X-Ray Fluorescence (XRF)

Analysis

Lucky Eka Irawibawa

PT. Lab-Sistematika Indonesia
Application Specialist
X-Ray Fluorescence Spectrometer

Introduction
 History

• 1895: Wilhelm Röntgen discovers X-ray

• 1910: Barkla found characteristic X-ray emission spectra
• 1913: Moseley established the relationship between frequency
and atomic number identification
• 1913: Max von Laue discovered the diffraction of X-rays
• Early 1950s: fluorescence approach, commercial
instrumentation for materials characterisation
• Until the late 1960s: wavelength dispersive spectrometers
• Early 1970s: energy dispersive spectrometers
• Late 1980s-early 1990s: synchrotron
 What are X-Rays?

Electromagnetic radiation

Wavelength 0.01 nm - 10.0 nm

Energy 124 keV - 0.124 keV

g -rays X-rays UV Visual

0.001 0.01 0.1 1.0 10.0 100 200 nm

1nm = 10Å = 10-9m = 10-6mm

What are X-Rays?
 What we can do in XRF analysis ?
XRF is an elemental analysis method
・We can find the elements contained.
⇒ Qualitative analysis
Element range : Be – U ( depends on XRF system )
・We can obtain concentration of the elements.
⇒ Quantitative analysis
Concentration range : ppm - 100 %
Coating analysis : nm, µg/cm2

・We can obtain concentration of the elements, with

No Calibration and Non Standard Analysis
⇒ Semi-Quantitative
Difference between XRF & XRD

result
substance XRD XRF
Fe2O3 Fe2O3 Fe, O

FeO FeO Fe, O

Fe3O4 Fe3O4 Fe, O

 Advantage of XRF
• Simple sample preparation
• Non destructive analysis
• Safe and lower analysis cost
• Short total analysis time
• High repeatability/reproducibility
• Analysis with high precision for 5B to 92U
• Small personal error
• Qualitative, semi-quantitative, and quantitative analysis
• Wide dynamic range (ppm to 100 mass%)
 Applications of XRF Analysis
Steel Nonferrous Environment Electronics Chemicals
metal Sewage
Special alloy Optical disk Catalyst
Aluminum can Sea/river water
Surface-treated Shape memory alloy
Magnetic head Polymer
Ferro-alloy Soil, Sludge LCD, CRT Medicine
Copper alloy
Cast iron Air dust Magnet Fertilizer
Noble metals
Iron ore Industrial waste LSI Paint/Pigment
Nickel alloy
Plating bath Coal fly ash Memory Detergent
Solder
WEEE/RoHS Cosmetics

Mining Petroleum Ceramics Cement Others

Soil
Mineral and ore Diesel fuel Silicon nitride Raw material Plants
Rocks Gasoline Alumina Slag Tissue
Volcanic ash Kerosene Glass Coal fly ash Food
Heavy oil Refractories Cultural
Bio-fuel Glaze properties
Lube oil/Grease Clay Atomic power
plant
Coal
 Production of x-rays fluorescent
Generation of fluorescent X-rays

Fluorescent X-rays
Electron Ka La
N-shell Kb1 1) Primary X-ray collides
M-shell with electron in an atom
L-shell 2) Photo-electron is emitted
K-shell
(excitation)
3) Electron moves down from
outer shell
Nucleus
4) Fluorescent X-rays is
generated
Primary x-rays
 Production of x-rays

Characteristic radiation
 Wide variety of XRF product lines
High powered system
High power Measurement range: Be-U
sequential Tube-above / Tube-below
Micro-area/Mapping
Wavelength ZSX Primus series
dispersive

Bench-top Bench-top system

sequential Measurement range: O-U
Power: 200 W
XRF
analyzer Supermini200

Multi-channel
simultaneous High powered systems

Energy
dispersive
Simultix 14

NEX QC Single-channel Benchtop systems

NEX CG
NEX QC+ QuantEZ
Micro-Z ULS Mini-Z series
 What is the difference between
WDX and EDX?

