroscopy