good diffraction data - profexprofex.doebelin.org/.../2019/10/good-diffraction-data.pdf · good...

$: Good Diffraction Data - Profexprofex.doebelin.org/.../2019/10/Good-diffraction-data.pdf · Good Diffraction Data Complex phase composition Large number of phases Low-symmetry phases$
Nicola Döbelin

RMS Foundation, Bettlach, Switzerland

Good Diffraction Data

BGMN/Profex User Meeting 2019

Good Diffraction Data2


• Very poor sample preparation

(a few grains on a piece of glass?)

• Problematic instrument setup

for low-angle range

• Texture

• «Refinable»

Good Diffraction Data

Complex phase composition

Large number of phases

Low-symmetry phases

Low crystallinity / glass / organic content

Fluorescence

Sample preparation

Graininess

Texture

Height displacements

Instrument setup

S/N ratio

Background

Angular resolution(peak width / asymmetry)

5

Inherent to the sample

Can (should)be minimized

Digital Diffractometers6

Transmission GeometryDebye-Scherrer Geometry

Glass Capillary

FoilFluid Cell

Capillaries are ideal for:• Light atoms (Polymers, Pharmaceuticals)• Small amounts• Hazardous materials• Air-sensitive materials

Use characteristic radiation withlow absorption coefficient

Flat powder sample

Reflective GeometryBragg-Brentano Geometry

Reflective Geometry is ideal for:• Absorbing materials (Ceramics, Metals)• Thin films• Texture analysis

Use characteristic radiation withhigh absorption coefficient

Bragg-Brentano Parafocusing Diffractometer7

Sample

X-raytube

Detector

Goniometercircle

Divergenceslit

Sollerslit

Sollerslit

Anti-Scatterslit

Beammask

Anti-Scatterslit

Receivingslit

Focusingcircle

Kβ Filter

Typical Configuration(with Kβ filter)

Bragg-Brentano Parafocusing Diffractometer8

Sample

Detector

Sollerslit Anti-Scatter

slit

Receivingslit

Secondarymonochromator

Typical Configuration(with secondary monochromator)

Modern instruments are modular.Configuration can be changed easily.PANalytical: «PreFIX»Bruker: «SNAP-LOCK»

Instrument Configuration

Many optical elements = many options to optimize data quality

How to find the best configuration?

9

Tube

Soller Slits

ProgrammableDivergence Slit

Beam Mask

Sample Stage«Spinner»

Anti-ScatterSlit

ProgrammableAnti-Scatter Slit

Soller Slits

Ni-Filter

Detector

Divergence Slit and Beam Mask10

Sample

X-raytube

Detector

Goniometercircle

Divergenceslit

Sollerslit

Sollerslit

Anti-Scatterslit

Beammask

Anti-Scatterslit

Receivingslit

Focusingcircle

Kβ Filter

Optimum Settings: Divergence Slit & Beam Mask11

Irradiated length

Beam divergence

Beam width

Controlled byDivergence slit

Controlled byBeam mask

θ

Optimum Settings: Divergence Slit12

20° 2θ

80° 2θ

Fix divergence slit:

Variable / Automatic divergence slit:

80° 2θ

- Constant intensity+ No mechanical parts

+ Gain of intensity at high 2θ- Backlash in slit mechanics

Fixed vs. Variable Divergence Slit13

20 30 40 50 60 70 80 90

Inte

nsity [a

.u.]

Diffraction Angle [°2]

fix

variable

More than 2x higher intensityat 90° 2θ with variable DS

Divergence Slit: Irradiated Length14

30.2 30.3 30.4 30.5 30.6 30.7

0

500

1000

1500

2000

2500

3000

Inte

nsity [co

un

ts]


5mm

10mm

15mm

Soller Slits: 0.02 rad, Beam Mask: 10mm

30.2 30.3 30.4 30.5 30.6 30.7

0

20

40

60

80

100

Inte

nsity [%

]


5mm

10mm

15mm

15 mm10 mm5 mm

Beam Mask15

30.2 30.3 30.4 30.5 30.6 30.7

0

500

1000

1500

2000

2500

3000

3500

Inte

nsity [co

un

ts]


5mm

10mm

20mm

Soller Slits: 0.02 rad, Irradiated Length: 10mm

30.2 30.3 30.4 30.5 30.6 30.7

0

20

40

60

80

100

Inte

nsity [%

]


5mm

10mm

20mm

20 mm10 mm5 mm


Sample holder

Sample surface

For phase quantifications:- Variable divergence slit- Adjust «irradiated length» and beam

mask for maximum illumination- Avoid beam spill-over!

Irradiated area

Correct! Wrong! Wrong!

Reduce «irradiated length»of divergence slit

Use a smaller Beam Mask

For structure refinements:- Fix divergence slit- Adjust «slit size» and beam mask for

maximum illumination at lowest 2θ- Avoid beam spill-over!


Using sample holders of various sizes?

Match your Divergence Slit and Beam Mask!

Or else: Waste of intensity Beam spill-overor

Soller Slits / Collimators18

Sample

X-raytube

Detector

Goniometercircle

Sollerslit

Sollerslit

Anti-Scatterslit

Anti-Scatterslit

Receivingslit

Focusingcircle

Kβ Filter

30.2 30.3 30.4 30.5 30.6 30.7

0

20

40

60

80

100

Inte

nsity [%

]


0.02rad

0.04rad

Soller Slits / Collimators19

30.2 30.3 30.4 30.5 30.6 30.7

0

1000

2000

3000

4000

5000

6000

7000

Inte

nsity [co

un

ts]


0.02rad

0.04rad

In primary & secondary beam, Beam Mask: 10mm, Irradiated Length: 10mm

0.04 rad0.02 rad

Asymmetry more pronounced at low 2θ, less at high 2θ

Anti-Scatter Slits20

Sample

X-raytube

Detector

Goniometercircle

Anti-Scatterslit

Anti-Scatterslit

Receivingslit

Focusingcircle

Kβ Filter

Anti-Scatter Slits and Beam Knifes21

Sample

Divergenceslit

Anti-Scatterslit

Removes radiation scatteredat air molecules

=Lower background signal

In primary beam:Same settings as divergence slit

Beam knife


Secondary beam

Point detector (0D):Same settings as divergence slit

Linear / area detector (1D / 2D):Wide open or removed


34.8 34.9 35.0 35.1 35.2 35.3 35.4 35.5

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

Inte

nsity (

Co

un

ts)

Diffraction Angle (°2theta)

9.00 mm

4.50 mm

2.25 mm

34.8 34.9 35.0 35.1 35.2 35.3 35.4 35.5

0

20

40

60

80

100

Inte

nsity (

%)


9.00 mm

4.50 mm

2.25 mm

Soller Slits: 2.5 deg, Irradiated Length: 15mm, LynxEye XE in 1D mode

Blocking detectorchannels

Receiving Slit24

Sample

X-raytube

Detector

Goniometercircle

Receivingslit

Focusingcircle

Kβ Filter

Receiving Slit25

34.8 34.9 35.0 35.1 35.2 35.3 35.4 35.5

0

200

400

600

800

1000

1200

1400

1600

1800

2000

2200

2400

Inte

nsity (

Co

un

ts)


0.075 mm

0.225 mm

0.525 mm

34.8 34.9 35.0 35.1 35.2 35.3 35.4 35.5

0

20

40

60

80

100

Inte

nsity (

%)



Detector Opening (PSD Opening)26

Sample

X-raytube

Detector

Goniometercircle

Focusingcircle

Kβ Filter


Maximum PSD opening

Openingangle

Minimum PSD opening

Position-sensitive detectors (1D, 2D)


34.8 34.9 35.0 35.1 35.2 35.3 35.4 35.5

0

20

40

60

80

100

Inte

nsity [%

]

Diffraction Angle [°2theta]

2.36°

1.00°

0.10°

0.01°

34.8 34.9 35.0 35.1 35.2 35.3 35.4 35.5

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

Inte

nsity [C

ou

nts

]

Diffraction Angle [°2theta]