WDX EDX
Wavelength Dispersive X-ray spectrometer Energy Dispersive X-ray spectrometer
Sample Detector Sample

X-ray tube
X-ray tube

Detector
Analyzing crystal
(Direct excitation)
WDX employs analyzing crystal with Fluorescent X-rays are directly
goniometer to detect specific wavelength of measured with higher resolution
fluorescent X-rays. detector.
 Analyzing Crystal in WD-XRF

Atomic N. 4 5 6 7 8 9 11 12 13 14 15 16 17 19 20 22 23 24 25 26 27 28 29 30 33 - 60
K Line Be B C N O F Na Mg Al Si P S Cl K Ca Ti V Cr Mn Fe Co Ni Cu Zn As - Nd
L Line 48Cd 56Ba 74W 82Pb
LiF(200)
LiF(220)
LiF(420)
PETH
GeH
RX25
RX35
RX40
RX45
RX61
RX61F
RX75
RX85

Good Available
Sequential System vs. Multi-channel System
WDXRF
Solar slit, Crystal,
selectable selectable

Scanning Fixed channels specialized to

Detectors,
Goniometer the element lines
selectable

Sequential system Multi-channel system

Flexible for wide range of elements Rapid analysis
 Sample Matrix
CHO matrix Oxides, light metals Heavy metals

Large
Precious metal Floor Standing XRF
Ore Iron ore Ferro alloy Special steel
Ash Glass
Spectral overlapping

Catalyst Cast iron, cast steel

Soil
Fertilizer Aluminum can
Pigment, paint
Waste water
WEEE/RoHS
Alumina BenchTop XRF /
Grease Biological Refractories Floor Standing XRF
Plating bath
Pharmaceutical Rock
Lubricating oil
Fuel oils Polymer
BenchTop XRF
Degree of heavy matrix Large
 X-Ray Fluorescence Spectrometer

Qualitative &
Semi-Quantitative Analysis
 Qualitative analysis

Fe-Kα Cr-Kα

Ni-Kα
Primary X-rays Sample contains Ni,Fe and Cr
Fluorescent
X-rays
Cr
Fe
Ni

Sample

Check energy/wavelength of peaks.

Identify what elements are contained.

 How to read/analyze qualitative
charts
1. Find the characteristic and Compton scattering
lines originating from the X-ray tube
2. Find peaks in large to small order and identify
spectrum name
3. Look for Kb when Ka is found, as well as La, Lb, Lg
4. Any information about sample can be helpful
in identifying peaks
 Qualitative chart

-Ka

Fe-Ka
1 h ② Identify next largest peak

Rh-Ka -Compton
R ① Rh characteristic and associated peaks (Kb)
and Compton lines
Rh-Kb Rh-K b1-Compton
Check if Ka/Kb intensity ratio
is reasonable (5/1)
③ Identify next largest
peak as well as associated
⑤ Continue to identify
X-ray intensity (a.u.)

peaks
next largest peak with
④ Is Ka/Kb ratio (5/1)?
associated peak
Sr-Kb Zr-Ka


⑥ Identify remaining peaks

Fe-Kb
Zr-Kb

Rb-Ka a

K
-
Pb-Lb

Zn-Ka
Zn-Kb

Cu-Ka
Pb-La

rS

Mn-Ka

Ti-Ka
0

2 angle
 Analysis of Dehydrated Cake
by Semi-Quantitavive

・We can obtain concentration of the elements, with

No Calibration and Non Standard Analysis
⇒ Semi-Quantitative

Dehydrated cake
Element Na Mg Al Si P S Cl K Ca Ti Cr
Method Balance Standard value 0.90 1.22 7.26 16.7 0.12 - - 1.14 3.90 0.40 -
Scatter FP Estimate Analysis value 0.69 1.28 8.20 15.9 0.12 0.19 0.06 1.12 3.69 0.38 0.02
FP C (carbon) Analysis value 0.51 0.93 5.89 11.4 0.08 0.14 0.05 0.80 2.61 0.27 0.01
FP O (oxygen) Analysis value 0.82 1.54 9.95 19.4 0.15 0.23 0.08 1.38 4.55 0.47 0.02

Element Mn Fe Cu Zn Br Rb Sr Y Zr Pb Balance
Method Balance Standard value 0.10 4.87 0.012 0.086 0.003 0.006 0.024 0.002 0.008 0.013 -
Scatter FP Estimate Analysis value 0.09 4.90 0.01 0.08 0.003 0.004 0.026 0.003 0.008 0.015 63.1
FP C (carbon) Analysis value 0.06 3.45 0.01 0.06 0.002 0.004 0.02 0.02 0.01 0.01 73.7
FP O (oxygen) Analysis value 0.11 6.04 0.02 0.10 0.004 0.006 0.03 0.00 0.01 0.02 54.9
 SQX Scatter FP Method

Without setting non-measurable elements, such as H, B, C or O, you can perform

accurate semi-quant analysis. Using the conventional FP software, accurate
analysis is difficult for industrial waste, such as soil or scale, unless you properly
set non-measurable elements. The Scatter FP Method estimates influence of non-
measurable elements using scatter X-ray intensities for correction. In addition,
theoretical intensity calculation of the Scatter FP Method incorporates “Geometry
Effect”, which corrects influence of a sample amount.

FP
method
Scatter FP
Known method
balance Unknown balance
component component
Soil
Oxide Scale
 Application: Rock
• Sample: Dolomitic Limestone
• Support film (Mylar) was used.
(C and O can not be analyzed.)
Unit: mass%

Analyte Std. value SQX Scatter FP FP (O balance)

Ca 21.4 21.2 24.0

Mg 12.7 11.4 12.7
Al 0.18 0.28 0.31
Si 0.52 0.70 0.79
K 0.09 0.10 0.12
Mn 0.012 0.013 0.015
Fe 0.19 0.21 0.24
Sr 0.006 0.006 0.007
 Application: Fly ash
• Sample: JCFA Coal Fly Ash
• Support film (polypropylene) was used.
• Loose powder method
Unit: mass%
Element Na Mg Al Si P S K Ca Ti
Std. value 1.69 1.27 12.8 23.6 0.26 0.196 1.05 6.36 0.79
SQX Scatter FP 2.14 1.08 12.6 20.3 0.25 0.28 1.04 6.23 0.79
FP (C balance) 1.20 0.59 6.74 10.7 0.13 0.14 0.534 3.19 0.40
Element V Cr Mn Fe Co Ni Cu Zn Ga
Std. value 0.0243 0.0075 0.05 3.64 0.0037 0.0032 0.0122 0.0063 (0.0045)

SQX Scatter FP 0.025 0.0112 0.0515 3.42 0.0049 0.0055 0.013 0.0064 0.0044
FP (C balance) 0.013 0.0057 0.0258 1.72 0.0025 0.0028 0.0066 0.0033 0.0022
Element As Rb Sr Y Zr Nb Ba Pb
Std. value 0.0029 0.0054 0.11 (0.0054) (0.0460) (0.0025) (0.1740) 0.0047
SQX Scatter FP 0.0035 0.0052 0.113 0.0046 0.031 0.0015 0.014 0.0038
FP (C balance) 0.0017 0.0027 0.0586 0.0024 0.017 0.0008 0.0687 0.0019
 Quick Registration of Matching Library

Merits of “Matching Library”

The analysis error caused by sample
conditions can be offset, to improve
analysis accuracy.
⚫ Chemical condition
⚫ Grain size effect
⚫ Mineralogical effect
⚫ Surface roughness

Easy Registration

25
Sample Preparation
Sample were ground using a ball mill to a powder of <200
mesh. The powder the converted into a pellet using a
hydraulic press at 20 tons.
X-Ray Fluorescence Spectrometer