2.36°

1.00°

0.10°

0.01°


Summary: Optical Elements29

Optical Element Effect (Too) Small (Too) Large

Divergence Slit Adjusts beam length on the sample

Loss of intensity Beam spills oversample

Beam Mask Adjusts beam width on the sample

Loss of intensity Beam spills oversample

Soller Slit Reduces peak asymmetry Loss of intensity,Better resolution

More asymmetry,Less resolution

Anti-Scatter Slit Reduces backgroundsignal

Loss of intensity High background

Receiving Slit Adjusts peak width / resolution

Loss of intensityBetter resolution

Loss of resolutionHigher intensity

PSD Opening Adjusts the number ofactive detector channels

Loss of intensity -

IntensityResolution

Filters and Monochromators30

Sample

X-raytube

Detector

Goniometercircle

Focusingcircle

Kβ Filter

Filters and Monochromators31

Secondarymonochromator

Kβ Filter

Kβ Filter

Energy-dispersivedetector

Summary: Filters and Monochromators32

Optical Element Filter effect Effect on Kα Intensity Comment

Primary beammonochromator

Eliminates Kβ Strong lossWorks with 1D / 2D

detectors

Secondary beam monochromator

Eliminates Kβ peaksEliminates Fluorescence

Strong loss 0D detectors only

Kβ Filter Reduces Kβ peaks Moderate loss

Can be combinedEnergy dispersive Detector

Reduces Kβ peaksEliminates Fluorescence

Little loss

31.5 34.5 35.0 35.5

0

200

400

600

800

1000

1200

1400

1600

1800

2000


Kbeta Filter 0.0125 mm

31.5 34.5 35.0 35.5

0

200

400

600

800

1000

1200

1400

1600

1800

2000


Digitally filtered

31.5 34.5 35.0 35.5

0

200

400

600

800

1000

1200

1400

1600

1800

2000


Digitally filtered + Kbeta Filter 0.0125 mm

+ =

Sample Preparation33

Potential Problems:- Graininess- Micro-absorption- Texture- Crystallite size- Sample height displacement- Surface roughness- Sample transparency

Graininess34

http://people.physics.anu.edu.au/~web107/

Single crystals generatespotty diffracted rays.

Fine powders generatesmooth diffraction rings.

Graininess35

Spotty diffraction rings

Very weak peak

Very strong peak

Grainy samples:- non-reproducible intensities- «phantom» peaks- «missing» peaks

The same sample, atthe same 2θ position,but different intensities!

Graininess: Rocks in Dust36

«Rocks in Dust»:A few large crystals ina fine matrix

Usually invisible, but ifscanned: Strong peaksout of nowhere!

http://www.diamond.ac.uk/

Graininess38

Reducing graininess:- Grinding / milling- Adjust divergence slit and beam mask for largest possible irradiated area

(= more particles contribute to diffraction pattern)- Use spinning sample stage

(= better randomisation)- Counting time per step ≥ 1 revolution of samples stage spinner

Few diffracting crystallites Many diffracting crystallites


topaz bearing granite, samples ball-milled < 63 µm and hand-ground < 20 µm

25 30 35 40

0

100

200

300

2 Theta / ° ( Scan Axis: 2:1 sym. const. area )

Inte

nsity /

cp

s

Comparison of 2 Scans

kfsp 002/040preferredorientation!

plag 002/040topaz 230

„rock in the dust“

< 63 µm

< 20 µm

Courtesy of R. Kleeberg

Micro-absorption40

Phase 1: High absorptioncoefficient for X-radiation

Phase 2: Low absorptioncoefficient for X-radiation

X-rays

X-rays

Micro-absorption41

Small particles absorb insignificantpart of the radiation. Volumes of interaction with

phases 1 & 2 are representativefor phase composition

Strong attenuation by phase 1Large particles absorb significantpart of the radiation. Small volume of interaction

Weak attenuation by phase 2 Large volume of interaction

Micro-absorption42

Reducing micro-absorption:- Grinding / milling to reduce particle size

Micro-absorption occurs in samples with…- … large particles (not crystallites!)- … phases with large contrast in absorption coefficients

Texture, Preferred Orientation43

SEM Images: L. Galea, RMS Foundation

Platelets, Needles, Fibers, Whiskers

Random orientation Preferred orientation

Texture, Preferred Orientation44

Try to avoid orientation at the surface of the sample:

- Press powder without «rubbing» the surface- Use back- or side-loading sample holder- Disorder surface with textured stamp

- Various creative solutions can befound on the internet (involvingVaseline, hair spray, spray pistols, …)

PO can be corrected mathematically,but phase quantification will be biased.

Photo: Kristin Unger

Photo: Aron Knoblich

Summary: The Perfect Sample45

The perfect sample for Bragg-Brentano diffractometers:- Crystallites and particles of 1-5 µm size- Perfectly random orientation- Perfectly flat surface- Surface precisely centered in the goniometer- Not transparent (absorb all primary radiation)

Industry standard for automated XRD sample milling:McCrone Micronizing Mill (now distributed by Retsch)

General Rules for Quantitative Phase Analysis

Instrumental background should be low and linear, no “bumps” or even edges

Force the peak/background ratio as high as you can

Peak/noise ratio on maximum as well, but a noise better than the quality of profile description doesn’t make sense

Irradiate maximum of sample surface (improve grain statistics and intensity)

Register as much diffracted intensity as you can get without significant loss of resolution

Parallel beam techniques and high-resolution configuration are bad choices for QPA purpose

Don’t measure/evaluate angular ranges containing no or potentially biased phase Intensity information (very low angles, high angles)

Choose step width of approximately 1/5 of FWHM of the sharpest peak

46

Courtesy of R. Kleeberg


Good sample preparation…

…and good instrument settings…

…are crucial for successful

Rietveld refinements

good diffraction data - profexprofex.doebelin.org/.../2019/10/good-diffraction-data.pdf · good...

Documents