Quantitative Analysis
 Quantitative XRF-Analysis

• Objective of quantitative X-Ray Fluorescence

Analysis: Determination of the elemental
composition of an unknown sample
(concentrations of elements and/or compounds)
• Spectrometer / measurement
process: "intensities" for X-ray
emission lines
e.g. Fe-Ka1: 37865 cps
• Last step to determine composition:
Convert measured intensities into concentrations
 Quantitative Analysis (empirical)
Calibration curve：
Intensity Coefficients a,b,c are determined
D
D
C
C

Intensity Unknown
of B B sample
unknown concentration
A A
sample

2θangle or energy concentration W (wt%)

W ：化学分析値
W =WaI=3 aI
+ bI+ +bIcI++cd
2 2 W：standard value
a,b,c ：係数
a, b, c,d ：coefficients
I
I：X-ray ：Ｘ線強度(kcps)
intensity (kcps)

Correlation between concentration and

intensity
 Selection of measurement lines
Element Line
4Be～47Ag Ka
48Cd～58Ce Ka, La
59Pr～92U La
1. Determine if overlapping lines exist from spectral chart
2. Choose lines without any overlap
3. When overlap can not be prevented, use overlap correction
4. Typically, L lines for thin films, K lines for bulk samples have better
sensitivity
 Sources of Error
In order to get a “perfect calibration” some potential sources of
error must be taken into account:
• Sample taking
• Sample preparation
• Instrument precision
• Chemical data of standards
• Inaccuracies during the regression analysis
(calibration line calculation process)
 Sources of Error
In order to get a “perfect calibration” some potential sources of
error must be taken into account:
• Sample taking

– Homogenity
– Representative of entire ship load composition
– Particle size
 Sample taking

Bulk
10x kg

Sampling

10 - 1000 g Pulverising,
Pelletising or
Fusing

Specimen
1 - 10 g

X- ray Spectrometer Precise and accurate result

Particle Size Effect
 Sources of Error
In order to get a “perfect calibration” some potential sources of
error must be taken into account:

• Sample preparation
 Sample Preparation

Solid (Metal / Polymer) Powder

Direct Polishing Direct Grinding Grinding

Pelletizing Fusion

Measurement Loose
Powder Briquette Fuse Bead

Measurement
Sample Preparation for Powder (Loose Powder Method)
Polyethylene sample cell Hold the cup and tap several opposite side: filter (plankton
CH1540 times it onto lightly on the net) HD-10 (Cat.No. 3399O271,
(Cat.No. 3399N187, 3399N197)
table to flatten the sample. 3399O181,
3399O182)
Sample amount
：more than 5g is
recommended

Polyester film (sample film)

(Cat.No. 3399G001)

Non-pulverized and pellet

samples can be measured
but with relatively large
analysis error.
Sample preparation for powders
(Briquette method)

Pressed
powder
4~5mm
Pressed
powder
4~5mm
Fuse Bead
Type of sample preparation
vs. Analytical precision
Low

w/o treatment

Analytical precision & homogeneity

Hot press
＋

Cryogenic gring +
Loose powder
＋
Cryogenic
gring + Hot
High press

Simple Complicated
Simplicity of sample preparation
 Sources of Error
In order to get a “perfect calibration” some potential sources of
error must be taken into account:

• Instrument precision

– Stability of X-ray source

– Stability of electronic circuit
– Counting loss
– Instrument calibration must be done periodically !!!
 Counting Loss

At high counting rates, detector displays counting loss

−N t 
N0 = N te
Measured count rate (No)

Counting Loss

N0(Nobs) ：measured count rate

(output) Nt(Ntrue) ：true count rate
Linear τ
(input) ：dead time
Range

SC 1000kcps
F-PC 2000kcps
True count rate
(Nt)
Peak Calibration
 Sources of Error
In order to get a “perfect calibration” some potential sources of
error must be taken into account:

• Chemical data of standards

– CRM ?
– Inhouse Standard ?
– Please be careful to input the number!!!
 Sources of Error
In order to get a “perfect calibration” some potential sources of
error must be taken into account:

• Inaccuracies during the regression analysis

(calibration line calculation process)
 Matrix effect (coexisting elements)
Primary X-rays

X-ray fluorescence from each element

A A C A C A A
B A B
B B
B
A
C B
B
A B
A
B C
B
Sample
A A A A
C C C C C A
C A

Characteristic X-rays from elements A，B，C

X-ray intensity correlates to concentration of
each element

However, X-ray intensity is effected by

sample matrix
 Effect of coexisting elements
Cd：100ppm

intensity
X-ray

Polyethylene (10mm)

Aluminum
alloy
Polyethylene
(2mm)
Copper
alloy

2θangle
Elements influencing each other
•No x-rays !
•Neighbor
•Neighbor

•Skip 1 el.
•Skip 2 el.
•Skip 3 el.

•Skip 4 el.
•Skip 5 el.
 Absorption Effect
Fe-Ni binary 2 element sample Mass absorption coefficient for Ni-Ka
Ni calibration curve

Mass absorption coefficient

500
FeKabs
NiKabs
Ni Kα intensity

(μ/ρ
)
Absorption by
Fe
Absorption by Ni
0
0 50 100 NiK CuK CoK NiK CoK
Ni mass% β α FeK
β
β α
α β
β α
Wavelength
 Enhancement effect
Enhancement
Fe-Cr 2 element sample Absorption
Cr calibration curve
effect Absorption effect (μ/ coefficient CrKab
s

Mass absorption coefficients

ρ)
500 FeKab
Cr only s
K
FeK
X-ray intensity

α α
Enhancement
effect CrK
Enhancement effect
Wave lengthα
Absorption effect only
0
NiK CoK FeK MnK CrK
CoK α FeK α MnK α CrKβα VK α
Cr mass% β β β β
wavelength
The calibration curve of Cr is concave down for Fe-Cr 2 element sample.
Because Cr-Ka is generated by both primary X-rays and Fe-Ka fluorescent X-
rays.
 Spectral Overlap
kcps
8.0

P-Ka is overlapped by Mo-Ll.

P-Kα
6.0

Mo-L l
4.0

2.0

0
138 142
 Overlap Correction
P in steel containing Mo

ΔP=0.019mass

% Mo 1.8mass% ΔP – WMo relationship

b ΔP: Deviation from
P Kα intensity

calibration curve
Mo 2.1mass% Δ
a P b a
0.9mass%
0.02 c
C Calibration
0.01 d
curve for P
0.4mass% d
0
W = aI- 0.010WMo+b 0 1 2
Mo mass%
ΔP = 0.01 WMo
0.01 : lj of overlapping
0 0.01 0.02 0.03 0.04 0.05 correction factor
P mass%
 Matrix correction: concept
For element i with true concentration value Wi and
apparent value Xi
When no effect of coexisting elements (matrix) exits, Wi = Xi

Wi／Xi = 1

When matrix effect exits、ratio deviates from 1. This deviation is

proportional to concentration Wj of coexisting element j.  is
for when multiple coexisting elements exist.
Xi：Quantitative value based on basis calibration
Wi／Xi = 1+∑ jdjWj curve (not corrected)
dj：absorption/enhancement correction coefficient
Wi：Concentration of element i
Wｊ：Concentration of element j
Effect of coexisting elements and matrix
correction
• When effect of coexisting elements exits
→matrix correction
Intensity

Intensity
Matrix
correction
concentration concentration
Lower Limit Detection
 Accuracy and Precision
Good trueness (small bias) Poor trueness (large bias)
Poor precision Good precision
Analysis
value
Mean
Mean
True value True value

Good trueness (small bias) Poor trueness (large bias)

Good precision Poor precision
Good accuracy

Mean
Mean
True value
True value